Modulation of Keratinocyte & Dermal Fibroblast Physiology by Selected Polysaccharides Thesis

MASARYKOVA UNIVERZITA

Přírodovědecká fakulta Ústav experimentální biologie

Disertační práce

Brno, 2011 Ema Ruszová Ema RUSZOVÁ Brno, 2011

MASARYKOVA UNIVERZITA

20B Přírodovědecká fakulta Ústav experimentální biologie

Modulation of Keratinocyte and Dermal Fibroblast

Physiology by Selected Polysaccharides

Disertační práce

Ema Ruszová

Školitel: Mgr. Lukáš Kubala, Ph.D. Brno, 2011

Bibliografický záznam

Jméno a příjmení autora: Ema Ruszová, Mgr.

Přírodovědecká fakulta Masarykovy univerzity

Ústav experimentální biologie

Název disertační práce: Modulace fyziologie keratinocytů a kožních

fibroblastů vybranými polysacharidy

Studijní program: Biologie

Studijní obor (směr), kombinace oborů: Fyziologie živočichů

Školitel: Mgr. Lukáš Kubala, Ph.D.,

Ústav experimentální biologie,

Přírodovědecká fakulta Masarykovy univerzity

Rok obhajoby: 2011

Klíčová slova v češtině: polysacharidy, keratinocyty, dermální

fibroblasty, stárnutí kůže, UV záření,

extracelulární matrix, zhášeči volných radikálů

Bibliographic entry

Author: Ema Ruszová, Mgr.

Faculty of Science, Masaryk University

Department of Experimental Biology

Title of dissertation: Modulation of Keratinocyte and Dermal

Fibroblast Physiology by Selected

Polysaccharides

Degree programme: Biology

Field of study: Animal physiology

Supervisor: Mgr. Lukáš Kubala, Ph.D.,

Faculty of Science, Masaryk University

Department of Experimental Biology

Year of defence: 2011

Keywords: polysaccharides, keratinocytes, dermal

fibroblasts, skin aging, UV-exposure,

extracellular matrix, ROS-scavengers

© Ema Ruszová, Masarykova univerzita, 2011

Poděkování

Ráda bych poděkovala všem, kdo mi pomohli napsat tuto práci. Především děkuji

mému školiteli RNDr. Lukášovi Kubalovi, PhD. za vedení nejen disertační práce,

ale i celým mým doktorantským studiem: za jeho podnětné konzultace, a za jeho

ochotu mi téměř kdykoliv a s čímkoli poradit. Kamarádce a kolegyni RNDr.

Veronice Hájkové, PhD. za její schopnost mi naslouchat v nelehkých chvílích

studia. Firmě CPN spol. s r.o., v čele s ředitelem RNDr. Vladimírem Velebným,

Csc. za možnost nejen studia, ale i využítí vybavení Laboratoře Dermálních a

Kosmetických Aplikací a v posledním roce i Martině Hašové, Mgr. a Iloně

Matějkové, Ing. za jejich technickou asistenci. Největší dík však patří mému

muži za jeho také neskonalou dobrotu a každodenní podporu a v neposlední řadě

chci velmi poděkovat jeho rodičům za pomoc při hlídání našich dětí a převzetí

povinností rodičů v případě naší nepřítomnosti.

Abstrakt

Již v dřívějších studiích bylo poukázáno na kapacitu nejrůznějších polysacharidů

inhibovat stárnutí buněk, respektive stárnutí kůže. Mezi hlavní popisované efekty patří

inhibice proteolytické degradace složek extracelulární matrix a tím inhibice produkce

pro-zánětlivých a pro-apoptických peptidů, dále inhibice Maillardovy reakce a přímý

pro-proliferativní účinek. Schopnost studovaných polysacharidů vychytávat volné

radikály je považována za jeden z klíčových mechanismů ochrany buněk exponovaných

ultrafialovému záření a taktéž významně zasahuje do procesů spojených se stárnutím

buněk indukovaným zářením. Poplysacharidy působí na buňky aktivně přes interakci s

několika typy receptorů rozpoznávajících polysacharidy, které byly prokázány také na

povrchu buněk kůže. Mezi běžné interagující receptory patří lektiny rozpoznávací

galaktózu, které jsou schopny vázat se na galaktosyl- i mannosyl- sacharidické

jednotky.

Hlavním úkolem této práce bylo zhodnotit vliv vybraných polysacharidů na

fyziologii keratinocytů a kožních fibroblastů v kontextu procesů spojených se stárnutím

buněk indukovaným zářením a celkovým stárnutím kůže. Pro studium byly vybrány tři

různé typy polysacharidů: glukomannany, oligo/polysacharidy bohaté na rhamnózu a

oligo/polysacharidy bohaté na fukózu. Jejich efekty byly studované jak na izolovaných

keratinocytech a dermálních fibroblastech in vitro tak na dobrovolnících in vivo.

Výsledky ukazují, že různé strukturní charakteristiky polysacharidů ovlivňují jejich

účinek na buňky, zvláště z hlediska modulace profilování genové exprese. Mimo jiné

výsledky naznačují možnost zapojení různých signálních drah vedoucích od aktivace

receptorů rozpoznávajících rhamnózu/fukózu s následnými dopady na modulaci

pojivové tkáně, extracelulární matrix a vlastností kůže obecně. Celkově práce prokazuje

významné účinky studovaných polysacharidů na keratinocytech a fibroblastech kůže a

jejich případné použití při ošetření pleti proti jejímu stárnutí nebo při její ochraně před

poškozením ultrafialovým světlem.

Abstract

The inhibition of harmful, age-related processes in skin and in cells in general

by different polysaccharides has been proposed by various authors. The main described

effects of polysaccharides include the inhibition of the production of detrimental

peptides during their proteolytic degradation, the inhibition of the Maillard reaction, and

direct pro-proliferative effects of polysaccharides. Furthermore, the ability of the

various polysaccharides to scavenge free radicals is also considered to significantly

protect ultraviolet-exposed cells and to intervene in the photoageing processes. Several

polysaccharide recognising receptors have been documented on the skin cell surface.

Galactose–recognising lectins are among the common interacting receptors, which are

able to bind to galactosyl as well mannosyl saccharidic units.

The main task of this thesis was to evaluate the effects of selected

polysaccharides on fibroblast and keratinocyte physiology in the context of skin ageing

and photoageing processes. Complex effects of three different types of polysaccharides,

glucomannan, rhamnose rich-oligo/polysaccharides and fucose-rich

oligo/polysaccharides have been studied both on keratinocytes and dermal fibroblasts in

vitro and on healthy volunteers in vivo. Results show that the different structural

features of studied polysacharides were crucial for their effects on cells, particularly the

modulation of gene expression profiles. Moreover, the results suggest involvement of

different signalling pathways delivering a signal from rhamnose/fucose–receptors to

intracellular targets, with potential consequences on anti-connective tissue ageing

properties.

Overall, the thesis presents data describing complex effects of different

polysaccharides on keratinocytes and skin fibroblasts and their potential use in skin

treatment to prevent skin ageing or damage by ultraviolet light.

viii

Content 1. Introduction ................................................................................................................... 1

2. Epithelial tissue ............................................................................................................. 3

2.1. The skin ....................................................................................................................... 5

2.1.1. The epidermis ................................................................................................ 5

2.1.2. The dermis...................................................................................................... 11

2.1.3. The hypodermis ............................................................................................. 14

3. Skin aging and UV-exposure ....................................................................................... 15

3.1. Ageing and oxidative stress ............................................................................. 15

3.2. Alternations of ECM during skin ageing ....................................................... 16

3.3. Postsynthetic mechanisms of skin ageing ....................................................... 17

4. Polysaccharides ............................................................................................................. 18

4.1. Structure of polysaccharides selected for the evaluation in this study ........ 20

4.1.1. Glucomannan (GM) ...................................................................................... 20

4.1.2. RROPs ............................................................................................................ 20

4.2. Recognition of polysaccharides by cell surface receptors ............................. 22

5. Previously reported effects of the polysaccharides tested in this study ................... 23

5.1. GM ..................................................................................................................... 23

5.2. RROPs ............................................................................................................... 23

5.3. FROPs ................................................................................................................ 24

7. Published articles .......................................................................................................... 26

7.1. Photoprotective effects of glucomannan isolated from Candida utilis ........ 26

7.2. Effect of advanced glycation endproducts on gene expression profiles of

human dermal fibroblasts ....................................................................................... 37

7.3. Rhamnose recognizing lectin site of human dermal fibroblasts functions as a

signal transducer ..................................................................................................... 43

7.4. Receptors and aging: Structural selectivity of the rhamnose-receptor on

fibroblasts as shown by Ca2+-mobilization and gene-expression profiles ......... 50

7.5. Pharmacological properties of rhamnose-rich polysaccharides, potential

interest in age-dependent alterations of connectives tissues ................................ 57

8. Discussion ...................................................................................................................... 63

8.1. UV irradiation crucially influences fibroblast response to GM ................... 63

8.2. Regulation of gene expression in fibroblasts by AGE ................................... 64

8.3. Effect of different MW of RROPs on fibroblast response ............................ 65

8.4. Effects of rhamnose/fucose rich-polysaccharides on different transduction

pathways ................................................................................................................... 66

9. Conclusion ..................................................................................................................... 67

10. References .................................................................................................................... 68

11. Abbreviations .............................................................................................................. 73

1

1. Introduction

Next to numerous, more investigated compounds extracted from plants, fungi, and

bacteria, polysaccharides have recently been growing in interest due to the wide variety

of physiological and biological activities of these polymers. Although numerous studies

have presented evidence showing their antimicrobicial activity associated with a strong

immunostimulatory effect, papers demonstrating their influence on non-immune cells

are still rare.

The biological activity of purified carbohydrates, including a decrease of infectious

complications and inhibiting tumour growth, is known to depend on their structure.

Parameters such as primary structure, degree of branching, molecular weight, solubility,

solution conformation, and ionic charge have been suggested to play a role in

determining the biological activity of these molecules.

This thesis will focus on polysaccharides of two different origins: the first,

glucomannan (GM) which pertains to carbohydrates isolated from the yeast Candida

utilis, and the second, Rhamnose Rich Oligo- and Polysaccharides (RROPs) and Fucose

Rich Oligo- and Polysaccharides (FROPs) that are polysaccharides obtained from the

Clebsiella bacterial strains particularly Clebsiella pneumonie or Clebsiella planticola.

Ultraviolet irradiation (UV) B light was used as a model for the evaluation of the

adverse effects of solar radiation on skin. Within the spectrum of sunlight, UVB light

has the lowest potential to penetrate skin, but is considered to be the main mediator of

cutaneous damage and inflammation leading to the formation of skin carcinomas and to

the alteration of cutaneous immune responses. The adverse effects of UVB on

cutaneous cells could be prevented by the activation of cellular mechanisms protecting

cells against the damaging effects of toxic agents such as UVB irradiation. The

cytoprotective mechanisms could be triggered by various factors including substances

recognised by receptors for microbial molecular patterns or Pathogen-Associated

Molecular Patterns (PAMPs), comprising lipids, carbohydrates, proteins, and nucleic

acids, because their molecular structure is distinct from those expressed on the surface

of mammalian cells.

The data obtained describes how epithelial cell response pivotally varies based

on embedded conditions such as the molecular weight (MW) of polysaccharides,

structure, and pretreatment conditions (e.g. UV light). The aim is to show how diverse

modulatory activity can be exhibited by these, with their tendency to restore physiology

2

status of epithelial cells and potency to work as “true” modulators of signal

transduction.

3

2. Epithelial tissue

Epithelia function to protect the underlying tissues from environmental influences

such as physical damage, bacterial infection, desiccation, ultraviolet radiation and heat

loss to maintain homeostasis (Presland and Dale 2000). The functions of Epithelia

include lining surfaces, forming glands and acting as receptor cells in sensory organs.

Epithelia tissues consist of squamous, cuboidal, columnar or polyhedral cells that are

attached to one another via short cell processes. These cells produce only a small

amount of intercellular substances (Bragulla and Homberger 2009). Cells of secretory

epithelia are polarized in order to direct secretory vesicles to the apical surface and

absorb metabolites on the basolateral surfaces (Frappier, 2006).

In general, epithelia are distinguished between simple, transitional, or stratified

according to (Bragulla and Homberger 2009):

Simple epithelia

Two types of simple epithelia are distinguished, i.e., single-layered and multi-

layered epithelia. In single-layered epithelia, all cells are attached to the base membrane

and extend to the surface of the epithelium (e. g. endothelium, mesothelium, epithelium

in the renal tubules and alveoli). In multi-layered epithelia (i.e., pseudostratified

epithelia), all cells are in contact with the base membrane but do not necessarily extend

to the surface of the epithelium (e. g. respiratory epithelium).

Transitional epithelia

In transitional epithelia, at least some cells attach to the base membrane and also

extend to the surface of the epithelium (e.g. epithelium of the renal pelvis, ureter and

urinary bladder (also called “urothelium”).

Stratified epithelia

In stratified epithelia, only the basal cells are attached to the base membrane and

only the most superficial of the suprabasal cell layers form the surface of these epithelia.

In the intermediate stratum or stratified epithelium, the cells undergo various processes

of terminal differentiation, such as keratinization. In general, two types of stratified

epithelia are distinguished, namely the stratified keratinized but non-cornified epithelia

4

(e.g., epithelium of the oral cavity, esophagus, vagina, urethra) and the stratified

keratinized-cornified epithelia.

Stratified soft-cornified epithelia

The epithelial cells in the superficial stratum (i.e the Stratum corneum, SC), the

corneocytes, are cornified and dead. Cornification requires the previous keratinization

of cells, including the addition of a proteinaceous layer (i.e., the cornified envelope CE),

on the cytoplasmatic surface of the cell membrane.

Stratified soft cornified epithelia, the suprabasal cell layers, include a Stratum

granulosum, and are characterized by the presence of basophilic keratohyalin granules.

In the process of soft cornification, a filament-matrix complex and proteinaceous

cellular envelope are formed on the inside of the cell membrane. The superficial

cornified cells of stratified non-modified soft-cornified epithelia (e.g., epidermis)

desquamate continuously and readily.

Stratified hard-cornified epithelia

In stratified, hard-cornified epithelia, the suprabasal cell layers, including the hair

cortex and human fingernail plate do not desquamate but are worn off.

From a functional point of view, the simple epithelia carry out the most diverse

activities, which include absorption, excretion, synthesis, secretion, and sensory

reception, whereas the stratified epithelia have protective functions, serve as conduits or

ducts, and produce reproductive cells. In order to serve their distinctive functional roles,

epithelial cells often display distinctive cell membrane or surface modifications and

appendages (Frappier, 2006).

5

2.1. The skin

The skin, or cutis, covers the entire outer surface of the body. The skin is arranged

in three layers. These are, from outside to in: a) the epidermis (and its associated

appendages, pilosebaceous follicles, and sweat glands); b) the dermis, separated from

the epidermis by the dermo-epidermal junction (DEJ); and c) the hypodermis (Kanitakis

2002).

Skin is capable of maintaining an anatomical barrier that protects the „inside“

from „outside“ dangers, i.e. pathogens, radiation (such as UV), toxic agents and

mechanical stress (Proksch, Brandner et al. 2008). Simultaneously, this barrier prevents

the loss of heat, fluid and electrolytes from the inside. The skin acts as a temperature

regulator, a sensory system, a sexual medium, excretes waste, and participate

significantly in vitamin D synthesis (Arck and Paus 2006; Paus, Theoharides et al.

2006; Reichrath 2007).

2.1.1. The epidermis

The epidermis is a keratinized stratified squamous epithelium (Figure 1). It

functions as a barrier protecting the body from dehydration and environmental insults

(Koster 2009). It is comprised of multiple cell types that are derived from different

embryonic origins: keratinocytes, melanocytes, Langerhans cells and Merkel cells. The

renewal of the human epidermis takes about 3 to 4 weeks (Haake, A., 2000).

Basal keratinocytes divide, with daughter cells migrating into the overlying

spinous cell layer. Keratinocytes in the spinous cell layer move sequentially into a

granular cell layer, and eventually move into the the outer cornified layer of the

epidermis. Basal, spinous and granular cell are alive, whereas the terminally

differentiated cell of the stratum corneum (called corneocytes) are no longer living.

Strong mechanical support in the epidermis is provided by an extensive intracellular

network of keratin filaments that connect to cell-to-cell junctions (desmosomes) and

basal cell-basal lamina attachments (hemidesmosomes) (Green and Simpson 2007;

Bolling and Jonkman 2009; Sandjeu and Haftek 2009).

6

Figure 1: Illustration of the skin - on the left side, the epidermis showing the different

layers and the dermis below, and on the right side the hemidesmosome-stable adhesion

complex connecting the basal cells to the dermal extracellular matrix (Bolling and

Jonkman 2009).

Structure of the epidermis

Five structurally different layers can be identified:

1. The stratum basale (stratum germinativum) is the deepest layer of the

epidermis (closest to the dermis). It consists of a single layer of columnar

or cuboidal cells which rest on the base membrane. The first

subpopulation of basal cells are the stem cells (transit-amplifying cells)

with high levels of β1- integrin (Koster 2009), proliferating cell nuclear

antigen (PCNA) and Ki67 (Kanitakis 2002). The most abundant

constitutive integrins in the epidermis are α2β1 (collagen receptor), αβ1

(predominantly laminin-5 receptor) and α6β4 (laminin receptor) (Watt

2002). Secondly, the basal cell layer consists primarily of mitotically

active keratinocytes and immigrant cells that contain fine bundles of K5

and K14 keratin filaments, which provide a cytoskeleton that has

sufficient flexibility to permit cell division and migration (Haake, A.,

2000).

7

2. In the stratum spinosum, the cells become irregularly polygonal. The

abundance of focal junctions (desmosomes) between adjacent

keratinocytes results in the formation of „spines“ due to the shrinkage of

artifacts during tissue processing. K1/K10 are newly synthesized in

spinous cells and often referred to as the differentiation-or keratinization-

specific keratins (Butnaru and Kanitakis 2002; Kanitakis 2002).

3. The stratum granulosum consists, in thick skin, of a few layers of

flattened cells. Only one layer may be visible in thin skin. The

cytoplasma of the cells contains numerous fine grains, keratohyalin

granules (composed primarily of an electron-dense protein, profilaggrin,

keratin intermediate filaments, and loricrin). The keratohyalin is not

located in membrane-bound organelles but forms "free" accumulations in

the cytoplasm of the cells. The cells begin to release the contents of the

lamellar granules. The lipids contained in the granules come to fill the

entire interstitial space, which is important for the function of the

epidermis as a barrier against the external environment. In the more

superficial layers of the stratum corneum, cholesterol sulfatase modifies

the lipids, which appears to be important for hydration and desquamation.

Involucrin, keratohyalin, loricrin, small proline-rich proteins (cornifin,

SPR1, SPR2) elafin, filagrin linker-segment peptides and envoplakin

have all been found as components of the cornified envelope (CE).

Proteins of the cornified envelope are rendered insoluble by

transglutaminase-catalyzed cross-linking to form a 7-15 nm thick layer,

which is evident morphologically only in cornified cells (Haake, A.,

2000, Butnaru and Kanitakis 2002; Kanitakis 2002).

4. The stratum lucidum (the transitional layer) consists of several layers of

flattened dead cells. Ultimately, the terminal differentiation program

results in the formation of corneocytes, keratinocytes which have lost

their nucleous and cytoplasmatic organelles (Koster 2009). This is driven

by several different degradative enzymes that have been identified in

granular cells and involves the destruction of cellular organelles. The

morphological and biochemical changes in the nucleus are characteristic

8

of an apoptotic form of cell death, including internucleosomal

fragmentation of DNA, and activation of caspases, although the other

morphological features of apoptosis are not apparent (Haake, A., 2000).

5. In the stratum corneum, cells are completely filled with keratin filaments

(horny cells) which are embedded in a dense matrix of proteins.

Individual cells are difficult to observe because (1) nuclei can no longer

be identified, (2) the cells are very flat and (3) the space between the cells

has been filled with lipids, which cement the cells together into a

continuous membrane. The corneocyte is the largest of the keratinocytes

and is a flattened, polyhedral shape. The shape and features of the

corneocyte are adapted to maintain the integrity of the stratum corneum,

yet allow for desquamation. Closest to the surface of the epidermis, the

stratum corneum has a somewhat looser appearance. Horny cells are

constantly shed from this part of the stratum corneum. The stratum

corneum provides the major skin barrier to water loss and permeation of

environmental substances, as well as contributing to mechanical

protection (Haake, A., 2000). The SC provides a barrier against

marauding pathogens (Marks 2004).

Specialized cells in epidermis

Melanocytes –pigment producing cell of the epidermis

The brown colour component is due to melanin, which is produced in the skin itself

in cells called melanocytes (typically 1000-2000 per sq. mm). These cells are located in

the epidermis and send fine processes between the other cells. In the melanocytes, the

melanin is located in membrane-bound organelles called melanosomes. Human skin

pigmentation varies among populations in a striking manner. This has sometimes led to

the classification of people(s) on the basis of skin colour (Costin and Hearing 2007).

Melanocytes can transfer melanin to keratinocytes - mainly to the basal cells. The

fine processes of melanocytes may invade keratinocytes and bud-off part of the

melanocyte cytoplasm, including the melanosomes, within the keratinocytes. Melanin

9

protects the chromosomes of mitotically active basal cells against light-induced damage

(Lin and Fisher 2007).

Langerhans Cells

Langerhans cells are another cell type found within the epidermis. Morphologically

they are not unlike melanocytes, but functionally they are more closely related to

macrophages. They are important in immune reactions of the epidermis. Their fine

processes form a network between the cells of the epidermis and phagocytose antigens

which have entered the epidermis. Langerhans cells may only be temporary residents of

the skin. If they have come into contact with an antigen, they can migrate to regional

lymph nodes, where they initiate an immune response (Toebak, Gibbs et al. 2009).

Langerhans cells express antigens conjugated with major histocompatibility complex

(MHC) class II positive molecules on their surfaces for presentation to T-helper

lymphocytes (Lipozencic and Ljubojevic 2004).

T-lymphocytes

T-lymphocytes, together with Langerhans cells, are sometimes referred to as SALT,

i.e., skin-associated lymphoid tissue (Lipozencic and Ljubojevic 2004).

Epidermal structural Proteins: Function in Health and Disease

Intermediate filaments in epithelia: keratin filaments

Intermediate filaments in epithelial cells (i.e., keratin filaments) are made of

keratins, and these keratins are found only in vertebrates. In human stratified epidermis,

keratins account for 25-35% of the extracted proteins. Different types of keratins are

distinguished according to various characteristics, such as physicochemical properties,

and according to the cells and tissues (Figure 2). Keratin filaments are attached to the

cell surface via specialized cell adhesion junctions termed desmosomes (Presland and

Jurevic 2002).

10

Figure 2: Keratin expressions in stem cells of the Stratum basale, in stratified keratinized,

and stratified keratinized-cornified epithelia (Presland and Jurevic 2002) .

The formation of the SC in the process of keratinization is complex providing

multiple opportunities for disorders to arise (Marks 2004).

Mutations in keratin genes have been found to cause several dominantly inherited

skin diseases, including epidermolysis bullosa (EB) simplex (involving mutations in K5

and K14), epidermolytic hyperkeratosis (mutations in KI and K10), ichthyosis bullosa

of Siemens (K2e mutations), pachyonychia congenita (mutations in K6, K 16, and K

17), and epidermolytic palmoplantar keratoderma (EPPK) that results from mutations in

K9, a keratin expressed in the palm and sole epidermis (Table 1).

Other disorders in which there are alterations in profilaggrin/filaggrin expression

include Harlequin ichthyosis and epidermolytic Hyperkeratosis.

The evidence that the desmosomal cadherins, and in particular desmoglein plays an

important role was found in studies to concern two diseases, pemphigus vulgaris (PV)

and pemphigus foliaceus.

To date, there are two skin diseases associated with CE formation and function:

lamellar ichthyosis, a recessive disorder involving mutations in the enzyme

transglutaminase 1, and loricrin keratoderma, a dominant disease resulting from

mutations in loricrin (Presland and Dale 2000).

11

Table. 1: Examples of epidermal proteins associated with skin diseases.

2.1.2. The dermis

The dermis is the connective tissue component of the skin and provides its pliability,

elasticity, and tensile strength. The dermis is less cellular than the epidermis, being

composed primarily of fibrous and amorphous cells surrounding the epidermally

derived appendages, neurovascular networks, sensory receptors and dermal cells.

The dermis is organized into two regions on the basis of the difference in connective

tissue density and arrangement: the uppermost papillary dermis and the lower reticular

region. The papillary dermis underlies the epidermis and is approximately two times its

thickness. The papillary dermis is characterized by small bundles of small-diameter

Mutated gene Associated skin

diseases

Source

Keratin 5, 14 Epidermolysis bullosa (Presland and Dale

2000)

Laminin-5 Epidermolysis bullosa (Presland and

Jurevic 2002)

Integrins (α6, β4) Epidermolysis bullosa (Presland and

Jurevic 2002)

K1, K10 Epidermolytic

hyperkeratosis

(Presland and Dale

2000)

Profillagrin/fillagrin Harlequin ichtyosis

Epidermolytic

hyperkeratosis

(Presland and

Jurevic 2002)

Desmosomal

proteins

Desmogleins,

desmoplakins

Palmoplantar

keratoderma

pemphigus vulgaris

(PV)

(Presland and

Jurevic 2002)

Transglutaminase-1

(CE)

Lamallar ichtyosis (Presland and

Jurevic 2002)

Loricrin (CE) Loricrin keratoderma (Presland and

Jurevic 2002)

12

collagen fibrils and oxytalan elastic fibres. The structural characteristics of the matrix in

papillary dermis permit the skin to accommodate mechanical stress. The papillary layer

supplies nutrients to select layers of the epidermis and regulates temperature. The cells

also produce larger amount of proteoglycans. The reticular dermis is composed

primarily of large-diameter collagen fibrils organized into large, interwoven fibre

bundles. Mature, band-like, branching elastic fibres form a superstructure around the

collagen fibre bundles. These two fibre systems are integrated, providing the dermis

with strong and resilient mechanical properties. It also supports other components of the

skin, such as hair follicles, sweat glands, and sebaceous glands (Haake, A., 2000,

Holbrook and Smith 2002).

The primary cell type of the dermis is fibroblast. Dermal fibroblasts are essential

component of skin; they not only produce and organize the extracellular matrix of the

dermis but they also communicate with each other cell types, playing a crucial role in

regulating skin physiology. Three subpopulations of fibroblasts reside in distinct dermal

layers: the papillary and reticular dermis and third group is associated with hair

follicles. Papillary fibroblasts divide at faster than do reticular fibroblasts. Reticular

fibroblasts contracs faster than do papillary fibroblasts (Sorrell and Caplan 2004).

Dermal connective tissue matrix

Collagen and elastic connective tissue are the main types of fibrous connective tissue

of the dermis. There are also non-fibrous, connective tissue molecules, including fine

filamentous glycoproteins and proteoglycans and glycosaminoglycans. The epidermis

and dermis are connected by the base membrane zone (BMZ), also called dermal-

epidermal junctions (DEJ), consisting of specialised aggregation of attachment and

signalling molecules (Table 2., Bolling and Jonkman 2009).

13

Table. 2. Distribution of selected extracellular matrix molecules in dermal

compartments by fibroblasts

Elastic fibres of the dermis

The elastic fibre matrix component, elastin, is processed from a secreted, soluble,

precursor tropoelastin molecule. A number of other molecules, including vitronectin, a

decay-accelerating factor, and fibronectin, are also associated with elastic fibres in the

skin. The turnover of elastic fibres is normally very slow in dermis, but may be

accelerated by ultraviolet light and inflammation. Dermal elastosis is one of the

hallmarks of photodamage in the skin (Muto, Kuroda et al. 2007).

Fibronectin (in the matrix), laminin, thrombospondin, vitronectin and tenascin are

glycoproteins found in the dermis and, like the proteoglycans and glycosaminoglycans,

they interact with other matrix components and with cells through specific integrin

receptors (Egles, Huet et al.). As a consequence of their binding to other glycoproteins,

collagen and elastic fibres, they are involved in, and in some cases mediate, cell

attachment (adhesion), migration, spreading (in vitro observations), morphogenesis and

differentiation (El Ghalbzouri, Lamme et al. 2002; Silvestre, Kenis et al. 2009).

Matrix

component

Papilary dermis Reticualr dermis

Collagen I and

III

High ratio III to I low ratio III to I

Collagen IV Present in base

membrane

absent

Collagen VI present in DEJ region weakly present

Collagen XII present present

Collagen XIV low to absent present

Collagen XVI present in DEJ region absent

Tenascin-C Present in DEJ region present

versican diffuse in DEJ associated with elastin

Laminin-1 present in DEJ region absent

14

2.1.3. The hypodermis

The hypodermis is not part of the skin, and lies below the dermis. Its purpose is to

attach the skin to underlying bone and muscle as well as supplying it with blood vessels

and nerves. The main cell types are fibroblasts, macrophages and adipocytes (the

hypodermis contains 50% of body fat). It plays an important role in thermoregualtion,

insulation, provision of energy and protection from mechanical injuries (Butnaru and

Kanitakis 2002; Kanitakis 2002).

15

3. Skin aging and UV-exposure

3.1. Ageing and oxidative stress

Skin changes are among the most visible signs of ageing. Unlike other organs, skin

is in direct contact with the environment, and therefore undergoes ageing as a

consequence of environmental damage. Hereditary genetic influences are now

considered to represent no more than 3% (Robert, Labat-Robert et al. 2009).

Skin ageing is characterized by signs like discolouration, wrinkles and texture loss.

The two major types of skin ageing are intrinsic ageing and extrinsic ageing. Intrinsic

ageing is genetically programmed ageing, also called chronological ageing, whereas

extrinsic ageing is due to environmental factors like sunlight, mainly UV rays

(commonly known as photoageing), stress and pollution (Chauhan and Shakya 2009).

Although both photoageing and chronological ageing is associated with wrinkles,

photoageing-induced wrinkles are regarded to be deeper and more coarse than those

associated with chronological ageing (Papanagiotou 2009).

Oxidative stress has been reported to play an important role in the development of

the various detrimental effects of UV (Callaghan and Wilhelm 2008). This increase in

reactive oxygen species (ROS) may not only alter the structure and function of many

genes directly (mitochondrial and nuclear DNA damage) (Greinert, Boguhn et al. 2000;

Ouhtit and Ananthaswamy 2001; Laga and Murphy 2009), proteins (Rocquet, Bonte,

2002), and membrane damage (eg. lipid peroxidation) (Sorg, Antille et al. 2006), but

may also modulate their expressions through signal transduction pathways and,

ultimately, lead to skin damage. Constitutive activation of cell surface receptors (EGFR,

IL-1R, TNFR), which trigger intracellular signalling via stress-associated mitogen-

activated protein kinases (e.g. p38, JNK) ultimately results in the nuclear transcription

of a complex known as activator-protein 1 (AP-1) and nuclear factor-kappa B (NF- B)

(Svobodova, Psotova et al. 2003; Laga and Murphy 2009).

Meanwhile, the keratinocytes are able to secrete a wide variety of proinflammatory

factors upon UV exposure, including NF-κB-driven production of tumor necrosis factor

(TNF)-α, interleukin (IL)-1, IL-6, and IL-8 (Basile, Eichten et al. 2003) which

significantly modulate physiology of other cell types present in skin including

fibroblasts.

16

Fibroblast extrinsic ageing is directly related with alternations in extracellular matrix

(ECM) structure and function.

3.2. Alternations of ECM during skin ageing

Extrinsic ageing is connected with changes in the composition and turnover of ECM.

Collagen content in skin ECM decreases during the processes of chronoageing and

photoageing (Chung, Seo et al. 2001; Kim, Cho et al. 2006; Fisher, Quan et al. 2009). A

decrease in collagen due to natural skin ageing may arise from its reduced synthesis and

increased degradation due to an elevation of matrix metalloprotease expression (MMP)

(Kim, Cho et al. 2006). The MMPs are a family of proteolytic enzymes that specifically

degrade collagen, elastin and other proteins. The transcription of the genes for the

MMPs can be triggered by increased oxidative stress, and is dependent

on the

transcription factors AP-1 and NF- B, as mentioned above (Kim, Kim et al. 2004;

Wenk, Schuller et al. 2004; Fisher, Quan et al. 2009). (Kim, Kim et al. 2004; Wenk,

Schuller et al. 2004). MMP-1 (collagenase 1) preferentially degrades fibrillar collagens,

which maintain the strength of connective tissues, whereas MMP-3 (stromelysin-1)

degrades a broad array of proteins. UVA/B is known to induce the expressions of

interstitial MMP-1 and -3 normal human dermal fibroblasts (NHDF) in vitro and normal

human epidermis in vivo.

Hyaluronic acid (HA) undergoes extensive metabolism in skin tissue. Hyaluronan

syntheses (HAS1, HAS2, and HAS3) are responsible for cell and tissue specific

regulated synthesis of HA (Girish, Kemparaju et al. 2009). Catabolism of HA is

achieved through the family of hyaluronidases (Hyal). In particular, Hyal 2 has been

suggested to be responsible for the degradation of HA in connective tissue ECM

(Jenkins, Thomas et al. 2004). To date, little is known about changes in HA metabolism

that occur as a consequence of UV-induced damage to the skin. In general, it is

suggested that UV-induced degradation of HA by Hyals or down-regulation of HA

synthesis due to inhibition of HAS result in a loss of HA from the skin during the

process of photoageing (Dai, Freudenberger et al. 2007).

Elastin is the third primary skin constituent. There are two pathological phenomena

relating to elastin on the circumstances of UV-exposure. At first, UV-mediated elafin

interacts with elastin and the elafin-elastin complex protects altered elastic fibres from

elastolytic degradation, leading to accumulation of elastic fibres in the actinic elastosis

17

of sun-damaged skin (Muto, Kuroda et al. 2007). These abnormal elastin fibres exhibit

less elasticity, and the elastin fibre system formed from these fails to have the same

quality of organization as found during periods of normal elastin production (Bernstein,

Brown et al. 1997). Secondly, UV-induced elastase-type activity promotes also elastin

synthesis possibly in an indirect way following the release of breakdown products of

elastotic fibres. When safeguards such as elafin or SLPI (skin derived

antileukoproteinase) fail, excess elastase activity cannot be controlled (Schalkwijk

2007). Authors report that skin fibroblast elastase play an essential part in the

degeneration and/or tortuosity of elastic fibres induced by cumulative ultraviolet B

irradiation (Tsukahara, Takema et al. 2001).

3.3. Postsynthetic mechanisms of skin ageing

Conclusively, loss of cells, loss of ECM and switching to senescent phenotype are

significant attributes of skin ageing. The next mechanism of skin ageing that can not be

excluded is postsynthetic mechanisms of skin ageing including glycation by reducing

sugars, proteolytic production of toxic peptides, and an accumulation of progerin, a

dominant negative form of lamin A occuring in skin during the ageing process

(Rodriguez, Coppede et al. 2009; Robert, Labat-Robert et al. 2009).

It has been shown that senescent fibroblasts, compared to young, proliferation

competent fibroblasts, contain a high steady-state level of FN mRNA (Kumazaki and

Mitsui 1995). Interestingly, increased levels of FN have also been detected in plasma

and intracellularly in vascular endothelial cells and skin fibroblasts from aged

individuals (Ksiazek, Mikula-Pietrasik et al. 2009). So, its production increases with

age. One fragment of fibronectin (FN) was shown to up-regulate the biosynthesis of FN.

Similar processes were found to be induced by degradation products of other ECM

macromolecules such as elastin peptides (Robert, Labat-Robert et al. 2009).

18

4. Polysaccharides

As a matter of fact, polysaccharides are biopolymers composed of differently linked

single sugar molecules. Polysaccharides are the major components of the

microbial/plant cell wall and provide multiple functions, ranging from the carriers of

immunochemical specificity and marker molecules, by which cells recognize each other

and interact with their environment, to the skeletal substances that define stability,

shape, and morphology of the cell (Kogan, Pajtinka et al. 2008). Parameters such as

primary structure, degree of branching, molecular weight, solubility, solution

conformation, and ionic charge have been suggested to play a role in determining the

biological activity of these molecules and the field of their potential use. The

biotechnological production of industrial polymers has received great attention over the

last two decades. Due to the variety of structures, microbial/plant polysaccharides have

been exploited in many different ways and a large number of microbial/plant

polysaccharides are currently being commercialized in products used by the general

public, and are also the subject of industrial or research study and development.

There have been studies concerned with the in vitro keratinocyte/fibroblasts cell

response to different polysaccharide treatment. Table 3. shows that except for heparin

analogs, alginate and pectine-like polysaccharides, all of them have positively affected

proliferation of epithelial cells.

Table. 3: Effects of different polysaccharides on physiology of skin cells

Beta- glucan Up-regulated

proliferation, increased

early differentiation

NHDF Lee, B. Ch. et al.,

2003

Fucosylated

oligo/polysaccharide

increased early

differentiation

NHK (Deters, Lengsfeld

et al. 2005)

Fucose rich oligo/

polysaccharides

(FROP)

Up-regulated

proliferation, antioxidant

activity, inhibition of

MMP2 and MMP9

expression

NHDF (Peterszegi, Isnard

et al. 2003)

Rhamnose rich

oligo/

polysaccharides

(RROPs)

Up-regulated

proliferation, ECM

synthesis

NHDF (Andres, Molinari et

al. 2006)

19

arabinogalactans Up-regulated cell viability

NHDF,

NHK

(Peterszegi, Isnard

et al. 2003; Zippel,

Deters et al. 2009;

Zippel, Deters et al.

2009)

arabinoxylans Up-regulated proliferation

increased early

differentiation

NHK, NHDF (Deters, Schroder et

al. 2005)

Fruit rhamno-

galactouronans

Up-regulated proliferation HaCaT,

NHK

(Deters, Lengsfeld

et al. 2005;

Gloaguen, Krausz et

al. 2008)

Pectin- like

polysaccharides

Up-regulated cell viability

Down-regulated

proliferation

HaCaT,

NHK

(Deters, Lengsfeld

et al. 2005)

CMBDS, heparin

analog

Down-regulated

proliferation

Down-regulated collagen

III synthesis

NHDF (Senni, K., et al.

1997)

Alginate

oligosaccharides

Down-regulated

proliferation

Down-regulated collagen

synthesis

NHDF (Tajima, Inoue et al.,

1999)

neutral type-II-

arabinogalactans,

and acidic

arabinorhamnogalac

turonans from plants

Up-regulated proliferation

Up-regulated collagen

synthesis

HaCaT/NHK

NHDF

(Deters, Schroder et

al. 2005)

Most significantly, β-glucans have been shown to increase collagen deposition in

rodent skin wounds and intestinal anastomoses (Kougias, Wei et al. 2001). Also van Tol

et al. have reported that bacterial peptidoglycan directly stimulated in-vitro expression

of collagen α1 (van Tol, Holt et al. 1999). Authors of another study demonstrated that in

wounds treated with β-glucan from Sparassis crispa, macrophage was significantly

increased, as was fibroblast migration, collagen regeneration and epithelization

compared to the control group (Kwon, Qiu et al. 2009).

In particular, the ability of polysaccharides of different origin to significantly

scavenge free radical formation (ROS production) can not be omitted (Kogan, Pajtinka

et al. 2008, Andres, Molinari et al. 2006). Moreover, the activities of catalase (CAT),

superoxide dismutase (SOD), glutathione peroxidase (GSH-Px) were increased

20

significantly by several documented polysaccharides with antioxidant properties (Li, Li

et al. ; Deng, Cui et al. 2003; Zhang, Li et al. 2003).

Non-enzyme inhibition of glycation by polysaccharides from Achyranthes

bidentata and Lycium barbarum was also demonstrated (Deng, Cui et al. 2003).

4.1. Structure of polysaccharides selected for the evaluation in

this study

4.1.1. Glucomannan (GM)

GM was isolated from Candida utilis by sodium hydroxide extraction. GM had a

molar ratio mass between 60-70 kDa and a narrow molecular distribution. The ratio of

mannan to glucan was 2.73:1.

4.1.2. RROPs

The tested RROPs were obtained from Solabia Bio-Europe (Pantin, France).

Two polysaccharides were obtained from two different Klebsiella strains: RROP-1

from a non pathogenic Klebsiella pneumoniae strain and RROP-2 from a Klebsiella

planticola strain. The first one, RROP-1, of about 50 kDa average molecular weight

(according to the producer, Solabia) has the following repeating structure:

- α-Rha-(1-3)-β-Gal-(1-2)-α-Rha-(1-4)-β-GlcA-(1-3)-β-Gal-

β2-Rha

21

The second, RROP-2, has an average molecular weight of about 45 kDa

(according to the producer, Solabia) with the following composition of the repeating

units:

---4 β-Glc-(1-2)-α-Rha-(1-4)-α-GlcA-(1-3-)β-Rha-1-

(1-3)

α-Rha

These two polysaccharides have slightly different compositions and structures.

RROP-1 (commercial name Rhamnosoft®

) contains L-α-rhamnose (αRha), β-L-

galactose (βGal) and β-L-glucuronic acid (βGlcA). Besides rhamnose, RROP-2 contains

α-L-glucuronic acid (αGlcA), β-D-glucose (βGlc), and no galactose. Both

polysaccharides contain branching α- or β-rhamnose (αRha or βRha) side chains. The

repeating unit of RROP-1 is a hexasaccharide (molecular weight ~957 Da), and that of

RROP-2, a pentasaccharide (molecular weight ~867 Da).

Further, RROP-2 was degraded by mild acid (HCl) hydrolysis to a 5 kDa

oligosaccharide (RROP-3).

4.1.3. FROPs

FROPs were polymers of repeating trisaccharidic units, composed of galactose,

4-O-acetyl-glucuronic acid and Fucose.

3)-α-Galp-(1→3)-α- GalpA

-4-O-Ac-(1→3)-αFucp-(1→

The starting material was a radiation-sterilized polysaccharide (Fucogel®, Solabia

BioEurope, Paris) with an apparent high average molecular weight of about 40 kDa,

produced by non-pathogenic strain Clebsiella pneumoniae.

22

4.2. Recognition of polysaccharides by cell surface receptors

Polysaccharides exhibit various biological activities, which are mediated by

interaction with cell surface receptors.

Glucomannan is recognized by several types of receptors including complement

receptor 3 (CR3), the mannose receptor (MR), toll-like receptors (TLRs), and other

other lectin receptors, which are widely expressed on leukocytes and mediate cellular

response to different types of PAMPs (Weindl, Wagener et al.).

However (mannan/ glucan/galactose) receptors and other non specified binding

sites have also been described to be present on other cell types including fibroblasts and

different epithelial cells. These receptors could also be assumed to be present on

keratinocytes. Indeed, fibroblasts and keratinocytes have already been shown to express

lectin receptors and toll-like receptor 4, which were suggested to be crucial for the

recognition of galactose, as were O-linked mannosyl polymers such as GMs (Song,

Park et al. 2002).

Galectin-3, a galectin expressed by macrophages, dendritic cells, and epithelial

cells, bind to Candida species that bear 1,2-linked oligomannans on the cell surface, but

did not bind to Saccharomyces cerevisiae that lacks 1,2 –linked oligomannans (Pinto,

Barreto-Bergter et al. 2008).

A surprising finding was the demonstration of an α-L-Rhamnose recognising

lectin on human keratinocytes by the team of Monsigny and Kieda, although Rhamnose

is only present in plant and prokaryote glycocojungates, but was not demonstrated in

vertebrate glycoconjugates. Some other monosaccharides such as glucose, fucose and

N-acetyl-glucosamine in complex with serum albumin could compete with rhamnose

for apparently the same lectin site, although with a much lower affinity. Galactose–

recognizing lectins include elastin-laminin receptors (Faury, Ruszova et al. 2008).

The functional subunit of the elastin-laminin receptor, a 67 kDa protein, has a

side recognizing amino acid sequences of elastin and another side interacting with

polysaccharides with an end-group of galactose conformation. The recent experiments

indicated the possibility that at least some of the reactions observed with these FROP-

type of oligo/polysaccharides can be attributed to interaction with the elastin-laminin

receptor (Peterszegi, Isnard et al. 2003).

23

5. Previously reported effects of the polysaccharides

tested in this study

5.1. GM

It is well established that water-soluble yeast wall GM possesses high

antioxidant and free-radical scavenging activity that may underlie its various protective

effects (see review Kogan, Pajtinka et al. 2008). GM from Candida utilis significantly

reduced mutagenicity of the diagnostic mutagens/carcinogens in Ames test (Vlckova,

Duhova et al. 2004) and also elicited anti-genotoxic and elicited anticlastogenic effect in

many model tests systems (Miadokova, Svidova et al. 2006). Earlier, the same authors

demonstrated that GM from Candida utilis exhibited significantly higher antioxidant

activity than mannans from Saccharomyces cerevisae or Candida albicans partially due

to differences in the general structure of GM from Candida utilis and interactions with

different receptors and/or modulation of postreceptor intracellular signalling pathways

(Vlckova, Duhova et al. 2004; Drabikova, Perecko et al. 2009).

5.2. RROPs

The tested RROPs were shown to stimulate fibroblast proliferation. RROP-3

revealed the ability to decrease elastase-type activity. RROP 2-3 were proven to

stimulate collagen biosynthesis and all RROPs studied displayed free radical scavenger

activity (Andres, Molinari et al. 2006). RROPs, particularly oligoRROPs (RROP-3)

were shown to protect fibroblasts against cell death induced by Maillard products

(advanced glycation end-products, AGE) (Ravelojaona, Molinari et al. 2006). Several

AGEs significantly up-regulated elastase-type activity when added to the culture

medium of fibroblasts. This effect appears to be mediated by some AGE-receptors, and

could be inhibited by a 5 kDa RROP-3 as well as by a FROP-3. When present in the

culture media, RROP-3 and FROP-3 efficiently inhibited the passage-dependent up-

regulation of elastase-type activity expressed by human skin fibroblasts (Robert,

Molinari et al. 2010). These substances might be also considered as of potential

therapeutical interest against hyperglycemia induced cytotoxic effects as in type II-

diabetes (Robert, Molinari et al. 2010).

24

5.3. FROPs

Along with the FROP properties described earlier, there is also the inhibition of

MMP-2 and MMP-9 activation by fibroblasts and keratinocytes in cell cultures (Isnard,

Peterszegi et al. 2002). The FROPs significantly inhibited pro-forms of MMP-2 and

MMP-9. Peterszegi et al. described the stimulatory effect of FROP on cell proliferation

of human skin fibroblasts and the protective effect against ascorbate-induced

cytotoxicity due to the release of ROS (Peterszegi, Isnard et al. 2003).

Further, Moon et al. showed that fucoidans, the sulphated polysaccharides

extracted from brown algae containing a significant amount of L-fucose and sulphate

ester groups, may inhibit MMP-1 expression in human skin fibroblasts by prevention of

the activation of ERK and JNK signalling pathways (Moon, Lee et al. 2008).

Taken together, FROPs and RROPs were proven to be active in inhibiting

cytotoxicity of AGE-products, the up-regulation of elastase activity and stimulating cell

proliferation and ECM-biosynthesis, in vitro as well in vivo (Robert, Labat-Robert et al.

2009).

25

6. Main goals of the thesis

Based on current knowledge and preliminary data, we hypothesized that

different polysaccharides elicit various protective effects on dermal fibroblast and

keratinocytes. According to literature, GM reveals anti-genotoxic and anti-clastogenic

effects, RROPs have been shown to stimulate dermal fibroblast proliferation, protect

against elastase-type activity, and stimulate ECM synthesis. Finally, FROP inhibits

different metalloprotease activity.

As the main goals of this thesis, we decided to evaluate the effects of GM,

RROP1-3 and FROP on complex physiological response of cells and particularly gene

expression profiles of keratinocytes and dermal fibroblasts. After the selection of GM,

RROP1-3 and FROP non-cytotoxic concentrations by viability assays, microarray

experiments were performed in order to explore modifications of gene expression as a

result of the interaction between cells and focused polysaccharides.

7. Published articles

7.1. Photoprotective effects of glucomannan isolated from

Candida utilis

26

Photoprotective effects of glucomannan isolated from Candida utilis

Ema Ruszova,a,* Stanislav Pavek,a Veronika Hajkova,a Sarka Jandova,a Vladimir Velebny,a Ivana Papezikovab and Lukas Kubalab

b CPN spol. s.r.o, 561 02 Dolni Dobrouc 401, Czech Republic Institute of Biophysics Academy of Sciences of the Czech Republic, Kralovopolska 135, 612 65 Brno, Czech Republic

Received 15 August 2007; received in revised form 8 November 2007; accepted 11 November 2007 Available online 19 November 2007

a

Abstract—Glucomannans belong to yeast and fungal cell wall polysaccharides with known immunostimulatory and radioprotective effects. However, glucomannan protective effects against pathological consequences of skin exposure to short wavelength solar light, ultraviolet (UV) radiation, are unclear. Herein, a highly branched glucomannan (GM) isolated from the cell wall of Candida utilis, a member of the a-(1!6)-D -mannan group, was tested for its photoprotective effects in an in vitro model of UVB-irradiated human keratinocytes and an in vivo model of UV-induced erythema formation in human volunteers. GM suppressed the UVB-induced decrease of keratinocyte viability, which was connected with the suppression of UVB-induced keratinocyte apoptosis. GM reduced UVB-mediated caspase activation together with suppression of DNA fragment release into the cytoplasm. Furthermore, GM suppressed UVB-induced gene expression of pro-inflammatory markers including nuclear factor kappa B, inducible nitric oxide synthase, interleukins 8 and 1, together with suppression of prostaglandin E2 and interleukin 1a protein release. In vivo, GM decreased UV-induced skin erythema formation, which was correlated with a decrease of phosholipase A2 activity within the stratum corneum. It could be concluded that GM isolated from C. utilis possesses significant photoprotective effects on human keratinocytes in vitro as well as in vivo.

Keywords: Glucomannan; Candida utilis; HaCaT keratinocytes; UV-protection; Polysaccharide; Apoptosis

1. Introduction

It is well recognized that skin exposure to solar radiation has detrimental consequences, both acute and chronic. In particular, the short wavelength part of solar light, ultraviolet (UV) radiation, contributes significantly to undesirable effects, which could lead to the development of cutaneous malignancies.1–3 The impact of cutaneous malignancies is significant since melanoma and non-

Abbreviations: aMSH, alpha melanocyte stimulating hormone; CR3, complement receptor 3; GM, glucomannan; iNOS, inducible nitric oxide synthase; IL-1a, interleukin 1 alpha; MR, mannose receptor; NO, nitric oxide; NFjB, nuclear factor jB; PAMPs, pathogen- associated molecular patterns; PLA2, phospholipase A2; PGE2, prostaglandin E2; TLR, toll-like receptor; UVB, ultraviolet light B * Corresponding author. Tel.: +420 465 519548; fax: +420 465 543 793; e-mail: [email protected]

melanoma skin cancer are the most abundant carcinomas in western countries.1,4 Therefore, development and improvement of skin protection strategies against solar radiation is a critical issue. Different mechanisms are suggested to contribute to the adverse effects of UV radiation on the skin. One of the most significant mechanisms is UV-induced suppression of immune functions connected with massive cutaneous cell death. Decrease of cutaneous cell viability is related to UV-induced cell damage due to the formation of free radicals that destruct cellular structures including lipids, nucleic acids, and proteins.1,2 Physiological functions associated with UV-induced impairment of skin are directly associated with the inflammatory response provoked by the damaged keratinocytes, which leads to the release of a wide range of inflammatory mediators. Cytokines released during this early phase of UV-induced skin reaction are considered to be impor-

27

E. Ruszova et al. / Carbohydrate Research 343 (2008) 501–511

tant mediators of the consequent immune suppression in skin.5 Therefore, the prevention of UV-induced keratinocyte injury could significantly reduce the detrimental effect of UV radiation on skin.1,2 Increased resistance of cells to different types of noxi- ous stimuli is connected with an increase in the activity of various intracellular protective and repairing mechanisms, which resemble adaptation to cellular stress.6,7

Activation of stress-induced signaling pathways by non-specific stressors, such as heat or radiation leads to the activation of cytoprotective responses.8 Similar cell responses could be induced through the activation of specific cell surface receptors including receptors for microbial molecular patterns (or pathogen-associated molecular patterns, PAMPs) comprising lipids, carbohydrates, proteins, and nucleic acids because their molecular structure is distinct from those expressed on the surface of mammalian cells.9–12 Polysaccharides that are part of the cell wall of yeasts and fungi are among the molecular structures recognized by these receptors.9–12 Therefore, biologically active polysaccharides with the ability to active cellular responses in skin could be skin-protective agents. Yeast and fungal polysaccharides consist of glucose and mannose units joined together by glycosidic linkages via different positions and in different ratios. The biological activity of purified glucans and glucomannans, including decrease of infectious complications and inhibition of tumor growth, is known to depend on their structure. Parameters such as primary structure, degree of branching, molecular weight, solubility, solution conformation, and ionic charge were suggested to play a role in determining the biological activity of these molecules.13,14 In a previous study, we observed strong immunostimulatory effects of two structurally different polysaccharides, schizophyllan, and carboxymethyl- glucan, which were isolated from Schizophyllum communae and Saccharomyces cerevisiae cell walls, respectively.15 However, the relationships between the structure of glucans and glucomannans and their stimulatory activities still remain unclear. Glucomannan (GM) isolated from Candida utilis consists of the a-(1!6)-D -mannopyranosyl backbone carrying mannooligosaccharidic side chains (1– 5 units) containing a-(1!2) linkages, where some of the side chains are terminated with non-reducing D -glucopyr- anosyl residues.16 Generally, the mass of this polysaccharide varies between 30 and 70 kDa and the mannose/ glucose ratio is 2–3:1 (Fig. 1). Glucans and glucomannans exhibit various biological activities, which are mediated by interaction with cell surface receptors. GM is recognized by several types of receptors including complement receptor 3 (CR3), the mannose receptor (MR), toll-like receptors (TLRs), and other lectin receptors, which are widely expressed on leukocytes and mediate cellular response to different types of PAMPs.10 However, mannan/glucan binding

Figure 1. Structure of GM isolated from Candida utilis, which consists of an a-(1-6)-D -mannopyranosyl backbone carrying mannooligosaccharidic side chains (1–5 units) composed of a-(1-2) linkages, where some of the side chains are terminated with non-reducing D -glucopyr- anosyl residues.

sites have also been described on other cell types including fibroblasts.10 and different epithelial cells.17,18 These glucan/mannan receptors could also be assumed to be present on keratinocytes. Indeed, keratinocytes were already shown to express lectin receptors and toll-like receptor 4, which were suggested to be crucial for the recognition of O-linked mannosyl polymers such as glucomannans.7,19,20 Given the above premises, it could be hypothesized that GM acts as a photoprotective agent and prevents UVB-induced damage of cutaneous cells. The photoprotective properties of the GM isolated from C. utilis were evaluated in vitro on UVB-irradiated primary human keratinocytes and the immortalized keratinocyte cell line HaCaT. Furthermore, the photoprotective effects of the GM were confirmed in vivo by the measurement of UV- induced erythema formation in human volunteers. Parameters of the skin inflammatory response (phospholipase A2 activity) were determined in stripped layers of stratum corneum. The data obtained show significant GM protective effects against UVB-induced death of human keratinocytes and suggest GM as a potent photoprotective agent.

2. Results

2.1. UVB exposure induced inflammatory response and apoptosis in keratinocytes

To explore the photoprotective properties of GM, model of primary human keratinocytes and human keratinocyte cell line HaCaT irradiated by UVB light was employed. Doses of UVB radiation (10 and 20 mJ/cm2) and concentrations of GM (50 and 500, 1000 μg/mL) were selected based on the preliminary results (data not shown). Upon irradiation with 10 mJ/ cm2 UVB, the viability of both keratinocyte cell cultures decreased significantly after 24 h. Thus, the dose 10 mJ/ cm2 of UVB was selected for cell viability evaluation and for the induction of the cell inflammatory response. A higher dose of UVB (20 mJ/cm2) was selected for the

28

E. Ruszova et al. / Carbohydrate Research 343 (2008) 501–511

examination of GM-dependent modulation of apoptotic processes due to significant induction of apoptosis by this dose up to 5 h after the treatment.

2.2. GM prevented the UVB-induced decrease of cell viability in primary keratinocytes

GM did not significantly modulate cell viability of cells not treated with UVB (Fig. 2a). However, in agreement with our hypothesis, GM at concentrations 50 and 500 μg/mL revealed a protective effect on the viability

of irradiated keratinocytes (Fig. 2b). Interestingly, comparison of GM isolated from C. utilis and GM isolated by the same method from S. cerevisiae showed the higher protective effect on keratinocyte viability of the former (data not shown). This confirmed our selection of GM isolated from C. utilis for further characterization of GM photoprotective effects.

2.3. GM prevented the UVB-induced apoptosis of primary keratinocytes

GM-mediated alternation of keratinocyte apoptotic cell death induced by UVB was evaluated to characterize the protective effects of GM on the viability of UVB-irradiated keratinocytes. The apoptotic process was characterized based on the determination of caspase 3 and pan- caspase activity, as well as the formation of low molecular weight DNA fragments (mono- and oligo-nucleosomes) in the cytoplasm of keratinocytes irradiated by UVB (20 mJ/cm2) 5 h after the treatment. GM significantly suppressed the activation of caspase 3 and the formation of low molecular weight DNA fragments in the keratinocyte cytoplasm (Fig. 3a and b). Similarly, UVB-induced pan-caspase activation was notably suppressed by GM treatment (Fig. 4). GM treatment alone did not induce any detectable increase of these apoptotic markers (data not shown).

2.4. GM decreased the formation of thymidine dimers in UVB-irradiated primary keratinocytes

The formation of thymidine dimers in UVB (10 mJ/cm2) irradiated primo-cultures of keratinocytes was evaluated since thymidine dimers are the most common cause of development of DNA lesions after the exposure of cells to UVB irradiation. Interestingly, treatment by GM decreased, in a dose dependent manner, the abundance of thymidine dimers in the irradiated keratinocytes 3 and 6 h after irradiation (Fig. 5a and b).

2.5. GM prevented an increase of pro-inflammatory markers in UVB-irradiated HaCaT

GM significantly down-regulated gene expression of nuclear factor jB, pro-inflammatory interleukin-1a (IL-1a), interleukin-8 (IL-8), and inducible nitric oxide synthase (iNOS) in UVB-irradiated (10 mJ/cm2) HaCaT (Table 1). Moreover, GM-dependent down regulation of IL-1a and prostaglandin E2 (PGE2) release from UVB- irradiated (10 mJ/cm2) HaCaT was observed already 24 h after cell irradiation (data not shown). This effect was more profound 48 h after the treatment (Fig. 6a and b). GM application alone did not induce a significant release of IL-1a and PGE2 from untreated keratinocytes (Fig. 6a and b).

a

1

0.9

0.8

0.7

0.6

control 500

50 1000

250 GM (μg/ml)

O.D. 0.5

0.4

0.3

0.2

0.1

0 24 48 72 96

time (hours)

b 120 % of non-

irradiated

control 100

80

60

40

20

0

*

control irradiated control

50 500

GM (μg/ml)

Figure 2. GM did not modulate the viability of intact human primary keratinocytes but did reduce UVB-induced decrease of keratinocyte viability. (a) Human primary keratinocytes were treated by GM (50, 250, 500, and 1000 μg/mL) for 24, 48, 72, and 96 h and the cell viability was determined by the XTT test. (b) Human primary keratinocytes were irradiated by UVB (10 mJ/cm2) and treated with GM (50 and 500 μg/mL) immediately after irradiation. Cell viability was determined 24 h after the UVB exposure by the XTT test. Non-irradiated keratinocytes served as the control. Three determinations were performed and data represent mean ± standard error of mean.

29

E. Ruszova et al. / Carbohydrate Research 343 (2008) 501–511 a

% of non-

irradiated

control 180 160

140

120

100

80

60

40

20

0

Consistently with the observed GM photoprotective effects, GM significantly potentiated the release of aMSH by UVB-irradiated primary keratinocytes (120–169% of irradiated control).

2.7. GM protected skin against UV-induced erythema formation in vivo

The photoprotective effect of GM against UV-induced erythema formation in human volunteers was studied. The application of GM (0.5% in emulsion) on skin visi- bly decreased UV-induced erythema in four of five volunteers (Fig. 7). The effects of GM were quantified by the evaluation of erythema indexes, which were significantly decreased [25.32 (±2.35) compared to 28.78 (±2.85), p < 0.05] and a* values of CIE L*a*b* color space [5.1 (±0.41) compared to 5.66 (±0.57), p < 0.05] in comparison to the control irradiated sites.

2.8. GM lowered phospholipase A2 (PLA2) activities in the stripped stratum corneum

control irradiated control

50 500

GM (μg/ml) b

1.2

1

0.8 *

0.6

0.4 **

0.2

0

**

To characterize the protective effect of GM on UV-irradiated skin, PLA2 activity in the stratum corneum stripped from irradiated sites was evaluated. GM-treated sites revealed lower PLA2 activity compared to control sites [160 (±27.5) vs 223.7 (±69)] suggesting a mechanism for PLA2 to be involved in the protective effect of GM in a model of UV-induced erythema in vivo.

O.D.

3. Discussion

irradiated control

500 1000 GM (μg/ml)

Figure 3. Reduction of (a) the caspase 3 activity and (b) low molecular weight DNA fragments (mono- and oligo-nucleosomes) in the cytoplasm of UVB-irradiated human primary keratinocytes by GM. Human primary keratinocytes were irradiated by UVB (20 mJ/cm2) and treated with GM (50, 500 and 1000 μg/mL) immediately after irradiation. (a) Caspase-3 activity was determined in cell lysate 5 h after UVB exposure. Non-irradiated keratinocytes served as the control. Three determinations were averaged and data represent mean ± standard error of mean. (b) Mono-and oligo-nucleosomes were determined 5 h (black columns) and 24 h (white columns) after the treatment. Non-irradiated keratinocytes served as the control. Four determinations were performed and data represent mean ± standard error of mean.

2.6. GM induced a-melanocyte stimulating hormone (aMSH) in UVB-irradiated primary keratinocytes

The release of aMSH by keratinocytes was evaluated by considering aMSH to be an autocrine mediator with a potent effect on cell protective mechanisms when released by keratinocytes in response to injurious stimuli.21,22

The aim of the study was to evaluate the photoprotective effects of biotechnologically prepared GM against UVB adverse effects on human keratinocytes using in vitro as well as in vivo approaches. GM isolated from C. utilis and from S. cerevisiae possessed the distinct ability to prevent cytotoxic effects of UVB radiation on human keratinocytes, with the former showing higher potential. This effect was shown to be connected with the reduction of UVB-induced keratinocyte apoptosis. Furthermore, GM isolated from C. utilis significantly decreased expression and release of selected pro-inflammatory markers and mediators. In contrast, GM potentiated the release of aMSH, a potent mediator of skin protection. Finally, the protective effect of GM against UVB-skin damage was confirmed by a reduction in UV-induced erythema formation on the skin of human volunteers. These effects correlated with the GM- mediated reduction of UV-induced increase of PLA2 in stratum corneum. These data suggest potent photoprotective effect of GM against UVB irradiation. Here, UVB light was used as a model for the evaluation of the adverse effects of sun radiation on skin. From

30

Figure 4. GM reduced pan-caspase activation in UVB-irradiated human primary keratinocytes as determined by fluorescence microscopy. Human primary keratinocytes were irradiated by UVB (20 mJ/cm2) and treated with GM (500 μg/mL) immediately after irradiation. Pan-caspase activity in the cytoplasm was determined by CaspACETM FITC-VAD-FMK. An in situ marker (Promega, Vienna, Austria) was added 24 h after the treatment. Magnification: 400·, used microscope Nikon Eclipse E400 (Tokyo, Japan) visualization software Lucia v. 4.82; blue color-lowest nuclei labeled using Hoechst. Left side: irradiated keratinocytes; right side: irradiated keratinocytes treated with GM (500 μg/mL). Four determinations were performed and typical figures were presented.

the spectrum of sunlight, UVB light has the lowest potential to penetrate skin, but it is considered to be the main mediator of cutaneous damage and inflammation leading to the formation of skin carcinomas and to the alteration of cutaneous immune responses.23 Adverse effects of UVB on cutaneous cells could be prevented by the activation of cellular mechanisms protecting cells against the damaging effects of toxic agents such as UVB irradiation. The cytoprotective mechanisms could be triggered by various factors including substances recognized by receptors for microbial molecular patterns or PAMPs. Therefore, GM could be suggested as a sub- stance with potential to activate cellular mechanisms leading to an increase of cutaneous cell resistance to UVB irradiation. GM revealed a protective effect on the viability of UVB-irradiated keratinocytes. This observation is in agreement with other authors who reported anti-muta- genic, anti-genotoxic, and anti-cancerogenic properties of GM in various model systems.24,25 Furthermore, direct photoprotective effects of polysaccharides were reported for carboxymethyl glucan, which increased resistance of corneal epithelial cells to UVB- and hydro- gen peroxide-induced cell death.11 The protective effect of GM against UVB-induced cell death was directly connected with reduced apoptosis of keratinocytes. GM significantly decreased the markers of early stages of apoptotic process: the activation of caspase 3, the activation of pan-caspase, and the release of mono- and oligo-nucleosomes into the keratinocyte cytoplasm.26–28 Caspases, a family of cysteine proteases, are the common executors of apoptosis and are induced by various stimuli.28 GM-mediated suppression of UVB-induced caspase activation could be connected with the suppression of a poly(ADP-ribosyl)polymerase

activation, as documented by the significantly lower levels of low molecular weight DNA fragments detected in the cytoplasm of GM-treated keratinocytes. Therefore, it could be speculated that GM protects cutaneous cells from UVB-induced apoptosis by the suppression of caspase activation and subsequent hindering of DNA fragmentation. Similar to keratinocytes, GM treatment of the UVB-irradiated Jurkat T lymphocyte cell line decreased the activation of pan-caspase (data not shown). GM was shown to significantly decrease detectable thymidine dimers in UVB-irradiated keratinocytes. The abundance of thymidine dimers was analyzed as it is considered to be the most common cause of development of DNA lesions after the exposure of cells to UVB irradiation.29,30 Although a large fraction of these DNA lesions are repaired enzymatically by nucleotide excision repair, the amount of unrepaired lesions is not negligible and could result in UVB-induced mutagenesis.29 The observed GM-mediated reduction of thymidine dimers suggests a potentiation of the DNA damage repair mechanisms in keratinocytes. This indicates another mechanism of GM photoprotective effect on UVB-irradiated human keratinocytes. The faster elimination of thymidine dimers in presence of GM provides other evidence for reparative effects evoked by GM. Upon UV irradiation, cutaneous cells including keratinocytes produce a wide range of inflammatory mediators.23,31 Interestingly, microarray analyses showed that GM treatment significantly suppressed UVB- induced NFκB gene expression in keratinocytes. This correlates with the observed GM-mediated down regulation of gene expression of the tagged pro-inflammatory markers IL-1a, IL-8, and iNOS. Furthermore, GM- dependent down regulation of IL-1a and PGE2 gene expression was confirmed by an observed decrease of

31

a % of

irradiated

control 120

100

80

60

Table 1. Gene expression profiling of GM effects on UVB-irradiated human keratinocyte cell line HaCaTa

Gene

PTGS2 FGF2 TP53 IL-8 NF-kB IL-1 iNOS TLR4 MMP3 FGF9 ARNT

a

Percentage of change against irradiated control (%)

-89 -85 -89.50 -84 -84 -83.70 -82.50 -82 -82 -80.40 -80

**

40

20

0 control 500

GM (μg/ml) 1000

b % of

irradiated

control

140

120

100

80

60

40

20

0

3 hours after the irradiation

6 hours after the irradiation

Human primary keratinocytes were irradiated by UVB (10 mJ/cm2) and treated with GM (500 μg/mL) immediately after irradiation. GM-mediated modification of gene expression profiles was evaluated by microarray assay 24 h after treatment. Up-regulation is indicated by +, down regulation by À, and intensity is expressed as % modification as compared to irradiated control without GM treatment. Gene abbreviations: PTGS2- cyclooxygenase 2; FGF- fibroblast growth factor, TP53- apoptotic protein p53; IL—interleukin; NFjB—nuclear factor jB; iNOS—inducible nitric oxide synthase, TLR—toll-like receptor, MMP3-matrix metalloprotease, ARNT— aryl hydrocarbon nuclear translocator.

* *

*

Together with the decrease of pro-inflammatory mediator expression, GM augmented the release of aMSH by irradiated keratinocytes.33 aMSH is tridecapeptide derived from propiomelanocortin and is involved in a variety of biological processes, such as pigmentation or extracellular matrix composition.34,35 aMSH possesses significant anti-inflammatory activities mediated through the activation of aMSH receptors resulting in a decrease of inflammatory mediator production or leukocyte migration.21,22,36 Therefore, it could be suggested that GM-mediated increase of aMSH formation may trigger keratinocytes in an autocrine fashion, which reduces inflammatory response after UVB irradiation. To confirm the photoprotective effects of GM observed in vitro, GM modulation of UVB-induced erythema was evaluated in vivo on the skin of healthy human volunteers. In agreement with in vitro data, the application of GM on the skin significantly decreased UV-induced erythema formation. Application of GM considerably attenuated the activity of PLA2, a pro- inflammatory enzyme activating leukotriene synthesis37 in the stripped stratum corneum. Interestingly, the activity of b-glucocerebrosidase, an enzyme contributing to the repair of UVB-damaged skin,38 was slightly increased by GM in comparison to the untreated sites. This suggests that GM improves a skin barrier function impaired by UV exposure (unpublished data). Another possible explanation for the reduced erythema formation could be connected with the ability of GM to down regulate UVB-induced iNOS expression in vitro. The reduction of iNOS expression and NO overproduction

control 500 1000

GM (μg/ml) Figure 5. GM decreased the formation of thymidine dimers in UVB- irradiated primary keratinocytes. Human primary keratinocytes were irradiated by UVB (10 mJ/cm2) and treated with GM (500 and 1000 μg/mL) immediately after irradiation. Levels of thymidine dimers in primary keratinocytes (a) and HaCaT (b) were determined in cell culture lysate after 3 h (black columns) and 6 h (white columns) after the treatment. Irradiated untreated keratinocytes served as the control. Four (the 3-h treatment) and two (the 6-h treatment) determinations were performed and data represent mean ± standard error of mean.

IL-1a and PGE2 protein release from UVB-irradiated keratinocytes. IL-1a and IL-8 are potent pro-inflammatory mediators and together with the increased production of NO, catalyzed by iNOS, contribute to acute inflammatory response to UV irradiation.31,32 Similar to the observed effects of GM, polysaccharides isolated from Grifola frondosa were reported to decrease expression and release of matrix metalloproteinase 1, which is another significant mediator contributing to the development of inflammatory reaction in UV-irradiated skin.9

32

a % of non-

irradiated

control 600

500

400

300

200

100

0 control 50 500 control

GM (μg/ml) 50 500

* *

b

% of non-

irradiated

control

3500

3000

2500

2000

1500

1000

500

0

**

**

Figure 7. GM-reduced skin erythema induced by UVA/B (1.25 MED). Typical results of erythema skin formation shown on 2 volunteers. Experiments were performed as described in Section 4.

control 50 500 control 50 500

non-irradiated GM (μg/ml)

irradiated cytes and skin. GM could therefore be suggested as a protective agent in cosmetic products, which prevents detrimental effects of solar radiation on skin.

Figure 6. GM reduced the release of IL-1a and PGE2 by UVB- irradiated human keratinocytes HaCaT into the culture media. Human primary keratinocytes were irradiated by UVB (10 mJ/cm2) and treated with GM (50 and 500 μg/mL) immediately after irradiation. Levels of (a) IL-1a and (b) PGE2 were determined in cell culture media 48 h after the treatment (black columns). Non-irradiated keratinocytes treated only by GM for 48 h served as the control (white columns). Four determinations were performed and data represent mean ± standard error of mean.

4. Experimental

4.1. Human subjects

The present study has been approved by the CPN, s.r.o. Institutional Review Board, and the Declaration of Hel- sinki protocols were followed. Patients gave their writ- ten, informed consent.

4.2. Glucomannan

Glucomannan was isolated from C. utilis by sodium hydroxide extraction (2% NaOH, 25 min at 95 °C). HCl was used to adjust pH to 5.5 and the supernatant was clarified by centrifugation. Glucomannan was precipitated by the addition of isopropanol. The precipitate was dissolved in water (40 °C) and filtered several times through paper and active charcoal filters. Finally, glucomannan was precipitated from an aqueous solution with isopropanol, the precipitate was dissolved in absolute isopropanol and dried under vacuum. GM had a molar

could reduce pathological vasodilatation coupled with UVB induced inflammation.23,32,39 Moreover, GM inhibition of PGE2 release that was observed in vitro could significantly reduce the skin inflammatory response to UVB due to a reduction of histamine release.31,40 Based on our data we could speculate that the GM effect on keratinocytes may evoke an adaptive cell response and may contribute to the activation of reparative processes leading to improved cell viability after the UVB exposure.6,7 In conclusion, GM isolated from C. utilis was shown to be a potent photoprotective compound preventing harmful effects of UVB exposure on human keratino-

33

mass between 60 and 70 kDa and a narrow molecular distribution (polydispersity index between 1.15 and 1.24) as determined by SEC/MALLS. The ratio of mannan to glucan was 2.73:1 in GM isolated from C. utilis. GM isolated from S. cerevisiae had a ratio of 30–53 mannans per glucan.

4.3. Cell cultures

The spontaneously immortalized human keratinocytes cell line HaCaT (kind gift of Prof. Dr. N. Fusenig41) was grown in DMEM supplemented with 10% of fetal bovine serum (Gibco, Carlsbad, California, USA), glutamine (0,25 mg/mL, Gibco), and gentamicin (50 μg/mL, Gibco) at 37 °C. Primary normal human epidermal keratinocytes were isolated from the skin removed during cosmetic plastic surgery. Keratinocyte cultures were prepared according to the method of Rheinwald and Green.42 The skin was minced and trypsinized (0.25% trypsin + 0.02% EDTA) at 37 °C for 16–18 h and keratinocytes were grown in 75 cm2 culture flasks with mitomycin-treated 3T3 fibroblasts used as a feeder. Keratinocytes were grown in cell culture media DMEM (Sigma–Aldrich, Vienna, Aus- tria), HAM-F-12 (Sigma–Aldrich, Vienna, Austria), newborn calf serum (10%, Gibco), hydrocortisone, adenine, choleratoxin, EGF, insulin, amphotericin B, penicillin, streptomycin, apotransferrin, and 3,3,5-tri- iodo-L -thyronine (Sigma–Aldrich, Vienna, Austria). The medium was changed every 2 days. Keratinocytes were cultivated until confluence was reached and passaged again at a density 1 · 104/cm2. After three passages keratinocytes were seeded at a concentration 1.55 · 104/cm2. Keratinocytes were cultivated overnight before the experimental treatment.

4.4. UVB irradiation and GM treatment in vitro

A 1000 W xenon arc solar UV-simulator (Oriel Instru- ments, Stratford, CT, USA) equipped with a dichroic mirror and a 300 ± 5 nm interference filter was used for cell irradiation in vitro. The UVB output power and radiation doses were determined by using photo- meter/radiometer PMA 2100 (Solar Light Co., Glenside, PA, USA). For the evaluation of cell viability and inflammatory response, a 10 mJ/cm2 dose was applied, whereas a 20 mJ/cm2 dose was used for the cell apoptosis evaluations. Before UVB treatment, the cell culture media were replaced by phosphate buffer saline (pH 7.4, PBS). Immediately after the UVB irradiation, PBS was replaced by cell culture media containing GM. The RNA was isolated by TrizolÒ Reagent (Invitrogen, Carlsbad, California, USA) 24 h after UVB treatment and stored at À80 °C in DEPC water. The cell culture media were collected 24 and 48 h after UVB treatment and stored at À20 °C until analysis. Cells were harvested

and lysed in different buffers according to the determined apoptotic markers 5 h after UVB treatment.

4.5. XTT cell viability assay

Cell viability was determined by the commercial XTT assay (RocheApplied Science, Meylan, France). Cells were incubated with medium containing 33% (v/v) XTT solution for 2 h. An absorbance of 150 lL media was determined by a Versamax microplate reader (Molecular Devices, Union City, California, USA) at a wavelength 450 nm.

4.6. Caspase activity assay

Cells were lysed in 200 μL of a lysis buffer containing 50 mM HEPES (pH 7.4), 0.2% CHAPS, 5 mM DTT, and 0.2 μM aprotinin for 20 min on ice and sonicated 3 times for 15 s. Samples were centrifuged (15,000g/ 10 min/4 °C), and the protein concentration in the supernatant was determined and adjusted to same concentration in all samples. Samples were incubated with assay buffer (20 mM HEPES with pH 7.4, 0.1% CHAPS, 5 mM DTT, and 2 mM EDTA) containing 10 mM caspase substrate (Ac-DEVD-AMC, Sigma–Aldrich, Vienna, Austria) and incubated for 2 h at 37 °C. The fluorescence of the substrate was quantified by an Infi- nite M200 (Tecan, Mannedorf, Switzerland) fluores-¨ cence reader with excitation at 380 nm and emission at 460 nm. Pan-caspase activation was detected using the commercial In Situ marker CaspACETM FITC-VAD- FMK (Promega, Vienna, Austria) kit and observed under fluorescent microscope Nikon E400 (Tokyo, Japan).

4.7. cDNA array

cDNA synthesis and labeling were performed by Revert- Aid H Minus MuLV-Reverse Transcriptase (Fermentas, Burlington, Ontario, Canada) and Biotin-11-dUTP (Fermentas). Hybridization was performed on the cDNA array system (Clondiag chip technologies GmbH, Jena, Germany). The amount of cDNA in each spot was detected by enzyme reaction of HRP using substrate True Blue (KPL, Gaithesburg, Maryland, USA). The developed color was determined by Reader ATR01 (Clondiag). IconoClust software (Clondiag) was used for spot proceeding and b-actin, GAPD, and histone H3 markers were used as housekeeping markers for the following equalization of the spot intensities.

4.8. Cytokine and DNA fragment ELISAs

PGE2 and IL-1a in cell culture media were determined in duplicate by commercial ELISA kits (KGE004 from RnD Systems Europe, Ltd, Abington, UK; and BMS 243/2MST from Bender MedSystems, Vienna, Austria,

34

respectively) according to the manufacturer manuals. Low molecular weight DNA fragments were determined in keratinocyte lysate by a Death Detection Kit following manufacturer instructions (Roche Applied Science, Meylan, France). Two different optimal dilutions were used to perform the assay: 1:40 (vol/vol) for 5 h and 1:67 (v/v) for 24 h intervals. Release of aMSH into the culture medium was determined by commercial aMSH EIA (Phoenix Pharmaceuticals, Inc., Karlsruhe, Germany).

4.9. Detection of thymidine dimers

DNA isolation from cells was performed according to a modified method described by Roza et al.43 Primary keratinocytes were irradiated by UVB (10 mJ/cm2), immediately treated with GM (500 and 1000 μg/mL), and incubated for 3 and 6 h before their lysis. Cells were lysed in 200 lL of a buffer containing 10 mM Tris–HCl (pH 8.0), 10 mM EDTA, 0.5% SDS, and Proteinase K (200 lg/mL). After a 2 h incubation at 55 °C, DNA was purified by phenol extraction, RNase H digestion, chloroform/isoamylalcohol (24:1) extraction and etha- nol precipitation. The concentration and the purity of DNA was assessed spectroscopically (UV–vis spectro- photometr UV-2401PC). Further, 50 ng of DNA was absorbed to each well of poly-L -lysine precoated plates (BD Biosciences, San Jose, CA, USA) by incubation at 4 °C overnight. Plates were washed (0.05% vol/vol Tween in PBS), blocked by BSA in PBS (1% wt/vol) for 1 h at room temperature. The primary antibody (anti-thymidine dimer KTM53) diluted 1:10,000 in PBS (Kamiya Biomedicals, Seattle, WA, USA) was added and the plates were incubated on orbital shaker (200 rpm) for 2 h at room temperature. After washing wells 3 times in wash buffer, biotinylated goat-anti- mouse IgG (Dako, Glostrup, Denmark) diluted 1:1000 in blocking buffer was added and incubated on orbital shaker for 1 h at room temperature. Next, plates were washed 3 times by wash buffer and streptavidin-poly HRP (Pierce, Rockford, IL, USA) diluted 1:10,000 in PBS with BSA (0.1% vol/vol) added and incubated on an orbital shaker for 30 min at room temperature. Plates were washed 3 times, TMB substrate (Sigma–Aldrich, Vienna, Austria) was added and the absorbance was read at 450 nm. Results were expressed as percentages of arbitrary absorbance units (AU) minus AU of the wells coated by DNA isolated from non-irradiated controls.

4.10. Skin erythema formation and determination of enzymatic activity in vivo

Seven female and male Caucasian volunteers with healthy skin (Fitzpatrick types II and III), aged 24–39 years, and in a good health were selected after a study

approval by the ethics committee. Each person signed an informed consent form; five completed the study. There were three fields, 4 · 3 cm, marked on the volar forearms of human volunteers. An emulsion with 0.5% C. utilis GM was applied 3 times a day to the test sites over a 7-day period, as well as a control emulsion without any active material to the control sites. The reference sites were left untreated. For the last three applications on the last day, the emulsions were replaced by a simple carbomer gel with or without 0.5% GM to avoid possible adverse skin reactions to some ingredient due to the UV irradiation. None of the preparations contained preservative and only freshly defrosted ones were used for the treatment every day. A 1000 W xenon arc solar UV-simulator (Oriel Instruments, Stratford, CT, USA) equipped with a dichroic mirror and a combination of UG5 (1 mm) and WG280 (2 mm) Schott filters giving a COLIPA compli- ant UV-spectrum was used for the irradiation of the test skin sites. The minimal UVA/UVB erythema doses for all volunteers were determined 2–4 days before testing individually. Skin redness before and 24 h after irradiation was measured using a reflectance spectrophotome- ter Spectrocam 75 RE (Spectrostar). The reflectance values within the range 380–750 nm were recorded, both a* values of L*a*b* color space and erythema indexes according Wagner et al.44 were calculated. On each test site, five determinations were performed and the mean of five values was calculated. Skin redness changes were expressed as the differences between the initial and final a* and erythema index values measured in the test, control, and adjacent untreated reference skin sites. After the pre-treatment period, the skin sites were cleansed gently and allowed to dry. The initial a* and erythema index values were measured and areas 3 · 1.2 cm located on the two pre-treated sites were irradiated with 1.25 MED UVA/UVB. Then, the treatment with GM and control gels continued three times during the following day. The skin was photographed (Coolpix 4500, Nikon, Japan) 24 h after irradiation, the redness was determined spectrophotometrically, and two groups of 11 strips (D-Squame, Cuderm Corp., Dallas, TX, USA) were taken from each skin site. These strips were immersed in 100 mM TRIS/HCl buffer, pH 7.5, containing 0.5% Triton X-100 for 1 h at room temperature. The extracts were collected and used for the enzyme activity determinations. b-Glucocerebrosidase activities were evaluated by fluorescence measurement of 4-methylum- belliferone released from 4-methylumbelliferyl b-D - glucopyranoside at 37 °C (Shimadzu RF-5301PC spectrofluorophotometer, excitation at 358 nm, emission at 448 nm). The PLA2 activities were estimated by the determination of fluorescent pyrene derivatives released from 1-hexadecanoyl-2-(1-pyrenedecanoyl)-sn-glycero- 3-phosphoglycerol at 37 °C (excitation at 344 nm, emission at 398 nm).

35

4.11. Statistical analysis

All experiments were carried out with 3–4 replicates in vitro, and with 5 replicates for in vivo observations. The results are reported as a mean ± standard error of mean (SEM) in the figures and mean ± (SEM) in the text. Statistical significance was analyzed by Mann and Whitney distribution-free-U test and considered significant at p < 0.05 (*), eventually p < 0.01 (**).

Conflict of interest

V.V. has a financial interest in CPN, Dolni Dobrouc, the company providing the tested glucomannan. E.R., S.P., V.H., S.J. are employees of CPN, Dolni Dobrouc. The research of I.P. and L.K. was partially supported by CPN, Dolni Dobrouc, Czech Republic.

Acknowledgments

The authors wish to thank S. Moravcikova for technical assistance with protein and enzymatic assays, M. Magdolenova for technical assistance with erythema measurements, and L. Balonova and A. Franzova for general technical assistance.

References

1. Assefa, Z.; Van Laethem, A.; Garmyn, M.; Agostinis, P. Biochim. Biophys. Acta 2005, 1755, 90–106. 2. Wang, L.; Eng, W.; Cockerell, C. J. Adv. Dermatol. 2002, 18, 247–286. 3. Fusenig, N. E.; Boukamp, P. Mol. Carcinogen. 1998, 23, 144–158. 4. Fourtanier, A.; Moyal, D.; Maccario, J.; Compan, D.; Wolf, P.; Quehenberger, F.; Cooper, K.; Baron, E.; Halliday, G.; Poon, T.; Seed, P.; Walker, S. L.; Young, A. R. J. Invest. Dermatol. 2005, 125, 403–409. 5. Norval, M. Bull. World Health Organ. 2002, 80, 906– 907. 6. Wang, L.; Lu, L. Invest. Ophthalmol. Vis. Sci. 2007, 48, 652–660. 7. Andres, E.; Molinari, J.; Peterszegi, G.; Mariko, B.; Ruszova, E.; Velebny, V.; Faury, G.; Robert, L. Pathol. Biol. (Paris) 2006, 54, 420–425. 8. Miura, Y. J. Radiat. Res. (Tokyo) 2004, 45, 357–372. 9. Bae, J. T.; Sim, G. S.; Lee, D. H.; Lee, B. C.; Pyo, H. B.; Choe, T. B.; Yun, J. W. FEMS Microbiol. Lett. 2005, 251, 347–354. 10. Kougias, P.; Wei, D.; Rice, P. J.; Ensley, H. E.; Kalbfleisch, J.; Williams, D. L.; Browder, I. W. Infect. Immun. 2001, 69, 3933–3938. 11. Porcu, M.; Guarna, F.; Formentini, L.; Faraco, G.; Fossati, S.; Mencucci, R.; Rapizzi, E.; Menchini, U.; Moroni, F.; Chiarugi, A. Cornea 2007, 26, 73–79. 12. Wei, D.; Zhang, L.; Williams, D. L.; Browder, I. W. Wound Repair Regen. 2002, 10, 161–168.

13. Bohn, J. A.; BeMiller, J. N. Carbohydr. Polym. 1995, 28, 3–14. 14. Kulicke, W. M.; Lettau, A. I.; Thielking, H. Carbohydr. Res. 1997, 297, 135–143. 15. Kubala, L.; Ruzickova, J.; Nickova, K.; Sandula, J.; Ciz, M.; Lojek, A. Carbohydr. Res. 2003, 338, 2835–2840. 16. Ogawa, K.; Matsuda, K.; Tamari, K.; Kiyo-oka, S. Agric. Biol. Chem. 1990, 54, 593–597. 17. Ahren, I. L.; Williams, D. L.; Rice, P. J.; Forsgren, A.; Riesbeck, K. J. Infect. Dis. 2001, 184, 150–158. 18. Rice, P. J.; Adams, E. L.; Ozment-Skelton, T.; Gonzalez, A. J.; Goldman, M. P.; Lockhart, B. E.; Barker, L. A.; Breuel, K. F.; Deponti, W. K.; Kalbfleisch, J. H.; Ensley, H. E.; Brown, G. D.; Gordon, S.; Williams, D. L. J. Pharmacol. Exp. Ther. 2005, 314, 1079–1086. 19. Cerdan, D.; Grillon, C.; Monsigny, M.; Redziniak, G.; Kieda, C. Biol. Cell. 1991, 73, 35–42. 20. Pivarcsi, A.; Bodai, L.; Rethi, B.; Kenderessy-Szabo, A.; Koreck, A.; Szell, M.; Beer, Z.; Bata-Csorgoo, Z.; Magocsi, M.; Rajnavolgyi, E.; Dobozy, A.; Kemeny, L. Int. Immunol. 2003, 15, 721–730. 21. Thody, A. J.; Graham, A. Pigment Cell Res. 1998, 11, 265–274. 22. Orel, L.; Simon, M. M.; Karlseder, J.; Bhardwaj, R.; Trautinger, F.; Schwarz, T.; Luger, T. A. J. Invest. Dermatol. 1997, 108, 401–405. 23. Clydesdale, G. J.; Dandie, G. W.; Muller, H. K. Immu- nol.Cell Biol. 2001, 79, 547–568. 24. Krizkova, L.; Durackova, Z.; Sandula, J.; Sasinkova, V.; Krajcovic, J. Mutat. Res. 2001, 497, 213–222. 25. Vlckova, V.; Duhova, V.; Svidova, S.; Farkassova, A.; Kamasova, S.; Vlcek, D.; Kogan, G.; Rauko, P.; Miadok- ova, E. Cell Biol. Toxicol. 2004, 20, 325–332. 26. Parat, M. O.; Richard, M. J.; Pollet, S.; Hadjur, C.; Favier, A.; Beani, J. C. J. Photochem. Photobiol. B 1997, 37, 101–106. 27. Shimizu, H.; Banno, Y.; Sumi, N.; Naganawa, T.; Kitajima, Y.; Nozawa, Y. J. Invest. Dermatol. 1999, 112, 769–774. 28. Wang, H. Q.; Quan, T.; He, T.; Franke, T. F.; Voorhees, J. J.; Fisher, G. J. J. Biol. Chem. 2003, 278, 45737– 45745. 29. Greinert, R.; Boguhn, O.; Harder, D.; Breitbart, E. W.; Mitchell, D. L.; Volkmer, B. Photochem. Photobiol. 2000, 72, 701–708. 30. Kappes, U. P.; Luo, D.; Potter, M.; Schulmeister, K.; Runger, T. M. J. Invest. Dermatol. 2006, 126, 667–675. 31. Rhodes, L. E.; Belgi, G.; Parslew, R.; McLoughlin, L.; Clough, G. F.; Friedmann, P. S. J. Invest. Dermatol. 2001, 117, 880–885. 32. Chang, H. R.; Tsao, D. A.; Wang, S. R.; Yu, H. S. Arch. Dermatol. Res. 2003, 295, 293–296. 33. Chakraborty, A. K.; Funasaka, Y.; Slominski, A.; Ermak, G.; Hwang, J.; Pawelek, J. M.; Ichihashi, M. Biochim. Biophys. Acta. 1996, 1313, 130–138. 34. Catania, A.; Colombo, G.; Rossi, C.; Carlin, A.; Sordi, A.; Lonati, C.; Turcatti, F.; Leonardi, P.; Grieco, P.; Gatti, S. Sci. World J. 2006, 6, 1241–1246. 35. Bohm, M.; Luger, T. A. Hautarzt 2004, 55, 436–445. 36. Corre, S.; Primot, A.; Sviderskaya, E.; Bennett, D. C.; Vaulont, S.; Goding, C. R.; Galibert, M. D. J. Biol. Chem. 2004, 279, 51226–51233. 37. Chen, X.; Gresham, A.; Morrison, A.; Pentland, A. P. Biochim. Biophys. Acta 1996, 1299, 23–33. 38. Hatano, Y.; Terashi, H.; Arakawa, S.; Katagiri, K. J. Invest. Dermatol. 2005, 124, 786–792.

36

42. Rheinwald, J. G.; Green, H. Cell 1975, 6, 331–343. 43. Roza, L.; van der Wulp, K. J.; MacFarlane, S. J.; Lohman, P. H.; Baan, R. A. Photochem. Photobiol. 1988, 48, 627–633. 44. Wagner, J. K.; Jovel, C.; Norton, H. L.; Parra, E. J.; Shriver, M. D. Pigment Cell Res. 2002, 15, 379–384.

39. Virag, L.; Szabo, E.; Bakondi, E.; Bai, P.; Gergely, P.; Hunyadi, J.; Szabo, C. Exp. Dermatol. 2002, 11, 189–202. 40. Pernet, I.; Mayoux, C.; Trompezinski, S.; Schmitt, D.; Viac, J. Exp. Dermatol. 2000, 9, 401–406. 41. Boukamp, P.; Petrussevska, R. T.; Breitkreutz, D.; Hor- nung, J.; Markham, A.; Fusenig, N. E. J. Cell Biol. 1988, 106, 761–771.

37

7.2. Effect of advanced glycation endproducts on gene expression

profiles of human dermal fibroblasts

Effect of advanced glycation endproducts on gene expression

profiles of human dermal fibroblasts

J. Molinari, E. Ruszova, V. Velebny, L. Robert

Received: 21 January 2008 / Accepted: 5 February 2008 / Published online: 23 February 2008 Ó Springer Science+Business Media B.V. 2008

Abstract The Maillard reaction and its end products, AGE-s (Advanced Glycation End products) are rightly considered as one of the important mechanisms of post-translational tissue modifications with aging. We studied the effect of two AGE-products prepared by the glycation of lysozyme and of BSA, on the expression profile of a large number of genes potentially involved in the above mentioned effects of AGE-s. The two AGE-products were added to human skin fibroblasts and gene expression profiles investigated using microarrays. Among the large number of genes monitored the expression of 16 genes was modified by each AGE-preparations, half of them only by both of them. Out of these 16 genes, 12 were more strongly affected, again not all the same for both preparations. Both of them upregulated MMP and serpin-expression and downregulated some of the collagen-chain coding genes, as well as the cadherin- and fibronectin genes. The BSA-AGE

preparation downregulated 10 of the 12 genes strongly affected, only the serpin-1 and MMP-9 genes were upregulated. The lysozyme-AGE preparation upregulated selectively the genes coding for acid phosphatase (ACP), integrin chain a5 (ITGA5) and thrombospondin (THBS) which were unaffected by the BSA-AGE preparation. It was shown previously that the lysozyme-AGE strongly increased the rate of proliferation and also cell death, much more than the BSA-AGE preparation. These differences between these two AGE-preparations tested suggest the possibility of different receptor-mediated transmission pathways activated by these two preparations. Most of the gene- expression modifications are in agreement with biological effects of Maillard products, especially interference with normal tissue structure and increased tissue destruction.

Keywords Advanced glycation end products, Aging, Fibroblasts, Gene expression, Maillard reaction

J. Molinari, L. Robert (&) Laboratoire de Recherche Ophtalmologique, Hotel Dieu, Universite Paris 5, 1 place du Parvis Notre Dame, 75181 Paris cedex 04, France e-mail: [email protected]

E. Ruszova, V. Velebny Laboratory of Dermal Applications, R&D Department- Contipro Group, Dolni Dobrouc 401, 561 02 Dolni Dobrouc, Czech Republic

Abbreviations AGEAdvanced glycation end-products NHDF Normal human dermal fibroblasts MMPMatrix metallo proteinases BSABovine serum albumin

Abbreviations of genes tested are explained in the text

38

Introduction

The Maillard reaction was shown to play an important role in aging and age-related pathologies at the cellular, tissue- and organismic levels (Ikan 1996; Baynes et al. 2005 for review). Most of these harmful effects could be attributed to a family of substances designated Advanced Glyca- tion Endproducts or AGE-s (Vlassara 2005; Uribarri et al. 2005). They were shown to accu- mulate in aging tissues, associated with several extracellular matrix (ECM) macromolecules, such as collagen and elastin fibres (Kohn 1971; Robert 2006). Other matrix- and cell membrane associated macromolecules, as fibronectin and others do also become glycated as a result of the direct (non-enzymatic) interaction of glucose and other reducing metabolites with amino groups on proteins and nucleotide bases. Besides their interaction with ECM-components, AGE-s were also shown to react with cell membrane receptors such as RAGE and others (Thornalley 1998). Among the receptor-mediated effects of AGE- products we could demonstrate for some of them an increased cell death in fibroblast cultures, accompanied by an accelerated proliferation (Peterszegi et al. 2006). The most intriguing result described in the above publication was the maintenance of increased cell death in subsequent subcultures of cells even in the absence of newly added AGE-s. Only antibody to FGFR-1 and FGFR-2 decreased cell death, antibodies to RAGE or to EGFR tended to increase it. The most efficient inhibitors of AGE-induced cell death were several ROS-scavengers and rhamnose-rich oligo- or polysaccharides, RROP-s (Peterszegi et al. 2006; Ravelojaona et al. 2006). For these reasons as well as for all previously described cytotoxic effects of AGE-s demonstrated in vitro and their involvement in age-related tissue toxi- city, it was decided to test the effect of some AGE-products on gene expression profiles. The number and variety of cell- and tissue actions ascribed to AGE-products raise the question of their action on the expression profile of genes, related directly or indirectly to the biological activities modified by AGE-products. We shall describe here our experiments carried out on this topic.

Materials and methods

Cell cultures

Normal human dermal fibroblasts (NHDF) were obtained from Cascade Biologics, UK. Cells were grown in DMEM Gibco containing 10% fetal calf serum, amphotericin B, penicillin, streptomycin. The cultures were kept at 37°C in a 7.5% CO2 atmosphere and the culture medium was changed every 2 days. After one passage cells were seeded on six-well plates (7 cm2) at a minimal density of 1x 106 cells per well for microarray experiments. At first, we used the MTT test (Mosmann 1983) to investigate the effect of AGE-products on the viability of cells at three concentrations (1, 5 and 10μM), each with two different incubation times of 24 and 48 h.

AGE-preparations

The preparations of glycated proteins followed previously described methods (Peterszegi et al. 2006; Ravelojaona et al. 2006). Two preparations were tested in these experiments: AGE-1: glycated lysozyme with glucose, and AGE-2: glycated BSA with glucose. For AGE-1, lysozyme (Sigma, Deis- enhofen, Germany) 1.65 mM was incubated with glucose 166 mM for 8 weeks at 37°C with constant shaking in sterile conditions. For AGE-2, BSA (Sigma) 3 mM was incubated with glucose, 1.67 M for 6 weeks at 37°C as for AGE-1. The incubated solutions were passed on a G-25 Sephadex (Phar- macia, Freiburg, Germany) column with D-PBS as solvent, containing 100 U/ml penicillin (Gibco, Heidelberg, Germany) and 100 μg/ml streptomycin (Gibco) for the elution. The first excluded fraction containing the glycated proteins as shown by fluorescence and SDS-PAGE, was used for the experiments and kept at -20°C. Concentrations used, expressed as μM glycated protein are indicated in Table 1. 10 μM of both glycated proteins, lysozyme and BSA were found to be suitable for microarray experiments. Control experiments with native, non-glycated proteins showed that only the glycated proteins exhibited cytotoxic effects on cell cultures (Peterszegi et al. 2006; Ravelojaona et al. 2006).

39

Microarray experiments

After 72 hrs incubation of cells with AGE-s, supernatants were removed and cells subjected to RNA isolation with Trizol Reagent (Invitrogen, Carlsbad, CA, USA), chloroform (Lachema) and 2-propanol (Sigma). The sediment was resuspended in 20 μl DEPC water, absorbances were measured by spectrophotometry (260 and 280 nm) and stored at -80°C. RNA purity was assessed by the 260/280 nm absorbance ratio and the amount of RNA calculated using the conversion factor of 40 μg RNA per optical density unit at 260 nm. Samples were stored at -80°C.

cDNA synthesis

RNA (*10 lg) was incubated with 5 μl of Random primers (Invitrogen at 3 μg/ml concentration) for 10 min at 65°C. First cDNA strand synthesis was carried out by RevertAid TM H Minus M-MUL V Reverse Transcriptase (Fermentas) and labeled with Biotin-11-dUTP (Fermentas). The cDNA purification was performed with YM-30 columns (Millipore, Bedford, MA, USA).

cDNA hybridisation in Array tubes (Clondiag, Chip Technologies GmbH, Jena, Germany)

Two sets CPN-6 and CPN-7 Array Tubes (AT) were purchased from Clondiag Chip Technologies GmbH with spotted 50-mer probes that we selected regarding the CPN spol. sr.o. needs (cat.N°: 2010.10, specification according to array layout CPN-6 and CPN-7, full list can be found at http://www. clondiag.com, http://www.cpn-contipro.cz.) AT-s covered markers, which include genes for cell cycle, cell migration, cell adhesion, homeostasis, cell contact with ECM, production of ECM components, angiogenesis, oxidative stress and inflammation. The chips on the bottom of microtubes contained 70, respectively, 127 different probes spotted in triplicates (Clondiag). Oligonucleotide probes were designed by using Oligoarray v.2.1 and input sequences were choosen in accordance with their GenBank reference accession numbers. The list of genes tested (70 and 127, respectively) can be obtained from [email protected] with the description of the CPN-6 and CPN-7 layouts.

Array hybridisation

The AT were placed in a thermomixer (Eppendorf, Hamburg, Germany), washed three times with 500 μl of hybridisation buffer (Clondiag Chip Technolo- gies), each time at 45°C for 5 min and 550 rpm. Usually, 10μl of labeled cDNA was diluted in 100 μl of hybridisation buffer (Clondiag Chip Technolo- gies), denatured at 94°C for 5 min, cooled down on ice for 2 min and added to the AT. The hybridisation was carried out at 45°C for 3 h by gentle shaking in a hybridisation oven. After hybridisation the AT were washed with 500 μl of 2x sodium chloride—sodium citrate (SSC)—0.2% sodium dodecyl sulfate (SDS) at 30°C for 5 min in the thermomixer at 550 rpm, followed by 500 μl of 2x SSC at 20°C for 5 min at 550 rpm, and finally with 0.2x SSC at 20°C for 5 min at 550 rpm. A blocking step was carried out with 2% (wt./vol) milk powder in 100 μl of 6x SSPE-0.005% Triton at 30°C for 15 min. Poly horseradishperoxidase-streptavidinconjugate (Pierce, Rockford, IL, USA, 100 pg/μl) was added in 100 μl of 6x SSPE-0.005% Triton and incubated at 30°C for 25 min at 550 rpm. The previous washing procedure was repeated after blocking with 29 SSC—0.01% Triton, 29 SSC and 0.29 SSC. AT were kept in the last buffer at 20°C until further processing. Visualisation of hybridisation was achieved by adding 100 μl of peroxidase substrate True Blue (KPL Lab.) to the AT, and detection of signals was done in the AT reader (ATR01, Clondiag Chip Technologies). Signals were recorded at 25°C for 30–40 min. The data were then analysed with the IconoClust version 2.2 software (Clondiag Chip Technologies).

Statistical methods

Normalisation was performed by comparing housekeeping gene transcripts in treated and control (untreated) samples: spot intensities for b-actin , Histone H3, GAPD were chosen as equalising markers. A Mann–Whitney non-parametric test was performed between the two series of values and significance was set at P \ 0.05 (Statgraphics, v.5). A value of 1.5 fold change was set as a sufficient threshold for consideration using a non-fluorescent detection with a colour-scale 0–1.

40

Results

Table 1 gives the results of the microarray experiments. The genes with modified expression profiles are listed for both AGE-preparations tested on Table 1. As can be seen, the expression of a number of genes was modified by the AGE-products tested. About 16 genes out of those tested showed significant modifications of their expression. Half of them was affected by both AGE-preparations tested, more or less to the same extent. Both preparations modified more significantly the expression of about 12 genes, but not always of the same genes. These differences between the two AGE- preparations tested suggest the possibility of different receptor-usage with different transmission pathways between the two AGE-products tested. Of particular

Table 1 List of modified gene expression profiles by fibroblasts incubated with AGE-products at a 10 μM concentration, as described in Methods

Affected genes abbreviations

ACP

FASN

ITGA5

Serpin F1

EFNA1

EFNB

MMP-8

MMP-9

THBS

HSPCA

FN-1 CAD

COL3A1

COL1A2

KDR/VEGFR

DCN -81%

+202%

+160%

+164%

-74%

-64% -45%

-87%

-82%

-64%

-51% -42%

-71%

-68%

-69%

-59%

+140%

AGE-1 (10 μM)

+174%

-81%

+261%

+140%

-64%

-65%

-78%

AGE-2 (10 μM)

interest is the up-regulation of MMP-s by both AGE- products tested, as well as the down-regulation of some of the genes coding for collagen chains. The down- regulation of the cadherin and the fibronectin genes, both involved in cell-cell and cell-matrix interactions is also of interest. Both effects might well loosen cell- matrix interactions and facilitate tissue degradation. It is also worth noticing, that AGE-2, the glycated BSA- preparation, possibly the most abundant glycated protein in the blood of diabetics, down-regulated 10 out of the 12 genes affected, only two were up regulated, serpine F1 and MMP-9, both potentially involved in catabolic processes. Among the up-regulated genes by AGE-1, ACP, ITGA5 and THBS are of interest, because unaffected by the BSA-preparation. In our previous experiments (Peterszegi et al. 2006) glycated lysozyme strongly increased cell death in fibroblast cultures (by 268% as compared to the control culture, P \ 0.0003) and increased the rate of cell- proliferation (by 101%, P \ 0.0065). Non-glycated lysozyme did not exhibit such effects. On the contrary, glycated BSA did not increase cell death and increased only moderately cell proliferation (+63%, N.S.). These differences between the two AGE-preparations tested suggest a considerable importance for the structural- conformational features of the glycated macromolecules as far as their biological activities are concerned. We can however conclude from the above results, that both AGE-preparations tested did significantly modify gene expression. Both AGE-s induced modifications in agreement with the harmful effects attributed to Maillard reaction products in pathological and age- associated modifications of tissues, essentially in diabetes (Kohn 1971; Baynes et al. 2005; Robert et al. 2007).

Discussion

The aim of the above-described experiments was to explore the possibility of AGE-induced modifications of the gene expression profile of human skin fibroblasts. As shown on Table 1, the expression of at least 16 genes out of a large number of genes observed was significantly modified by the two AGE- products investigated. Glycated BSA (AGE-2) but also glycated lysozyme (AGE-1) modified the expression of genes corresponding to the harmful effects of the Maillard reaction (Ikan 1996; Baynes et al. 2005;

Up-regulated genes are marked + and down-regulated genes -. The figures indicate % modifications as compared to control cultures set at 100%. Spot intensities were normalised according to b-actin, GAPD and Histone H3 housekeeping gene intensity in the control samples

ACP, acid phosphatase; FASN, fatty acid synthase; ITGA5, integrin chain a5; F1; EFNA1, ephrin A1; EFNB, ephrin B; MMP-8 and -9, matrix metallo proteinases; THBS, thrombospondins; HSPCA, heat shock protein A; FN-1, fibronectin; CAD, cadherin; Col3A1, collagen type III chain a1; Col1A2, collagen type I chain a2; KDR/VEGFR, receptor for KD and for VEGF; DCN, decorin

41

L. Robert and A. M. Robert 2007 for review). The up-regulation of some MMP-s and the down- regulation of genes coding for ECM-components as some collagen chains and fibronectin are good exemples. Some genes were affected only by one of the two AGE-s investigated. Glycated lysozyme increased acid phosphatase (ACP) and MMP-8 expression, glycated BSA did not. This AGE-2 did however modify several genes (serpin F1, EFNA1, EFNB) not affected by AGE-1 (see Table 1). Fatty acid synthase (FASN) was down-regulated in presence of both AGE-s. Aging in rats is associated with significant reduction of lipogenic enzymes gene expression and the rate of fatty acids synthesis in white adipous tissue (Nogalska and Swierczensky 2004). MRNA concentrations of FASN were compared between young and old rats, and the levels in old animals were lower, mostly 40–70% of those in young animals. It is suggested that the age-dependent decrease of FASN can mainly be ascribed to the transcriptional steps (Fukuda and Iritani 1992). In our experiments FASN was down-regulated with both AGE preparations, but more with AGE-2. AGE-1 and AGE-2 were monitored on the transcriptional level also to determine their influence on changes in gene expression of markers of the CPN-7 array. It was shown previously that aging is accompanied by a radical change in fibroblast morphology— from a phenotype associated with active ECM production to one with arrest of mitotic activity and increased catabolism (Perrier et al. 2004). This passage from the mitotic to the senescent phenotype can be attributed to (anti-) oncogene-mediated switch of gene-usage (Campisi 2005; Labat-Robert and Robert 2007). The role of fibronectin is crucial in maintaining tissue architecture as a real ‗‗biological cement‘‘ (Hynes 1990). We observed the attenuation of the following genes at the mRNA levels: fibronectin (FN-1), cadherin (CAD) and heat-shock protein A (HSPCA). A reduction of dermal fibronectin level was reported during skin aging (Vitellaro-Zuccarello et al. 1992; Pieraggi et al. 1984; Perrier et al. 2004). Decrease and/or fragmentation of FN is probably involved in the morphological modifications of dermal connective tissue with age (Labat-Robert 2004). Immunostaining for fibronectin showed a decrease in the wounds of old mice, with delay of the inflammatory response of re-epithelisation, and neosynthesis of ECM components (Ashcroft et al. 1997).

Compared to control levels, MMP-8 and MMP-9 gene expression was increased in presence of both AGE preparations, suggesting enhanced ECM degradation (Lindsey et al. 2005). Besides alterations in MMP-9 gene expression levels, upregulation of Serpin F1 (PEDF) transcription was observed also further enhancing degradative activity of cells. Advancing age was shown to be associated with greater alpha-1 and alpha-5 integrin expression, suggesting that they undergo coordinated regulation in the aging heart (Burgess at al. 2001). The existence of coordinated changes in the expression of functionally related groups of genes with age was proposed including decreased expression of heat-shock genes (HSPCA, 90 kDa heat shock protein; Melov and Hublard 2004). Decorin (DCN), downregulated by both AGE-s tested (Table 1) belongs to SRP family, shown to associate with the C-terminus of collagen type I. Its deficiency affects fibril formation and fibril stability in tissues due to lack of lateral association of collagen fibres. Decorin also associates with EGF. Decorin mRNA level in the rat skin decreases until 90 days of age, although Western blotting showed that the amount of decorin increased with age on a protein level (Nomura et al. 2003). Serpin F1 (PEDF), upregulated only by AGE-2, can decrease the generation of reactive oxygen species (ROS) in AGE-exposed microvascular endothelial cells by suppressing NADPH oxidase activity via downregulation of mRNA levels of p22phox and gp91phox. This leads to a blockade of the AGE- elicited Ras activation and NF-jB-dependent VEGF gene induction (Yamagishi et al. 2006). The CAD encoded protein, downregulated by both AGE-s, is a calcium-dependent cell–cell adhesion glycoprotein composed of five extracellular cadherin repeats, a transmembrane region and a highly conserved cytoplasmic tail. It functions as a classical cadherin by imparting to cells the ability to adhere in a homophilic manner, and as such it may play an important role through the control of cohesion and organisation of the intercellular junctions. Of interest are also the differences between the gene-expression modulation of the two AGE-s examined. These differences suggest that in accord with the multiple receptors shown to be involved in the mediation of AGE-action on cells (Thornalley 1998),

42

7

different types of actions can be expected and found with different glycated tissue components. As however serum albumin is one of those proteins reaching potentially high diffusible tissue concentration, its action on gene expression as exemplified by the results presented on Table 1 might well represent the most frequent in vivo outcome. It has to be mentioned however that the nature of the glycation reaction used for the production of AGEs with or without accelerated glycoxidation by the addition of Fe to the protein-reducing sugar mixture also influences the outcome of AGEs on cells as shown by our previous investigations (Peterszegi et al. 2006).

Melov S, Hubbard A (2004) Microarrays as a Tool to Inves- tigate the Biology of Aging: a Retrospective and a Look to the Future. http://sageke.sciencemag.org/cgi/content/full/ 2004/42/re7 Mosmann T (1983) Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Meth 65:55–63 Nogalska A, Swierzczynski J (2004) Potential role of high serum leptin concentration in age-related decrease of fatty acid synthase gene expression in rat white adipous tissue. Exp Gerontol 39:147–150 Nomura Y, Abe Y, Ishii Y, Watanabe M, Kobayashi M, Hattori A, Tsujimoto M (2003) Structural changes in the gly- cosaminoglycan chain of rat skin decorin with growth. J Dermatol 30:655–664 Perrier E, Pivard F, Grenier S, Andre V (2004) The fibronectin network during aging: a missing cell connectivity. J Cosmet Sci 55:215–216 Peterszegi G, Molinari J, Ravelojaona V, Robert L (2006) Effect of advanced glycation end-products on cell proliferation and cell death. Path Biol 54:396–404 Pieraggi MT, Julian M, Bouissou H, Stocker S, Grimaud JA (1984) Dermal aging. Immunofluorescence study of collagens I and III and fibronectin. Ann Pathol 4:185–194 Ravelojaona V, Molinari J, Robert L (2006) Protection by rhamnose-rich polysaccharides against the cytotoxicity of Maillard reaction products. Biomed Pharmacother 60:359–362 Robert L (2006) Fritz Verzar was born 120 years ago: his contribution to experimental gerontology through the collagen research as assessed after half a century. Arch Gerontol Geriat 43:13–43 Robert L, Robert AM (2007) La reaction de Maillard—Role physiopathologique et approche pharmacologique. J Soc Biol 201:167–174 Thornalley J (1998) Cell activation by glycated proteins. AGE receptors, receptor recognition factors and functional classification of AGE-s. Cell Mol Biol 44:1013–1023 Uribarri J, Cai W, Sandu O, Peppa M, Goldberg T, Vlassara H (2005) Diet-derived advanced glycation end products are major contributors to the body‘s AGE pool and induce inflammation in healthy subjects. Ann NY Acad Sci 1043:461–466 Vitellaro-Zuccarello L, Garbelli R, Rossi VD (1992) Immu- nocytochemical localization of collagen types I, II, IV and fibronectin in the human dermis. Modifications with aging. Cell Tissue Res 268:505–511 Vlassara H (2005) Advanced Glycation in Health and Disease. Role of the modern environment. Ann NY Acad Sci 1043:452–460 Yamagishi S, Nakamura K, Matsui T, Inagaki Y, Takenaka K, Jinnouchi Y, Yoshida Y, Matsuura T, Narama I, Motomiya Y, Takeuchi M, Inoue H, Yoshimura A, Bucala R, Imaizumi T (2006) Pigment epithelium derived factor (PEDF) inhibits advanced glycation end-product-induced retinal vascular hyperpermeability by blocking ROS- mediated endothelial growth factor (VEGF) expression. J Biol Chem 281:20213–20220

Acknowledgments Supported by Institut DERM, Paris. We thank Professor Gilles Renard, Head of Ophthalmology Department at the Hotel Dieu Hospital for his generous hospitality and interest in our experiments.

References

Ashcroft GS, Horan MA, Fergusson MW (1997) Aging is associated with reduced deposition of specific extracellular matrix components, an upregulation of angiogenesis, and an altered inflammatory response in a murine incisional wound healing model. J Invest Dermatol 108:430–437 Baynes JW, Monnier VM, Ames JM, Thorpe SR (eds) (2005) The Maillard Reaction. Chemistry at the interface of nutrition, aging and disease. Annals of the New York Academy of Sciences, vol 1043. The New York Academy of Sciences, New York Burgess ML, McCrea JC, Hedrick HL (2001) Age-associated changes in cardiac matrix and integrins. Mech Ageing Dev 122:1739–1756 Campisi J (2005) Senescent cells, tumor suppression and organismal aging: good citizens, bad neighbors. Cell 120:513–522 Fukuda H, Iritani N (1992) Effects of aging on gene expression of acetyl-CoA carboxylase and fatty acid synthase in rat liver. J Biochem 12:277–280 Hynes RO (1990) Fibronectins. Springer, Heidelberg, Germany Ikan R (ed) (1996) The Maillard reaction. Consequences for the chemical and life sciences. Wiley, Chichester, UK Kohn RR (1971) Principles of mammalian aging. Prentice Hall, Inc., NJ Labat-Robert J (2004) Cell-matrix interactions in aging: role of receptors and matricryptins. Ageing Res Rev 3:233–247 Labat-Robert J, Robert L (2007) The effect of cell-matrix interactions and aging on the malignant process. Adv Cancer Res 98:221–259 Lindsey ML, Goshorn DK, Squires CE, Escobar GP, Hendrick JW, Mingoia JT, Sweterlitsch SE, Spinale FG (2005) Age- dependent changes in myocardial matrix metalloproteinase/ tissue inhibitor of metalloproteinase profiles and fibroblast function. Cardiovasc Res 66:410–419

43

7.3. Rhamnose recognizing lectin site of human dermal fibroblasts

functions as a signal transducer

Biochimica et Biophysica Acta

The α-L-Rhamnose recognizing lectin site of human dermal fibroblasts functions as a

signal transducer Modulation of Ca2+ fluxes and gene expression

Gilles Faury a,b,c,d,⁎, E. Ruszova e, J. Molinari f, B. Mariko a,b,c,d, S. Raveaud a,b,c,d, V. Velebny e, L. Robert f

a Laboratoire Physiopathologies vasculaires: interactions cellulaires, signalisation et vieillissement, iRTSV-APV, CEA-Grenoble, 17 rue des Martyrs, 38054 Grenoble Cedex 9, France Université Joseph Fourier, Grenoble, F-38041, France cCEA, Grenoble, F-38054, France dINSERM, U882, Grenoble, F-38054, France eLaboratory of Dermal Applications, R & D Department, CPN s.r.o. Dolni Dobrouc, 401, 561 02, Czech Republic fLaboratoire de Recherche Ophtalmologique, Université Paris 5, Hôpital Hôtel Dieu, 1 place du Parvis Notre Dame, 75181 Paris Cedex 04, France

b

a r t i c l e i n f o a b s t r a c t

An α-L-Rhamnose specific lectin site was described on human skin keratinocytes and fibrobasts. The addition of Rhamnose-rich oligo- and polysaccharides (RROPs) to fibroblasts has been shown to stimulate cell proliferation and increase extracellular matrix biosynthesis, suggesting that this lectin site functions as a ―true‖ receptor transmitting messages to the cell interior. It was confirmed here that addition of the Rhamnose-rich polysaccharide, RROP-1, to normal human dermal fibroblasts (NHDFs) and human endothelial cells produced a dose-dependent stimulation of the calcium-signaling pathway, inducing fast and transient increases in Ca2+ influx and intracellular free Ca2+ level. The Rhamnose-rich oligosaccharide RROP-3 as well as L-Rhamnose alone were also able to trigger similar intracellular free Ca2+ concentration increases in NHDFs. Moreover, the recording of the RROP-1-induced modification of the gene-expression profile in fibroblasts showed that this polysaccharide triggered a down-regulation of the expression of several growth factors, adhesion molecules and extracellular matrix proteins involved in pro-tumoral activity and/or fibrotic processes. These results further support the hypothesis of a receptor function for the Rhamnose-recognizing lectin site in fibroblasts. Anti-fibrotic and anti-tumoral potential of RROP-1 remains to be further explored. © 2008 Elsevier B.V. All rights reserved.

Article history: Received 24 December 2007 Received in revised form 5 July 2008 Accepted 10 July 2008 Available online 28 July 2008

Keywords: Rhamnose Lectin Receptor Calcium channel Gene regulation Human endothelial cell Human dermal fibroblast

1. Introduction

Oligo- and polysaccharides are ubiquitous components of biological macromolecules, present as side-chains on glycoproteins or as polysaccharides linked to proteins as in proteoglycans. Many studies have described the structure and function of the glycan components of glycoproteins and proteoglycans [1,2 for reviews]. Two major discoveries greatly accelerated these advances: i) blood group substances, identified with oligosaccharides [3 for review], and ii) lectins, mostly of plant origin and later identified in vertebrate tissues also [4 for review]. Interactions involving blood group substances or lectins, or sugars and proteins, were supposed to result in formation of non-covalent although strong complexes, such as in agglutination of

⁎ Corresponding author. Laboratoire Physiopathologies vasculaires: interactions cellulaires, signalisation et vieillissement, INSERM U882, iRTSV-APV, CEA-Grenoble, 17 rue des Martyrs, 38054 Grenoble Cedex 9, France. Tel.: +33 4 38 78 23 48; fax: +33 4 38 78 49 64. E-mail address: [email protected] (G. Faury).

red cells or other cells, with no other functional consequences. Only more recent studies of animal lectins showed that carbohydrate– lectin interactions might result in message transmission to the cell interior, raising the hypothesis that some lectins could be considered as «true» receptors [5–7 for reviews]. Such interactions mainly concerned glycan components, currently found in vertebrate glycoconjugates, especially Galactose-recognizing lectins [8,9], which include the elastin-laminin receptor [10]. A surprising finding was the demonstration of an α-L-Rhamnose recognizing lectin on human keratinocytes by the team of Monsigny and Kieda [11], although Rhamnose is only present in plant and prokaryote glycoconjugates, but was not demonstrated in vertebrate glycoconjugates. These authors showed that the α-L-Rhamnose recognizing lectin presents only a relative specificity for rhamnose and binds also, although with a lower affinity, to at least glucose and fucose [11]. Previous experiments indicated that Rhamnose-rich oligo- and polysaccharides (RROPs), used as ligands for the α-L-Rhamnose recognizing lectin, did interact with normal human dermal fibroblasts (NHDFs), produced modifications of cell proliferation and matrix biosynthesis

44

[12]. The suggestion of an α-L-Rhamnose recognizing lectin-mediated message transduction [12] was confirmed and extended here by showing that RROP-1, RROP-3 and L-Rhamnose rapidly and significantly increased calcium concentration in NHDFs. RROP-1 also changed the gene-expression profile of these cells.

2. Materials and methods

Rhamnose-rich polysaccharide-1 (commercial name: Rhamno- soft®, abbreviated RROP-1) was obtained from Klebsellia pneumoniae, a non pathogenic bacterial strain. RROP-1 has an average molecular weight of about 50 kDa and consists in repeating pentasaccharide units composed of α-L-Rhamnose (Rha), β-D-Galactose (Gal) and β-D- glucuronic acid (GlcA), with one Rhamnose per pentasaccharide protruding from the polysaccharide chain:

2.3. Intracellular free calcium concentration in NHDFs and HUVECs

Fluorescence microscopy experiments were performed on adhering NHDFs or HUVECs by using the calcium-sensitive dye FLUO3 (excitation: 488 nm; emission: N510 nm), as previously described [15]. The cells were washed twice with physiological saline solution (PSS) containing: 125 mM NaCl; 5.6 mM KCl; 2.4 mM CaCl2; 1.2 mM MgCl2; 10 mM HEPES (pH 7.4). 0.75 ml PSS was then added followed by addition of 0.25 ml of a 2% bovine serum albumin and 0.08% Pluronic F127 solution in PSS containing 5 μl of a 0.5 mM FLUO3/AM solution in DMSO (final concentrations: 2.5 μM FLUO3/AM, b1% DMSO). After a 45 minute incubation of NHDFs at 37 °C (25 min at room temperature for HUVECs) in the dark, the cells were washed twice and bathed in 1.5 ml PSS 10 min before and during the measurements (22 °C). Fluorescence microscopy recordings were performed on a Cell-R system (Olympus, Rungis, France). In each experiment, fluorescence was recorded for about 15 min (1 image each 5 s). F/Fo is the ratio of current fluorescence to initial fluorescence and is assumed to be indicative of [Ca2+]i. RROP-1 was added 2 min after recording has started. Experiments were at least triplicated for each RROP-1 concentration.

2.4. Patch-clamp experiments — Ca2+ influx in NHDFs

Single-channel currents were recorded at room temperature (22 °C) from cell-attached patches on NHDF membranes (Holding Potential = +20 mV) and analyzed using previously described procedures and instrumentation [15]. Before recording, the cells were washed twice then bathed in PSS. The patch pipette was filled with 90 mM Ba(CH3COO)2 and 10 mM HEPES (pH 7.4). Barium was used instead of calcium in order to improve signal resolution, because calcium channels are generally more permeable to this ion, and barium is known to inhibit potassium currents [17]. RROP-1 was added to the bath during the recording. NHDF membrane resting potential (Vm) was assumed to be −109 ± 3 mV [18]. The recorded transmembrane currents were integrated and analyzed using the software Biopatch (Biologic, Claix, France).

2.5. Microarray and ELISA experiments in NHDFs

First, the MTT viability method of Mossmann [13] was used to investigate the effect of four concentrations of RROP-1 (10, 100, 250, and 500 μg/ml) on NHDFs. 10 and 100 μg/ml RROP-1 were found to be suitable for microarray experiments. After incubation of cells with RROP-1 for 48 h , supernatants were removed and cells were subjected to RNA isolation with Trisol Reagent (Invitrogen, Carlsbad, CA, USA). RNA (20 μg) was first incubated with random primers (Invitrogen, Carlsbad, CA, USA) for 10 min at 65 °C. cDNA strand synthesis was carried out by RevertAid H Minus M-MULV Reverse Transcriptase (Frementas, Berlington, Ontario, CA) and labelled with Biotin-11-dUTP (Fermentas, Berlington, Ontario, CA). Array microtubes contained chips with 127 50-mer probes covering genes for cell cycle, cell migration, cell adhesion, homeostasis, cell adhesion to ECM, production of ECM components, angiogenesis, and inflammation (Clondiag, Jena, Germany). Oligonucleotide probes were designed by using Oligoarray Designer 2.2 and input sequences were chosen in accordance with their GenBank reference accession numbers. Usually, 10 μl labeled genomic DNA were diluted in 100 μl hybridization buffer (Clondiag, Jena, Germany), denatured at 94 °C for 5 min , cooled down on ice for 2 min and added to the array tubes. The hybridization was carried out at 45 °C for 3 h and the chips were subsequently washed. A blocking step was carried out with milk powder, before incubation with poly-horseradish peroxidase-streptavidin conjugate. Hybridized probes were revealed by addition of 100 μl peroxidase substrate True Blue (KPL Lab., Gaithesburg, Maryland, USA) to the array tubes. Detection of signals was done in an Array Tube Reader (ATR01,

Rhamnose-rich oligosaccharide-3 (RROP-3) resulted from acid hydrolysis of a precursor (RROP-2) obtained from Klebsellia planticola. RROP-3 has a molecular weight of 5 kDa and consists in repeating tetrasaccharide units composed of α-L-Rhamnose (Rha), β-D-Glucose (Glc) and β-D-glucuronic acid (GlcA), with one Rhamnose per tetrasaccharide protruding from the oligosaccharide chain:

RROP-1 and RROP-3 were furnished by SOLABIA (Pantin, France) as sterile, lyophilised powder and kept at −20 °C. Sterile solutions in PBS were prepared for the experiments.

2.1. Normal human dermal fibroblasts (NHDFs)

Cells were obtained from Cambrex (Emerainville, France) derived from a 39 year old woman (code CC-2511, lot n°4F1293) and cultured in a Dulbecco-modified Eagle's essential medium (DMEM glutamax, GIBCO- Invitrogen, Eragny, France) with penicillin 100 U/ml, Streptomycin 100 U/ml penicillin, 100 µg/ml streptomycin, 0.25 µg/ml Amphotericin B, with 10% fetal calf serum (FCS, GIBCO) in humidification-controlled culture flasks. Culture medium was changed every 2–3 days. Cells were seeded in 12, 24 or 96 well Nunc-plates, at a density of 5.104 cells/ml, or in 35 mm culture dishes. Cell proliferation was controlled by the MTT- test [13] and by counting trypsinized cells in a Coulter-counter.

2.2. Human umbilical venous endothelial cells (HUVEC)

HUVECs were obtained according to a protocol derived from Jaffe et al. [14,15]. Under a sterile hood the umbilical vein was cannulated and perfused with a physiological buffer solution containing: 140 mM NaCl, 4 mM KCl, 7.6 mM D-glucose, 15 mM HEPES, plus 0.1 mg/ml-1 streptomycin, 100 U/ml-1 penicillin, and 0.1% phenol red (pH 7.4). HUVECs were collected after perfusion of the umbilical vein with 0.1% collagenase 1A at 37 °C for 10 min. The cells were cultured in medium 199 containing 22% human serum, 20 mM HEPES, 10 mM NaHCO3, 2 mM L-glutamine, 0.075 mg/ml streptomycin, 75 units/ml penicillin and 0.1% phenol red (37 °C, 5% CO2, humid atmosphere), in 0.25 mg/ml fibronectin-coated dishes [16]. The culture medium was replaced after 2 h and the cells were grown in the same conditions. The cells were used before the second passage as primary culture.

45

Fig. 2. Representative images of the evolution of Fluo-3 fluorescence over time in NHDFs stimulated with 10 μg/ml RROP-1 at 2 min . (A) Time 0 min: before RROP-1 addition. (B) Time 3.4 min: some cells already exhibited elevated fluorescence (left). (C) Time 11.3 min: some cells still presented elevated fluorescence, while some others just started to respond (right) and some other cell responses had ended (extreme upper left). Some cells did not respond at all to RROP-1 in the time-frame of the experiment and underwent a normal decrease in fluorescence over time because of Fluo-3 photobleaching (cell at the upper right of the images).

Clondiag Chip Technologies, Jena, Germany). The data were then analyzed with the IconoClust version 2.2 software (Clondiag Chip Technologies, Jena, Germany). A Mann–Whitney non parametric test was performed between the two series of values and significance was set at p b 0.05. The two normalization markers used were β-actin and GAPD. A more detailed protocol can be found in the online Supplementary material-1.

Fig. 1. Evolution of Fluo-3 fluorescence in NHDFs in response to RROP-1 added at 120 s: 0 μg/ml — solvent PBS (A), 0.01 μg/ml (B), 0.1 μg/ml (C), 1 μg/ml (D), 10 μg/ml (E). In each graph, one tracing corresponds to the response of one single cell (n N 30 cells for each concentration).

46

3. Results

3.1. Effect of RROP-1 on intracellular free calcium concentration

In NHDFs, no response of the solvent (PBS) of RROP-1 could be detected, while a dose-dependent increase in [Ca2+]i was observed in the 0.01–10 μg/ml RROP-1 range. Only some cells, not all, responded to RROP-1, leading to fluorescence increases up to +50–150% at 10 μg/ml RROP-1 concentration (Fig. 1A–E). The evolution pattern of fluorescence was variable in responding cells: most of them exhibited fast transient elevation of fluorescence, while a few others presented a slower and long-lasting elevation of fluorescence within the time- frame of the experiments (Figs. 1 and 2, and online Supplementary materials-(movie)-2–3). There is no known specific blocker of Rhamnose-dependent binding to the Rhamnose-recognizing-lectin. Therefore, in order to verify that the RROP-1-induced elevation of NHDF [Ca2+]i could be attributed to the activation of the Rhamnose-recognizing-lectin, we then stimulated NHDFs with different agonists. It is difficult, if not impossible, to find a ―control‖ polysaccharide inactive on cells, since any carbohydrate chain could potentially bind to one of the wide

Fig. 3. Evolution of Fluo-3 fluorescence in NHDFs in response to solvent (A), 1.3 μg/ml RROP-3 (B), 10 μg/ml L-Rhamnose (C) and 10 μg/ml D-Galactose + 10 μg/ml D-Glucuronic acid (D). Agonists were added at 120 s. In each graph, one tracing corresponds to the response of one single cell (n N 30 cells for each concentration).

ELISA experiments for PDGF determination were performed according to standard technique. ELISA kit was purchased from R&D Systems (Mineapolis, MN, USA). After application or not of 10 μg/ml RROP-1 to NHDFs, absorbance of 75 μl cell culture media was determined by using a microplate reader Versa max (Molecular devices, Sunnyvale, CA, U.S.A.) at a wavelength 450 nm, with the correction wavelength set at 570 nm.

Fig. 4. Evolution of Fluo-3 fluorescence in HUVECs in response to RROP-1 added at 120 s. (A) solvent of RROP-1 (PBS). (B) 0.1 μg/ml RROP-1. (C) 10 μg/ml RROP-1. Beside cells responding to RROP-1, some other cells did not respond at all in the time-frame of the experiments. In each graph, one tracing corresponds to the response of one cell. n = 14– 30 cells for each concentration.

47

variety of lectins present on cell membranes, including galectins and selectins, and induce a variety of reactions [4–7,19–21]. For instance, fucose/lectin interaction triggers stimulation of fibroblast proliferation and inhibition of MMP-expression and activation [22,23] and lactose, galactose or N-acetylgalactosamine binding to the elastin receptor (a galactolectin) promotes dissociation of the receptor from elastin and the cell [24,25]. We therefore chose to study the effects on NHDF [Ca2+]i of: i) a Rhamnose-rich oligosaccharide (RROP-3), ii) the monosaccharide L-Rhamnose, and separately iii) a mix of D-Galactose and D-Glucuronic acid (since RROP-1 is composed of L-Rhamnose, D- Galactose and D-Glucuronic acid). Our results showed that RROP-3 triggered mainly early/fast/transient and, to a lower extent, late/long- lasting elevations of NHDF [Ca2+]i, with a pattern similar to that induced by RROP-1. L-Rhamnose mainly induced early and transient increases in NHDF [Ca2+]i, and no late/long-lasting elevations. The differences between the effects of L-Rhamnose, RROP-3 and RROP-1 are likely due to the structural differences between these molecules: a 164 Da monosaccharide for Rhamnose vs. a 5 kDa oligosaccharide containing rhamnose, glucose and glucuronic acid for RROP-3 or a 50 kDa polysaccharide containing rhamnose, galactose and glucuronic acid for RROP-1. The D-Galactose and D-Glucuronic acid mix induced little effect with a different pattern, i.e. a slight, late and long-lasting increase in NHDF [Ca2+]i in only few responding cells (Fig. 3). Complementary experiments performed in HUVECs showed that the effect of 0.1–10 μg/ml RROP-1 was also dose-dependent and heterogeneous, including the absence of response in some cells and, in responding cells, fast transient elevation or, to a lower extent, slow and long-lasting elevations of [Ca2+]i (+30–200%) (Fig. 4).

3.2. Effect of RROP-1 on calcium channel activity in NHDFs

In order to elucidate the question of the origin of calcium in the above cited elevation of [Ca2+]i, we have performed patch-clamp experiments enabling to study the response of membrane calcium channels when NHDF were stimulated with RROP-1. The basal activity of NHDF calcium channels was low (Figs. 5A). Addition of 10 μg/ml RROP-1 induced a strong activation of calcium channels (Fig. 5B) in all the cells tested. Integration of the calcium charges crossing the cell

membrane over time allowed the calculation of the transmembrane calcium current, which increased up to 4 pA (Fig. 5C) shortly after application of the polysaccharide. The conductance of the RROP-1- activated calcium channels was measured and found in the range of 23 pS (Fig. 5D).

3.3. Effect of RROP-1 on gene expression and protein release in NHDFs Microarray experiments were carried out in order to explore modifications of gene expression as a result of the interaction between fibroblasts and RROP-1. Expression of twelve genes was found to be significantly modified by addition of 10 μg/ml RROP-1, all of them being down-regulated by 20–30% (to about 70–80% of the control value): SERPINE1: plasminogen activator inhibitor-1 (PAI-1), STAB2: stabilin-2, FGF1: fibroblast growth factor-1, FGF9: fibroblast growth factor-9, TGFB3: transforming growth factor beta-3, ITGB1: integrin beta-1, COL1A1: collagen-I-alpha-1, NID: nidogen, IL-8: interleukin-8, PDGFC: platelet-derived growth factor-C, FGFR2: fibroblast growth factor receptor-2, LAMB1: laminin-beta-1 chain (Fig. 6A, B). In order to verify how the RROP-1-induced down-regulation of gene expression impacted the protein production level, control measurements of PDGF release by NHDFs were performed. Compared to control, the ELISA experiments showed a 20% decrease in PDGF release after cell stimulation by RROP-1 (Fig. 6C), in good agreement with the microarray experiments.

4. Discussion

The purpose of these experiments was to further explore the possibility of a receptor/signal transducer function for the α-L- Rhamnose recognizing lectin site, demonstrated to be present on human skin keratinocyte and fibroblast membranes [11,12]. Prelimin- ary results were in favour of this contention, since they suggested that Rhamnose-rich oligo- and polysaccharides (RROPs) enhanced fibroblast proliferation and collagen production, increased intracellular calcium, and modulated gene expression [12]. Consistently, our present results showed that the Rhamnose-rich polysaccharide RROP-1 was able to induce a dose-dependent elevation of [Ca2+]i in

Fig. 5. Membrane calcium channel activation induced by RROP-1 in NHDFs demonstrated with the patch-clamp technique. (A) addition of the solvent (PBS). (B) addition of 10 μg/ml RROP-1. (C) integration over time of the calcium charges passing through the cell membrane before and after addition of 10 μg/ml RROP-1. (D) Current–membrane potential (I/V) curve, leading to calculation of the conductance of the RROP-1-activated calcium channels: 23 pS. n = 7 in each group.

48

Fig. 6. Microarray and ELISA experiments in NHDFs. Scatter-plot of the changes in mRNA expression following application of 10 μg/ml RROP-1 (A). In panel B are represented the 10 μg/ml RROP-1-induced down-regulations of the following genes: SERPINE1: plasminogen activator inhibitor-1 (PAI-1), STAB2: stabilin-2, FGF1: fibroblast growth factor-1, FGF9: fibroblast growth factor-9, TGFB3: transforming growth factor beta-3, ITGB1: integrin beta-1, COL1A1: collagen-I-alpha-1, NID: nidogen, IL-8: interleukin-8, PDGFC: platelet-derived growth factor-C, FGFR2: fibroblast growth factor receptor-2, LAMB1: laminin-beta-1 chain. The decrease in the release of PDGF, measured by ELISA, following addition of 10 μg/ml RROP-1 is presented in panel C.

NHDFs, and to a lower extent in HUVECs, with a maximum effect at 10 μg/ml, equivalent to about 20 nM. The Rhamnose-rich oligosaccharide RROP-3 also induced calcium responses of a similar pattern, which were fast/transient or (to a lower extent) slow/long lasting, of heterogeneous amplitude, asynchronous, not all cells reacting at the same time and with the same amplitude. Similar findings were reported by us during our studies on the elastin-laminin receptor on HUVECs [15] and by others, for instance regarding the effect of ATP and shear stress on cell calcium mobilization [26]. However, activation of the elastin-laminin receptor was somehow different since: i) no slow/long-lasting increase in calcium was detected and ii) this receptor contains two binding sites, the lectin site only modulating fixation of elastin on the protein-binding site. Verification of the roles of Rhamnose and the α-L-Rhamnose-recognizing lectin in triggering calcium signaling was achieved using the monosaccharides contained

in RROP-1, i.e. L-Rhamnose, D-Galactose and D-Glucuronic acid. These experiments showed that only RROP-1, RROP-3 and L-Rhamnose triggered fast/transient elevations of [Ca2+]i while the D-Galactose/D- Glucuronic acid mix triggered only occasionally a slight late/slow/ long-lasting increases in [Ca2+]i in NHDFs. This suggests that the calcium-signaling events produced by RROP-1 are largely mediated by the Rhamnose-recognizing-lectin (fast response). However, the slight cell responses triggered by D-Galactose and D-Glucuronic acid also suggest that the α-L-Rhamnose-recognizing lectin presents only a relative specificity, consistent with previous experiments showing that Glucose-, Fucose- and N-Acetyl-Glucosamine-Proteins also bind to this lectin although with a lower affinity [11]. This situation, an only relative specificity, is frequent for lectins: for instance, lactose, galactose and N-acetylgalactosamine-containing chondroitin sulfate bind to the elastin-laminin receptor [24,25], multiple glycans bind to galectins [19], and multiple glycans and oligosaccharides bind to selectins [21]. RROP-1-induced increase in [Ca2+]i was, at least in part, of extracellular origin since patch-clamp experiments in NHDFs demonstrated that this molecule triggered a substantial activation of membrane calcium channels, and subsequently calcium influx, in all the cells studied. The conductance of the activated calcium channels was in the range of 23 pS, once again close to the value observed with calcium channel activation following stimulation of the elastin- laminin receptor: 16 pS [15]. RROP-1 induced an activation of membrane calcium channels in all the NHDFs studied, [Ca2+]i increase was however heterogeneous, as far as the proportion of the cells and the intensity of the reaction is concerned. This might indicate that the RROP-1-triggered calcium influx could either be dispatched in the whole cytosolic compartment, participating in [Ca2+]i elevation, or compartmentalized in sub-membrane spaces, with no or very limited influence on [Ca2+]i. An alternative explanation could be that RROP-1- induced calcium influx triggered the release of intracellular calcium stores (calcium-induced calcium release), resulting in an observable [Ca2+]i increase only when calcium influx is of sufficiently high amplitude. Lower calcium influxes might be unable to induce mobilization of the calcium stores. Our present results on calcium fluxes also suggest that action of RROP-1 is not tissue-specific since we demonstrated its efficiency on intracellular calcium regulation in two very different cell types, i.e. human dermal fibroblasts (NHDFs) and endothelial cells (HUVECs), indicating that the Rhamnose-recognizing lectin appears to have a rather general distribution and signaling role in the organism. Among the important roles of intracellular calcium signaling are the regulation of cell cycle and gene expression [27,28]. It was already demonstrated that RROP-1 could increase proliferation of NHDFs [12]. Here, we further investigated the action of RROP-1 on gene expression using microarrays. All 12 genes whose expression was significantly modified by 10 μg/ml RROP-1 were down-regulated: PAI-1, IL-8, 4 growth factors and at least 1 growth factor receptor (FGF1, FGF9, FGFR2, TGFB3, PDGFC), 2 adhesion molecules (STAB2, ITGB1) and 3 extracellular matrix (ECM) proteins (LAMB1, COL1A1, NID). We also verified if RROP-1-induced changes in mRNA levels could result in corresponding protein level variations. Because the expression of several genes was modulated by RROP-1, we chose to address this question regarding only one of the most important down-regulated genes: PDGF. At the protein level, RROP-1-induced a lowering of PDGF release by NHDFs in the same range as the down-regulation of the PDGF gene expression (−20%), confirming the inhibitory role of RROP- 1. Some consistency could be found from the literature regarding the simultaneous repression of these genes, all implicated in angiogenesis, cancer cell transformation, proliferation or invasiveness [29–41]. Also, RROP-1 decreased expressions of several ECM proteins and growth factors stimulating ECM production by fibroblasts. Similar effects were demonstrated by an anti-fibrotic drug, Imatinib, on the production of the same proteins as those modified by RROP-1 [42–44]. Further

49

References

studies should uncover the early signaling steps, downstream of the lectin site, leading to intracellular calcium mobilization and regulation of gene expression and determine whether RROP-1 might present some anti-tumoral and/or anti-fibrotic properties.

[20] E. Duverger, N. Frison, A.C. Roche, M. Monsigny, Carbohydrate–lectin interactions assessed by surface plasmon resonance, Biochimie 85 (2003) 167–179. [21] A. Varki, Selectin ligands, Proc. Natl. Acad. Sci. U. S. A. 91 (1994) 7390–7397. [22] N. Isnard, G. Peterszegi, A.M. Robert, L. Robert, Regulation of elastase-type endopeptidase activity, MMP-2 and MMP-9 expression and activation in human dermal fibroblasts by fucose and a fucose-rich polysaccharide, Biomed. Pharmac- other. 56 (2002) 258–264. [23] G. Peterszegi, I. Fodil-Bourahla, A.M. Robert, L. Robert, Pharmacological properties of fucose. Applications in age-related modifications of connective tissues, Biomed. Pharmacother. 57 (2003) 240–245. [24] A. Hinek, D.S. Wrenn, R.P. Mecham, S.H. Barondes, The elastin receptor: a galactoside-binding protein, Science 239 (1988) 1539–1541. [25] A. Hinek, J. Boyle, M. Rabinovitch, Vascular smooth muscle cell detachment from elastin and migration through elastic laminae is promoted by chondroitin sulfate- induced ―shedding‖ of the 67-kDa cell surface elastin binding protein, Exp. Cell Res. 203 (1992) 344–353. [26] D. Hong, K.A. Barbee, D.G. Buerk, D. Jaron, Heterogeneous cytoplasmic calcium response in microvascular endothelial cells, Conf. Proc. IEEE Eng. Med. Biol. Soc. 7 (2005) 7493–7496. [27] A. Ben-Ze'ev, Cytoarchitecture and signal transduction, Crit. Rev. Eukaryot. Gene Expr. 2 (1992) 265–281. [28] P. Nicotera, B. Zhivotovsky, S. Orrenius, Nuclear calcium transport and the role of calcium in apoptosis, Cell Calcium 16 (1994) 279–288. [29] S. Matsumoto-Yoshitomi, J. Habashita, C. Nomura, K. Kuroshima, T. Kurokawa, Autocrine transformation by fibroblast growth factor 9 (FGF-9) and its possible participation in human oncogenesis, Int. J. Cancer 71 (1997) 442–450. [30] T. Todo, T. Kondo, T. Kirino, A. Asai, E.F. Adams, S. Nakamura, K. Ikeda, T. Kurokawa, Expression and growth stimulatory effect of fibroblast growth factor 9 in human brain tumors, Neurosurgery 43 (1998) 337–346. [31] J.P. Zwerner, W.A. May, PDGF-C is an EWS/FLI induced transforming growth factor in Ewing family tumors, Oncogene 20 (2001) 626–633. [32] N.A. Lokker, C.M. Sullivan, S.J. Hollenbach, M.A. Israel, N.A. Giese, Platelet-derived growth factor (PDGF) autocrine signaling regulates survival and mitogenic pathways in glioblastoma cells: evidence that the novel PDGF-C and PDGF-D ligands may play a role in the development of brain tumors, Cancer Res. 62 (2002) 3729–3735. [33] G. Giannelli, E. Fransvea, F. Marinosci, C. Bergamini, S. Colucci, O. Schiraldi, S. Antonaci, Transforming growth factor-beta1 triggers hepatocellular carcinoma invasiveness via alpha3beta1 integrin, Am. J. Pathol. 161 (2002) 183–193. [34] M. Granerus, W. Engstrom, Effects of fibroblast growth factor 8, fibroblast growth factor 9 and keratinocyte growth factor on multiplication and locomotion in human teratocarcinoma cells in vitro, Anticancer Res. 23 (2003) 1313–1316. [35] J.W. Martens, A.M. Sieuwerts, J. Bolt-deVries, P.T. Bosma, S.J. Swiggers, J.G. Klijn, J.A. Foekens, Aging of stromal-derived human breast fibroblasts might contribute to breast cancer progression, Thromb. Haemost. 89 (2003) 393–404. [36] I.A. Steele, R.J. Edmondson, H.Y. Leung, B.R. Davies, Ligands to FGF receptor 2-IIIb induce proliferation, motility, protection from cell death and cytoskeletal rearrangements in epithelial ovarian cancer cell lines, Growth Factors 24 (2006) 45–53. [37] S.J. Kim, H. Uehara, T. Karashima, M. McCarty, N. Shih, I.J. Fidler, Expression of interleukin-8 correlates with angiogenesis, tumorigenicity, and metastasis of human prostate cancer cells implanted orthotopically in nude mice, Neoplasia 3 (2001) 33–42. [38] T. Nagatoro, K. Fujita, E. Murata, M. Akita, Angiogenesis and fibroblast growth factors (FGFs) in a three-dimensional collagen gel culture, Okajimas Folia Anat. Jpn. 80 (2003) 7–14. [39] M. Tunyogi-Csapo, T. Koreny, C. Vermes, J.O. Galante, J.J. Jacobs, T.T. Glant, Role of fibroblasts and fibroblast-derived growth factors in periprosthetic angiogenesis, J. Orthop. Res. 25 (2007) 1378–1388. [40] J.H. Martens, J. Kzhyshkowska, M. Falkowski-Hansen, K. Schledzewski, A. Gratchev, U. Mansmann, C. Schmuttermaier, E. Dippel, W. Koenen, F. Riedel, M. Sankala, K. Tryggvason, L. Kobzik, G. Moldenhauer, B. Arnold, S. Goerdt, Differential expression of a gene signature for scavenger/lectin receptors by endothelial cells and macrophages in human lymph node sinuses, the primary sites of regional metastasis, J. Pathol. 208 (2006) 574–589. [41] M. Rusnati, M. Presta, Fibroblast growth factors/fibroblast growth factor receptors as targets for the development of anti-angiogenesis strategies, Curr. Pharm. Des. 13 (2007) 2025–2044. [42] L. Yu, W.A. Border, Y. Huang, N.A. Noble, TGF-beta isoforms in renal fibrogenesis, Kidney Int. 64 (2003) 844–856. [43] J.S. Campbell, S.D. Hughes, D.G. Gilbertson, T.E. Palmer, M.S. Holdren, A.C. Haran, M.M. Odell, R.L. Bauer, H.P. Ren, H.S. Haugen, M.M. Yeh, N. Fausto, Platelet-derived growth factor C induces liver fibrosis, steatosis, and hepatocellular carcinoma, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 3389–3394. [44] J.H. Distler, A. Jungel, L.C. Huber, U. Schulze-Horsel, J. Zwerina, R.E. Gay, B.A. Michel, T. Hauser, G. Schett, S. Gay, O. Distler, Imatinib mesylate reduces production of extracellular matrix and prevents development of experimental dermal fibrosis, Arthritis Rheum. 56 (2007) 311–322.

Acknowledgements

This work was supported by Institut DERM, Paris. We thank Professor Gilles Renard, Head of the Ophtalmology Department at the Hôtel Dieu Hospital, Paris 5 University, for his generous hospitality.

[1] E.M. Culav, C.H. Clark, M.J. Merrilees, Connective tissues: matrix composition and its relevance to physical therapy, Phys. Ther. 79 (1999) 308–319. [2] F.T. Bosman, I. Stamenkovic, Functional structure and composition of the extracellular matrix, J. Pathol. 200 (2003) 423–428. [3] K.O. Lloyd, The chemistry and immunochemistry of blood group A, B, H, and Lewis antigens: past, present and future, Glycoconj. J. 17 (2000) 531–541. [4] H.J. Gabius, S. Gabius (Eds.), Lectins and Glycobiology, Springer Verlag, Berlin, 1993. [5] M. Monsigny, C. Kieda, A.C. Roche, Membrane glycoproteins glycolipids and membrane lectins as recognition signals in normal and malignant cells, Biol. Cell 47 (1986) 95–110. [6] K. Drickamer, M.E. Taylor, Biology of animal lectins, Ann. Rev. Cell Biol. 9 (1993) 237–264. [7] J.P. Zanetta, S. Kuchler, S. Lehmann, A. Badache, S. Maschke, D. Thomas, P. Dufourcq, G. Vincendon, Glycoproteins and lectins in cell adhesion and cell recognition processes, Histochem. J. 24 (1992) 791–804. [8] M. Caron, D. Bladier, R. Joubert, Soluble galactoside-binding vertebrate lectins: a protein family with common properties, Int. J. Biochem. 22 (1990) 1379–1385. [9] K. Smetana Jr., B. Dvorankova, M. Chovanec, J. Boucek, J. Klima, J. Motlik, M. Lensch, H. Kaltner, S. Andre, H.J. Gabius, Nuclear presence of adhesion-/growth-regulatory galectins in normal/malignant cells of squamous epithelial origin, Histochem. Cell Biol. 125 (2006) 171–182. [10] L. Robert, W. Hornebeck, Interaction between elastin fibers and cells, in: J. Labat- Robert, R. Timpl (Eds.), Structural Glycoproteins in Cell–Matrix Interaction. Frontiers of Matrix Biology, vol 11, Karger Basel, 1986, pp. 58–77. [11] D. Cerdan, C. Grillon, M. Monsigny, G. Redziniak, C. Kieda, Human keratinocyte membrane lectins: characterization and modulation of their expression by cytokines, Biol. Cell 73 (1991) 35–42. [12] E. Andres, J. Molinari, G. Peterszegi, B. Mariko, E. Ruszova, V. Velebny, G. Faury, L. Robert, Pharmacological properties of rhamnose-rich polysaccharides, potential interest in age-dependent alterations of connectives tissues, Pathol. Biol. (Paris) 54 (2006) 420–425. [13] T. Mossmann, Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays, J. Immunol. Meth. 65 (1983) 55–63. [14] E.A. Jaffe, R.L. Nachman, C.G. Becker, C.R. Minick, Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immuno- logic criteria, J. Clin. Invest. 52 (1973) 2745–2756. [15] G. Faury, Y. Usson, M. Robert-Nicoud, L. Robert, J. Verdetti, Nuclear and cytoplasmic free calcium level changes induced by elastin peptides in human endothelial cells, Proc. Natl. Acad. Sci. U. S. A. 95 (1998) 2967–2972. [16] K.M. Yamada, K. Olden, Fibronectins—adhesive glycoproteins of cell surface and blood, Nature 275 (1978) 179–184. [17] R.W. Tsien, P. Hess, E.W. McCleskey, R.L. Rosenberg, Calcium channels: mechanisms of selectivity, permeation, and block, Annu. Rev. Biophys. Biophys. Chem. 16 (1987) 265–290. [18] S.B. Russell, J.D. Russell, J.S. Trupin, Hydrocortisone induction of system A amino acid transport in human fibroblasts from normal dermis and keloid, J. Biol. Chem. 259 (1984) 11464–11469. [19] J. Hirabayashi, T. Hashidate, Y. Arata, N. Nishi, T. Nakamura, M. Hirashima, T. Urashima, T. Oka, M. Futai, W.E. Muller, F. Yagi, K. Kasai, Oligosaccharide specificity of galectins: a search by frontal affinity chromatography, Biochim. Biophys. Acta 1572 (2002) 232–254.

50

7.4. Receptors and aging: Structural selectivity of the rhamnose-

receptor on fibroblasts as shown by Ca2+-mobilization and gene-

expression profiles

Archives Gerontology and Geriatrics

Receptors and aging: Structural selectivity of the rhamnose-receptor on

fibroblasts as shown by Ca2+-mobilization and gene-expression profiles

aLaboratoire Physiopathologie Vasculaire: Interactions Cellulaires, Signalisation et Vieillissement, iRTSV-APV, CEA-Grenoble, 17 rue des Martyrs, F-38054 Grenoble Cedex 9, France b Universite Joseph Fourier, F-38041 Grenoble, France cINSERM, U882, F-38054 Grenoble, France dˆpital Ho Dieu, Universite Paris 5, Paris, Franceˆtel´Laboratoire de Recherche Ophtalmologique, Ho eInstitut DERM, 42 rue Monge, F-75005 Paris, France fLaboratory of Dermal Applications, R&D Dept, Contipro Group, Dolni Dobruc, 401, 561 02 Dolni Dobruc, Czech Republic

G. Faury a,b,c, J. Molinari d,e, E. Ruszova f, B. Mariko a,b,c, S. Raveaud a,b,c, P. Huber a,b,c, V. Velebny f, A.M. Robert d,e, L. Robert d,e,*

A R T I C L E I N F O A B S T R A C T

Article history: Received 22 January 2010 Received in revised form 12 May 2010 Accepted 14 May 2010

Keywords: Receptors and aging Rhamnose-rich oligo- and polysaccharides Lectins Signal transmission Intracellular calcium Gene expression

Qualitative and quantitative modifications of receptors were shown to play a key role in cell and tissue aging. We recently described the properties of a rhamnose-recognizing receptor on fibroblasts involved in the mediation of age-dependent functions of these cells. Using Ca2+-mobilization and DNA- microarrays we could show in the presence of rhamnose-rich oligo- and polysaccharides (RROPs) Ca2+- mobilization and changes in gene regulation. Here, we compared the effects of several RROPs, differing in their carbohydrate sequence and molecular weights, in normal human dermal fibroblasts (NHDFs). It appeared that different structural features were required for maximal effects on Ca2+-mobilization and gene-expression profiles. Maximal effect on Ca2+ influx and intracellular free calcium regulation was exhibited by RROP-1, a 50 kDa average molecular weight polysaccharide, and RROP-3, a 5 kDa average molecular weight oligosaccharide with a different carbohydrate sequence. Maximal effect on gene- expression profiles was obtained with RROP-3. These results suggest the possibility of several different transmission pathways from the rhamnose-receptor to intracellular targets, differentially affecting these two intracellular functions, with potential consequences on aging. Although of only relative specificity, this receptor site exhibits a high affinity for rhamnose, absent from vertebrate glycoconjugates. The rhamnose-receptor might well represent an evolutionary conserved conformation of a prokaryote lectin. ß 2010 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Receptors were shown to undergo age-dependent modifications such as loss of receptor density in cells and tissues and uncoupling of the receptor from the normal (‗‗young‘‘) transmission pathway. Since the original propositions of Paul Ehrlich attributing to receptors a key role in the mediation of cell interactions with ‗‗foreign‘‘ molecules (Robert et al., 2009a, for review), a great deal of progress was accomplished for the characterization of a large variety of receptors and their transmission pathways (Bradshaw and Dennis, 2004, for review). Among these receptors, lectin-type molecules play a special role by mediating the interaction of cells with oligo- and polysaccharides. Oligo- and polysaccharides are attached to numerous biological macromolecules (glycoconjugates) found in virtually all kinds of

⁎ Corresponding author. Laboratoire Physiopathologies vasculaires: interactions cellulaires, signalisation et vieillissement, INSERM U882, iRTSV-APV, CEA-Grenoble, 17 rue des Martyrs, 38054 Grenoble Cedex 9, France. Tel.: +33 4 38 78 23 48; fax: +33 4 38 78 49 64. E-mail address: [email protected] (G. Faury).

living species: prokaryotes and eukaryotes, animals and plants. Carbohydrates can attach to proteins to form glycoproteins and are also present as polysaccharides linked to proteins in proteoglycans (Kennedy and White, 1983; Zanetta et al., 1992; Varki et al., 1999, for reviews). Lectins were characterized as specifically interacting with carbohydrates, such as in red cell aggregation or leukocyte- endothelial cell interaction during the early steps of extravasation. Until recently, interactions involving carbohydrates and lectins were supposed to result only in the formation of molecular complexes, for instance producing cell aggregation, without further consequence on cell function. However, recent results showed, in animal cells, that carbohydrate binding to lectins was able to trigger message transduction to the cell interior, suggesting that, at least, some lectins could function as ‗‗true‘‘ receptors (Monsigny et al., 1986; Hinek et al., 1988; Faury et al., 2008). Using synthetic neoglycoproteins (monosaccharides covalently bound to BSA) the team of Monsigny–Kieda demonstrated a lectin site on human skin keratinocytes exhibiting a relative specificity for a-L-rhamnose (Cerdan et al., 1991), findings confirmed in our laboratory (Andres et al., 2006). These results were somewhat

A R T I C L E I N F O A B S T R A C T

Keywords: Receptors and aging Rhamnose-rich oligo- and polysaccharides Lectins Signal transmission Intracellular calcium Gene expression

51

Table 1 The list of the RROPs.

2.3. Measurements of transmembrane Ca2+ fluxes

Fluorescence microscopic experiments were performed using the calcium-sensitive dye FLUO3 (excitation: 488 nm; emission: >510 nm), following a procedure deriving from a previously described method already in use in the laboratory (Faury et al., 1998, 2008). Adhering NHDFs were washed twice with PSS. PSS (0.75 ml) followed by addition of 0.25 ml of a 2% bovine serum albumin and 0.08% Pluronic F127 solution in PSS containing 5 ml of a 0.5 mM FLUO3/AM solution in DMSO (final concentrations: 2.5 mM FLUO3/AM, <1% DMSO). NHDFs were incubated in this solution for 45 min at 37 8C in the dark. The cells were then washed twice and bathed in 1.5 ml PSS 10 min before and during the measurements (22 8C). Fluorescence microscopic recordings were performed on a Cell-R system (Olympus, Rungis, France). In each experiment, fluorescence was recorded for about 15 min (1 image every 5 s). F/Fo is the ratio of current fluorescence to initial

2. Materials and methods

2.1. Cells

unexpected since rhamnose has been demonstrated to be present in bacteria and plants, but was never detected in vertebrate glycoconjugates. Some other monosaccharides bound to BSA such as glucose, fucose and N-acetyl-glucosamine could compete with rhamnose for apparently the same lectin site, but with a much lower affinity (Monsigny et al., 1986; Figs. 1 and 2 in Andres et al., 2006). Ligands of this lectin, i.e. Rhamnose-rich oligo- and polysaccharides (RROPs), were shown to regulate several age-dependent cell functions in fibroblasts and related cells, such as collagen and elastase-type endopeptidase productions, as well as protection against free radical-mediated degradation of hyaluronan and advanced glycation end-products-induced accumulation of senescence-associated b-galactosidase (SA-b-gal) and cell death (Andres et al., 2006; Ravelojaona et al., 2006, 2008a,b; Robert et al., 2009b). Further investigation of the related signaling mechanisms in NHDFs has shown that the rhamnose-recognizing lectin site exhibited the characteristics of a ‗‗true‘‘ receptor by transmitting ‗‗messages‘‘ to the cell interior. In particular, a $50 kDa RROP-1 induced increases in Ca2+-fluxes and intracellular [Ca2+]i concentration, as well as modulation of gene-expression profiles (Faury et al., 2008). This finding was further supported by the fact that rhamnose alone triggered changes in NHDF [Ca2+]i similar to those induced by RROP-1, while the other components of RROP-1 (galactose and glucuronic acid) induced only slight modifications of [Ca2+]i (Faury et al., 2008). These results also confirmed the relative specificity of the rhamnose-recognizing lectin, suggested by the binding of glucose and fucose by this lectin, although with a lower affinity (Monsigny et al., 1986). In the present study we compared the action of several different RROPs using again the above-cited two tests, i.e., Ca2+-mobilization and modification of gene-expression profiles, in order to further explore the specificity of the rhamnose-recognizing receptor.

52 52 6 15 5

Designation Approximate average mol. wt.a (kDa)

Average no. of penta- or hexasaccharide subunits per average mol. wt.

The cells used were NHDF obtained from Cambrex (Emerain- ville, France) and cultured as previously described (Andres et al., 2006).

2.2. RROPs

50 45 5 14.5 5

Single-channel currents were recorded from cell-attached patches on NHDF membranes (holding potential = +20 mV) using previously described patch-clamp procedures and instrumentation (Faury et al., 1998, 2008; Garnier-Raveaud et al., 2001). All the experiments were performed at room temperature (22 8C). The cells were washed twice with a physiological saline solution (PSS) containing: 125 mM NaCl; 5.6 mM KCl; 2.4 mM CaCl2; 1.2 mM MgCl2; 10 mM HEPES (pH 7.4), then bathed in PSS during the recording. The patch pipette was filled with 90 mM Ba(CH3COO)2 and 10 mM HEPES (pH 7.4). Barium was used instead of calcium in order to improve signal resolution, because calcium channels are generally more permeable to this ion, and barium is known to inhibit potassium currents (Tsien et al., 1987). RROPs were added to the bath during the recordings. NHDF resting membrane potential (mV) was assumed to be À109 Æ 3 mV, according to the literature (Russell et al., 1984). The recorded transmembrane currents were analyzed and integrated using the software Biopatch (Biologic, Claix, France). Experiments were at least triplicated in each case.

2.4. Intracellular free calcium concentration in NHDFs

The RROPs used were obtained from Solabia Bio-Europe (Pantin, France). Two polysaccharides were obtained from two different Klebsiella strains: RROP-1 from a non-pathogenic Klebsiella pneumoniae strain and RROP-2 from a Klebsiella planticola strain. The first, RROP-1, of about 50 kDa average molecular weight (according to the producer, Solabia) has the following repeating structure:

The second, RROP-2, with an average molecular weight of about 45 kDa (according to the producer, Solabia) with the following composition of the repeating units:

Note: RROP-1, 4 and 5 are derived from Klebsiella pneumoniae, and RROP-2 and 3 from Klebsiella planticola. aAs communicated by the producer, Solabia Bio-Europe.

acid (bGlcA). Besides rhamnose, RROP-2 contains a-L-glucuronic acid (aGlcA), b-D-glucose (bGlc), and no galactose. Both polysaccharides contain branching a- or b-rhamnose (aRha or bRha) side chains. The repeating unit of RROP-1 is a hexasaccharide (molecular weight $957 Da), and that of RROP-2, a pentasaccharide (molecular weight $867 Da). RROP-2 was degraded by mild acid (HCl) hydrolysis to a 5 kDa oligosaccharide (RROP-3). RROP-4 (14.5 kDa) and RROP-5 (5 kDa) were obtained from RROP-1 by acid hydrolysis (Andres et al., 2006). All these substances were obtained as sterile, lyophylized white powders and were kept at À20 8C. Solutions were prepared in sterile phosphate-buffered saline (PBS) just before use and kept at +4 8C for a short time. Table 1 lists these five RROPs with their essential characteristics.

These two polysaccharides have slightly different compositions and structures. RROP-1 (commercial name Rhamnosoft1) contains a-L-rhamnose (aRha), b-L-galactose (bGal) and b-L-glucuronic

RROP-1 RROP-2 RROP-3 RROP-4 RROP-5

52

fluorescence and is assumed to be indicative of [Ca2+]i. Experi- ments were at least triplicated in each case.

2.5. Cell-viability and/or proliferation test

The methyl-thiazol tetrazolium (MTT)-test was used according to the method described by Mossmann (1983). This test gives an indication of mitochondrial function and, indirectly, of cell number.

2.6. Microarray tests

First, we used MTT viability method to investigate the effect of RROP-1 and RROP-3 on NHDF cell viability. Four concentrations were tested (10, 50, 200 and 500 μg/ml), each with two different durations of treatment (24 and 48 h). The following concentrations were found to be suitable for microarray experiments on NHDF: RROP-2 at 10 and 100 mg/ml and RROP-3 at 10 mg/ml.

2.6.1. The mRNA extraction NHDF were seeded on 6-well plates (7 cm2) in the concentration of minimally 1x 10

6 cells/well for microarray experiments.

After incubation with RROPs, supernatants were removed and cells subjected to RNA isolation with Trisol reagent (Invitrogen, Carlsbad, CA, USA), chloroform (Lachema, Brno, Czech Republic) and 2-propanol (Sigma–Aldrich, St. Louis, MO, USA). The sediment was resuspended in 20 ml DEPC water. RNA purity was assessed by the 260/280 nm absorbance ratio and the amount of RNA was calculated using the conversion factor of 40 mg RNA per optical density unit at 260 nm. Samples were stored at À80 8C.

2.6.2. The cDNA synthesis RNA (10 mg) was first incubated with Random primers (5 ml, Invitrogen) for 10 min at 65 8C. First, cDNA strand synthesis was carried out by Revert Aid H Minus M-MULV Reverse Transcriptase (Fermentas, Berlington, Ontario, CA, USA) and labeled with Biotin- 11-dUTP (Fermentas). The cDNA purification was performed with YM-30 columns (Millipore, Billerica, MA, USA).

2.6.3. cDNA hybridization in array tubes (AT) One set of CPN-7 AT was purchased from Clondiag1 Chip Technologies GmbH (Jena, Germany) with spotted 50-mer probes that we selected regarding the CPN spol. S.R.O. needs (cat. no.: 2010.10, specification according to array layout CPN-7, full list can be found at http://www.clondiag.com, http://www.cpn-contipro.cz). AT-s covered markers, which include genes for cell-cycle, cell migration, cell adhesion, homeostasis, cell contact with the extracellular matrix (ECM), production of ECM components, angiogenesis, and inflammation. The chips on the bottom of microtubes contained 127 different probes spotted in triplicates (Clondiag). Oligonucleotide probes were designed using Oligoarray Designer 2.2 and input sequences were chosen in accordance with their GenBank reference accession numbers. The list of genes tested (127) can be obtained from [email protected] with the description of the CPN-7 layouts.

2.6.4. Array hybridization The AT were placed in a thermomixer (Eppendorf, Hamburg, Germany), washed three times with 500 μl of hybridization buffer (Clondiag Chip Technologies), each time at 45 8C for 5 min and 550 rpm. Usually, 10 μl of cDNA was diluted in 100 μl of hybridization buffer (Clondiag Chip Technologies), denatured at 94 8C for 5 min, cooled down on ice for 2 min and added to the AT. The hybridization was carried out at 45 8C for 3 h by gentle shaking in a hybridization oven. After hybridization the AT were washed with 500 μl of 2Â sodium chloride–sodium citrate (SSC)–0.2%

sodium dodecyl sulfate (SDS) at 30 8C for 5 min in the thermomixer at 550 rpm, followed by 500 μl of 2x SSC at 20 8C for 5 min at 550 rpm, and finally with 0.2x SSC at 20 8C for 5 min at 550 rpm. A blocking step was carried out with 2% (w/v) milk powder in 100 μl of 6x SSPE-0.005% Triton at 30 8C for 15 min. Poly-horseradish peroxidase–streptavidin conjugate (100 pg/ml) was added in 100 ml of 6x SSPE-0.005% Triton and incubated at 30 8C for 25 min at 550 rpm. The previous washing procedure was repeated after blocking with 2x SSC-0.01% Triton, 2x SSC, and 0.2x SSC. AT were kept in the last buffer at 20 8C until further processing. Visualization of hybridization was achieved by adding 100 μl of peroxidase substrate True Blue (KPL Lab.) to the AT, and detection of signals was done in the AT-reader (ATR01, Clondiag Chip Technologies). Signals were recorded at 25 8C for 30–40 min. The data were than analyzed with the IconoClust version 2.2 software (Clondiag Chip Technologies).

2.7. ELISA tests

Some of the modifications of gene expressions were controlled by direct determination of the gene-product by ELISA, using anti- matrix metalloproteinase-9 (MMP-9) (Bender MedSystems, Vienna, Austria) and anti-platelet-derived growth factor = PDGF- AA (R&D Systems, Minneapolis, MN, USA) antibodies.

2.8. Statistics

All microarray experiments were performed in three to six parallel cultures. Normalization was performed by comparing housekeeping gene transcripts in treated and control (untreated) samples: the intensities of b-actin (ACTB), Histone H3, GAPD were chosen as equalizing markers. A Mann–Whitney non-parametric U-test was performed between the two series of values and significance was set at p < 0.05 (Statgraphics, v.5). Figures were created in Statistica (v.5, Statsoft, USA).

3. Results

3.1. Effects of RROPs on gene regulation

Table 2 summarizes the results of the microarray experiments carried out using RROP-1, RROP-2 and RROP-3. Only significantly modified gene expressions are shown. Fig. 1 shows the scattergram of the gene-expression modifications produced by RROP-2 and RROP-3. Regarding RROP-1, down-regulations of 13 genes were observed. Several of these genes are involved in cell–matrix interactions, such as integrin or collagen genes, others in cell proliferation (growth factors) or immune-inflammatory reactions. Table 2 shows also the results on gene regulations obtained with RROP-2 and RROP-3, derived from another strain of Klebsiella than RROP-1. With these polysaccharides, the results are quite different than those obtained with RROP-1. Although most genes whose expression was modified by RROP-2 were down-regulated, 2 of the 7 genes affected showed, however, a significant upregulation, i.e., nidogen-2 and MMP-9. The results were completely different with the RROP-3 oligosaccharide. Out of 14 significantly affected genes, only 5 were down-regulated while all the others were significantly upregulated, including MMP-9, growth factors and adhesion molecules. RROP-3 proved to be the most active in gene-expression regulations. The two other oligosaccharides mentioned in Table 1, i.e., RROP- 4 and RROP-5, were not tested for gene regulations. Some of the gene-expression modifications were further controlled by direct ELISA determination of the gene-products.

53

Table 2 Effects of RROPs on gene expression in NHDF.

Design I I I I I II II II II II II II III III III III III III III III III III III III III IV IV

Gene Collagen, type 1, alpha 1 Collagen, type 4, alpha 1 Nidogen Nidogen-2 Laminin beta 1 PECAM Integrin alpha 4 Integrin, beta 1 Stabilin-2 Thrombospondin-1 Cadherin 5, type 2 (VE) EPHB-4 FGF-1 FGF-9 FGF-R2 FGF-R4 TGF-beta 3 TGFR-1 PDGF-A PDGF-C IL-1 IL-8 LIF VEGF-R2 IGF-1 SERPINE-1 MMP-9

RROP-1

-21.1 -38 -21.1 -31 -22.7

-21 -20.1

-40 -50

-30 -20.3 -23

+46 -20.5

+12 +45

-23 +6

-21.5 -30 -2

-20 +50

+30

+30 +7

-22

-23 +66 -48

+50 --40 - 40

+33 -35 -32 +42

RROP-2 RROP-3

Notes: Variation (in percent) of the gene-expression levels, compared to the corresponding gene-expression level in control/untreated cells, measured by microarray experiments in NHDFs treated with 10 μg/ml RROPs. À, down- regulation; +, up-regulation. Designations: I, Extracellular matrix components; II, mediators of cell–matrix or cell–cell interactions; III, growth factors, growth factor receptors and cytokines; IV, matrix modifying enzymes. Abbreviations used in this table: R, receptor; SERPINE-1, plasminogen activator inhibitor 1 (PAI-1); FGF, fibroblast growth factor; TGF, transforming growth factor; IL, interleukin. Fig. 1. Scatter plots of the results from microarray experiments. Gene-expression

levels after cell stimulation by 10 mg/ml RROP-2 (A) or RROP-3 (B) are represented by comparing the expression of the same genes in control/non-stimulated cells.

MMP-9 gene-expression up-regulation by RROP-2 (mRNA: +50%, Table 2) was confirmed by the ELISA test which also showed an increase in the release of the MMP-9 protein by 110%. Similarly, the increased expression of the PDGF-AA gene by RROP-3 (mRNA: +45%, Table 2) could also be confirmed by the ELISA test which showed that the PDGF-AA protein product was increased by 640%. The stimulation of NHDF by RROP-2 or RROP-3 led to regulation of genes coding for a limited number of protein families: ECM- related, mediating cell–cell interaction and growth factor-related proteins. The affected ECM-related proteins can be classified into three categories: some ECM proteins such as laminin, nidogen, collagen, an MMP-9 and ECM binding receptors, integrin chains (b2, a4). The expression of some proteins mediating cell–cell interactions (ephrin B4, EPHB4, platelet/endothelial cell adhesion molecule 1 = PECAM, and cadherin) as well as growth factors (PDGF, insulin-like growth factor = IGF), growth factor receptors (TGFb-R, FGF-R, vascular endothelial growth factor = VEGF-R) are also modified, as well as the genes of leukemia inhibitory factor (LIF), IL-1 and TSP-1.

3.2. Effects of RROPs on membrane Ca2+-channel activity and [Ca2+]i concentration

Previous results demonstrated that RROP-1 triggers in NHDFs a strong activation of membrane calcium channels and subsequent Ca2+ influx, as well as an increase in intracellular free calcium concentration [Ca2+]i (Andres et al., 2006). Here, RROP-2 and RROP- 3 were also shown to activate membrane calcium channels and increase Ca2+ influx in NHDFs, although RROP-3-induced channel activations were of larger amplitude than those produced by RROP-2 (Fig. 2). Addition of RROP-2 and RROP-3 also triggered

substantial increases in [Ca2+]i in NHDFs, although the patterns of response to RROP-2 and RROP-3 were different. While RROP-2 induced slow and long lasting increases in [Ca2+]i, RROP-3 mainly triggered the appearance of transient [Ca2+]i peaks, although some slow elevations were also noted. It should be noted, however, that only some cells, not all, responded to RROP-2 and RROP-3 by calcium level elevations, leading to fluorescence increases up to +50–80% at 1–10 mg/ml RROP concentration (Fig. 3).

4. Discussion

The here-presented results confirm and extend previous demonstrations showing that carbohydrate-recognizing lectin sites can act as receptors and transduce ‗‗messages‘‘ (Monsigny et al., 1986; Zanetta et al., 1992; Drickamer and Taylor, 1993; Smetana et al., 2006). Our previous experiments carried out on the interaction of RROPs with several different human cell types, skin derived fibroblasts, human umbilical venous endothelial cells (HUVECs) and corneal keratocytes revealed several reactions mediated by the rhamnose-receptor (Andres et al., 2006; Ravelojaonaet al., 2006, 2008a,b; Faury et al., 2008; Robert et al., 2009b). The most important findings were the increase in cell proliferation, collagen biosynthesis and protection of hyaluronan against free radical-mediated degradation (Andres et al., 2006). RROPs were shown to protect fibroblasts against Maillard- products (advanced glycation end-products, AGE)-induced cell death (Ravelojaona et al., 2006). By comparing fibroblasts and keratocytes, it was shown that RROPs increased much more the collagen biosynthesis in keratocytes than in fibroblasts (Fig. 2 in Ravelojaona et al., 2008a). This effect was also shown to be passage-dependent. This stimulation was also dependent on the

54

Fig. 2. Membrane Ca-channel activation induced by RROP-2 and RROP-3 in NHDFs using the patch-clamp technique. (A) Representative patch-clamp recording after addition of the solvent (PBS). (B) Representative patch-clamp recording after addition of 10 mg/ml RROP-2 (left) and corresponding current integration over time (right). (C) Representative patch-clamp recording after addition of 1.3 mg/ml RROP-3 (left) and corresponding current integration over time (right). The experiments were at least triplicated in each group.

structure and composition of the RROPs used. These rhamnose- rich substances could also efficiently protect fibroblasts against the accumulation of SA-b-gal and even more efficiently against the AGE-induced accumulation of SA-b-gal (Ravelojaona et al., 2008b). Previously, we showed that fibroblasts increase their elastase- type endopeptidase production with increasing passage number. This effect could also be efficiently counteracted by RROPs (Robert et al., 2009b). Using Ca2+-signal analysis, by patch-clamp and fluorescence techniques, and gene arrays, we could confirm that the RROP-mediated signal transduction did modify intracellular Ca2+- and gene-expression profiles (Faury et al., 2008, and the present results). We thereby confirmed and extended our previous results (Andres et al., 2006), showing that RROPs react with a relatively specific lectin site on human skin fibroblasts, HUVECs and keratinocytes, resulting in modifications of gene expressions, Ca2+-fluxes and [Ca2+]i. Although of a relative specificity (reacting to a lesser extent with glucose and fucose: about 60% of rhamnose-

affinity), the existence of a lectin/receptor transmitting a signal to the cell interior in response to rhamnose is still intriguing because of the absence of rhamnose in vertebrate glycans and glycoconjugates. A further interesting finding reported in the above described experiments is the difference between two different RROPs, as far as calcium mobilization and action on gene expression are concerned. RROP-1, a 50 kDa polysaccharide with a repeating sequence, and RROP-3, a 5 kDa oligosaccharide with a different sequence, where galactose is replaced by glucose, were the most active on calcium mobilization (Andres et al., 2006; Faury et al., 2008, and present results). RROP-3 proved to be the most active on gene mobilization, as shown in Table 2. As shown also on the scattergrams (Fig. 1), different genes are differently affected by these two rhamnose-rich preparations. The rhamnose-rich polysaccharide RROP-2 and the rhamnose- rich oligosaccharide RROP-3 induced intracellular calcium elevations in NHDFs, although these responses presented different

55

Fig. 3. Evolution of FLUO3 fluorescence in NHDFs in response to RROP-2 and RROP-3 added at 120 s after the beginning of the recording: control–solvent PBS (A), 10 μg/ ml RROP-2 (B), 1.3 μg/ml RROP-3 (C). Representative tracings are presented. In each graph, one tracing corresponds to the response of one single cell (n > 30 cells in each case).

patterns. While response to RROP-2 consisted essentially in slow/ long lasting increases in [Ca2+]i, the intracellular calcium increases triggered by RROP-3 gave mixed fast/transient [Ca2+]i peaks in some cells, and slow/long lasting [Ca2+]i increases in other cells. For both preparations, RROP-2 or RROP-3, not all cells responded to the stimulations. The NHDF response to RROP-3 resembled the response induced by the RROP-1, and partly to the response induced by free rhamnose (Faury et al., 2008). The discrepancy between the cell responses to the different rhamnose-containing molecules, including the response to the monosaccharide rhamnose, suggests the existence of ligand structure-determined differences in the binding to the rhamnose-recognizing lectin as well as in the cell reactions triggered by these preparations. This can be the result of different affinities of the lectin for these ligands producing different cell responses, in particular different types of calcium responses. This hypothesis is supported by the different levels of calcium channel activation triggered by the different rhamnose-containing molecules, as seen in the patch-clamp experiments performed here and in previous studies (Faury et al., 2008). The RROP-induced calcium channel activation also shows that the calcium transients induced by RROPs are mediated, at least in part, by extracellular calcium influx. While RROP-1 and

RROP-3 induced strong activations of membrane calcium channels, RROP-2 triggered a slighter activity of these channels. This suggests that the calcium signaling events are modulated by the affinity of the rhamnose-preparation to the rhamnose-recognizing lectin. This is consistent with the presence of intracellular calcium transient peaks in NDHFs stimulated by RROP-1 and RROP-3, and the absence of these peaks in cells stimulated by RROP-2. To explain this discrepancy, it could be hypothesized that the particular structure of RROP-2 renders more difficult for the rhamnose side chains to bind to the rhamnose-recognizing lectin. Other constituents present in RROP-2, i.e., glucose, galactose or glucuronic acid, could also bind to the rhamnose-recognizing lectin, although with a lower affinity, and produce the slighter cell response as observed. When RROP-2 is cleaved by acid hydrolysis into RROP-3 fragments, the rhamnose side chains could become more available to react with the lectin, leading to stronger responses, as observed with RROP-3. This could be triggered by the carbohydrates present in RROP-1 other than rhamnose, i.e., D- galactose and D-glucuronic acid. This is in agreement with studies by the Monsigny–Kieda team showing that the a-L-rhamnose- recognizing lectin reacts also with glucose and fucose, although with a lower specificity (Cerdan et al., 1991; Andres et al., 2006). A relative specificity is frequently observed for lectins: lactose, galactose and N-acetyl-galactosamine containing chondroitine sulphates were shown to bind to the elastin–laminin receptor (Hinek et al., 1988, 1992), multiple glycans can bind to galectins (Hirabayashi et al., 2002) and multiple glycans and oligosaccharides can bind to selectins (Varki, 1994; Varki et al., 1999). When calcium signaling is compared to the gene-expression regulations induced by RROP-2 and RROP-3, a certain consistency is found with previous results regarding the effects of RROP-1 on NHDFs (Andres et al., 2006), although we noticed that RROP-2 (the test molecule which induced the lowest cell responses) regulates also a lower number of genes than do RROP-1 and RROP-3. Cell stimulation by RROP-1, RROP-2 and RROP-3 led to modulation of the expression of genes coding for a limited number of protein families, such as ECM-related, cell–cell interaction mediating and growth factor-related ones. However, the similarity between the effects on gene expression of RROP-1, RROP-2 and RROP-3 is limited to the implicated genes, with only slight differences in Ca2+-regulations, while the degree of expression of the affected genes is different. RROP-1 induced only down-regulations of gene expressions, whereas RROP-3 mainly up-regulated the expression of several genes of the same families. While RROP-1 and RROP-2 down-regulated the expressions of genes involved in angiogenesis, cancer cell transformation, proliferation or invasiveness (growth factors, growth factor receptors, cytokines, adhesion molecules), RROP-2 up- or down-regulated these genes, and RROP-3 mostly activated this type of genes. Also, RROP-1 decreased the expressions of several ECM protein coding genes and growth factors stimulating ECM production, resembling the effects of an anti- fibrotic drug, Imatinib (Distler et al., 2007), while RROP-2 and RROP-3 increased and sometimes decreased the expression of these gene families. These findings might be interpreted as the result of different transmission pathways from the same rhamnose-recognizing receptor to cell interior. Alternatively, it could be hypothesized that several, although slightly different, rhamnose-recognizing receptors coexist in the same vertebrate cell type. It is interesting to note, however, that several reactions mediated by the rhamnose-receptor protect cells and tissues against age-related harmful reactions.

56

Acknowledgements

Supported by Institut DERM, Paris. We thankfully acknowledge the hospitality and helpful discussions with Prof. Gilles Renard, Head of the Department of Ophthalmology Hospital.

Andres, E., Molinari, J., Peterszegi, G., Mariko, B., Ruszova, E., Velebny, V., Faury, G., Robert, L., 2006. Pharmacological properties of rhamnose-rich polysaccharides, potential interest in age-dependent alterations of connective tissues. Pathol. Biol. 54, 420–425. Bradshaw, R.A., Dennis, E.A. (Eds.), 2004. Handbook of Cell Signaling, vol. 1. Academic Press/Elsevier, Amsterdam. Cerdan, D., Grillon, C., Monsigny, M., Redziniak, G., Kieda, C., 1991. Human keratinocyte membrane lectins: characterisation and modulation of their expression by cytokines. Biol. Cell 73, 35–42. Distler, J.H., Jungel, A., Huber, L.C., Schulze-Horsel, U., Zwerina, J., Gay, R.E., Michel, B.A., Hauser, T., Schett, G., Gay, S., Distler, O., 2007. Imatinib mesylate reduces production of extracellular matrix and prevents development of experimental dermal fibrosis. Arthritis Rheum. 56, 311–322. Drickamer, K., Taylor, M.E., 1993. Biology of animal lectins. Ann. Rev. Cell Biol. 9, 237–264. Faury, G., Usson, Y., Robert-Nicoud, M., Robert, L., Verdetti, J., 1998. Nuclear and cytoplasmic free calcium level changes induced by elastin peptides in human endothelial cells. Proc. Natl. Acad. Sci. U.S.A. 95, 2967–2972. Faury, G., Ruszova, E., Molinari, J., Mariko, B., Raveaud, S., Velebny, V., Robert, L., 2008. The alpha-1-rhamnose recognising lectin site of human dermal fibroblasts functions as a signal transducer. Modulation of Ca++ fluxes and gene expression. Biochim. Biophys. Acta 1780, 1388–1394. Garnier-Raveaud, S., Usson, Y., Cand, F., Robert-Nicoud, M., Verdetti, J., Faury, G., 2001. Identification of membrane calcium channels essential for cytoplasmic and nuclear calcium elevations induced by vascular endothelial growth factor in human endothelial cells. Growth Factors 19, 35–48. Hinek, A., Wrenn, D.S., Mecham, R.P., Barondes, S.H., 1988. The elastin receptor: a galactoside-binding protein. Science 239, 1539–1541. Hinek, A., Boyle, J., Rabinovitch, M., 1992. Vascular smooth muscle cell detachment from elastin and migration through elastic laminae is promoted by chondroitin- sulfate induced ‗‗shedding‘‘ of the 67-kDa cell surface elastin binding protein. Exp. Cell Res. 203, 344–353. Hirabayashi, J., Hashidate, T., Arata, Y., Nishi, N., Nakamura, T., Hirashima, M., Urashima, T., Oka, T., Futai, M., Muller, W.E., Yagi, F., Kasai, K., 2002. Oligosac-

charide specificity of galectins: a search by frontal affinity chromatography. Biochim. Biophys. Acta 1572, 232–254. Kennedy, J.F., White, C.A. (Eds.), 1983. Bioactive Carbohydrates in Chemistry, Biochemistry and Biology. Ellis Horwood Ltd./J. Wiley & Sons, New York, USA. Monsigny, M., Kieda, C., Roche, A.C., 1986. Membrane glycoproteins, glycolipids and membrane lectins as recognition signals in normal and malignant cells. Biol. Cell 47, 95–110. Mossmann, T., 1983. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J. Immunol. Methods 65, 55–63. Ravelojaona, V., Molinari, J., Robert, L., 2006. Protection by rhamnose-rich polysaccharides against the cytotoxicity of Maillard reaction products. Biomed. Pharmacother. 60, 359–362. Ravelojaona, V., Robert, A.M., Robert, L., Renard, G., 2008a. Collagen biosynthesis in cell culture: comparison of corneal keratocytes and skin fibroblasts. Effect of rhamnose-rich oligo- and polysaccharides. Pathol. Biol. 56, 66–69. Ravelojaona, V., Robert, A.M., Robert, L., 2008b. Expression of senescence-associated b-galactosidase (SA-b-Gal) by human skin fibroblasts, effect of advanced glycation end-products and fucose or rhamnose-rich polysaccharides. Arch. Gerontol. Geriatr. 48, 151–154. Robert, L., Labat-Robert, J., Robert, A.M., 2009a. Receptors and aging: dedicated to the memory of Paul Ehrlich for the 100th anniversary of his Nobel Prize. Arch. Gerontol. Geriatr., doi:10.1016/j.archger.2009.11.013. Robert, L., Molinari, J., Ravelojaona, V., Andres, E., Robert, A.M., 2009b. Age- and passage-dependent upregulation of fibroblast elastase-type endopeptidase activity. Role of advanced glycation end-products, inhibition by fucose and rhamnose-rich oligosaccharides. Arch. Gerontol. Geriatr., doi:10.1016/j. archger.2009.05.006. Russell, S.B., Russell, J.D., Turpin, J.S., 1984. Hydrocortisone induction of system A amino acid transport in human fibroblasts from normal dermis and keloid. J. Biol. Chem. 259, 11464–11469. Smetana Jr., K., Dvorankova, B., Chovanec, M., Boucek, J., Klima, J., Motlik, J., Lensch, M., Kaltner, H., Andre, S., Gabius, H.J., 2006. Nuclear presence of adhesion-/ growth-regulatory galectins in normal/malignant cells of squamous epithelial origin. Histochem. Cell Biol. 125, 171–182. Tsien, R.W., Hess, P., McCleskey, E.W., Rosenberg, R.L., 1987. Calcium channels: mechanisms of selectivity, permeation, and block. Annu. Rev. Biophys. Biophys. Chem. 16, 265–290. Varki, A., 1994. Selectin ligands. Proc. Natl. Acad. Sci. U.S.A. 91, 7390–7397. Varki, A., Cummings, R., Esko, J., Freeze, H., Hart, G., Marth, J. (Eds.), 1999. Essentials of Glycobiology. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, USA. Zanetta, J.P., Kuchler, S., Lehmann, S., Badache, A., Maschke, S., Thomas, D., Dufourcq, P., Vincendon, G., 1992. Glycoproteins and lectins in cell adhesion and cell recognition processes. Histochem. J. 24, 791–804.

References

57

7.5. Pharmacological properties of rhamnose-rich polysaccharides,

potential interest in age-dependent alterations of connectives tissues

Les oligo- et polysaccharides riches en rhamnose (RROPs) ont été testés pour leurs propriétés pharmacologiques avec des fibroblastes de peau humaine cultivés en série. Des activités biologiques comme la stimulation de la prolifération, la diminution de l’activité élastasique et la protection de l’acide hyaluronique contre sa dégradation face aux radicaux libres ont été misent en évidence. Ces effets biologiques semblent être induits via une lectine reconnaissant l’α-L-rhamnose spécifiquement et qui agirait comme un récepteur en transmettant des signaux à la cellule. L’augmentation rapide du calcium libre intracellulaire après ajout du RROP-1, ainsi que des données préliminaires obtenues par micropuces à ADN semblent confirmer cette hypothèse. © 2006 Published by Elsevier SAS.

ARTICLE IN PRESS

Pathologie Biologie ■■ (2006) ■■■

Pharmacological properties of rhamnose-rich polysaccharides, potential interest in age-dependent alterations of connectives tissues

Propriétés pharmacologiques des polysaccharides riches en rhamnose

et intérêt potentiel dans les altérations des tissus conjonctifs relatives à l‘âge

a Laboratoire de recherches ophtalmologiques, Hôtel-Dieu, université Paris-V, 1, place Parvis-Notre-Dame, 75181 Paris cedex 04, France bLaboratoire de développement et vieillissement de l‘endothélium, Inserm EMI 0219/université Joseph-Fourier/CEA, DRDC, CEA, 17, rue des Martyrs, 38054 Grenoble cedex 09, France cLaboratory of Dermal Applications, R and D Department, Contipro Group, Dolni Dobrouc 401, 56102 Dolni Dobrouc, Czech Republic

Received 20 June 2006; accepted 4 July 2006

Abstract

Résumé

Keywords: Rhamnose; Polysaccharide; Intracellular calcium; Proliferation; Elastase-type activity; ROS; Scavenging; Lectin; Microarray

1. Introduction

Abbreviation: RROP-s, rhamnose-rich oligo- and polysaccharides.. author. E-mail addresses: [email protected] (E. Andrès), [email protected] (L. Robert). 1New address from September 2006: Laboratoire de recherches NATURA, 55, avenue Victor-Hugo, 75016 Paris, France.

α-L-rhamnose-rich polysaccharides, obtained from Kleb- siella strains as well as derived oligosaccharides were tested

Rhamnose-rich oligo- and polysaccharides (RROPs) were tested for their potential pharmacological properties using human skin fibroblasts in serial cultures. The substances tested were shown to stimulate cell proliferation, decrease elastase-type activity, stimulate collagen biosynthesis, and protect hyaluronan against free radical mediated degradation. These reactions appear to be triggered by the mediation of a specific α-L-rhamnose recognizing lectin-site acting as a receptor, transmitting signals to the cell-interior. The rapid increase of intracellular free calciumafter addition of RROP-1 and preliminary data using micro arrays appear also to confirm this contention. © 2006 Published by Elsevier SAS.

for their pharmacological properties. Both parent polysaccharides as well as the derived oligosaccharides are rich in α- L-rhamnose (50–60%). Rhamnose-rich oligo- and polysaccharides (RROP)-1 (Rhamnosoft®) was previously shown to pos- sess anti-inflammatory properties triggered by cytokines, IL-1α and TNFα released by activated keratinocytes [1,2]. This same polysaccharide reduced also the adhesion of bacteria to the skin and protected against UV-induced erythema [1]. The anti- inflammatory effect was obtained by the inhibition of adhesion of polynuclear leukocytes to keratinocytes, as well as by the inhibition of phospholipase A2 and the release of prostaglan-

E. Andrèsa,1, J. Molinaria,1, G. Péterszegia, B. Marikob, E. Ruszovac, V. Velebnyc G. Fauryb, L. Roberta,*

58

It has an average molecular weight of 50 kDa with about a 50% of α-L-rhamnose content, the two other constituents are D-galactose and 3-O-acetyl-D-glucuronic acid.

The second polysaccharide RROP-2 (commercial name: BEC-291 polysaccharide) has the following composition of its repeating subunits:

Its average molecular weight is 45 kDa. The oligosaccharides are obtained from these polysaccharides by acid hydrolysis, RROP-3, a 5 kDa oligosaccharide

Fig. 1. Demonstration of specific recognition of α-L-rhamnose containing neoglycoproteins by a lectin-site on keratinocytes. Experiments of the Kieda– Monsigny team (reproduced from Ref. [1] with permission). (Rham-P: Rhamnose-Protein; Glc-P: Glucose-Protein; Fuc-P: Fucose-Protein; N-Ac-Glc-P: N-Acetyl-Glucose-Protein).

from RROP-2, RROP-4 (14.5 kDa) and RROP-5 (5 kDa) from RROP-1. These preparations were obtained as sterile, lyophi- lized powders and dissolved in sterile PBS at concentrations from 0.1 to 0.5 mg/ml for further use. The polysaccharide preparations were stored at –20 °C.

Fig. 2. Fixation of increasing concentrations of neoglycoproteins containing different monosaccharides on keratinocytes. The α-L-rhamnose-containing neoglycoprotein reaches the highest saturation density (except for fucose at the highest concentration used). Experiments of the Kieda–Monsigny team (reproduced from Ref. [1] with permission).

dins as PGE3 [1]. RROP-1 was also shown to decrease the adhesion of bacteria to skin surface [1]. The action of RROP- s was claimed to be mediated by an α-L-rhamnose recognizing lectin-site on human skin keratinocytes by the team of Cerdan et al. [3]. As shown on Fig. 1, the keratinocyte lectin-site recognized preferentially the α-L-rhamnose containing neoglycoproteins as compared to similar glycoconjugates containing other monosaccharides such as glucose, fucose or N- acetyl-glucosamine [3]. This interaction was shown by the Kieda–Monsigny team to be saturable (Fig. 2) and relatively specific. There appear to be however some interferences from the α-L-fucose rich glycoproteins at least at high concentrations. Our experiments aimed to extend the above-cited findings by testing the effect of several rhamnose-rich polysaccharides and derived oligosaccharides on human skin fibroblasts using tests related to cell-functions known to decline with age. Such functions are cell proliferation, collagen biosynthesis, and free radical scavenging activity [4]. Another function tested is the increased production of elastase-type endo-

peptidases with chronological age [5] and cell proliferation [6]. These results will be described because of their potential interest for the treatment of age-related functional decline and related pathologies.

2. Material and methods

2.1. Polysaccharides tested

All L-rhamnose-rich oligo- and polysaccharide preparations (RROPs) were obtained from Solabia (Pantin, France). RROP-1 is obtained from Klebsiella pneumoniae strains. The other polysaccharide RROP-2 is obtained from Klebsiella planticola. RROP-1 (commercial name: Rhamnosoft®) has the following structure of its repeating subunits:

59

2.2. Cell culture

The human skin fibroblasts were obtained from Cambrex (NHDF, from a 39 years old, female donor). Cells were cultured in Dulbecco‘s modified Eagles culture medium (DMEM- glutamax, Gibco, Invitrogen) supplemented with 10% (v/v) foetal calf serum (FCS, Gibco), antibiotics: penicillin (100 U/ ml; Gibco) and streptomycin (100 μg/ml; Gibco), and an anti- fongic (0.25 μg/ml Amphotericin B, Gibco). This was designated as the complete culture medium. Cells were incubated in a temperature-controlled, humidified incubator with 5% CO2 at 37 °C. Cells were grown in the complete culture medium, in 75 cm2 surface ventilated culture flasks (Nunc) and subcultured by trypsinization (0.05% trypsin, Gibco). Culture medium was changed every 2–3 days. Cells were used between the 3rd and 19th passages for our experiments, seeded in complete culture medium in 12, 24 or 96-well culture plates, at a density of 5 × 104 cells per ml.

2.3. Cell proliferation

The MTT test was used according to the method described by Mosmann [7]. This test gives an indication of mitochondrial function and indirectly on cell number. The MTT test consists of measuring the mitochondrial succinate dehydrogenase activity of cells. This enzyme, by cutting the tetrazolium cycle, makes the yellow MTT (3- (4,5-dimethyl-2-thiazolyl)-2,5-diphenyl 2H-tetrazolium bro-mide), turn into blue Formazan crystals. The optical densities obtained are directly proportional to the number of living cells. Therefore, the viable cells will produce an intense blue color and the dead ones a very light stain. Electronic cell counting (Coulter Corp., Hialech, FL, USA) was performed after trypsinization of cell layer and in the culture medium to estimate cell detachment.

3. Collagen biosynthesis assays

3.2. Determination of elastase-type endopeptidase activity

The cells are incubated for 48 h with the plant extracts and then, their elastase activity was tested by using a chromophoric synthetic substrate, the N-succinyl-trialanyl-paranitroanilide (N-Suc) [10]. The hydrolysis of this substrate by elastase-type endopeptidases produces a chromogenic compound, absorbing at 410 nm. The cells were washed and the extracts recovered with a Triton ×100 1/100 treatment (500 μl/10 min at 37 °C), then extracts were kept at –40 °C. Reaction was initiated with 100 μl extract and 100 μl substrate and DO at 410 nm was measured for 24 hours on a 96-well plate reader Metertech Σ960.

3.3. Free radical scavenging assays

3.1. Intracellular free calcium measurements

All data were obtained and analyzed with the Olympus CellR microscopic system. After culture, the cells were washed with a Tyrode‘s buffer (NaCl 125 mM, KCl 5.6 mM, CaCl2 2.4 mM, MgCl2 1.2 mM, Hepes 10 mM, Glucose 11 mM, pH 7.4) and loaded in this same buffer in the presence of 0.5% BSA, 0.02% Pluronic F127 and 2.5 μM Fluo 3/AM at 37 °C/5% CO2 for 45 min, this loading protocol is derived and adapted from an original protocol using others cells [9]. This indicator detects Ca2+ concentration via its fluorescent spectral changes upon Ca2+ binding. The Fluo-3 is used in its membrane-permeable AM ester form. The AM ester of the

Evaluation of collagen biosynthesis was obtained by the Sir- ius red staining method [8]. Cells were cultured in presence of RROP-s for 3 days and collagen accumulation was measured by spectrophotometry.

Free radical scavenging was determined with a previously described viscosimetric procedure [11]. Kinetics of viscosity changes of a 1 mg/ml hyaluronan solution in a rotating visc- ometer, alone and after addition of a highly diluted (1/1000) ascorbate–EDTA–FeCl2 mixture, the Udenfriend reagent, were recorded. The Udenfriend reagent liberates essentially OH● radicals, which degrade rapidly hyaluronan and produce a rapid fall of its viscosity. In presence of adequate scavenging agents, this fall of viscosity is slowed down. The rotating visc- ometer used is from Fungilab S.A, model « Visco Basic Plus » giving directly viscosity readings in centipoises.

3.4. Assays of lectin-mediated intracellular effects

Microarrays were also used to characterize the modification of gene expression [12], only preliminary data will be given.

3.5. Statistics

All experiments were carried out with four to six parallels; significance was calculated with the Mann and Whitney distribution-free U-Test.

Fluo-3 is membrane-permeable and thus can be loaded into cells by simple incubation of the cell in a buffer containing the AM ester and Pluronic F-127, a mild non-ionic detergent that can facilitate the loading of the AM ester. The AM ester itself does not bind Ca2+. However, once it has entered the cells, it is readily hydrolyzed by intracellular esterases into the non-AM ester form, thus becoming responsive to Ca2+. After loading, the cells were washed again with Tyrode‘s buffer and then introduced in the CellR microscopic system [9]. This system enables to record, as a function of time, images of the fluorescence released by the Fluo-3 probe when calcium concentration increases in the cell cytoplasm , for example, after the addition of a well known transductional activator such as the agonist of a receptor. After acquisition, the image sequence was analyzed with a specific software that gives the kinetics of fluorescence intensity in each cell. The image acquisition rate was one image every 8 s.

60

4. Results

4.1. Effect of RROPs on the biological activities of fibroblasts

Table 1 shows the results obtained with the rhamnose-rich preparations on the properties of fibroblasts such as proliferation, collagen biosynthesis, elastase-type endopeptidases activity, and free radical scavenging. The + signs for cell proliferation signify stimulation of proliferation as shown by the MTT procedure and cell counting. All these results were significant (P < 0.001). All the tested rhamnose-rich preparations stimulated cell proliferation between + 40% for the 5 kDa RROP-5 preparation and up to + 80% for the 5 kDa RROP-3 preparation. These differences between their activity suggests that besides the rhamnose end-groups, others details of the polysaccharides structure might also play a role. Collagen biosynthesis was stimulated only by some of the preparations, strongest by the RROP-4 and somewhat less by RROP-2 and RROP-3 poly- and oligosaccharides. Curiously, elastase-type endopeptidases activity was down regulated only by the RROP-3 oligosaccharide, none of the other preparations showed significant activity. Fig. 3 shows a typical experiment showing the free radical quenching activity of some of the preparations tested. In presence of the diluted Udenfriend reagent, there is a rapid fall in viscosity because of the liberation of OH● radicals [11]. In

Table 1 Biological activity of RROPs on parameters linked to cell and tissue aging

RROPs tested Cell proliferation (%) +60 +80 +80 +60 +40

Collagen biosynthesis (%) No effect +20 +20 +40 NT

Elastase activity (%)

NT NT –60 NT NT

Free radical scavenging (%) +40 +40 +20 +40 +20

presence of the rhamnose-rich preparations, the loss of viscosity was significantly reduced by 20–40%. Except free radical scavenging, all other activities shown on Table 1 are supposed to be mediated by intracellular mechanisms triggered by the activation of the α-L-rhamnose recognizing lectin-site detected by the team of Cerdan et al. [3]. This lectin-site functions apparently as a pharmacologically active receptor. In order to confirm this proposition, additional experiments were carried out.

4.2. Effect on intracellular calcium

Fig. 4 show the response of NHDFs to stimulation by RROP-1. Fig. 4a presents the evolution over time of the fluorescence signal in the absence of RROPs, after addition of physiological solution alone for 120s. In this case, an illumination- dependent bleaching of the probe induced a slight decline of the fluorescence intensity over time, and no significant devia- tion from the basic signal can be seen. By contrast to the control experiments (Fig. 4a), addition of RROP-1 (Fig. 4b) induced a rapid increase in [Ca2+]i in most cells over a period of 10 min and more. One can notice that the intensity of the reaction is variable from one cell to the next: some cells hardly reacted, while others exhibited a clear increase in [Ca2+]i (up to 50–100%) following addition of the rhamnose-rich-polysaccharides RROP-1. Addition of RROP-2 and RROP-3 induced similar effect (data not shown).

4.3. Effect of rhamnose-rich preparations on gene regulation

The results to be mentioned have to be considered as preliminary, experiments are still actively performed in order to control and substantiate the most crucial results [12]. It is however quite logical that the above presented biological activities are hard to understand without a signal transmission from the

RROP-1 RROP-2 RROP-3 RROP-4 RROP-5

NT: not tested.

Fig. 3. Decrease of viscosity of a hyaluronan solution in presence of a free-radical generating reagent (see Section 2) and protection by RROP-s.

61

Table 2 Microarrays experiments showing the modification of gene expression profiles by RROPs used at 10 or 100 μg/ml. Upregulation is indicated by +, down regulation by –, and intensity is expressed as % modification as compared to controls [12]

Gene

Nidogen2 MMP-9 LIF LAMß PECAM TSP-1 EphrinB PDGFA INTα4 CAD FGFR-4 VEGFR

RROP-1 (10 ug/ ml)

RROP-1 (100 ug/ml)

+150 +150 –69 –65 –65 –65 –48

RROP-3

+134 +134

–65 –50 +145 +142 +167 +146 +133

–68 –66 –64 –48

MMP: matrix metallo-proteinase; LIF: lymphocyte inducing factor; LAMβ: laminin β; PECAM: platelet endothelial cell adhesion molecule; TSP: thrombospondin; PDGF: platelet derived growth factor; INTα: integrin α; FGFR: fibroblast grow factor receptor; VEGFR: vascular epithelium grow factor receptor.

Fig. 4. A. Variation of fluorescence in individual fibroblasts as a sign of oscillation of intracellular free calcium (control). B. Variation of fluorescence in individual fibroblasts as a sign of oscillation of intracellular free calcium after addition of 10 μg/ml of RROP-1.

Fig. 5. Scatter plot obtained by microarrays experiments [12] showing the modulation of expression of certain genes by RROP-3.

agonist-activated lectin-site to the cell interior. A ―crosstalk‖ had to be assumed between the rhamnose-recognizing lectin- site and the regulatory pathways controlling cell proliferation, collagen biosynthesis and for RROP-3, MMP expression and regulation [13]. Some of the preliminary results obtained by

the microarrays experiments are shown on Fig. 5 and Table 2. Fig. 5 shows the scattergram for gene-up or down regulations by the RROP-3 oligosaccharide. Table 2 shows the most important modifications of genes expression by two concentrations of RROP-1 and by RROP-3. These preliminary results tend to confirm the above contention by showing the modification of expression of several key-genes involved in the regulation of cell proliferation and of some cytokines involved in the inflammatory pathway. If confirmed by the ongoing control experiments, a more detailed description of the intracellular transmission pathways of the α-L-rhamnose recognizing receptor will become available.

5. Discussion

The above-described results strongly suggest that the α- L-rhamnose recognizing lectin-site detected first on keratinocytes [3] is present on human skin fibroblasts also. They also show that the addition of the RROPs to fibroblast cultures induced several modifications of cell behavior. The most interesting effect, as far as the pharmacological activities of these preparations are concerned, are their effect on cell proliferation and collagen biosynthesis (Table 1). All five preparations were active in these tests. Only RROP-3, a 5 kDa derivative from the RROP-2 polysaccharide showed activity on the down- regulation of elastase-type endopeptidases. Interestingly, this preparation turned out to be the most active in the microarrays experiments also (Fig. 5). All the preparations tested protected hyaluronan against free radical mediated degradation. This effect might be attributed to the preferential attraction of free radicals to the rhamnose-rich polysaccharides, protecting thus hyaluronan from degradation. It is however also possible that the rhamnose-rich preparations increase the scavenging potential of cells against free radical damage, this hypothesis deserves to be further explored. Further experiments are clearly

62

Supported by Institut Derm, Paris and Natura Ltd., São Paulo, Brazil. We thankfully acknowledge the hospitality of Professor Gilles Renard, head of the department of ophthalmology of Hôtel Dieu, as well as fruitful discussions during these experiments, also with Eduardo Luppi, Daniel Gonzaga and Jean-Luc Gesztesi from the R&D Department of Natura Ltd.

necessary in order to precise the details of the intracellular transmission pathways triggered by the action of the rhamnose-rich preparations on fibroblasts. As shown by the chemical composition of the tested preparations, mentioned in Section 2, there is about one end-standing rhamnose residue for every hexasaccharides subunit (see page 5). For a 5 kDa oligosaccharide, this would represent about 5 rhamnose residues per molecule, enough to trigger for instance a coordinated assembly of several lectin sites on the cell membrane, a necessary step for the activation demonstrated for many cell membrane receptors [14]. The most intriguing aspect of these results is the absence of rhamnose in vertebrate glycoconjugates. The presence of this methyl pentose was demonstrated only in prokaryotes and mainly in plants. It is a component of the plant cell wall pectic polysaccharides rhamnogalacturonan I and rhamnogalacturonan II and is also present in diverse secondary metabolites including anthocyanins, flavonoids and triterpenoids. Further studies conducted in our laboratory should confirm and extend the above results, substantiating the claim those rhamnose-rich preparations might well be considered as of pharmacological interest in the age-related decline of tissue and cell-functions and related pathologies.

Acknowledgements

References

[1] Rhamnosoft® polysaccharide atténuateur de l‘inflammation. Solabia Report No. AC046. Pantin, France: Solabia BioEurope; 2001. [2] Barker JN, Mitra RS, Griffiths CE, Dixit VM, Nickoloff BJ. Keratino- cytes as initiators of inflammation. Lancet 1991;337(8735):211–4. [3] Cerdan D, Grillon C, Monsigny M, Redziniak G, Kieda C. Human keratinocyte membrane lectins: characterization and modulation of their expression by cytokines. Biol Cell 1991;73:35–42. [4] Robert L. Le vieillissement. Paris: CNRS-Belin; 1994. [5] Robert L, Hornbeck W. Elastin and elastases I–II. Boka Raton, USA: CRC Press; 1989. [6] Macieira-Coelho A. Biology of normal proliferating cells in vitro relevance for in vivo aging. Basel: Karger; 1998. [7] Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Meth 1983;65(1–2):55–63. [8] Tullberg-Reinert H, Jundt G. In situ measurement of collagen synthesis by human bone cells with a Sirius Red-based colorimetric microassay: effects of transforming growth factor β2 and ascorbic acid-2-phosphate. Histochem Cell Biol 1999;112:271–6. [9] Faury G, Usson Y, Robert-Nicoud M, Robert L, Verdetti J. Nuclear and cytoplasmic free calcium level changes induced by elastin peptides in human endothelial cells. Proc Natl Acad Sci USA 1998;95:2967–72. [10] Bieth J, Spiess B, Wermuth CG. The synthesis and analytical use of a highly sensitive and convenient substrate of elastase. Biochem Med 1974;11(4):350–7. [11] Deguine V, Menasche M, Fraisse L, Ferrari P, Pouliquen Y, Robert L. Determination of extracellular matrix degradation by free radicals using viscosity measurement of hyaluronan. Clin Chim Acta 1997;262(1–2): 147–52. [12] Molinari J, Andrès E, Faury G, Mariko B, Ruszova E, Velebny V, et al. Pharmacological properties of rhamnose-rich polysaccharides. Confirma- tion of microarrays experiments. In preparation. 2006. [13] Ravelojaona V, Molinari J, Robert L. Protection by rhamnose-rich polysaccharides against the cytotoxicity of Maillard reaction products, Bio- medicine and pharmacotherapy. In print. 2006. [14] Bradshaw R, Dennis E. Handbook of cell signalling. San Diego: Aca- demic Press; 2004.

63

8. Discussion

8.1. UV irradiation crucially influences fibroblast response to

GM

It was shown that GM acts very well as a photoprotective agent (Article n.1). It

protects against the harmful effect of UV irradiation both in vitro as well in vivo. GM

supresses the UV-induced decrease of keratinocyte viability. It could be speculated that

GM protects cutaneous cells from UV-induced apoptosis by the suppression of caspase

activatin and subsequent hindering of DNA fragmentation. The observed GM-mediated

reduction of thymine dimers also suggests a potentiation of the DNA damage repair.

The faster elimination of thymine dimers in the presence of GM provides other evidence

for reparative effects evoked by GM.

In agreement with in vitro data, the application of GM on skin significantly

decreased UV-induced erythema formation.

Further, as yet undiscussed results, include the fact that GM application did not

induce a significant release of IL-1α and PGE2 from untreated keratinocytes, however

minor the immunostimulatory effect in the non-irradiated HaCaT sample is detectable

(Article n. 1, IL-1α , fig. 6A). On the contrary, under the UVB-conditioning, the

microarray analyses showed that GM treatment significantly suppressed UVB induced

NF-κB gene expression in keratinocytes and NF-κB-driven gene expression of several

pro-inflammatory markers IL-1α, IL-8, and iNOS. Thus, the inclusion of different

conditioning as UVB treatment indicates how different signal transduction elicited by

GM can be asserted.

64

8.2. Regulation of gene expression in fibroblasts by AGE

In the next part of the theses the effects of two different AGEs (lysozym-

glucose and BSA-glucose) on global gene expression were monitored in human dermal

fibroblasts (Article n.2). It was shown that AGE modified expression of more than

dozens of genes particularly MMPs and serpin-expressions were upregulated and the

collagen-chain coding genes, as well as the cadherin- and fibronectin genes were

downregulated. Observed differences among tested AGE-preparations tested suggest the

possibility of different receptor-mediated transmission pathways. The results intimate

that fibroblasts try to compensate last two mentioned detrimental effects of AGE

formation on separate, transcriptional level.

The advanced glycation end products (AGEs) are one of the important

mechanism of post-translational tissue modifications with aging. AGEs and/or

hypeglycemic conditions are shown to up-regulate receptor for AGE (RAGE) in various

cell types. RAGE with AGEs elicits oxidative stress generation, thus participating in

activation of MAPK (ERK1 and p38), pro-inflammatory response (NF-κB, MCP-1),

angiogenesis via VEGF overexpression and TGFb1 (CTGF)-driven fibrogenic reactions

subscribed by accumulation of collagen type IV, laminin B1, fibronectin (FN) (Wang,

Liu et al. ; Yamagishi and Matsui) .

Other mechanisms as impairing of degradation of various types of AGEs by

even up-regulated degradative enzymes may be proposed, finally leading to basement

membrane (BMZ) thickening (Mott, Khalifah et al. 1997; Tamarat, Silvestre et al.

2003). And AGE formation, structural modifications of various AGE substrate

modifications (matrix proteins), is also responsible for abrogation of focal adhesion

formation, therefore cell attachment and spreading (McDonald, Coleman et al. 2009).

65

8.3. Effect of different MW of RROPs on fibroblast response

The earliest results strongly suggest that α- L-rhamnose recognizing lectin-site

detected first on keratino-cytes is present on human skin fibroblasts also (Article n. 5).

The RROP-1 induced modifications of the gene expression profile in fibroblasts showed

that this polysaccharide triggered a down-regulation of the expression of several growth

factors, adhesion molecules and extracellular matrix proteins involved in pro-tumoral

activity and/or fibrotic processes (Article n. 3). In contrast, RROP-2 induced the lowest

cell response. It could be speculate that in the case of RROP-2 it is more difficult for the

rhamnose side chains to bind to rhamnose-recognising lecitin side on fibroblasts. When

RROP-2 is cleaved by acid hydrolysis into RROP-3 fragments, the rhamnose side

chains could become more available to react with the lectin, leading to stronger cell

response.

Previously, another polysaccharide isolated from Cordyceps militaris also

showed a monosaccharide composition of rhamnose, mannose, and galactose, as

revealed by a GS chromatogram (Yan, H. et al 2008). This polysaccharide possessed

antioxidant activity and it also works well as an anti-liver fibrosis agent (Yan, H. et al

2008).

Moreover, in addition to published results, we also obtained data regarding N-

cadherin. N-cadherin is found predominantly in neural tissues and fibroblasts where it is

thought to mediate a less stable and more dynamic form of cell-cell adhesion (Mary,

Charrasse et al. 2002). The observed down-regulation in N-cadherin extracellular part

shedding found in samples treated with RROP-3 corresponds to enhancement in the

strength of cell-cell interactions. This is in agreement with another study by Liu, L. et

al., which examined herbal Chinese polysaccharides (SJZD) comprising also galactose

and rhamnose monosaccharide units (Liu, Han et al. 2005). They explored the proposal

that raising a cytosolic free Ca2+

concentration leads to modulation of cell-cell

interaction via a mediating cadherin-catenin–associated interaction (Guo, Lee et al.

2003).

The MMP-9 gene expression was up-regulated by both RROP-2 and RROP-3.

Confirmatory ELISA tests also showed an increase in MMP-9 by RROP-2 and -3 and

PDGF-AA release by RROP-3.

The various findings might be interpreted as the result of different signalling

pathways from the same rhamnose-recognising receptor to the cell interior or,

66

alternatively, that several similar but not identical rhamnose-recognising receptors

coexist on fibroblasts or their different ligation according to MW of polysaccharide

inducing different signalling pathways.

8.4. Effects of rhamnose/fucose rich-polysaccharides on

different transduction pathways

The last part of the study was dedicated to comparing the rhamnose/fucose rich-

polysaccharides and the co-existence of different transduction pathways. FROP (100

ug/ml) increased collagen IV alpha chain by 158% (unpublished data). Collagen IV is

fundamental for the maintenance of the integrity and function of base membranes and

under conditions of increasing mechanical demands it has been shown to inhibit

induced vessel growth (Koskimaki, Karagiannis et al. ; Poschl, Schlotzer-Schrehardt et

al. 2004).

The transcription levels of heparan sulfate proteoglycan, transforming the

growth factor of beta2, and nidogen-2 were discovered to be induced (173, 162, or,

159%) by FROP at a concentration of 10 µg/ml (unpublished data). The base membrane

heparan sulfate proteoglycan or perlecan is able to bind fibronectin and is involved in

cell adhesion, participating in cell-matrix and/or cell-cell contacts. Perlecan, is now

known to be an important component of all base membranes (along with collagen type

IV and laminin), and is thought to play a role in wound healing and angiogenesis

(DeCarlo and Whitelock 2006).

Therefore, while RROP-1 had an effect and decreased the expressions of gene

coding for several ECM and growth factors stimulating ECM production (Article n. 3,

4), FROP regulation of gene expression possesses anti-connective tissue ageing

properties linked with a stimulatory effect on ECM compartment synthesis and renewal

of base membrane components. These present results suggest that several similar but

not identical rhamnose/fucose recognising receptors coexist on fibroblasts, triggering

different cell responses.

67

9. Conclusion

In contrast to extensive literature describing the effects of different polysaccharides

on the modulation of immune cells, information about their impact on non immune cells

are still limited. Therefore, this thesis was focused on cell response of keratinocytes and

dermal fibroblasts to three particular polysaccharides GM, RROPs, and FROPs.

Although all the polysaccharides mentioned above are able to react with cells (due

to common receptors), their response differs significantly depending on the primary

conditions of the experiments.

(i) GM reveals a significant anti-inflammatory and anti-apoptotic effect; as well it

displays photoprotective properties in vivo.

(ii) RROPs differ in their modulatory effect on fibroblast gene regulation in the

dependence on their structure and branching. Meanwhile, RROP-1 carries some

antifibrotic/anti-tumorogenic features; RROP-3 reinforces cell-cell interactions.

(iii) FROP can be linked with its effect on the renewal of the ageing responsible

interface, which includes the lower epidermis, upper dermis and the region of the

dermo-epidermal junction.

Continuous exposure to UV light leads to extensive damage to the dermal

connective tissue, which is a hallmark of photoaged skin. This thesis shows that

polysaccharides, or similarly fucose/rhamnose/mannose-rich oligo/polysaccharides,

perhaps due to their high free radical scavenging activity, might well be considered of

interest in the age-related decline of connective tissue and cell-functions and related

pathologies (photoageing). In light of the present sudy, it is worthwhile to state that

anti-photoageing and photoprotective natural compounds, in particular polysaccharides,

will pave a way for novel products, which find immense use in the pharmaceutical and

cosmetic industry.

68

10. References

Andres, E., J. Molinari, et al. (2006). "Pharmacological properties of rhamnose-rich

polysaccharides, potential interest in age-dependent alterations of connectives

tissues." Pathol Biol (Paris) 54(7): 420-5.

Arck, P. and R. Paus (2006). "From the brain-skin connection: the neuroendocrine-

immune misalliance of stress and itch." Neuroimmunomodulation 13(5-6): 347-

56.

Assefa, Z., A. Van Laethem, et al. (2005). "Ultraviolet radiation-induced apoptosis in

keratinocytes: on the role of cytosolic factors." Biochim Biophys Acta 1755(2):

90-106.

Basile, J. R., A. Eichten, et al. (2003). "NF-kappaB-mediated induction of

p21(Cip1/Waf1) by tumor necrosis factor alpha induces growth arrest and

cytoprotection in normal human keratinocytes." Mol Cancer Res 1(4): 262-70.

Bernstein, E. F., D. B. Brown, et al. (1997). "Evaluation of sunscreens with various sun

protection factors in a new transgenic mouse model of cutaneous photoageing

that measures elastin promoter activation." J Am Acad Dermatol 37(5 Pt 1):

725-9.

Bolling, M. C. and M. F. Jonkman (2009). "Skin and heart: une liaison dangereuse."

Exp Dermatol 18(8): 658-68.

Bragulla, H. H. and D. G. Homberger (2009). "Structure and functions of keratin

proteins in simple, stratified, keratinized and cornified epithelia." J Anat 214(4):

516-59.

Butnaru, C. A. and J. Kanitakis (2002). "Structure of normal human skin." Eur J

Dermatol 12(6): II-IV.

Callaghan, T. M. and K. P. Wilhelm (2008). "A review of ageing and an examination of

clinical methods in the assessment of ageing skin. Part 2: Clinical perspectives

and clinical methods in the evaluation of ageing skin." Int J Cosmet Sci 30(5):

323-32.

Costin, G. E. and V. J. Hearing (2007). "Human skin pigmentation: melanocytes

modulate skin color in response to stress." FASEB J 21(4): 976-94.

DeCarlo, A. A. and J. M. Whitelock (2006). "The role of heparan sulfate and perlecan in

bone-regenerative procedures." J Dent Res 85(2): 122-32.

Deng, H. B., D. P. Cui, et al. (2003). "Inhibiting effects of Achyranthes bidentata

polysaccharide and Lycium barbarum polysaccharide on nonenzyme glycation

in D-galactose induced mouse ageing model." Biomed Environ Sci 16(3): 267-

75.

Deters, A. M., C. Lengsfeld, et al. (2005). "Oligo- and polysaccharides exhibit a

structure-dependent bioactivity on human keratinocytes in vitro." J

Ethnopharmacol 102(3): 391-9.

Drabikova, K., T. Perecko, et al. (2009). "Glucomannan reduces neutrophil free radical

production in vitro and in rats with adjuvant arthritis." Pharmacol Res 59(6):

399-403.

Egles, C., H. A. Huet, et al. "Integrin-blocking antibodies delay keratinocyte re-

epithelialization in a human three-dimensional wound healing model." PLoS

One 5(5): e10528.

El Ghalbzouri, A., E. Lamme, et al. (2002). "Crucial role of fibroblasts in regulating

epidermal morphogenesis." Cell Tissue Res 310(2): 189-99.

69

Faury, G., E. Ruszova, et al. (2008). "The alpha-L-Rhamnose recognizing lectin site of

human dermal fibroblasts functions as a signal transducer: modulation of Ca2+

fluxes and gene expression." Biochim Biophys Acta 1780(12): 1388-94.

Fisher, G. J., T. Quan, et al. (2009). "Collagen fragmentation promotes oxidative stress

and elevates matrix metalloproteinase-1 in fibroblasts in aged human skin." Am

J Pathol 174(1): 101-14.

Frappier BL (2006). "Epithelium." In: Eurell JA, Frappier BL, editors. Dellmann's

Textbook of Veterinary Histology. 6th edn. Ames, Iowa, USA: Blackwell

Publishing; 17–30.

Girish, K. S., K. Kemparaju, et al. (2009). "Hyaluronidase inhibitors: a biological and

therapeutic perspective." Curr Med Chem 16(18): 2261-88.

Green, K. J. and C. L. Simpson (2007). "Desmosomes: new perspectives on a classic." J

Invest Dermatol 127(11): 2499-515.

Greinert, R., O. Boguhn, et al. (2000). "The dose dependence of cyclobutane dimer

induction and repair in UVB-irradiated human keratinocytes." Photochem

Photobiol 72(5): 701-8.

Guo, H. B., I. Lee, et al. (2003). "N-acetylglucosaminyltransferase V expression levels

regulate cadherin-associated homotypic cell-cell adhesion and intracellular

signaling pathways." J Biol Chem 278(52): 52412-24.

Haake, A., G.A. Scott, K.A. Holbrook (2000). "Structure and function of the skin:

overview of the epidermis and dermis", in The Biology of the Skin (Freinkel RK

and Woodley DT eds), pp 19–45, The Parthenon Publishing Group, New York.

Chauhan, P. and M. Shakya (2009). "Modeling signaling pathways leading to wrinkle

formation: identification of the skin ageing target." Indian J Dermatol Venereol

Leprol 75(5): 463-8.

Isnard, N., G. Peterszegi, et al. (2002). "Regulation of elastase-type endopeptidase

activity, MMP-2 and MMP-9 expression and activation in human dermal

fibroblasts by fucose and a fucose-rich polysaccharide." Biomed Pharmacother

56(5): 258-64.

Jenkins, R. H., G. J. Thomas, et al. (2004). "Myofibroblastic differentiation leads to

hyaluronan accumulation through reduced hyaluronan turnover." J Biol Chem

279(40): 41453-60.

Kanitakis, J. (2002). "Anatomy, histology and immunohistochemistry of normal human

skin." Eur J Dermatol 12(4): 390-9; quiz 400-1.

Kim, H. H., S. Cho, et al. (2006). "Photoprotective and anti-skin-ageing effects of

eicosapentaenoic acid in human skin in vivo." J Lipid Res 47(5): 921-30.

Kim, S. Y., S. J. Kim, et al. (2004). "Protective effects of dietary soy isoflavones against

UV-induced skin-ageing in hairless mouse model." J Am Coll Nutr 23(2): 157-

62.

Kogan, G., M. Pajtinka, et al. (2008). "Yeast cell wall polysaccharides as antioxidants

and antimutagens: can they fight cancer?" Neoplasma 55(5): 387-93.

Koskimaki, J. E., E. D. Karagiannis, et al. "Pentastatin-1, a collagen IV derived 20-mer

peptide, suppresses tumor growth in a small cell lung cancer xenograft model."

BMC Cancer 10: 29.

Koster, M. I. (2009). "Making an epidermis." Ann N Y Acad Sci 1170: 7-10.

Kougias, P., D. Wei, et al. (2001). "Normal human fibroblasts express pattern

recognition receptors for fungal (1-->3)-beta-D-glucans." Infect Immun 69(6):

3933-8.

Ksiazek, K., J. Mikula-Pietrasik, et al. (2009). "Senescent peritoneal mesothelial cells

promote ovarian cancer cell adhesion: the role of oxidative stress-induced

fibronectin." Am J Pathol 174(4): 1230-40.

70

Kumazaki, T. and Y. Mitsui (1995). "[Enhanced expression of fibronectin during

cellular ageing of endothelial cells and fibroblasts]." Nippon Ronen Igakkai

Zasshi 32(5): 322-5.

Kwon, A. H., Z. Qiu, et al. (2009). "Effects of medicinal mushroom (Sparassis crispa)

on wound healing in streptozotocin-induced diabetic rats." Am J Surg 197(4):

503-9.

Laga, A. C. and G. F. Murphy (2009). "The translational basis of human cutaneous

photoageing: on models, methods, and meaning." Am J Pathol 174(2): 357-60.

Li, X. T., H. C. Li, et al. "Protective effects on mitochondria and anti-ageing activity of

polysaccharides from cultivated fruiting bodies of Cordyceps militaris." Am J

Chin Med 38(6): 1093-106.

Lin, J. Y. and D. E. Fisher (2007). "Melanocyte biology and skin pigmentation." Nature

445(7130): 843-50.

Lipozencic, J. and S. Ljubojevic (2004). "[Identification of Langerhans cells in

dermatology]." Arh Hig Rada Toksikol 55(2-3): 167-74.

Marks, R. (2004). "The stratum corneum barrier: the final frontier." J Nutr 134(8

Suppl): 2017S-2021S.

Mary, S., S. Charrasse, et al. (2002). "Biogenesis of N-cadherin-dependent cell-cell

contacts in living fibroblasts is a microtubule-dependent kinesin-driven

mechanism." Mol Biol Cell 13(1): 285-301.

McDonald, D. M., G. Coleman, et al. (2009). "Advanced glycation of the Arg-Gly-Asp

(RGD) tripeptide motif modulates retinal microvascular endothelial cell

dysfunction." Mol Vis 15: 1509-20.

Miadokova, E., S. Svidova, et al. (2006). "Diverse biomodulatory effects of

glucomannan from Candida utilis." Toxicol In Vitro 20(5): 649-57.

Moon, H. J., S. R. Lee, et al. (2008). "Fucoidan inhibits UVB-induced MMP-1

expression in human skin fibroblasts." Biol Pharm Bull 31(2): 284-9.

Mott, J. D., R. G. Khalifah, et al. (1997). "Nonenzymatic glycation of type IV collagen

and matrix metalloproteinase susceptibility." Kidney Int 52(5): 1302-12.

Muto, J., K. Kuroda, et al. (2007). "Accumulation of elafin in actinic elastosis of sun-

damaged skin: elafin binds to elastin and prevents elastolytic degradation." J

Invest Dermatol 127(6): 1358-66.

Ouhtit, A. and H. N. Ananthaswamy (2001). "A Model for UV-Induction of Skin

Cancer." J Biomed Biotechnol 1(1): 5-6.

Papanagiotou, V. D. (2009). " Skin aging nad photoaging. " ΔέρμαDerma 4: 57-65.

Paus, R., T. C. Theoharides, et al. (2006). "Neuroimmunoendocrine circuitry of the

'brain-skin connection'." Trends Immunol 27(1): 32-9.

Peterszegi, G., N. Isnard, et al. (2003). "Studies on skin ageing. Preparation and

properties of fucose-rich oligo- and polysaccharides. Effect on fibroblast

proliferation and survival." Biomed Pharmacother 57(5-6): 187-94.

Pinto, M. R., E. Barreto-Bergter, et al. (2008). "Glycoconjugates and polysaccharides of

fungal cell wall and activation of immune system." Brazilian Journal of

Microbiology 39: 195-208.

Poschl, E., U. Schlotzer-Schrehardt, et al. (2004). "Collagen IV is essential for base

membrane stability but dispensable for initiation of its assembly during early

development." Development 131(7): 1619-28.

Presland, R. B. and B. A. Dale (2000). "Epithelial structural proteins of the skin and oral

cavity: function in health and disease." Crit Rev Oral Biol Med 11(4): 383-408.

Presland, R. B. and R. J. Jurevic (2002). "Making sense of the epithelial barrier: what

molecular biology and genetics tell us about the functions of oral mucosal and

epidermal tissues." J Dent Educ 66(4): 564-74.

71

Proksch, E., J. M. Brandner, et al. (2008). "The skin: an indispensable barrier." Exp

Dermatol 17(12): 1063-72.

Ravelojaona, V., J. Molinari, et al. (2006). "Protection by rhamnose-rich

polysaccharides against the cytotoxicity of Maillard reaction products." Biomed

Pharmacother 60(7): 359-62.

Reichrath, J. (2007). "Vitamin D and the skin: an ancient friend, revisited." Exp

Dermatol 16(7): 618-25.

Robert, L., J. Labat-Robert, et al. (2009). "Physiology of skin ageing." Pathol Biol

(Paris) 57(4): 336-41.

Robert, L., J. Molinari, et al. "Age- and passage-dependent upregulation of fibroblast

elastase-type endopeptidase activity. Role of advanced glycation end products,

inhibition by fucose- and rhamnose-rich oligosaccharides." Arch Gerontol

Geriatr 50(3): 327-31.

Rocquet C, Bonte F (2002). "Molecular aspects of skin ageing: recent data." Acta

Dermatoven APA. 3:71-94.

Rodriguez, S., F. Coppede, et al. (2009). "Increased expression of the Hutchinson-

Gilford progeria syndrome truncated lamin A transcript during cell ageing." Eur

J Hum Genet 17(7): 928-37.

Sandjeu, Y. and M. Haftek (2009). "Desmosealin and other components of the

epidermal extracellular matrix." J Physiol Pharmacol 60 Suppl 4: 23-30.

Schalkwijk, J. (2007). "Cross-linking of elafin/SKALP to elastic fibres in

photodamaged skin: too much of a good thing?" J Invest Dermatol 127(6): 1286-

7.

Silvestre, J., P. J. Kenis, et al. (2009). "Cadherin and integrin regulation of epithelial

cell migration." Langmuir 25(17): 10092-9.

Song, P. I., Y. M. Park, et al. (2002). "Human keratinocytes express functional CD14

and toll-like receptor 4." J Invest Dermatol 119(2): 424-32.

Sorg, O., C. Antille, et al. (2006). "Retinoids in cosmeceuticals." Dermatol Ther 19(5):

289-96.

Sorrell, J. M. and A. I. Caplan (2004). "Fibroblast heterogeneity: more than skin deep."

J Cell Sci 117(Pt 5): 667-75.

Stern, R. and H. I. Maibach (2008). "Hyaluronan in skin: aspects of ageing and its

pharmacologic modulation." Clin Dermatol 26(2): 106-22.

Svobodova, A., J. Psotova, et al. (2003). "Natural phenolics in the prevention of UV-

induced skin damage. A review." Biomed Pap Med Fac Univ Palacky Olomouc

Czech Repub 147(2): 137-45.

Tajima, S., H. Inoue, et al. (1999). "Alginate oligosaccharides modulate cell

morphology, cell proliferation and collagen expression in human skin fibroblasts

in vitro." Arch Dermatol Res 291(7-8): 432-6.

Tamarat, R., J. S. Silvestre, et al. (2003). "Blockade of advanced glycation end-product

formation restores ischemia-induced angiogenesis in diabetic mice." Proc Natl

Acad Sci U S A 100(14): 8555-60.

Toebak, M. J., S. Gibbs, et al. (2009). "Dendritic cells: biology of the skin." Contact

Dermatitis 60(1): 2-20.

Tsukahara, K., Y. Takema, et al. (2001). "Selective inhibition of skin fibroblast elastase

elicits a concentration-dependent prevention of ultraviolet B-induced wrinkle

formation." J Invest Dermatol 117(3): 671-7.

Vlckova, V., V. Duhova, et al. (2004). "Antigenotoxic potential of glucomannan on four

model test systems." Cell Biol Toxicol 20(6): 325-32.

Wang, C. Y., H. J. Liu, et al. "AGE-BSA down-regulates endothelial connexin43 gap

junctions." BMC Cell Biol 12: 19.

72

Watt, F. M. (2002). "Role of integrins in regulating epidermal adhesion, growth and

differentiation." EMBO J 21(15): 3919-26.

Weindl, G., J. Wagener, et al. "Epithelial cells and innate antifungal defense." J Dent

Res 89(7): 666-75.

Wenk, J., J. Schuller, et al. (2004). "Overexpression of phospholipid-hydroperoxide

glutathione peroxidase in human dermal fibroblasts abrogates UVA irradiation-

induced expression of interstitial collagenase/matrix metalloproteinase-1 by

suppression of phosphatidylcholine hydroperoxide-mediated NFkappaB

activation and interleukin-6 release." J Biol Chem 279(44): 45634-42.

Yamagishi, S. and T. Matsui "Advanced glycation end products, oxidative stress and

diabetic nephropathy." Oxid Med Cell Longev 3(2): 101-8.

Zhang, Q., N. Li, et al. (2003). "In vivo antioxidant activity of polysaccharide fraction

from Porphyra haitanesis (Rhodephyta) in ageing mice." Pharmacol Res 48(2):

151-5.

73

11. Abbreviations

AGEs advanced glycation end products

BMZ basement membrane zone

CE cornified envelope

CR3 complement receptor 3

CTGF connective tissue growth factor

DEJ dermo-epidermal junction

FROP fucose-rich oligo/polysaccharide

GM glucomannan

IL-1α interleukin-1α

IL-8 interleukin-8

iNOS inducible nitric oxide synthase

MAPK mitogen-activated protein kinases

MMP matrix metalloprotease

MR mannose receptor

MW molecular weight

MCP-1 monocyte chemoattractant protein-1

NF-κB nuclear factor- κB

NHDF normal human dermal fibroblasts

NHK normal human keratinocytes

PAMP pathogen recognition patterns

PGE2 prostaglandin E2

RROP rhamnose-rich oligo/polysacharide

ROS reactive oxygen species

SC Stratum corneum

TLR toll-like receptors

TGFb1 transforming growth factor, beta 1

UVB ultraviolet B light

VEGF vascular endothelial growth factor

Documents

Modulation of Keratinocyte & Dermal Fibroblast Physiology by Selected Polysaccharides Thesis