Exposure to Probiotic Lactobacillus acidophilus L-92 ModulatesGene Expression Profiles of Epithelial Caco-2 Cells

Sae Yanagihara,1,2 Shinji Fukuda,2–4 Hiroshi Ohno,2–4 and Naoyuki Yamamoto1,2

1Microbiology and Fermentation Laboratory, Calpis Co., Ltd.,Sagamihara, Kanagawa, Japan.

2Intestinal Microbe Symbiosis Laboratory, RIKEN, Wako, Saitama, Japan.3Laboratory for Epithelial Immunobiology, Research Center for Allergy and Immunology, RIKEN, Yokohama, Kanagawa, Japan.

4Supramolecular Biology, International Graduate School of Bionanoscience, Yokohama City University,Yokohama, Kanagawa, Japan.

ABSTRACT To understand host gastrointestinal response after exposure to probiotic Lactobacillus acidophilus L-92,

microarray analysis of cultured epithelial Caco-2 cells was performed. Of the 187 genes down-regulated after 4 h treatment

with L-92, 25 were involved in RNA splicing; 12, in cell cycle; 8 were transcriptional regulators; 2 were involved in ubiquitin

proteolysis; 2, in adhesion; 2, in meiosis; 2, in splicing; and 2 encoding cytokines. In the RNA splicing group, genes encoding

small nuclear RNAs, nuclear pore complex interacting proteins, RNA binding motif proteins, and SMG1 homologs (phos-

phatidylinositol 3-kinase-related kinase) were identified. Among the only 13 genes up-regulated by the treatment, 5 were

involved in histone structure, and 2 were involved in metabolism. Genes belonging to cell adhesion, transmembrane proteins,

mitogen-activated protein kinase, immune response, DNA binding, inflammation, and protein synthesis groups were mainly

up-regulated after 20 h of treatment, whereas no significantly down-regulated genes were observed. In the present tran-

scriptome analysis, during the early stage of treatment (four hours of treatment) with L-92, genes involved in cell growth and

cell meiosis were mainly repressed. During the late phase of treatment (20 h of treatment), the expression of the genes linked

to cell adhesion activity and metabolism for cell growth was enhanced. From the present transcriptome analysis, we suggest

that Caco-2 cells slow down cell death and turnover of RNA synthesis as an early response to L-92 treatment; at the late stage

of treatment, the genes involved in cell proliferation, transcriptional activity, and apoptosis are activated.

KEY WORDS: � apoptosis � cell proliferation � epithelial Caco-2 cells � gastrointestinal response � Lactobacillus

acidophilus L-92 � transcriptome analysis

INTRODUCTION

Many industrial probiotic bacterial strains have beendeveloped that help manage a number of health issues,

such as gastrointestinal (GI) disorders, inflammatory boweldiseases, allergic symptoms,1–3 and pathogenic infections.Among these probiotic strains, Lactobacillus acidophilus L-92 has been developed with anti-allergic effects for pollenallergy,4 perennial allergy,5 atopic dermatitis,6,7 and theability to control gastro-intestinal disorders.8 These resultsare based on clinical evidence from Japanese subjects. Thesehealth benefits, especially immunomodulatory effects, arethought to be caused by interactions between L-92 and thehost, which are followed by secondary responses, mainlythrough the immune network. The adhesion of L-92 to epi-

thelial cells of the GI tract may also be an important event fortheir communication with host GI cells and survival in the GItract.

Recently, transcriptome analyses of human intestinalsamples, collected from biopsies during clinical trials, wereconducted to study host GI events.9–12 Studies have soughtto understand the host response after the intake of probioticcells. However, the collection of human samples duringclinical trials is not an easy task; GI cells are strongly af-fected by diet, but it is difficult to control the subjects’ diet.

Caco-2 cells are a well-established human epithelial cellline and have been used in many studies pertaining totransepithelial transport, metabolism, and disease.13–15 Arecent study analyzed changes in the host response throughthe analysis of Caco-2 cells using microarray analysis.16–18

Caco-2 cells could be used to compare differences in pro-biotic characteristics of different bacterial strains. Of par-ticular interest, Caco-2 cells overlayed on macrophage-likeRAJI cells showed cross-communication in the transwellsystem.19–22 A transwell system using co-cultivating Caco-2

Manuscript received 3 February 2012. Revision accepted 2 March 2012.

Address correspondence to: Naoyuki Yamamoto, Ph.D., Microbiology and FermentationLaboratory, Calpis Co., Ltd., 11-10, 5-Chome, Fuchinobe, Sagamihara-shi, Chuo-ku,Kanagawa 252-0206, Japan, E-mail: naoyuki.yamamoto@calpis.co.jp

JOURNAL OF MEDICINAL FOODJ Med Food 15 (6) 2012, 511–519# Mary Ann Liebert, Inc., and Korean Society of Food Science and NutritionDOI: 10.1089/jmf.2012.0040

511

cells with macrophage-like cells, such as THP-1 cells andRAJI cells, serves as a useful tool for studying cross-talkbetween epithelial cells and immune cells.23,24 In the use ofthe transwell system, inflammatory damage caused bypathogenic bacterial treatment and its rescue by some pro-biotics has been reported.24

Based on a comparative proteome analysis using L.acidophilus L-92 and the less adhesive CP23 strain,25 it wassuggested that surface layer protein A (SlpA) of L. acid-ophilus L-92 could serve as an adhesion molecule withCaco-2 cells to aid the communication between L-92 cellsand host cells. Results show L-92 to be more effective inreleasing an inflammatory cytokine, IL-12, from spleno-cytes compared with CP23, which has less SlpA. Changes insome cytokines and chemokines after uptake of L-92 werereported in animal studies.6,26 However, information pro-cesses, in addition to host metabolism, is required.

In the present study, using microarray analysis, we in-vestigated changes in gene expressions link to metabolismoccurring in GI epithelial cells after exposure to L-92 cells.

MATERIALS AND METHODS

Bacterial strain

L. acidophilus L-92 strain was used from our stock cul-ture collection and was maintained anaerobically at 37�C for20 h. in Man Rogosa Sharpe (MRS) broth (Difco Labora-tories, Detroit, MI).

Caco-2 cell culture

The human colonic cell line Caco-2 was obtained fromthe RIKEN Cell Bank (Tsukuba, Japan) and maintained inDulbecco’s modified Eagle’s medium (Sigma, St. Louis,MO) containing 10% (V/V) of heat-inactivated fetal bovineserum (Hana-Nesco Bio, Tokyo, Japan), 1% (W/V) aminoacids (Sigma), streptomycin (100 lg/mL), and penicillin(100 U/mL) (GIBCO, Grand Island, NY) in a 100 cm2 dish(Corning Glass Works, Corning, NY) at 37�C in 5% CO2/95% air. The culture medium was replaced every other dayto maintain the cells. Monolayered cells were collected byadding 0.25% trypsin (Sigma) for 5 min, and were seeded at4.5 · 105 cells/well in 3.0 mL of the cell suspension.

L-92 treatment of Caco-2 cells

For the transcriptome analysis assay, Caco-2 cells weregrown in six-well tissue plates (Asahi Glass Co., Ltd., Tokyo,Japan) at 37�C. The medium was changed every other day.The assay was performed on Caco-2 cells on day 21. On day21, the L-92 cells cultured in MRS medium were collected bycentrifugation at 13,000 g for 10 min, and resuspended inphosphate-buffered saline (PBS) (8.15 mM Na2HPO412H2O,1.47 mM KH2PO4 2.68 mM KCl, and 0.137 M NaCl, pH 7.4)containing 0.137 M NaCl. The L-92 cells were collectedagain by centrifugation at 13,000 g for 10 min. The washedcells were resuspended in PBS and heat treated at 100�C for10 min. Then, 1 · 1010 cells of heat-killed L-92, suspended in

3 mL of culture buffer were added to each well of the cultureplate. After centrifugation of the plate at 3000 g for 1 min toaccelerate the association of L-92 and Caco-2 cells, the mixedcells were incubated at 37�C in 5% CO2/95% air. After 4 and20 h of incubation, the supernatant from the mixed culturedcells was discarded, and the Caco-2 cells were washed twicewith PBS to remove nonbound cells. Finally, the Caco-2 cellswere harvested using a cell scraper after the addition of coldPBS. After centrifugation, 1.5 mL of RNAlater solution(QIAGEN K. K., Tokyo, Japan) was quickly added to thecollected cell pellet to extract RNA. The suspension wasmaintained at 4�C for 12 h, and then stored at - 30�C. TotalRNA was purified and used for cDNA synthesis after analysisof its quality. Untreated cells were used as controls. Threereplicate samples were prepared for each group.

Microarray analysis

Total RNA was extracted using an RNeasy Mini Kit(QIAGEN, Valencia, CA). Double-stranded cDNA wassynthesized from 5 lg of total RNA, and the cDNA wassubjected to in vitro transcription in the presence of bioti-nylated nucleotide triphosphates. Human genome-widegene expression was examined using the GeneChip� HumanGene 1.0 ST Array (Affymetrix, Santa Clara, CA) con-taining oligonucoleotide probe sets for approximately28,869 full-length genes and expressed sequence tags. Thebiotinylated cRNA was hybridized with the probe array for16 h at 45�C. The hybridized products were stained withstreptavidin-phycoerythrin, and then scanned with a Hew-lett–Packard Gene Array Scanner (Palo Alto, CA). Thefluorescence intensity of each probe was quantified usingGeneChip Analysis Suite 5.0 software (Affymetrix). Thelevel of gene expression was determined as the averagedifference using the GeneChip software.

Statistical and functional analysis of microarray data

Data analysis was performed using Genespring GX soft-ware, version 11.5.1 (Silicon Genetics, San Carlos, CA).Expression data were considered significant when theydiffered by at least 1.25 fold between the L-92 treatment andthe PBS treatment, as P values of < .05 were observed whenstatistically analyzed by T test unpaired with Benjamini–Hochberg as post test.

RESULTS

Changed gene expressions after four hoursof treatment of L-92 cells

Most of the changes in gene expression in Caco-2 cellswere relatively mild after treatment with L. acidophilus L-92. However, significant changes were observed. After the4 h of treatment with L. acidophilus L-92, 187 genes weresignificantly down-regulated by < 0.77-fold, and 12 geneswere up-regulated by > 1.25-fold in Caco-2 cells comparedwith nontreated cells, as illustrated in Figure 1. After 20 h oftreatment, 45 genes were significantly up-regulated in Caco-2 cells by > 1.25-fold (Fig. 1). For a detailed understanding

512 YANAGIHARA ET AL.

of the response of the cultured Caco-2 cells, up-regulatedand down-regulated genes were categorized based on thecommon pathways in KEGG analysis (www.genome.jp/kegg/pathway.html), as shown in Table 1. As illustrated inFigure 1, most of the characteristic changes in gene expres-sion observed at four hours pertained to down-regulation. Ofthe genes down-regulated after the 4 h of treatment, 25 geneswere classified as RNA splicing, 6 genes as transcriptionalregulators, 2 genes as ubiquitin proteolytics, 2 genes as ad-hesion, 2 genes as meiosis, 2 genes as cytokines, and 2 genesas splicing (Fig. 1). In the RNA splicing group, genes en-coding small nuclear RNAs, nuclear pore complex interact-ing proteins, RNA binding motif proteins, and SMG1homologs (phosphatidylinositol 3-kinase-related kinase)were identified. After the 4 h of treatment, only five geneswere involved in histone structure, two genes were involvedin metabolism, and a further five other genes were up-regulated (Table 1).

Changed gene expressions after 20 hof treatment of L-92 cells

After the 20 h of treatment, genes belonging to the cate-gories of cell adhesion, mitogen-activated protein (MAP)kinase, transmembrane function, immune response, metab-olism, and others were mainly up-regulated (Fig. 1). Amongthe up-regulated genes, two dual-specificity phosphatasegenes and two small Cajal body-specific RNA genes inthe MAP kinase pathway, four carcinoembryonic antigen-related cell adhesion molecule genes, two amphiregulingenes, and two integrin genes for cell adhesion were up-regulated (Table 2). No significantly down-regulated geneswere observed in the 20 h treatment group.

Changed gene expressions after 4 and 20 hof treatment of L-92 cells

Only two genes in the metabolism and nucleosomestructure group were significantly up-regulated after boththe 4 and 20 h of treatments. Genes for meiosis, cytosolic

DNA sensing, cell cycle, and leucine rich repeat were down-regulated after 4 h but up-regulated after the 20 h of treat-ment. Throughout the present transcriptome analysis, duringthe early stage of treatment with L-92 (four hours of treat-ment), many genes linked to cell growth and cell meiosiswere mainly repressed, as shown in Figure 1. During the latephase of the treatment (20 h), many genes linked to celladhesive activity and metabolism of cell growth, such asDNA synthesis, protein synthesis, and kinase activity, wereincreased (Fig. 1).

Summary of the changed gene expressions in L-92 cells

To help understand the events occurring in the Caco-2cells after L-92 treatment, many of the genes listed in Tables1 and 2 are illustrated in Figures 2 and 3. During the earlystage of the treatment (4 h), genes such as FBX, RNF, ZNF,and ATF (Table 1) were down-regulated, and two metabo-lism-related genes, both involved in apoptosis, were up-regulated (Fig. 2). Moreover, genes involved in RNAsplicing and RNA degradation were down-regulated, andgenes encoding RNA synthesis were up-regulated (Fig. 2).Genes encoding the adhesive molecule mucin and somegenes linked to transcriptional regulation and MAP kinasewere down-regulated (Fig. 2). In addition, genes linked tophagocytosis and the G-protein-coupled receptor were alsodown-regulated. These results suggest that Caco-2 cellsmight respond to L-92 treatment by slowing down cell deathand RNA degradation, and accelerating RNA synthesis as anearly response. After the 20 h of treatment, most of the geneslinked to cell proliferation, gene regulation, MAP kinase,apoptosis, and some adhesive molecules were up-regulated(Fig. 3). These results suggest that long-term treatment ofL-92 (20 h) might activate cell proliferation, transcriptionalactivity, and apoptosis in Caco-2 cells.

DISCUSSION

Host-bacterial cross-talk between probiotic bacteria andGI epithelial cells is thought to be an important process in

FIG. 1. Profile of the changed genesin the transcriptome analysis of Caco-2cells after 4- and 20-h treatment withLactobacillus acidophilus L-92. Genescategorized by KEGG analysis areshown in the square.

CACO-2 CELL RESPONSE AFTER L-92 TREATMENT 513

Table 1. Changed Genes in Caco-2 Cells After Four Hours of Treatment of Lactobacillus acidophilus L-92

CategoryGene description Gene symbol Genbank Fold change P value

Up-regulated genesNucleosome structure

Histone cluster 1, H2bf HIST1H2BF BC056264 1.40 0.003Histone cluster 1, H2bh HIST1H2BH BC096116 1.34 0.005Histone cluster 1, H1c HIST1H1C AK290947 1.30 0.006Histone cluster 1, H4d HIST1H4D BC128104 1.31 0.001Histone cluster 1, H1b HIST1H1B BC069101 1.31 0.010

MetabolismCytochrome P450, family 1, subfamily B, polypeptide 1 CYP1B1 U03688 3.51 0.000TIMP metallopeptidase inhibitor 3 TIMP3 1.46 0.007

RNA synthesisPolymerase (RNA) I polypeptide E, 53 kDa POLR1E BC001337 1.33 0.000

PhagosomeCoronin, actin binding protein, 2A CORO2A BC011690 1.46 0.000

MAPKK familyPDZ binding kinase PBK AF237709 1.34 0.000

TransmembranePhosphatidylinositol glycan anchor biosynthesis, class K PIGK AF022913 1.31 0.002

UbiquitinationUbiquitin associated and SH3 domain containing B UBASH3B BC007541 1.40 0.006

Down-regulated genesRNA spliceosome, survailance

Small nucleolar RNA, H/ACA box 16A SNORA16A AK092096 - 0.49 0.001Small nucleolar RNA, C/D box 13 pseudogene 2 SNORD13P2 X58060 - 0.58 0.002Small nucleolar RNA, H/ACA box 38B (retrotransposed) SNORA38B - 0.59 0.000Small nucleolar RNA, H/ACA box 75 SNORA75 - 0.61 0.002Small nucleolar RNA, C/D box 15A SNORD15A - 0.67 0.001Small nucleolar RNA, H/ACA box 70 SNORA70 AK027197 - 0.75 0.001Small nucleolar RNA, H/ACA box 70B SNORA70B - 0.76 0.004Small nucleolar RNA, H/ACA box 61 SNORA61 AK092096 - 0.76 0.042Small nucleolar RNA, H/ACA box 51 SNORA51 - 0.76 0.042Small nucleolar RNA, H/ACA box 67 SNORA67 AK296664 - 0.77 0.006Nuclear pore complex interacting protein NPIP AK294177 - 0.70 0.001Nuclear pore complex interacting protein NPIP AF132984 - 0.74 0.002Nuclear pore complex interacting protein NPIPL4 AK296338 - 0.75 0.011Nuclear pore complex interacting protein NPIP AF132984 - 0.76 0.003Nuclear pore complex interacting protein NPIPL5 BC144439 - 0.77 0.020RNA binding motif protein 14 RBM14 BC007641 - 0.66 0.001RNA binding motif protein 14 RBM14 BC007641 - 0.71 0.005RNA binding motif protein 5 RBM5 AF091263 - 0.74 0.001SMG1 homolog, phosphatidylinositol 3-kinase-related kinase SMG1 AB061371 - 0.74 0.009SMG1 homolog, phosphatidylinositol 3-kinase-related kinase SMG1 AF395444 - 0.76 0.006SMG1 homolog, phosphatidylinositol 3-kinase-related kinase SMG1 AF395444 - 0.77 0.009Coiled-coil domain containing 82 CCDC82 AK313893 - 0.67 0.000WD repeat domain 52 WDR52 AK002004 - 0.70 0.022Pyridoxal-dependent decarboxylase domain containing 2 PDXDC2 AK294177 - 0.75 0.008Coiled-coil domain containing 142 CCDC142 BC143399 - 0.75 0.001

Cell cycleChromosome 12 open reading frame 27 C12orf27 - 0.47 0.001Chromosome 3 open reading frame 42 C3orf42 AF280797 - 0.68 0.001Chromosome 6 open reading frame 134 C6orf134 BC047303 - 0.70 0.009Chromosome 17 open reading frame 91 C17orf91 BX648321 - 0.70 0.001Chromosome 20 open reading frame 29 C20orf29 BC043344 - 0.71 0.001Chromosome 15 open reading frame 28 C15orf28 - 0.73 0.002Chromosome 2 open reading frame 49 C2orf49 AK127661 - 0.73 0.004Chromosome X open reading frame 40B CXorf40B L43578 - 0.73 0.005Chromosome 19 open reading frame 6 C19orf6 DQ005958 - 0.74 0.000Chromosome 6 open reading frame 64 C6orf64 BC022007 - 0.75 0.004Chromosome 2 open reading frame 14 C2orf14 AK093281 - 0.77 0.039Chromosome 17 open reading frame 88 C17orf88 AF143236 - 0.76 0.016

(continued)

514 YANAGIHARA ET AL.

eliciting the host response, including defensive responsesagainst pathogenic bacteria and immune responses. To un-derstand the probiotic roles of the host GI cells, tran-scriptome analyses for human GI cell samples collected bybiopsy are indicated. However, this is not an easy task.Separations of the epithelial cells and lamina propria im-mune cells, which may play different roles in host defenseand immunomodulatory effects, are also difficult to achieve.

To mimic the in vivo gut environment, human intestinalepithelial Caco-2 cells have been utilized to study GI events,epithelial disease status, transport of nutrients, and bioactivecomponents.27–29 In the present study, we performed atranscriptome analysis of epithelial Caco-2 cells to analyzethe host GI response after the uptake of probiotic L-92 cells.Specific binding of the L-92 cells to Caco-2 cells was ex-pected via a cell-wall-associated adhesive molecule, SlpA,

Table 1. Continued

CategoryGene description Gene symbol Genbank Fold change P value

Transcriptional regulatorZinc finger protein 778 ZNF778 AK295122 - 0.63 0.000Zinc finger and SCAN domain containing 16 ZSCAN16 BC004255 - 0.76 0.006Zinc finger family member 767 ZNF767 BC047675 - 0.76 0.013Zinc finger, BED-type containing 3 ZBED3 BC007239 - 0.77 0.005Activating transcription factor 4 ATF4 D90209 - 0.76 0.000Activating transcription factor 5 ATF5 AB021663 - 0.74 0.004Ring finger protein 5 pseudogene 1 RNF5P1 BC119741 - 0.59 0.000Ring finger protein 5 pseudogene 1 RNF5 BC111392 - 0.65 0.000

Taste transduction (Bitter taste)Taste receptor, type 2, member 14 TAS2R14 BC103699 - 0.62 0.017Taste receptor, type 2, member 20 TAS2R20 BC100915 - 0.65 0.013Taste receptor, type 2, member 31 TAS2R31 BC117421 - 0.67 0.023Taste receptor, type 2, member 4 TAS2R4 BC130439 - 0.68 0.039Taste receptor, type 2, member 19 TAS2R19 BC101804 - 0.74 0.031Taste receptor, type 2, member 3 TAS2R3 BC095523 - 0.76 0.041

Adhesion pathwayMucin 12, cell surface associated MUC12 - 0.74 0.026Mucin 3B, cell surface associated MUC3B AB038783 - 0.72 0.008

PeroxisomeSolute carrier family 7, (cationic amino acid transporter, y + system) SLC7A11 AF252872 - 0.69 0.001Solute carrier family 7 (cationic amino acid transporter, y + system) SLC7A6 BC028216 - 0.71 0.000Solute carrier family 25, member 32 SLC25A32 BC021893 - 0.74 0.000

MeiosisG protein-coupled receptor 21 GPR21 BC066885 - 0.68 0.002G protein-coupled receptor 75 GPR75 AF072693 - 0.74 0.016

Splicesome pathwaySplicing factor, arginine/serine-rich 1 SFRS1 BC033785 - 0.77 0.001Splicing factor, arginine/serine-rich 6 SFRS6 AK300411 - 0.74 0.003

Cytosolic DNA sensing pathwayTransmembrane protein 80 TMEM80 BC008671 - 0.63 0.000Transmembrane protein 88 TMEM88 BC057812 - 0.76 0.009

Cytokine productionInterleukin 17 receptor B IL17RB AF208111 - 0.77 0.024

Neurone interactionGamma-aminobutyric acid (GABA) A receptor, epsilon GABRE BC026337 - 0.65 0.011

OthersProtease, serine, 35 PRSS35 AY358661 - 0.44 0.001FRSS1829 LOC100132099 AY358798 - 0.46 0.005Oculomedin OCLM AF142063 - 0.51 0.003MicroRNA 221 MIR221 - 0.58 0.013Nuclear receptor coactivator 5 NCOA5 BC140836 - 0.60 0.000Ankyrin repeat domain 36B ANKRD36B BC125132 - 0.60 0.002Transient receptor potential cation channe LOC100133315 - 0.60 0.007Family with sequence similarity 106, member A FAM106A - 0.62 0.001IGYY565 LOC100130428 BC040288 - 0.62 0.005Family with sequence similarity 106, member A FAM106A - 0.62 0.001F-box protein 9 FBXO9 AK095307 - 0.63 0.003Family with sequence similarity 106, member C pseudogene FAM106C - 0.63 0.002GALI1870 LOC100133299 AY358688 - 0.66 0.004F-box protein 17 FBXO17 AK021860 - 0.75 0.008

CACO-2 CELL RESPONSE AFTER L-92 TREATMENT 515

Table 2 Up-Regulated Gene Expressions in Caco-2 cells After 20 H of Treatment of Lactobacillus acidophilus L-92

CategoryGene description Gene symbol Genbank Fold change P value

Up-regulated genesCell adhesion

Carcinoembryonic antigen-related cell adhesion molecule 7 CEACAM7 X98311 1.615 0.0082Carcinoembryonic antigen-related cell adhesion molecule 1 CEACAM1 J03858 1.509 0.0012Carcinoembryonic antigen-related cell adhesion molecule 5 CEACAM5 M29540 1.452 0.0093Carcinoembryonic antigen-related cell adhesion molecule 6 CEACAM6 BC005008 1.401 0.0074Amphiregulin AREG BC009799 1.567 0.0053Amphiregulin AREG BC009799 1.315 0.0041Integrin, alpha 2 ITGA2 1.310 0.0071Integrin, alpha 6 ITGA6 BC136455 1.265 0.0119Similar to keratin 18 LOC442249 1.359 0.0009Mannose-binding lectin (protein C) 2, soluble (opsonic defect) MBL2 X15422 1.279 0.0069Leucine rich repeat containing 66 LRRC66 1.360 0.0081Keratin 20 KRT20 BC031559 1.330 0.0043

MAP kinase pathwayDual specificity phosphatase 6 DUSP6 BC005047 1.309 0.0029Dual specificity phosphatase 5 DUSP5 BC062545 1.304 0.0006Small Cajal body-specific RNA 22 SCARNA22 1.460 0.0003Small Cajal body-specific RNA 13 SCARNA13 1.301 0.0047Cytochrome P450, family 1, subfamily A, polypeptide 1 CYP1A1 BC023019 2.076 0.0003G protein-coupled receptor, family C, group 5, member A GPRC5A AK122672 1.266 0.0055Sprouty-related, EVH1 domain containing 1 SPRED1 BC137480 1.433 0.0009

Transmembrane proteinMatrix metallopeptidase 15 (membrane-inserted) MMP15 BC036495 1.389 0.0029Transmembrane channel-like 7 TMC7 BC036205 1.373 0.0089Homeobox B9 HOXB9 BC015565 1.495 0.0004Transmembrane protein ENSP00000340100 LOC100129969 1.337 0.0132

Immune related pathwayFc fragment of IgE FCER1G M33195 1.428 0.0048Immunoglobulin heavy constant delta IGHD BC021276 1.356 0.0016Immediate early response 3 IER3 BC005080 1.274 0.0140ERBB receptor feedback inhibitor 1 ERRFI1 BC025337 1.272 0.0026

MetabolismChromosome 10 open reading frame 54 C10orf54 AY358379 1.305 0.0119Chromosome 17 open reading frame 78 C17orf78 BC034672 1.275 0.0013Histone cluster 1, H2ai HIST1H2AI BC112254 1.332 0.0090

DNA bindingets variant 4 ETV4 BC016623 1.325 0.0098

InflammationAcyloxyacyl hydrolase (neutrophil) AOAH BC025698 1.311 0.0146

Protein synthesisRibosomal protein S3A RPS3A L13802 1.652 0.0099

Others20,50-oligoadenylate synthetase 1, 40/46 kDa OAS1 AY730627 1.273 0.0002Ankyrin repeat domain 1 (cardiac muscle) ANKRD1 BC018667 1.428 0.0065Tescalcin TESC AK000614 1.255 0.0048Glutathione peroxidase 2 (gastrointestinal) GPX2 BC005277 1.350 0.0024Carbonic anhydrase XII CA12 AF051882 1.328 0.0002WNK lysine deficient protein kinase 4 WNK4 BC136664 1.392 0.0040Serpin peptidase inhibitor, clade E (nexin, plasminogen activator inhibitor type 1) SERPINE1 BC010860 1.333 0.0060ATP-binding cassette, sub-family B (MDR/TAP), member 1 ABCB1 M14758 1.466 0.0097Transforming, acidic coiled-coil containing protein 1 TACC1 AF049910875 1.285 0.0030Similar to hCG2042717 LOC100293211 1.565 0.0038Recombination signal binding protein for immunoglobulin kappa J region RBPJ D14041 1.279 0.0108Serpin peptidase inhibitor, clade E (nexin, plasminogen activator inhibitor type 1) SERPINE2 BC042628 1.499 0.0056Mannosyl (beta-1,4-)-glycoprotein beta-1,4-N-acetylglucosaminyltransferase MGAT3 BC113383 1.456 0.0023

516 YANAGIHARA ET AL.

present in L-92 cells.25 Changes in gene expression profilein Caco-2 cells were relatively mild, but many significantchanges were observed by microarray analysis after 4 and20 h of treatment, as summarized in Figures 2 and 3. Theresults suggest that Caco-2 cells are prevented from celldeath, categorized as apoptosis, and cell proliferation(growth) during the early phase of the treatment (Fig. 2).Based on these results, Caco-2 cells seem to respond first tothe stress from L-92 binding, then adjust their activities tocell proliferation during the late phase of the treatment. Byrepeated oral uptakes of probiotic L-92 cells, responses

observed in Caco-2 cells at 20 h would be most likely ex-pected in the epithelial GI tract in humans.

Probiotic strains are thought to communicate with intes-tinal epithelial cells in a strain-specific manner. Some of theresponses observed in this study were previously reportedusing different lactic acid bacterial strains in animal, human,and cultured cell studies. Among these previous studies,Lactobacillus GG was shown to suppress cytokine-inducedapoptosis in in vitro cultured epithelial cells.30 By contrast,up-regulation of anti-apoptotic and cytoprotective geneswas confirmed in the animal study that used orally

FIG. 2. Genes mainly changed inCaco-2 after 4-h exposure to L. acid-ophilus L-92. Down-regulated genes areindicated with blue, while up-regulatedgenes are red. Categorized networks arecircled.

FIG. 3. Genes mainly changed inCaco-2 cells after 20-h exposure toL. acidophilus L-92. Up-regulatedgenes are indicated with red. Catego-rized networks are circled.

CACO-2 CELL RESPONSE AFTER L-92 TREATMENT 517

administered Lactobacillus GG.31 The repressive effect onapoptosis was observed in previous studies32 and in thepresent study during the early stage. The induction of MAPkinases, observed in this study after the 20 h of treatment,was also detected in intestinal epithelial murine colonicYAMC cells after Lactobacillus GG treatment (32). Chan-ges in gene expression for G-protein coupled receptor, cy-tochrome P450, and zinc finger protein were observed in thepresent study and was also previously reported.17

Differential responses to probiotic bacteria between thepresent study and previous in vivo studies have been ob-served. In a previous study on the human intestinal tract, up-regulation of a G-protein coupled receptor was observedafter 1 h of treatment in a clinical trial, but it was only up-regulated after 20 h of treatment in the present study.10

In a human challenge trial, various genes linked to the im-munoregulatory response were also changed.10 In contrast,changes in the expression of genes involved in the immu-nomodulatory response were small in the present study.These results obtained in the previous in vivo studies mayinvolve responses not only from epithelial cells but alsofrom immune cells, such as dendritic cells and T cells.Moreover, treatment with Lactobacillus GG over a longperiod in the human challenge trial increased the expressionof many immunomodulatory genes, such as cytokines andchemokines.9 However, this was not observed in the presentstudy. In contrast, increases in the gene expression of mucin,integrin, cytochrome P450, and ubiquitin observed in this invitro study were also detected throughout the continuousuptake of GG in the previous study.9 Moreover, human colonepithelial cells, such as T84 cells, HT29 cells, and Caco-2cells, secreted pro-inflammatory and chemo-attractant cyto-kines in response to bacterial invasion treatment in a previ-ous study.33 However, the expression levels of cytokinegenes were largely unchanged in the present study.

Many studies have focused on host GI responses to pro-biotic lactic acid bacteria, yet details of host-probiotic cross-talk are not fully understood. Probiotic strains are thought tospecifically communicate with GI epithelial cells in a strain-dependent manner. Throughout the present transcriptomeanalysis of cultured Caco-2 cells, the host epithelial re-sponse to specific exposure with L-92 was mainly discussedto understand its impact on the GI tract. Our previous study,involving comparative proteome analysis with the L. acid-ophilus CP23 strain, which lacks SlpA, revealed that thebinding of L-92 cells to Caco-2 cells might be strongly as-sociated with the cell-surface SlpA protein (25). By at-tachment of the L-92 strain to the epithelial cells, most likelyvia the SlpA protein, Caco-2 cells seemed to respond toL-92 by accelerating cell proliferation after the first com-munication, during the early stage. It is also suggested thatL-92 might have an in vivo effect on epithelial cells of the GItract by specific binding. Information obtained in the presentstudy could be used in considering the impact of L-92 on theGI tract and host health. In addition, comparative tran-scriptome analyses using different probiotic strains on Caco-2 cells would be of interest to help understand the potentialof probiotics. Moreover, transcriptome analysis of immune

cells, such as dendritic cells, should be addressed next inorder to further understand gastric immunity.

AUTHOR DISCLOSURE STATEMENT

The authors confirm that no competing financial interestsexist.

REFERENCES

1. Kalliomaki M, Salminen S, Poussa T, Arvilommi H, Isolauri E:

Probiotics and prevention of atopic disease: 4-year follow-up of a

randomised placebo-controlled trial. Lancet 2003;361:1869–

1871.

2. Kalliomaki MA, Isolauri E: Probiotics and down-regulation of

the allergic response. Immunol Allergy Clin North Am 2004;24:

739–752.

3. Kalliomaki M, Antoine JM, Herz U, Rijkers GT, Wells JM,

Mercenier A: Guidance for substantiating the evidence for ben-

eficial effects of probiotics: prevention and management of al-

lergic diseases by probiotics. J Nutr 2010;140:713S–721S.

4. Ishida Y, Nakamura F, Kanzato H, Sawada D, Yamamoto N,

Kagata H, Oh-Ida M, Takeuchi H, Fujiwara S: Effect of milk

fermented with Lactobacillus acidophilus strain L-92 on symp-

toms of Japanese cedar pollen allergy: a randomized placebo-

controlled trial. Biosci Biotechnol Biochem 2005;69:1652–1660.

5. Ishida Y, Nakamura F, Kanzato H, Sawada D, Hirata H., Nish-

imura A, Kajimoto O, Fujiwara S: Clinical effects of Lactoba-

cillus acidophilus strain L-92 on perennial allergic rhinitis: a

double-blind, placebo-controlled study. J Dairy Sci 2005;88:

527–533.

6. Shah MM, Miyamoto Y, Yamada Y, Yamashita H, Tanaka H,

Ezaki T, Nagai H, Inagaki N: Orally supplemented Lactobacillus

acidophilus strain L-92 inhibits passive and active cutaneous

anaphylaxis as well as 2,4-dinitrofluorobenzene and mite fecal

antigen induced atopic dermatitis-like skin lesions in mice. Mi-

crobiol Immunol 2010;54:523–533.

7. Torii S, Torii A, Itoh K, Urisu A, Terada A, Fujisawa T, Yamada

K, Suzuki H, Ishida Y, Nakamura F, Kanzato H, Sawada D,

Nonaka A, Hatanaka M, Fujiwara S: Effects of oral administra-

tion of Lactobacillus acidophilus L-92 on the symptoms and

serum markers of atopic dermatitis in children. Int Arch Allergy

Immunol 2010;154:236–245.

8. Ikeda N, Kaneko K, Ejiri M, Mizutani J, Ohshima K, Kajimoto

O, Takahashi T, Yamamoto N: Improvement of the constipation

of healthy volunteers by uptake of lactic acid bacteria fermented

milk. J Nutr Food (in Japanese) 2000;3:59–68.

9. Di Caro S, Tao H, Grillo A, Elia C, Gasbarrini G, Sepulveda AR,

Gasbarrini A: Effects of Lactobacillus GG on genes expression

pattern in small bowel mucosa. Dig Liver Dis 2005;37:320–329.

10. Troost FJ, van Baarlen P, Lindsey P, Kodde A, de Vos WM,

Kleerebezem M, Brummer RJ: Identification of the transcrip-

tional response of human intestinal mucosa to Lactobacillus

plantarum WCFS1 in vivo. BMC Genomics 2008;9:374.

11. van Baarlen P, Troost F, van der Meer C, Hooiveld G, Boek-

schoten M, Brummer RJ, Kleerebezem M: Human mucosal

in vivo transcriptome responses to three lactobacilli indicate how

probiotics may modulate human cellular pathways. Proc Natl

Acad Sci USA 2011;108:4562–4569.

12. van Baarlen P, Troost FJ, van Hemert S, van der Meer C, de Vos

WM, de Groot PJ, Hooiveld GJ, Brummer RJ, Kleerebezem M:

518 YANAGIHARA ET AL.

Differential NF-jB pathways induction by Lactobacillus plan-

tarum in the duodenum of healthy humans correlating with im-

mune tolerance. Proc Natl Acad Sci USA 2009;106:2371–2376.

13. Adibi SA: Regulation of expression of the intestinal oligopeptide

transporter (Pept-1) in health and disease. Am J Physiol Gas-

trointest Liver Physiol 2003;285:G779–G788.

14. Borrelli F, Izzo AA: Role of acylethanolamides in the gastro-

intestinal tract with special reference to food intake and energy

balance. Best Pract Res Clin Endocrinol Metab 2009;23:

33–49.

15. van de Kerkhof EG, de Graaf IA, Groothuis GM: In vitro

methods to study intestinal drug metabolism. Curr Drug Metab

2007;8:658–675.

16. Anderson RC, Cookson AL, McNabb WC, Park Z, McCann MJ,

Kelly WJ, Roy NC: Lactobacillus plantarum MB452 enhances

the function of the intestinal barrier by increasing the expression

levels of genes involved in tight junction formation. BMC Mi-

crobiol 2010;10:316.

17. Panigrahi P, Braileanu GT, Chen H, Stine OC: Probiotic bacteria

change Escherichia coli-induced gene expression in cultured

colonocytes: implications in intestinal pathophysiology. World J

Gastroenterol 2007;13:370–6378.

18. Wang M, Zhang G, Han X, Yao L, Zhou Y, Jiang Y: Expression

profile analysis of intestinal Caco-2 cells treated with Lactoba-

cillus acidophilus NCFM. Wei Sheng Wu Xue Bao 2009;49:1247–

1252. (In Chinese.)

19. Fang HW, Fang SB, Chiang Chiau JS, Yeung CY, Chan WT,

Jiang CB, Cheng ML, Lee HC: Inhibitory effects of Lactoba-

cillus casei subsp. rhamnosus on Salmonella lipopolysaccharide-

induced inflammation and epithelial barrier dysfunction in a co-

culture model using Caco-2/peripheral blood mononuclear cells.

J Med Microb 2010;59:573–579.

20. Haller D, Holt L, Parlesak A, Zanga J, Bauerlein A, Sartor RB,

Jobin C: Differential effect of immune cells on non-pathogenic

Gram-negative bacteria-induced nuclear factor-jB activation and

pro-inflammatory gene expression in intestinal epithelial cells.

Immunology 2004;112:310–320.

21. Leonard F, Collnot EM, Lehr CM: A three-dimensional coculture

of enterocytes, monocytes and dendritic cells to model inflamed

intestinal mucosa in vitro. Mol Pharm 2010;7:2103–2119.

22. Parlesak A, Haller D, Brinz S, Baeuerlein A, Bode C: Modula-

tion of cytokine release by differentiated Caco-2 cells in a

compartmentalized coculture model with mononuclear leuco-

cytes and nonpathogenic bacteria. Scand J Immunol 2004;60:

477–485.

23. Ishimoto Y, Nakai Y, Satsu H, Totsuka M, Shimizu M: Transient

up-regulation of immunity- and apoptosis-related genes in Caco-

2 cells cocultured with THP-1 cells evaluated by DNA micro-

array analysis. Biosci Biotech Biochem 2010;74:437–439.

24. Tanoue T, Nishitani Y, Kanazawa K, Hashimoto T, Mizuno M:

In vitro model to estimate gut inflammation using co-cultured

Caco-2 and RAW264.7 cells. Biochem Biophys Res Commun

2008;374:565–569.

25. Ashida N, Yanagihara S, Shinoda T, Yamamoto N: Character-

ization of adhesive molecule with affinity to Caco-2 cells in

Lactobacillus acidophilus by proteome analysis. J Biosci Bioeng

2011;112:333–337.

26. Torii A, Torii S, Fujiwara S, Tanaka H, Inagaki N, Nagai H: Lac-

tobacillus acidophilus strain L-92 regulates the production of Th1

cytokine as well as Th2 cytokines. Allergol Int 2007;56:293–301.

27. Gratz S, Wu QK, El-Nezami H, Juvonen RO, Mykkanen H,

Turner PC: Lactobacillus rhamnosus strain GG reduces aflatoxin

B1 transport, metabolism, and toxicity in Caco-2 Cells. Appl

Environ Microbiol 2007;73:3958–3964.

28. Kwon O, Eck P, Chen S, Corpe CP, Lee JH, Kruhlak M, Levine

M: Inhibition of the intestinal glucose transporter GLUT2 by

flavonoids. FASEB J 2007;21:366–377.

29. San Martin CD, Garri C, Pizarro F, Walter T, Theil EC, Nunez

MT: Caco-2 intestinal epithelial cells absorb soybean ferritin by

mu2 (AP2)-dependent endocytosis. J Nutr 2008;138:659–666.

30. Yan F, Polk DB: Probiotic bacterium prevents cytokine-induced

apoptosis in intestinal epithelial cells. J Biol Chem 2002;277:

50959–50965.

31. Lin PW, Nasr TR, Berardinelli AJ, Kumar A, Neish AS: The

probiotic Lactobacillus GG may augment intestinal host defense

by regulating apoptosis and promoting cytoprotective responses

in the developing murine gut. Pediatr Res 2008;64:511–516.

32. Tao Y, Drabik KA, Waypa TS, Musch MW, Alverdy JC,

Schneewind O, Chang EB, Petrof EO: Soluble factors from

Lactobacillus GG activate MAPKs and induce cytoprotective

heat shock proteins in intestinal epithelial cells. Am J Physiol

Cell Physiol 2006;290:C1018–C1030.

33. Jung HC, Eckmann L, Yang SK, Panja A, Fierer J, Morzycka-

Wroblewska E, Kagnoff MF: A distinct array of proinflammatory

cytokines is expressed in human colon epithelial cells in response

to bacterial invasion. J Clin Invest 1995;95:55–65.

CACO-2 CELL RESPONSE AFTER L-92 TREATMENT 519