interleukin-15 increases hepatic regenerative activity

www.elsevier.com/locate/jhep

Journal of Hepatology 45 (2006) 410–418

Interleukin-15 increases hepatic regenerative activityq

Ayako Suzuki1, Shannon McCall2, Steve S. Choi1, Jason K. Sicklick1,5, Jiawen Huang1,Yi Qi1, Marzena Zdanowicz1, Terese Camp1, Yin-Xiong Li1,3,4, Anna Mae Diehl1,*

1Division of Gastroenterology and Department of Medicine, Duke University Medical Center, Durham, NC, USA2Department of Pathology, Duke University Medical Center, Durham, NC, USA

3Department of Cell Biology, Duke University Medical Center, Durham, NC, USA4Department of Pediatrics, Duke University Medical Center, Durham, NC, USA

5Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, MD, USA

See Editorial, pages 347–349

Background/Aims: Interleukin-15 (IL-15) is expressed in many organs. It generally inhibits apoptosis and increases cel-

lular proliferation and differentiation. However, IL-15’s roles in liver are unknown. We aimed to determine if IL-15 influ-ences hepatic integrity and regenerative activity.

Methods: Expression of IL-15 and its receptors was evaluated in several liver injury models, primary hepatocytes, and

two liver cell lines. Effects of IL-15 on viability, proliferation, and apoptosis were assessed in cultured liver cells, and also in

the livers of healthy mice.

Results: IL-15 and its receptors are expressed constitutively in healthy livers, and ligand expression is induced in injured

livers. Cultured primary hepatocytes and liver cell lines express IL-15 and its receptors. Administration of IL-15 has min-

imal effects on cultured liver cells, but significantly up-regulates oval cell accumulation, cyclin mRNA expression, and

mature hepatocyte replication in healthy mice. These effects are associated with focal hepatic inflammation and increasedexpression of TNF-a and IFN-c, but not with increased cell death or aminotransferase release.

Conclusions: IL-15 expression increases during liver injury and IL-15 treatment induces a wound healing-type response

in healthy adult mice. These findings suggest that IL-15 may contribute to regenerative activity in damaged liver.

� 2006 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.

Keywords: Apoptosis; Cytokine; Liver; Mitosis; Oval cell

1. Introduction

Interleukin-15 (IL-15) is a constitutively expressedcytokine in both immune and non-immune cells [1]. Itis a survival factor for lymphocytes, NK, and NKT cells

0168-8278/$32.00 � 2006 European Association for the Study of the Liver.

doi:10.1016/j.jhep.2006.04.008

Received 6 December 2005; received in revised form 17 March 2006;

accepted 9 April 2006; available online 22 May 2006q The authors who have taken part in the research of this paper have

no relationship with the manufactures of the drug involved either inthe past or present. The authors state that they did not receive fundingfrom the manufactures to carry out their research. The authorsreceived funding from NIH which enabled them to carry out theirstudy.

* Corresponding author. Tel.: +1 919 684 4173; fax: +1 919 6844183.

E-mail address: [email protected] (A.M. Diehl).

[1–4], and also increases cellular proliferation [5–7],inhibits apoptosis [6,8–11], and promotes cellular differ-entiation [12,13] in various non-immunological cells. Ininflammatory conditions, such as rheumatoid arthritisand psoriasis, IL-15 is detected in inflamed tissues[4,14–16]. Serum levels of IL-15 are directly associatedwith severity of liver injury in patients with chronic Hep-atitis C infection [17].

Whether IL-15 participates in damage or repair ofinjured livers is uncertain. IL-15 promotes the expansionof NK cells, which can be directly cytotoxic to liver cells[1–3]. However, given its generally trophic actions inmany cell types [5–7,12,13] and evidence that IL-15dose-dependently inhibited hepatic apoptosis and devel-opment of hepatic failure in a mouse model of Fas-

Published by Elsevier B.V. All rights reserved.

mailto:[email protected]

A. Suzuki et al. / Journal of Hepatology 45 (2006) 410–418 411

induced liver injury [10], it is conceivable that IL-15’sprimary action in injured livers is to promoteregeneration.

IL-15 is expressed in the liver [18], but whether liverepithelial cells generate IL-15 or express any of the threedifferent IL-15 receptor subunits, IL-15 receptor a(IL-15Ra), IL-2/IL-15Rb (IL-15Rb) and IL-2/IL-15Rcc (IL-15Rc) [1], is unknown. Although transcriptsfor IL-15Ra [19] have been detected in extracts of wholeliver tissue, co-expression of all three receptor subunitsis obligatory for active IL-15 signaling in target cells[1]. Hence, it is uncertain if liver cells themselves canrespond to IL-15.

In intestinal and skin epithelial cells IL-15 inducesproliferation [5–7]. Therefore, we hypothesized that liverepithelial cells might also be IL-15 targets and that inju-ry-related increases in hepatic IL-15 may promote liverrepair by directly inducing expansion of mature hepato-cytes or hepatic progenitor populations. To test thishypothesis, we analyzed several models of liver injury,characterized IL-15 and its receptors in primary hepato-cytes and hepatic cell lines, and studied the effects ofin vitro and in vivo administration of IL-15 on liver cellproliferation and viability.

Herein, we demonstrate that hepatic IL-15 is inducedduring liver injury and that IL-15 and its receptors arepresent in mature hepatocytes and hepatic progenitors.Although IL-15 has no effect on the viability or prolifer-ative activity of cultured liver cells, IL-15 administrationto healthy adult mice profoundly up-regulates hepato-cyte mitosis and progenitor accumulation, and some-what increases hepatic lobular inflammation, withoutinducing significant cell death. These findings supportthe concept that IL-15 has hepatotrophic actions, butsuggest that its regenerative effects are mediatedindirectly.

2. Methods

2.1. Experimental designs

Adult (aged 10–12 weeks) male, lean C57BL/6 mice and ob/ob micewere obtained from Jackson Laboratory (Bar Harbor, ME). To assessthe effects of chronic liver injury and repair on hepatic expression levelsof IL-15 ligand and its receptors, liver RNA was pooled from three micein each of the following three models of chronic fatty liver disease: ob/obmice, lean mice fed with a methionine–choline deficient diet (for 4 weeks)[20], or lean mice fed an ethanol-supplemented diet (for 9 weeks)[21].Effects of acute liver injury and repair were evaluated by assessingIL-15 and IL-15 receptor expression in RNA pooled from 3 lean mice48 h after 70% partial hepatectomy [22]. Analysis of liver RNA pooledfrom three lean mice on standard chow diet was used as a control.

Another group of adult (aged 10 weeks) male C57BL/6 mice wereadministered either 10 lg/day of recombinant human IL-15 (Amgen,Seattle, WA) in 0.5% bovine serum albumin (BSA) diluted in PBS(N = 11) or 0.5% BSA–PBS vehicle (N = 11) by daily intraperitonealinjection for 3 (N = 5) or 7 (N = 6) days. Following the treatments,mice were weighed and anesthetized. Livers were harvested, weighed,and processed: snap frozen in liquid nitrogen for subsequent RNAanalysis and fixed in formalin and embedded in paraffin for histologi-

cal evaluation. All animal experiments fulfilled NIH and approvedDuke University Institutional Animal Care and Use Committee Proto-cols for humane care of laboratory animals.

2.2. Histological evaluations

Formalin-fixed, paraffin-embedded liver sections were stained withhemotoxylin and eosin (H&E) and then assessed for tissue architec-ture, lobular inflammation, and hepatocyte mitoses. Lobular inflam-matory foci were defined by their location in the lobule (lack ofassociation with portal triads or terminal hepatic venules) and by theaccumulation of multiple intra-acinar neutrophils and/or lymphocytesassociated with injured or dead hepatocytes, regardless of total focussize [23,24]. Inflammatory foci were counted and expressed as the num-ber per 10 high power fields (hpf). The hepatocyte mitotic index wasderived by counting the numbers of mitotic figures in 500 hepato-cytes/tissue section. Results were expressed as a percentage.

To assess proliferative activity, apoptotic activity, and oval cells,we performed immunohistochemistry for proliferative cell nuclear anti-gen (PCNA) [25,26], Caspase-3 [27,28], terminal deoxyribonucleotidyltransferase (TdT)-mediated dUTP-digoxigenin nick end labeling(TUNEL) [29], and polyclonal cytokeratin [30] as previously describedby using antibodies shown in Table 1. For TUNEL staining, TdT reac-tion mixture (Biotin-16-dUTP with TdT, Roche Diagnostic), streptavi-din–HRP (1:500, Vector Laboratories), and DAB chromagen (DakoCytomation) were employed. Counting of positive cells for PCNA,Caspase-3, or TUNEL was similar to our method for quantitatingmitotic nuclei. For oval cell counting, pancytokeratin-positive cellswithin the interlobular bile ducts were excluded from counts, accordingto previously reported methods [30]. The oval cell counts wereexpressed as the number of pancytokeratin positive cells per 10 hpf.

2.3. Culture of murine cell lines and primary hepatocytes

A well-differentiated murine hepatocyte cell line (AML-12, Hep)was obtained from American Type Culture Collection (Manassas,VA) and cultured according to their instructions [31]. A murine hepaticprogenitor cell line (OV) was kindly provided by Dr. B.E. Petersen(University of Florida; Gainesville, FL) and cultured as previouslydescribed [32,33]. Primary hepatocytes were isolated from healthyadult male mice (C57BL/6); cells from two mice were pooled and cul-tured as reported before [34]. The viability of primary hepatocytes afterthe isolation was over 90%.

2.4. RNA isolation, semi-quantitative RT-PCR, real-time

RT-PCR, and ribonuclease protection assay

Total RNA was extracted from frozen liver samples and the celllines using RNeasy Mini kits (Qiagen, Valencia, CA) followed byRNase-free DNase I treatment (Qiagen). cDNA was then synthesizedfrom 1 lg of total RNA using random primers (Invitrogen, Carlsbad,CA) and a SuperScriptII kit (Invitrogen).

For semi-quantitative RT-PCR, cDNA was amplified in a reactionmixture (50 lL) containing 1· PCR buffer, 3 mM MgCl2, 0.4 mMdNTP, 0.4 mM primers (sense and antisense), 0.1 U/lL platinumTaq DNA polymerase (Invitrogen), and 5% of the first-strand reaction.The primers employed are summarized in Table 2 [35,36].The sampleswere amplified in an iCycler iQ (Bio-Rad, Hercules, CA) for 40 cycles.Each cycle consisted of denaturation at 94 �C for 15 s, annealing at 60–61 �C for 21 s, and elongation at 72 �C for 35 s. The 40 cycles were pre-ceded by an initial denaturation at 94 �C for 3 min, and were followedby a final extension at 72 �C for 2 min. Aliquots of PCR products wereelectrophoresed on 1.5% agarose gel buffered with 0.5· TBE.

For real-time RT-PCR, triplicate samples of the 5% of the first-strand reactions were amplified using iQ-SYBR Green Supermix(Bio-Rad, Hercules, CA), an iCycler iQ Real-Time Detection System(Bio-Rad, Hercules, CA), and 400 nM of primers for IL-15 and theb-glucuronidase (Gus) housekeeping gene.

To evaluate Cyclin, TNF-a, and IFN-c gene expression in the liverwith and without IL-15 treatment, we performed ribonuclease protec-tion assays as previously described [37].

Table 1

Antibodies for immunoblot, immunofluorescent, and immunohistochemistry analyses

Antibodies Dilution Source

Primary antibodies

Anti-IL-15 1:200 Santa Cruz Biotechnology, H-114Anti-IL-15Ra [35] 1:200 Santa Cruz Biotechnology, N-19Anti-IL-2Rb/IL-15Rb [45] 1:200 Santa Cruz Biotechnology, M-20Anti-IL-2Rc/IL-15Rc [5] 1:200 Santa Cruz Biotechnology, M-20Anti-b-actin 1:200 Santa Cruz Biotechnology, H-300Anti-PCNA 1:2000 Santa Cruz Biotechnology, PC 10Anti-cleaved Caspase-3 1:100 Cell Signaling Technology, Asp 175Polyclonal anti-human callus cytokeratin 1:200 Dako Cytomation

Secondary antibodies

Donkey anti-rabbit peroxidase 1:10,000 Amersham BioscienceDonkey anti-goat peroxidase 1:10,000 Jackson ImmunoresearchGoat anti-rabbit Alexa 488 1:100 Molecular ProbesDonkey anti-goat rhodamine 1:100 Jackson ImmunoresearchBiotinylated horse anti-mouse IgG 1:300 Vector LaboratoriesBiotinylated goat anti-rabbit IgG 1:300 Vector LaboratoriesAnti-rabbit Envision [30] – Dako Cytomation

412 A. Suzuki et al. / Journal of Hepatology 45 (2006) 410–418

2.5. Immunoblot

Total cell lysates were prepared by RIPA lysis buffer (Santa CruzBiotechnology, Santa Cruz, CA) containing protease inhibitors at4 �C. After quantification, equal amount of the proteins (35 lg/lane)was separated by polyacrylamide gel electrophoresis and electropho-retically transferred to Immobilon-P membranes (Millipore, Bedford,MA). Membranes were then blocked and exposed to the antibodiesshown in Table 1. The antigens were then demonstrated by enhancedchemiluminescence (Amersham Bioscience, Piscataway, NJ).

2.6. Immunofluorescent staining of cells

The cell lines were cultured overnight using the Lab-Tek� ChamberSlide� System (Nalge Nunc International Corp., Naperville, IL). Afterthe fixation with 4% paraformaldehyde, the cells were blocked andexposed to the antibodies shown in Table 1. After incubating with DAPI,the cells were evaluated using a fluorescent microscope. Negative con-trols were performed by omitting the primary antibodies from theprotocol.

2.7. Cell viability, proliferation, and apoptosis assays

The cell lines were plated on 96-well plates (5000 cells per 7 mm-di-ameter well, 130 cells/mm3). After overnight incubation, serum wasremoved from the medium for 48 h and then cells were treated with

Table 2

RT-PCR primers for analysis

Gene Direction Sequence

IL-15 Forward ATGTGAGGAGCTGGReverse AGCTTAGTTTGCCCA

IL-15Ra [35] Forward AACATCCACCCTGATReverse GTTTCCATGGTTTCC

IL-15Rb [35] Forward GTCGACGCTCCTCTCReverse GGATCCCAGAAGAC

IL-15Rc [35] Forward GTCGACAGAGCAAGReverse GGATCCTGGGATCA

b-Actin [35] Forward GTGGGGCGCCCCAGReverse CTCCTTAATGTCACG

Gus [36] Forward GCAGTTGTGTGGGTReverse GGGTCAGTGTGTTG

vehicle control or recombinant human IL-15 (0.01–10 ng/mL) for24 h. The primary hepatocytes were plated on collagen I coated 96-wellplates (5000 cells per 7 mm-diameter well, 130 cells/mm3). After over-night incubation, the cells were removed from serum-containing medi-um and cultured in DMEM containing EGF (10 ng/ml), penicillin(50 U/ml), streptomycin (50 lg/ml), L-glutamine (2 mM), and 1%ITS and treated with vehicle control or human recombinant IL-15(0.1–10 ng/ml) for 40 h. To assess the effect of inhibiting intrinsicIL-15, we also used a neutralizing anti-mouse IL-15 antibody(AF447, R& D system, Minneapolis, MN) on primary hepatocytes cul-tured with the same conditions. Cell viability was measured with theCell Counting Kit-8 (Dojindo Molecular Technologies, Gaithersburg,MD) [38] according to the manufacturer’s instructions. Proliferativeand apoptotic activity were assayed in parallel experiments using theCell Proliferation ELIZA, BrdU (chemiluminescence) kit (Roche, Indi-anapolis, IN) and the Apo-ONE Homogeneous Caspase 3/7 ApoptosisAssay (Promega, Madison, WI, USA) [39].

2.8. Statistical analysis

Data are expressed as means ± standard error of the mean (SEM).For the cell viability, proliferation, and apoptosis assays, the readingsat each concentration were normalized to the appropriate controls andcompared using single sample t-tests. Liver weights-to-body weightratios, histological scores, and quantification of positively stained cellswere compared using the t-test. Significance was accepted at the 5%level. The a-levels were adjusted for the number of comparisons.

Product size (bp)

AGGAGA 189GCAGATGAGTGT 525ACCTCAAAGCTGTAGTGGCTACCATA 506GTCTACGGGCCTCAAATCCCAACACCATGTTGAAACTA 506CAAGATTCTGTAGGTTGCACCA 506CACGATTTC

GAATGG 142TTGATGG


3. Results

3.1. Expression of hepatic IL-15 mRNA is induced during

regenerative responses to acute and chronic liver injury

Because levels of IL-15 in sera increase during liverinjury [17], we evaluated the effects of liver injury andregeneration on hepatic expression of IL-15 and itsreceptors. RT-PCR analysis was used to compare thehepatic expression of these genes during the regenerativeresponse to acute liver injury (i.e., partial hepatectomy)and during several forms of chronic liver injury (i.e., fat-ty liver induced by methionine–choline deficient diets,leptin deficiency, or alcohol consumption). Comparedwith the livers of control mice, hepatic IL-15 mRNAwas induced in acutely regenerating livers, as well as inall chronic liver injury models (Fig. 1). In contrast, theexpression of hepatic IL-15R subunits was similaramongst the samples.

3.2. Hepatocyte and oval cell lines express IL-15 and IL-

15R subunits

Increased proliferative activity of mature hepatocytesdrives liver regeneration after partial hepatectomy [40],while hepatic progenitors play a major role in replacingdead hepatocytes in chronic fatty liver disease [21]. Inorder to determine if either cell type might produceIL-15 or be a direct target for IL-15 actions, we evaluat-ed mature hepatocytes and hepatic epithelial progenitors(i.e., oval cells, OV) for expression of IL-15 and itsreceptors. Primary hepatocytes harvested from healthyadult mice (data not shown), a mature murine hepato-cyte line (AML-12 cells (Hep)) and a murine oval cell

Fig. 1. Hepatic IL-15 induction during acute liver regeneration and

chronic liver injury. Murine models of repair from acute liver injury

(partial hepatectomy) or chronic liver injury (left to right: fatty liver due

to methionine–choline deficient diets, leptin-deficiency, or chronic ethanol

feeding) were evaluated. Total RNA was isolated from the livers of

control and injured mice. Agarose gel electrophoresis of the RT-PCR

amplicons demonstrates expression of IL-15, IL-15Ra, IL-15Rb, and

IL-15Rc, as well as the b-actin housekeeping gene.

line (OV) expressed mRNA for IL-15 and the IL-15R

subunits (Fig. 2A). Semi-quantitative comparison ofmRNA expression in the lines demonstrated that expres-sion of IL-15 and IL-15Rb was higher in OV whileexpression of IL-15Rc was higher in Hep. In real-timeRT-PCR analysis, expression of IL-15 in OV was2.5-fold higher than in Hep (Fig. 2B).

Despite differential mRNA expression of IL-15 andits receptors, immunoblots using equal amounts of totalcellular protein demonstrated relatively equivalentamounts of IL-15R subunits in Hep and OV (Fig. 2C).Immunofluorescent staining of Hep (Fig. 2D–G) andOV (Fig. 2H–K) for IL-15 ligand (Fig. 2D and H),IL-15Ra (Fig. 2E and I), IL-15Rb (Fig. 2F and J),and IL-15Rc (Fig. 2G and K) demonstrated cytoplas-mic, nuclear, and/or membrane localization of IL-15ligand and its receptor subunits. Thus, both maturehepatocytes and hepatic progenitors express IL-15 andall of the IL-15 receptor subunits that are necessaryfor IL-15-initiated signaling.

3.3. In vitro administration of IL-15 had minimal effectsupon cultured liver cells

Since IL-15 increases in injured livers and bothmature hepatocytes and hepatic progenitors expressIL-15 receptors, it is conceivable that injury-relatedincreases in IL-15 might regulate hepatic regeneration.To evaluate this possibility, we determined the effectsof exogenous IL-15 on the viability, proliferative activ-ity, and apoptotic activity of Hep, OV, and primaryhepatocytes. Treatment with recombinant IL-15 didnot influence any of these parameters in primary hepa-tocytes (data not shown) or the mature Hep line(Fig. 3A–C). IL-15 also had no effect on the overallviability or apoptotic activity of the OV line (Fig. 3Aand C), but it tended to increase the proliferative activ-ity of these hepatic progenitors slightly (13–17% greaterBrdU incorporation than vehicle treated controls,Fig. 3B). Since our immunocytochemistry findings sug-gested that liver cells produce IL-15, we also culturedprimary hepatocytes in various concentrations of neu-tralizing IL-15 antibody (0.1–10 lg/ml) to determinehow blocking endogenous IL-15 might influence livercells. No consistent effects on proliferative activity orviability were observed.

3.4. In vivo administration of IL-15 increased oval cell

accumulation, mature hepatocyte proliferation and

hepatic inflammation

Although our studies of cultured liver cells suggestthat IL-15 is unlikely to have major, direct mitogenicactions on either hepatocytes or their progenitors, thepossibility remains that IL-15 might promote liver repairindirectly via its effects on other cells that produce

Fig. 2. Hepatocyte and oval cell lines express IL-15 and IL-15R subunits. (A) RNA was isolated from a hepatocyte (Hep) and an oval cell (OV) line.

Expression of IL-15 and its receptor subunits was evaluated using RT-PCR analysis. Representative agarose gel electrophoresis of RT-PCR products

from Hep and OV demonstrated expression of IL-15, IL-15Ra, IL-15Rb, and IL-15Rc, as well as the b-actin housekeeping gene. (B) Real-time RT-PCR

analysis was performed to compare expression of IL-15 in the two cell lines. (C) Total protein lysate was isolated from the cell lines, quantified, and equal

amounts of protein were electrophoresed on a polyacrylamide gel. IL-15Ra, IL-15Rb, and IL-15Rc subunits were subsequently detected. Similar to other

type of cells [35,46], both liver cell lines expressed multiple isoforms of IL-15Ra including wild-type, splice variants (D4, D2, D3,4), as well as glycosylated

isoform (i.e., 54–57 kDa). Immunofluorescence microscopy was used to study the Hep (D–G) and OV (H–K) cell lines. Representative staining for IL-15

(D and H), IL-15Ra (E,I), IL-15Rb (F and J), and IL-15Rc (G and K) is displayed. For each antibody, negative controls were performed by omitting the

primary antibody from the staining protocol as displayed.


growth-regulatory cytokines. To evaluate this possibili-ty, we treated healthy adult mice with recombinantIL-15 or vehicle and examined the effects on hepatocyteturnover and hepatic cytokine production. Liver histol-ogy and immunohistochemistry, liver weight-to-bodyweight (LW/BW) ratios, serum aminotransferases, andhepatic cytokine mRNA expression were evaluated 3and/or 7 days after treatment initiation.

After three days of IL-15 treatment, there was abouta 30% increase in the number of periportal cells thatstained positive for polyclonal cytokeratin, a well-ac-cepted marker for OV (P < 0.01 for IL-15-treated vs.

control) (Fig. 4A–C). This OV accumulation was notassociated with an appreciable increase in apoptosis orproliferation of mature hepatocytes, hepatic inflamma-tory foci, or serum AST or ALT (data not shown).

After 7 days, the number of periportal cells thatstained positive for polyclonal cytokeratin was about50% greater in the IL-15-treated mice than controls(84 vs. 125/10 hpf in control vs. IL-15-treated, respec-tively, P < 0.01, Fig. 5G and H). At this time point,IL-15-treated mice also had more hepatocyte mitoses(P < 0.03 for IL-15 group vs. controls, Fig. 4C andD). This increase in hepatocyte proliferative activity

Fig. 3. IL-15 administration caused minimal effects in vitro. The

hepatocyte (Hep) and oval cell (OV) cell lines were treated with

recombinant human IL-15 in a dose-dependent fashion. (A) Cell viability

was assessed by tetrazolium salt metabolism. (B) In parallel cultures,

chemiluminescent BrdU incorporation was used to assess cellular

proliferative activity (P = 0.033 for 5 ng/ml and P = 0.25 for 10 ng/ml).

(C) Similarly, apoptotic activity was evaluated by fluorometric assess-

ment of Caspase-3/7 activity. For all experiments, the values in the

treatment groups were normalized to those of control cultures. Data are

presented means ± SEM.

Fig. 4. IL-15 administration increased oval cells around periportal areas

after 3 days. After 3-day administration of recombinant IL-15 (IL-15) or

control vehicle (Control), oval cells counted as polyclonal cytokeratin-

positive cells were significantly increased in IL-15-treated livers (B)

compared to control livers (A). The scores of polyclonal CK-positive cells

(black arrow) after excluding the interlobular bile duct cells (white

arrows) are displayed as the numbers per 10 hpf (*P < 0.007) (C).


was confirmed by an increase in the number of PCNA-positive hepatocytes (0.4% in controls vs. 1.4% in IL-15group, P = 0.06, Fig. 5E and F), and induction of Cyc-lins A1, A2, B1, D1, and D2 mRNA in IL-15-treated liv-ers (Fig. 6). Transcripts of TNF-a and IFN-c, potentregulators of both proliferation and viability in hepato-cytes [31,41–43], were also induced in the IL-15-treatedlivers (Fig. 6).

Interestingly, in the IL-15 group, some hepatocytesundergoing mitosis also stained positive for Caspase-3or TUNEL at day 7, suggesting that some mitotic hepa-tocytes were simultaneously undergoing apoptosis(Fig. 5I and K). Consistent with this finding, mice thatwere treated with IL-15 for 7 days had slightly moreCaspase-3-positive cells than controls (0% in controlsvs. 0.13% in IL-15 group, P < 0.02, Fig. 5I and J).

However, as at 3 days, at 7 days, TUNEL stainingwas not significantly different between the two groups(P > 0.10, Fig. 5K–L). Also, although the IL-15-grouphad more hepatic inflammatory foci than controls after7 days of treatment (0.4 in controls vs. 2.9/10 hpf inIL-15-treated mice, P < 0.01, Fig. 5A and B), serum levelsof aspartate aminotransferase (AST) and alanine amino-transferase (ALT) remained similar in the two groups.

Finally, despite clear evidence for IL-15-relatedincreases in hepatic progenitor numbers and hepatocyteproliferative activity, the LW/BW ratios were virtuallyidentical in the two groups at both 3 days (Control,4.9 ± 0.2% vs. IL-15, 5.0 ± 0.2%, P > 0.05) and 7 days(Control, 4.3 ± 0.2% vs. IL-15, 4.6 ± 0.1%, P > 0.05).Taken together, these findings suggest that IL-15 mobi-lized hepatic progenitors and increased proliferation ofmature hepatocytes, but this growth was counterbal-anced by subsequent increases in hepatocyte death,resulting in tissue turnover without a net gain in livermass.

4. Discussion

In this study we demonstrated that healthy liversexpress mRNA for IL-15 and all three receptor sub-units, and that hepatic expression of IL-15 is inducedas the liver attempts to recover from either acute orchronic liver injury. In vitro studies of primary hepato-cytes and cell lines of mature hepatocytes and hepaticprogenitors showed that both mature hepatocytesand their progenitors express IL-15 and its receptors.However, supplemental IL-15 had little, if any, effect

Fig. 5. IL-15 administration induced profound effects in healthy livers

after 7 days. Healthy, wild-type mice underwent daily intraperitoneal

injection with recombinant IL-15 (IL-15) or control vehicle (control) for 1

week. Livers were harvested and histologically evaluated. Representative

photomicrographs of IL-15-treated livers are displayed at 40· magnifica-

tion (A, C, E, G, I, and K). The results of histologic scoring are presented as

means ± SEM (B, D, F, H, J, and L). Following H&E staining, the number

of inflammatory foci (A, arrow) was quantified per 10 hpf (B, *P < 0.004).

In the same sections, mitotic nuclei (C, black arrow) and daughter cells (C,

white arrow) were assessed. (D) The number of mitotic nuclei per 500 nuclei

counted are displayed as a percentage (*P = 0.02). (E) PCNA-positive

staining (arrow) was used to assess proliferative activity in the livers. (F)

PCNA-positive cells were scored similar to mitotic nuclei (#P = 0.06). (G)

Polyclonal cytokeratin (CK) staining was employed to demonstrate oval

cells (arrow) in the liver. (H) Oval cells were counted as polyclonal

CK-positive cells. Scoring excluded the interlobular bile duct cells. Results

are displayed as the numbers per 10 hpf (*P < 0.008). (I) Apoptotic,

Caspase-3-stained cells (arrow) were assessed. (J) Caspase-3-positive cells

were scored similar to mitotic nuclei (*P < 0.02). (K) TUNEL-stained cells

(arrow) were assessed. (L) Similar to Caspase-3 scoring, TUNEL-positive

cells were counted (P > 0.10). Of note, Caspase-3-positive (I, arrow) and

TUNEL-positive (K, arrow) mitotic figures were observed.

Fig. 6. Hepatic expression of cyclins and cytokines were induced in IL-

15-treated mice. Total RNA was isolated from the livers of control

(N = 2) and the IL-15-treated (N = 2) mice. For these ribonuclease

protection assays, we used a commercial kit (Pharmingen, San Diego,

CA) containing probes for Cyclins, TNF-a, and IFN-c mRNAs, as well

as the internal standard and housekeeping gene, glyceraldehyde-3-

phosphate dehydrogenase (Gapdh). For each experiment, 20 lg of total

RNA was employed. After separation on a 5% acrylamide gel, the

mRNA hybridization signals were detected using a phosphoimager and

then visualized on X-ray film for analysis. The mRNA expression

patterns of (A) Cyclins (A1, A2, B1, C, D1, D2, and D3) and (B)

cytokines (TNF-a and IFN-c) were evaluated. Gapdh was employed as

the housekeeping gene. Representative blots are displayed.


on proliferation, apoptosis, or overall viability of eithermature or immature hepatocytes cultured with ourconditions.

Despite having minimal effects on cultured liver cells,IL-15 treatment of healthy adult mice significantlyexpanded hepatic OV populations and transientlyincreased proliferative activity in mature hepatocytes.Increases in OV were noted within 3 days of initiatingIL-15 treatment and occurred without evident increasesin liver injury, as assessed by hepatic histology, histo-chemistry, and serum aminotransferase levels. OV cellscontinued to increase during protracted IL-15 adminis-tration, and by 7 days progenitor expansion was accom-panied by increased mRNA expression of cell cycleregulatory genes, up-regulation of PCNA expression inmature hepatocytes, and increased hepatocyte mitoses.

Given that our study did not show dramatic effects ofIL-15 on cell numbers in culture, IL-15 appears to haveincreased proliferative activity of mature hepatocytes bymore complex mechanisms. Evidence that increasedproliferative activity in mature hepatocytes followedhepatic accumulation of OV raises the intriguingpossibility that IL-15 promoted OV accumulation anddifferentiation into more mature hepatocytes that


subsequently proliferated. Our experimental approachdid not permit us to test this hypothesis directly. How-ever, we did demonstrate IL-15-dependent increases inTNFa and IFN-c, two cytokines that are known to reg-ulate proliferative activity in both hepatic progenitorsand mature hepatocytes. Fausto and colleagues recentlydemonstrated that in combination, these two factorsinduce oval proliferation, but reversibly repress prolifer-ation of more mature hepatocytes, without provokingapoptosis in either cell type [31]. Liver NKT cells areimportant sources on IFN-c and TNFa, and we previ-ously showed that IL-15 administration enhancedhepatic accumulation of NKT cells [37]. Others havealso suggested that increases in liver NKT cells promotehepatic regeneration [44]. Taken together, these resultssuggest that IL-15 might enhance liver repair by pro-moting hepatic accumulation of certain types of immunecells that produce growth-stimulatory cytokines for liverepithelial cells. Experiments are planned to evaluate thishypothesis in the future, but such studies are beyond thescope of the present work.

As mentioned earlier, circulating levels of IL-15 cor-relate with the severity of liver damage in patients withchronic hepatitis C. Thus, it is important to considerthe possibility that IL-15-mediated liver injury mighthave been the major factor that triggered liver regenera-tion in our study. We think that this is unlikely becauseoval cell numbers increased before increases in caspaseactivity or inflammatory foci were observed, and wedid not document increases in TUNEL staining or ami-notransferase release after either 3 or 7 days of IL-15treatment. We did find occasional mitotic hepatocytesthat seemed to be undergoing apoptosis after 7 days ofIL-15 exposure. This observation is intriguing because itsuggests that healthy adult livers carefully monitor livermass, up-regulating mechanisms that delete extraneoushepatocytes. Evidence that liver to body weight ratioswere similar in IL-15-treated mice and controls, despitedefinite IL-15-mediated increases in hepatocyte prolifera-tion, is consistent with this concept. This interpretationwould also reconcile apparent discrepancies betweenour evidence that IL-15 treatment slightly increased livercell apoptosis in healthy mice, while it has been shown toprevent Fas-mediated apoptosis in injured livers [10].

In summary, our findings suggest that IL-15 pre-dominantly promotes a hepatotrophic, wound heal-ing-type response in the liver. Thus, injury-relatedincreases in hepatic IL-15 may enhance regenerationof damaged livers. These findings have important ther-apeutic implications for patients with various types ofliver diseases. The identification of IL-15 as a meansto increase hepatic regenerative activity is particularlyimportant because it provides a new therapeuticapproach for patients with liver failure. On the otherhand, blocking IL-15 may be a potential target in livercancers. Actualization of these putative therapeutic

advances will require a better delineation of how IL-15 interfaces with other autocrine and paracrine signalsthat control the fate of cells in the liver during healthand disease.

Acknowledgements

The authors thank Amgen for the kind gift of recom-binant human IL-15. This work was supported by theNational Institutes of Health Grants RO1 AA010154(AMD), RO1 DK053792 (AMD), RO1 AA012059(AMD), and T32 DK007713 (JKS).

References

[1] Fehniger TA, Caligiuri MA. Interleukin 15: biology and relevanceto human disease. Blood 2001;97:14–32.

[2] Kennedy MK et al. Reversible defects in natural killer andmemory CD8 T cell lineages in interleukin 15-deficient mice. J ExpMed 2000;191:771–780.

[3] Fehniger TA et al. Fatal leukemia in interleukin-15 transgenicmice. Blood Cells Mol Dis 2001;27:223–230.

[4] Kirman I, Vainer B, Nielsen OH. Interleukin-15 and its role inchronic inflammatory diseases. Inflamm Res 1998;47:285–289.

[5] Kuniyasu H et al. Production of interleukin 15 by human coloncancer cells is associated with induction of mucosal hyperplasia,angiogenesis, and metastasis. Clin Cancer Res 2003;9:4802–4810.

[6] Yano S et al. Interleukin 15 induces the signals of epidermalproliferation through ERK and PI 3-kinase in a human epidermalkeratinocyte cell line, HaCaT. Biochem Biophys Res Commun2003;301:841–847.

[7] Reinecker HC et al. Intestinal epithelial cells both express andrespond to interleukin 15. Gastroenterology 1996;111:1706–1713.

[8] Shinozaki M et al. IL-15, a survival factor for kidney epithelialcells, counteracts apoptosis and inflammation during nephritis.J Clin Invest 2002;109:951–960.

[9] Bulfone-Pau SS et al. Death deflected: IL-15 inhibits TNF-alpha-mediated apoptosis in fibroblasts by TRAF2 recruitment to theIL-15Ralpha chain. FASEB J 1999;13:1575–1585.

[10] Bulfone-Paus S et al. Interleukin-15 protects from lethal apopto-sis in vivo. Nat Med 1997;3:1124–1128.

[11] Hjorth-Hansen H, Waage A, Borset M. Interleukin-15 blocksapoptosis and induces proliferation of the human myeloma cell lineOH-2 and freshly isolated myeloma cells. Br J Haematol1999;106:28–34.

[12] Quinn LS, Haugk KL, Damon SE. Interleukin-15 stimulates C2skeletal myoblast differentiation. Biochem Biophys Res Commun1997;239:6–10.

[13] Ogata Y et al. A novel role of IL-15 in the development ofosteoclasts: inability to replace its activity with IL-2. J Immunol1999;162:2754–2760.

[14] Ruckert R et al. Inhibition of keratinocyte apoptosis by IL-15: Anew parameter in the pathogenesis of psoriasis?. J Immunol2000;165:2240–2250.

[15] Maiuri L et al. Interleukin 15 mediates epithelial changes in celiacdisease. Gastroenterology 2000;119:996–1006.

[16] Pavlakis M et al. Intragraft IL-15 transcripts are increased inhuman renal allograft rejection. Transplantation 1996;62:543–545.

[17] Kakumu S et al. Serum levels of IL-10, IL-15 and soluble tumournecrosis factor-alpha (TNF-alpha) receptors in type C chronicliver disease. Clin Exp Immunol 1997;109:458–463.

[18] Grabstein KH et al. Cloning of a T cell growth factor thatinteracts with the beta chain of the interleukin-2 receptor. Science1994;264:965–968.


[19] Giri JG et al. Identification and cloning of a novel IL-15 bindingprotein that is structurally related to the alpha chain of the IL-2receptor. EMBO J 1995;14:3654–3663.

[20] Gao D et al. Oxidative DNA damage and DNA repair enzymeexpression are inversely related in murine models of fatty liverdisease. Am.J Physiol Gastrointest Liver Physiol2004;287:G1070–G1077.

[21] Yang S et al. Oval cells compensate for damage and replicativesenescence of mature hepatocytes in mice with fatty liver disease.Hepatology 2004;39:403–411.

[22] Higgins GM, Anderson RM. Experimental pathology of liver:restoration of liver of the white rat following partial surgicalremoval. Arch Pathol 1931;12:186–202.

[23] Kleiner DE et al. Design and validation of a histological scoringsystem for nonalcoholic fatty liver disease. Hepatology2005;41:1313–1321.

[24] Brunt EM et al. Nonalcoholic steatohepatitis: a proposal forgrading and staging the histological lesions. Am J Gastroenterol1999;94:2467–2474.

[25] Yu CC et al. Haemangiopericytomas: the prognostic value ofimmunohistochemical staining with a monoclonal antibody toproliferating cell nuclear antigen (PCNA). Histopathology1991;19:29–33.

[26] Hall PA et al. Proliferating cell nuclear antigen (PCNA) immu-nolocalization in paraffin sections: an index of cell proliferationwith evidence of deregulated expression in some neoplasms.J Pathol 1990;162:285–294.

[27] Di Cunto F et al. Defective neurogenesis in citron kinase knockoutmice by altered cytokinesis and massive apoptosis. Neuron2000;28:115–127.

[28] Gown AM, Willingham MC. Improved detection of apoptotic cellsin archival paraffin sections: immunohistochemistry using antibod-ies to cleaved caspase 3. J Histochem Cytochem 2002;50:449–454.

[29] Gavrieli Y, Sherman Y, Ben-Sasson SA. Identification ofprogrammed cell death in situ via specific labeling of nuclearDNA fragmentation. J Cell Biol 1992;119:493–501.

[30] Roskams T et al. Oxidative stress and oval cell accumulation inmice and humans with alcoholic and nonalcoholic fatty liverdisease. Am J Pathol 2003;163:1301–1311.

[31] Brooling JT et al. Differential regulation of rodent hepatocyteand oval cell proliferation by interferon gamma. Hepatology2005;41:906–915.

[32] Petersen BE et al. Bone marrow as a potential source of hepaticoval cells. Science 1999;284:1168–1170.

[33] Petersen BE et al. Mouse A6-positive hepatic oval cells alsoexpress several hematopoietic stem cell markers. Hepatology2003;37:632–640.

[34] Herman B et al. Irreversible injury in anoxic hepatocytes precip-itated by an abrupt increase in plasma membrane permeability.FASEB J 1988;2:146–151.

[35] Bulanova E et al. Mast cells express novel functional IL-15receptor alpha isoforms. J Immunol 2003;170:5045–5055.

[36] Sicklick JK et al. Role for Hedgehog signaling in hepatic stellatecell activation and viability. Lab Invest 2005.

[37] Li Z et al. Murine leptin deficiency alters Kupffer cell productionof cytokines that regulate the innate immune system. Gastroen-terology 2002;123:1304–1310.

[38] Itano N et al. Abnormal accumulation of hyaluronan matrixdiminishes contact inhibition of cell growth and promotes cellmigration. Proc Natl Acad Sci USA 2002;99:3609–3614.

[39] Takao T et al. Effect of 17beta-estradiol on tumor necrosis factor-alpha-induced cytotoxicity in the human peripheral T lympho-cytes. J Endocrinol 2005;184:191–197.

[40] Fausto N, Campbell JS. The role of hepatocytes and oval cellsin liver regeneration and repopulation. Mech Dev 2003;120:117–130.

[41] Yamada Y et al. Initiation of liver growth by tumor necrosis factor:deficient liver regeneration in mice lacking type I tumor necrosisfactor receptor. Proc Natl Acad Sci USA 1997;94:1441–1446.

[42] Lowes KN et al. Oval cell-mediated liver regeneration: Role ofcytokines and growth factors. J Gastroenterol Hepatol 2003;18:4–12.

[43] Diehl AM. Cytokine regulation of liver injury and repair.Immunol Rev 2000;174:160–171.

[44] Minagawa M et al. Intensive expansion of natural killer T cells inthe early phase of hepatocyte regeneration after partial hepatec-tomy in mice and its association with sympathetic nerve activa-tion. Hepatology 2000;31:907–915.

[45] Davoodi-Semiromi A et al. A mutant Stat5b with weaker DNAbinding affinity defines a key defective pathway in nonobesediabetic mice. J Biol Chem 2004;279:11553–11561.

[46] Dubois S et al. Natural splicing of exon 2 of human interleukin-15 receptor alpha-chain mRNA results in a shortened form with adistinct pattern of expression. J Biol Chem 1999;274:26978–26984.

interleukin-15 increases hepatic regenerative activity

Documents