Novel functional hepatocyte cell line derived from spontaneous dwarf rat: Model of growth hormone function in vitro

CELL LINEhuc_97 164..172

Novel functional hepatocyte cell line derived fromspontaneous dwarf rat: Model of growth hormonefunction in vitro

Mayumi ISHIKAWA,1 Toshiaki TACHIBANA,2 Gen YOSHINO,1 Hisashi HASHIMOTO2 andToshiaki TANAKA3

1The Division of Diabetes, Metabolism and Endocrinology, Department of Internal Medicine, Toho University Schoolof Medicine, 2Department of Anatomy, Jikei University School of Medicine, and 3Tanaka Growth Clinic, Tokyo,Japan

Abstract

Currently there is no good hepatocyte model for studying growth hormone (GH) function thatreflects its normal physiological roles. Here we report the establishment of a functionalhepatocyte cell line, SDRL-1, from the liver of young male spontaneous dwarf rats (SDR), withisolated GH deficiency. This line has been maintained in Dulbecco’s Modified Eagle Medium(DMEM)/F12 medium supplemented with 10% fetal bovine serum (FBS) with retention of a neardiploid karyotype for extended periods of time. When grown as a monolayer sheet, it displayeda pavement-like appearance and contact inhibition. These cells have a poorly developed roughendoplasmic reticulum (r-ER), few mitochondria and glycogen granules, and produce a smallamount of albumin and a-fetoprotein, that is enhanced when grown on a collagen gel sponge.Human recombinant GH stimulated JAK2 and STAT5b tyrosine phosphorylation and IGF-Iproduction in a concentration-dependent manner. When the cells were cultured withGH-supplemented medium, the number of mitochondria and glycogen granules increasedtogether with the r-ER and Golgi apparatus. A number of microvilli were observed on thesurface of the cultured cells, further suggesting that this cell line is composed of normallyfunctioning hepatocytes. In summary, we established a novel hepatocyte cell line (SDRL-1), thatappears to display normal function, which we propose can serve as a good in vitro model forstudying GH-target organ interactions.

Key words: growth hormone, hepatocyte cell line, spontaneous dwarf rat.

INTRODUCTION

The liver is a major target of growth hormone (GH)action, and there are several reports that describe GHaction on cultured hepatocytes or hepatoma cell lines.

Braneld et al.1 examined GH control of IGF-I and GHreceptor (GHR) mRNA expression using porcine hepa-tocytes, while Scott et al.2 determined GH effects on rathepatocyte IGF-I. Additionally, Conover et al.3 studiedGH control over the synthesis and secretion of IGF-I, aprohormone, using human hepatoma cells. However,in these studies, the GH concentrations used to acti-vate IGF-I production were 20 ng/mL to 5000 ng/mL,which exceeded physiological concentrations.

We have established a novel hepatocyte cell line fromthe spontaneous dwarf rat (SDR), which has isolated GHdeficiency.4,5 The SDR was isolated in 1977 from a

Correspondence: Dr Mayumi Ishikawa, The Division ofDiabetes, Metabolism and Endocrinology, Department ofInternal Medicine, Toho University School of Medicine,Tokyo 143-8541, Japan. Email: [email protected]

Received 11 September 2010; accepted 23 September 2010

Human Cell 2010; 23: 164–172 doi: 10.1111/j.1749-0774.2010.00097.x

© 2010 The Authors164 Human Cell © 2010 Japan Human Cell Society

closed colony of the Sprague-Dawley (SD) rats.6 Thebody weights of these animals were reported to increasedose-dependently in response to administration of ratGH, ovine GH or recombinant human GH (hGH).7,8

Since the SDR is growth hormone deficient (GHD), theirhepatocytes are expected to be more sensitive to GHthan those from normal SD rats. In the present study, weshow that a hepatocyte cell line derived from the SDRdisplays normal hepatic function, including protein pro-duction, glycogen and glucose synthesis and an intacturea cycle.

METHODS

Materials

Human recombinant 22 kilodalton GH (hGH) was pro-vided by Novo Nordisk (Bagsværd, Denmark) (Nor-ditropin 4 IU), while the human IGF-I was supplied byFugisawa Pharmaceutical Company (Osaka, Japan).

Animals

Experiments were approved by the Committee onAnimal Care and Use of the Jikei University School ofMedicine.

Primary hepatocyte culture andestablishment of the novelhepatocyte line

Young male adult SDRs were anesthetized with Nemb-utal (35 mg/kg) and blood was removed through the leftventricle (LV) using a 5-mL syringe. Blood in the liverwas removed with Hank’s solution by injecting this intothe LV and draining it from the right jugular vein. Theliver was then perfused with 0.1% trypsin and 0.02%ethylenediaminetetraacetic acid (EDTA) in Ca++-freephosphate buffered saline (PBS [(–]) via the LV. The liverwas then removed, sliced into small pieces with razorblades and dispersed in the abovementioned digestivesolution for 30 min at 37 °C. Cells were then dissociatedwith vigorous pipetting. In order to remove the digestiveenzymes and EDTA, single cells and small cell aggre-gates were washed several times with growth medium(GM: Dulbecco’s Modified Eagle Medium[DMEM]/Ham’s F 12 medium [containing 1:1] supplementedwith 10% FBS [Cellular Mate, Gunma, Japan], 50 U/mLpenicillin – 50 mg/mL streptomycin [Gibco, GrandIsland, NY, USA], 1% non-essential amino acids

[Gibco], and 2000 mg glucose per 100 mL medium)and then centrifuged (350 g for 3 min). Pellets of thedissociated liver cells were then resuspended with GMand approximately 3 ¥ 106 cells were cultured per60 mm dishes (Falcon, Franklin Lakes, NJ, USA) in GM.The GM was changed twice weekly. Most of the cellsdied; however, some small epithelial colonies enlargedduring the 3-month culture period. Two types of epi-thelial cells were observed in these colonies by phasecontrast microscopy; one was a small angular cell andthe other flattened. The remaining fibroblasts wereremoved by cell scraper (Narge Nunc International,Rochester, NY, USA) The novel cell line, SDRL-1, wasthen established. From the beginning of the culture,1 ng/mL leukemia inhibitory factor (LIF) or FGF-2 wasplaced into the GM to preserve the cells in a normalundifferentiated stage. A primary liver culture fromnormal SD rats was also established by the same methodas described above.

Cytological study

The SDRL-1 cell line was cultured with or without 22 K-hGH or IGF-I and was observed by phase contrastmicroscopy, transmission electron microscopy, scanningelectron microscopy and immunocytochemistryusing an anti-albumin or anti-alpha-fetoprotein (AFP)antibody.

Karyotype

A chromosome preparation was performed followingthe procedure of Rooney et al.9

Cell stimulation for RT-PCR andWestern blot

Cells were incubated with 10-6 M dexamethasone and10–8 M insulin for 48 h, and then were starved for 24 hbefore hGH treatment. The cells were treated with hGHfor various concentrations and time.

Treated cells were resuspended in lysis buffer (50 mMTris, 250 mM NaCl, 2 mM ethylene glycol tetraaceticacid (EGTA), 1 mM Na3VO4, 50 mg of PNPP(P-Nitrophenylphosphoric Acid Disodium Salt), 0.1%Triton X, and protease inhibitor (complete Mini; RocheApplied Science, Indianapolis, IN, USA, which wasdiluted according to the manufacturer’s instructions)and incubated with anti-JAK2 (Cat No; # 06-255,Upstate Biotechnology Inc., Lake Placid, NY, USA) or

Novel functional hepatocyte cell line

© 2010 The Authors165Human Cell © 2010 Japan Human Cell Society

STAT5b antibody (Cat No; # 9363, Upstate Biotechnol-ogy Inc.) for 3 h on 4 °C. Antibody dilutions were as permanufacturer’s instructions. The cell lysate was incu-bated with Protein A-Sepharose (Amasham PharmaciaBiotech, Uppsala, Sweden) for 2 h on 4 °C. Beads werethen washed and eluted by boiling in SDS sample buffer.Samples were then electrophoresed and membranesprobed with 4G10 (Upstate Biotechnology Inc.), orJAK2 (Cat No; # 06-255, Upstate Biotechnology Inc.) orSTAT5b (Cat No; # 9363, Upstate Biotechnology Inc.).

For reverse transcription-polymerase chain reaction(RT-PCR), the cells were treated by various concentra-tions of hGH for 24 h. The cells were homogenized inIsogen (Nippon Gene, Toyama-ken, Japan) and RNAprepared as per the manufacturer’s instructions. DNasetreated RNA (Turbo DNA-free kit [Applied Biosystems,Foster City, CA, USA]) was reverse transcribed usingSuperscript III (Invitrogen, Carlsbad, CA, USA), andused for PCR with IGF-I forward primers: AAGCCTA-CAAAGTCAGCTCG, IGF-I reverse primer: GGTCT-TGTTTCCTGCACTTC,10 b-actin forward primer:GATGACGATATCGCTGCGCTCG, b-actin reverseprimer: ACGGTTGGCCTTAGGGTTCAGAG.11

Secretion of IGF-I

Conditioned medium was collected after 3 days incu-bation with DMEM/F12 (1:1) without serum, supple-mented with 50 U/mL penicillin – 50 mg/mL streptomy-cin and 1% non-essential amino acids for IGF-Imeasurement. The concentration of IGF-I in the condi-tioned medium was measured by formic acid-acetoneextraction12 and RIA (EIKEN Chemical Co. Ltd, Tokyo,Japan) using rat IGF-I as a standard.

Production of albumin and AFP

The SDRL-1 cells were suspended in GM and rotated intriangular flasks (30 mL) at 80 rpm for 48 h until theymade spheroids. The spheroids in the GM were injectedinto a narrow glass tube containing a type I collagen-sponge (Stem Co. Ltd., Tokyo, Japan) that had beensoaked in the GM 48 h before use. After the cells grewfor 36 h in the presence or absence of various concen-trations of hGH, the conditioned media was removedand kept at -20 °C for assay of albumin and AFP bynephelometry (Dade Behring Co., Deerfield, IL, USA)and chemiluminesent immunoassay (Dainabot, Tokyo,Japan), respectively. The collagen-sponge was fixed witha solution of 1% paraformaldehyde-aqueous saturatedpicric acid solution (15% [v/v]) or 2.5% glutaraldehyde

in phosphate buffer and post-fixed with osmium tetrox-ide, then examined either by immunocytochemistry orelectron microscopy.

Glycogen and glucose synthesis

Glycogen and glucose synthesis was determined with orwithout 0.1 mM glucagon. For glycogen synthesis, thehepatocytes were washed three times with cold salineand immediately frozen in liquid nitrogen. They werethen digested in 0.2N NaOH and an aliquot wasremoved. The amount of glycogen and protein weredetermined by the Roehring and Allred13 and Lowrymethods,14 respectively. To determine glucose synthesis,the hepatocytes were incubated with or without 0.1 mM glucagon in Hanks medium that did not containglucose. The glucose released into the medium was thenestimated enzymatically with glucose oxidase.13

Measurement of GHR

Growth hormone receptor concentrations in the SDRL-1cells and in hepatocytes from SD rats were measured bythe homologous radioimmunoassay method15 usingrecombinant rat GH receptor (GHR)/Fc chimera (Cat;1211-GR-050, R&D Systems, Minneapolis, MN, USA)and anti-rat GHR antibody (Cat; AF1211, R&DSystems).

Urea cycle of SDRL-1

The urea cycle of the SDRL-1 cells was determined usinga slight modification of the method of Nagamori et al.16

Briefly, the cells were cultured with 5 mM NH4Cl usinga Rose circumfusion apparatus and the concentration ofNH3 and urea was measured in the condition mediumusing the methods of Okuda-Fujii17 and urease UV,18

respectively.

Statistics

Comparisons of the two groups only used unpairedStudents t-test. Multiple comparisons were undertakenby ANOVA with Tukeys multiple comparison test. Analy-sis of glucose synthesis was undertaken by two-wayANOVA. The values are shown as the mean � standarddeviation (SD).

RESULTS

Cytological analysis of SDRL-1 cell line

Three months after establishing the primary culture, thefibroblasts degenerated and decreased in number. In

M Ishikawa et al.


contrast, the hepatocytes increased and the SDRL-1 cellline was established with cells growing as a monolayerpavement-like sheet. The population doubling time ofthe SDRL-1 cell line was approximately 96 h and thesaturation density 107 000/cm2. After they reached con-fluence, meshwork-like canaliculi were observedbetween the cells and two types of cells were observed;one was small and angular and the other large and flat(Fig. 1a). Electron microscopy revealed that the SDRL-1cells had elongated mitochondria, a number of freeribosomes and poorly developed rough endoplasmicreticulum (r-ER) (Fig. 1b). Also, there were manytonofilaments in the cytoplasm that were orientedtowards the periphery of the cells. Treatment with 22 K-hGH or IGF-I led to an enlargement of the cisterna ofthe r-ER (Fig. 1c,d). Multivesicular bodies are frequentlypresent along with desmosomes. The number of

microvilli varied from a few on the surface of the large,flat cells to many on the surface of the smaller, moreangular cells. A long cilia-like structure was consistentlypresent in the center of the large flat cells (Fig. 1e).

Karyotype of SDRL-1 cell line

A histogram of the number of chromosomes in SDRL-1cells demonstrated that 88% of the SDRL-1 cells showeda normal diploid karyotype, with the distributionof chromosome number 42 (3), 43 (2), 44 (44), and45 (1).

JAK2 and STAT5 phosphorylation, IGF-ImRNA expression, and IGF-I production

JAK2 and STAT5b tyrosine-phosphorylation werestimulated by GH treatment in a concentration-

Figure 1 Cytological examination ofSDRL-1. Small dark and large lightcells are observed by light microscopy(bar, 100 mm) (a). Electron micro-graphs are shown in (b–e). (b) A welldeveloped Golgi apparatus (G), elon-gated mitochondria, poorly devel-oped rough endoplasmic reticulum(r-ER) and a number of microvilli areobserved (untreated [bar, 5 mm]).(c,d) After treatment with recombi-nant human growth hormone (hGH)(c) or IGF-I (d), a well developedGolgi apparatus (G) and enlarged r-ERare observed (bar, 1 mm). A scanningmicrograph illustration is shown (e).A large, long cilia is observed near thecenter of some cells and the numberof the microvilli are different betweenthe two cell types. Generally, the smallcells have significantly more microvillithan the larger ones (bar, 10 mm).



dependent manner (Fig. 2). IGF-I mRNA and IGF-Iproduction conditioned media were increased byGH treatment in a concentration-dependent manner(Fig. 3).

Albumin and AFP production of theSDRL-1 cell line

No AFP or albumin was released by monolayer SDRL-1cells. In contrast, after reconstruction of the liver in thetype-1 collagen sponge system, AFP and albumin weresecreted (Fig. 4a,b) and their production was stimulatedby hGH in a concentration-dependent manner(Fig. 4c,d).

Glycogen and glucose synthesis

Both glycogen storage and glucose release in condi-tioned media were observed in the SDRL-1 cell line.Following glucagon administration, the storage of gly-cogen was decreased while glucose release was increased(Fig. 5a,b).

Urea cycle of the SDRL-1 cell line

The concentration of NH3 decreased significantly ondays 20 and 30 while that of urea, the metabolite ofNH3, increased (Fig. 5c).

GHR concentration

Growth hormone receptor protein expression in theSDRL-1 cells was significantly higher than that presentin hepatocytes from primary cultures of SD rats (Fig. 6).

DISCUSSION

The development of several hepatoma cell lines has beenreported19–22 but there are only a few that are suitable tostudy GH function, and these require supra physiologi-cal sensitivity GH doses for effect. Here we reportestablishment of a hepatocyte line with physiologicalsensitivity to exogenous GH. Given that the body weightand serum levels of IGF-I in SDR increases markedlyafter treatment with hGH,8 we hypothesized that their

a

JAK 2-y

JAK 2

0 1 2 5 10 30

minutes

b

JAK 2-y

JAK 2

0 10-9 10-8 10-7

GH concentration

c

* P<0.05 vs GH 0 M

STAT5b-y

STAT5b

d

0 1 2 5 10 30

minutes

STAT5b-y

STAT5b

0 10-9 10-8 10-7

GH concentratione

f

* P<0.01 vs GH 0 M

Figure 2 Tyrosine phosphorylationof JAK2 and STAT5b by growthhormone (GH) treatment. Human GH(10–7 M) stimulated tyrosine phos-phorylation of JAK2 and STAT5b, andthe peaks were 5–10 min (a,d). 5 minhuman growth hormone (hGH)also stimulates the phosphorylationconcentration-dependently (b,c,e,f).Three samples were graphed in eachexperiment (mean � standard devia-tion [SD]).

M Ishikawa et al.


hepatocytes would likewise be sensitive to GH. It isknown that GH stimulates IGF-I production throughJAK2 and STAT5 phosphorylation.23,24 With our SDRL-1line, hGH stimulates JAK2 and STAT5 phosphorylation,together with IGF-I mRNA expression. The SDRL-1 cellline also released IGF-I following exposure to hGH in amanner that was similar to that noted in vivo. Moreover,the response of these cells to physiological concentra-tions of hGH was stronger than other hepatocyte cell

lines, such as HepG2.3 This increased sensitivity to GHcan be explained by a higher number of GHR on hepa-tocytes derived from SDR than what has been notedfrom cells derived from SD rats. The abundance of GHRwas observed not only on hepatocytes but also on otherorgans harvested from SDRs (data not shown). Theresult differs from previous reports. Maiter D showedthat hypophysectomy in female rats decreased liverGHR mRNA abundance by 30–35% compared with

a

IGF-I

β-actin

b

* P<0.05 vs GH 0M

# P<0.01 vs GH 0M

c

* P<0.05 vs GH 0 ng/mL

Figure 3 IGF-I mRNA expressionand secretion by human growthhormone (hGH) treatment. IGF-ImRNA expression was increasedby hGH treatment for 24 h (a,b).IGF-I secretion was stimulated byhGH treatment in a concentration-dependent manner after 72 h ofculture. Three samples were graphedin each experiment (mean � SD).

* P<0.0001 vs GH 0 ng/mL * P<0.01 vs GH 0 ng/mL

a

c

b

dFigure 4 Albumin and alpha-fetoprotein (AFP) secretion withSDRL-1 cell line. SDRL-1 cells werecultured in collagen sponge (a,b).The cells secreted albumin (a)and AFP (b) (bar, 30 mm). The secre-tions were stimulated by hGH treat-ment concentration-dependently (c:albumin; d: AFP). Three samples weregraphed in each experiment (mean �SD).



controls, and they were not reversed by T4 and corti-sone treatment.25 Butler et al. showed hepatic GHR waslower in dw/dw rat compared with Lewis rats.26 TheSDRs do not have bioactive GH but have 1–72 amino-acids of GH.5 The mutant GH cannot bind GHR,27 butthere is the possibility that mutant GH induced theincreased GHR expression, or some other strain-specificfactor may be involved.

Substantial albumin and AFP were also producedby the SDRL-1 cells, contrasting with other hepatoma

lines.19–22 These two proteins were stimulated in a dose-dependent manner by GH and IGF-I and as shown inFigure 4, the r-ER reflected the enhanced stimulation byenlarging following exposure to these two hormones.While the activities of the r-ER in this cell line have notbeen fully elucidated, it is speculated that enlargementof the r-ER is attributed to the production of manyproteins including IGF-I, albumin or AFP. The ureacycle was also investigated and its response resembledthat of the normal hepatocyte. Thus, these resultssuggest that this cell line may be useful not only for thestudy of GH function but also for the study of normalhepatocyte protein production.

The cell line has been cultured with LIF for morethan 3 years and the karyotype and function as deter-mined by IGF-I production has not changed. LIF,co-discovered in the conditioned medium from buffalorat liver cells,28 is a pleiotrophic molecule with multipleeffects on a broad range of cell lines. Its functions havebeen reported to substitute for differential inhibitoryactivities in the maintenance of totipotent embryonicstem cell lines29 or to in ML-1 and KG-1 cells.30 LIF’ssignaling system may also be involved in the expansionand differentiation of the liver stem cell compartment.31

From these reports, we expected LIF to inhibit the trans-formation of the hepatocytes during long term culture.Thus by using GM containing LIF, we succeeded in

a b

c

* P<0.0001 vs glucagon (-)

* P<0.0005 vs 10d

# P<0.0001 vs 10d

Figure 5 Glycogen storage, glucosesecretion, and urea cycle activity ofthe SDR-L1 cell line. Glycogen storagewas decreased by glucagon treatment(a). The glucose released in condi-tioned media was higher in the gluca-gons treatment group compared withthe control group (b). Analysis of ureacycle activity was performed as thecells incubated in the Rose’s circum-fusion apparatus in the presence of5 mM NH4Cl, and then the concentra-tions of NH3 and urea were measuredin the conditioned medium. NH3concentration in conditioned mediawas increased initially, but decreasedafter 15 days, while the urea concen-tration was increased (c).

Figure 6 Growth hormone receptor (HR) concentration ofSDRL-1. The GHR concentrations in the SDRL-1 cell line,found to be higher than that found in primary culture oflivers from Sprague-Dawley (SD) rats.

M Ishikawa et al.


culturing the SDRL-1 cell line for more than 3 yearswithout transformation. Other reports demonstratedthat the condition medium of cell lines, such as HepG2,was necessary to establish lines from normal livers.32

However, the use of LIF is convenient and more effectivethan other techniques used to inhibit transformation.

In the cytological study conducted on the SDRL-1 cellline, two types of cells were observed by light micros-copy. Both the small angular and the rounded, flattenedcells produced similar concentrations of IGF-I, albuminand AFP; additionally, no differences were noted in thekaryotypes (data not shown). Changes in the number ofmicrovilli were found as well as the presence of a longcilia-like structure in the flattened cells. Its role isunclear as well as any functional difference betweenthese two cell types.

Although the SDRL-1 cells do not secrete albuminand AFP in monolayer cultures; they did so on thecollagen type-I sponge. The results can be explained bythe fact that hepatic functions were maintained orenhanced once the organ was reconstructed in a formthat more resembled that found in vivo. These findingsare supported by Kono et al.33 who found that estab-lished hepatocyte cell lines maintained their functionbetter in a collagen gel sandwich culture system.

In summary, we established a novel hepatocyte cellline with almost normal function using tissue obtainedfrom the SDR, which serves as a good in vitro model forstudying GH, normal hepatic function and target organinteractions.

ACKNOWLEDGMENTS

The authors thank Dr M. J. Waters (Institute for Molecu-lar Biology, University of Queensland, Qld, Australia)for his help during the preparation of this paper.

REFERENCES

1 Brameld JM, Weller PA, Saunders JC, Buttery PJ,Gilmour RS. Hormonal control of insulin-like growthfactor-I and growth hormone receptor mRNA expressionby porcine hepatocytes in culture. J Endocrinol 1995;146: 239–45.

2 Scott CD, Martin JL, Baxter RC. Rat hepatocyte insulin-like growth factor I and binding protein: effect of growthhormone in vitro and in vivo. Endcrinology 1985; 39:1102–7.

3 Conover CA, Baker BK, Bale LK, Clarkson JT, Liu F,Hintz RL. Human hepatoma cells synthesize and secrete

insulin-like growth factor Ia prohormone under growthhormone control. Regul Pept 1993; 48: 1–8.

4 Okuma S. Study of growth hormone in spontaneousdwarf rat. Nippon Naibunpi Gakkai Zasshi 1984; 60:1005–14.

5 Takeuchi T, Suzuki H, Sakurai S, Nogami H, Okuma S,Ishikawa H. Molecular mechanism of growth hormone(GH) deficiency in the spontaneous dwarf rat: detectionof abnormal splicing of GH messenger ribonucleic acidby the polymerase chain reaction. Endcrinology 1990;126: 31–8.

6 Okuma S, Kawashima S. Spontaneous dwarf rat. ExpAnim 1980; 29: 301–4.

7 Nogami H, Watanabe T, Takeuchi T. Effect of growthhormone (GH) on the promotion of body weight gain inthe spontaneous dwarf rat: a novel experimental modelfor isolated GH deficiency. Horm Metab Res 1992; 24:300–1.

8 Ishikawa M, Tachibana T, Kamioka T, Horikawa R, Kat-sumata N, Tanaka T. Comparison of the somatogenicaction of 20 KDa – and 22 KDa – human growth hor-mones on spontaneous dwarf rats. Growth Horm IGF Res2000; 10: 199–206.

9 Rooney DE, Czepulkowski BH. Human Cytogenetics: APractical Approach. Oxford: IRL Press at Oxford Univer-sity Press, 1986.

10 Zhang J, Whitehead RE, Underwood LE. Effect of fastingon insulin-like growth factor (IGF)-IA and IGF-IB mes-senger ribonucleic acids and prehormones in rat liver.Endocrinology 1997; 138: 3112–18.

11 Hoshida S, Nishida M, Yamashita N et al. Hemeoxygenase-1 expression and its relation to oxidativestress during primary culture of cardiomyocytes. J MolCell Cardiol 1996; 28: 1845–55.

12 Bowsher RR, Lee W, Apathy JM et al. Measurement ofinsulin-like growth factor-II in physiological fluids andtissues. I An improved extraction procedure and radio-immunoassay for human and rat fluids. Endcrinology1991; 128: 805–14.

13 Roehring KL, Allres JB. Direct enzymatic procedure fordetermination of liver glycogen. Anal Biochem 1974; 58:414–21.

14 Lowry OH, Rosnbrough NJ, Farr AL, Randall RJ. Proteinmeasurement with the Folin phenol reagent. J Biol Chem1951; 193: 265–71.

15 Camarillo IG, Throdarson G, Ilkbahar YN, Talamantes F.Development of a homologous radioimmunoassay formouse growth hormone receptor. Endcrinology 1998;139: 3585–9.

16 Nagamori S, Hasumura S, Matsuura T, Aizaki H, KawadaM. Developments in bioartificial liver research: concepts,performance, and applications. J Gastroenterol 2000; 35:493–503.

17 Okuda H, Fujii S. Direct colorimetry of ammonia in theblood. Saishin Igaku 1996; 21: 622–7.



18 Morishita Y, Nakane K, Fukatsu T et al. Kinetic assay ofserum and urine for urea with use of urease and leucinedehydrogenase. Clin Chem 1997; 43: 1932–6.

19 Aden DP, Fogel A, Plotkin S, Damjanov I, Knowles BB.Controlled synthesis of HbsAg in a differentiated humanliver carcinoma-derived cell line. Nature 1979; 282:615–16.

20 Dippold WG, Dienes HP, Knuth A et al. HepatocellularCarcinoma after thorotrast exposure: establishment of anew cell line (Mz-Hep-1). Hepatology 1985; 5: 1112–19.

21 Stevenson D, Lin JH, Tong MJ, Marshall GJ. Character-istics of a cell line (Tong/HCC) established from a humanhepatocellular carcinoma. Hepatology 1987; 7: 1291–5.

22 Sing GK, Pace R, Prior S et al. Establishment of a cell linefrom a hepatocellular carcinoma from a patient withhemochromatosis. Hepatology 1994; 20: 74–81.

23 Postel-Vinay MC, Finidori J. Growth hormone receptor:structure and signal transduction. Eur J Endocrinol 1995;133: 654–9.

24 Lanning NJ, Carter-Su C. Recent advances in growthhormone signaling. Rev Endocr Metab Disord 2006; 7:225–35.

25 Maiter D, Walker JL, Adam E et al. Differential regulationby growth hormone (GH) of insulin-like growth factor Iand GH receptor/binding protein gene expression in ratliver. Endocrinology 1992; 130: 3257–64.

26 Butler AA, Funk B, Breier BH, LeRoith D, Roberts CT Jr,Gluckman PD. Growth hormone (GH) status regulatesGH receptor and GH binding protein mRNA in a

tissue- and transcript-specific manner but has no effecton insulin-like growth factor-I receptor mRNA in the rat.Mol Cell Endocrinol 1996; 116: 181–9.

27 Walsh ST, Sylvester JE, Kossiakoff AA. The high- andlow-affinity receptor binding sites of growth hormoneare allosterically coupled. Proc Natl Acad Sci USA 2004;101: 17078–83.

28 Smith AG, Heath JK, Donaldson DD et al. Inhibition ofpluripotential embryonic stem cell differentiation bypurified polypeptides. Nature 1998; 336: 688–90.

29 Williams RL, Hilton DJ, Pease S et al. Myeloid leukemiainhibitory factor maintains the developmental potentialof embryonic stem cells. Nature 1998; 336: 684–7.

30 Samal BB, Arakawa T, Boone TC et al. High level expres-sion of human leukemia inhibitory factor (LIF) from asynthetic gene in Escherichia coli and the physical andbiological characterization of the protein. Biochim BiophysActa 1995; 1260: 27–34.

31 Omori N, Evarts RP, Omori M, Hu Z, Marsden ER,Thorgeirsson SS. Expression of leukemia inhibitoryfactor and its receptor during liver regeneration in theadult rat. Lab Invest 1996; 75: 15–24.

32 Roberts EA, Letarte M, Squire J, Yang S. Characterizationof human hepatocyte lines derived from normal livertissue. Hepatology 1994; 19: 1390–9.

33 Kono Y, Yang S, Letarte M, Roberts EA. Establishment ofa human hepatocyte line derived from primary culture ina collagen gel sandwich culture system. Exp Cell Res1995; 221: 478–85.

M Ishikawa et al.


Documents

Novel functional hepatocyte cell line derived from spontaneous dwarf rat: Model of growth hormone function in vitro