| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1-Deficient Mice Requires CD4+ T Cells1
,
Departments of
*
Pathology and
Microbiology and Immunology, and
The Norris Cotton Cancer Center, Dartmouth Medical School, Lebanon, NH 03756
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1 spontaneously develop necroinflammatory liver disease, but the immune basis for the development of this pathology has not been demonstrated. Here, we show that BALB/c-TGF-1-/- mice exhibit abnormal expansion in hepatic mononuclear cells (MNCs) compared with wild-type littermate control mice, particularly in the T cell and macrophage lineages. To test whether lymphocytes of the adaptive immune system are required for the spontaneous development of necroinflammatory liver disease, BALB/c-TGF-1-/- mice were rendered deficient in B and T cells by crossing them with BALB/c-recombinase-activating gene 1-/- mice. BALB/c-TGF-1-/-/recombinase-activating gene 1-/- double-knockout mice showed extended survival and did not develop necroinflammatory liver disease. The cytolytic activity of BALB/c-TGF-1-/- hepatic lymphocytes was assessed using an in vitro CTL assay. CTL activity was much higher in BALB/c-TGF-1-/- hepatic MNCs compared with littermate control hepatic MNCs and was particularly pronounced in the CD4+ T cell subset. Experimental depletion of CD4+ T cells in young BALB/c-TGF-1-/- mice prevented the subsequent development of necroinflammatory liver disease, indicating that CD4+ T cells are essential for disease pathogenesis in vivo. These data definitively establish an immune-mediated etiology for necroinflammatory liver disease in BALB/c-TGF-1-/- mice and demonstrate the importance of CD4+ T cells in disease pathogenesis in vivo. Furthermore, TGF-1 has a critical role in homeostatic regulation of the hepatic immune system, inhibiting the development or expansion of hepatic cytolytic CD4+ T cells. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and TNF-
when stimulated ex vivo (6), suggesting that CD4+ T cells contribute to the development and/or progression of hepatocellular damage. The etiology of AIH remains poorly understood, however, and evidence supporting a pathogenic role for CD4+ T cells remains circumstantial. Advances in the understanding of the pathogenesis of AIH have been realized slowly, partly because of the lack of a small animal model system. An appropriate model system should demonstrate certain relevant characteristics, including the spontaneous development of necroinflammatory disease, an influence of genetic background, and an immune etiology, with involvement of type 1 T cells.
We have recently reported that BALB/c mice rendered deficient in the immunoregulatory cytokine TGF-1 spontaneously develop progressive severe necroinflammatory liver disease in the second week of life (7). Disease is restricted to susceptible genetic backgrounds. Whereas TGF-1 knockout mice on 129-based backgrounds develop a multifocal inflammatory disease affecting several vital organs (8, 9), the development of necroinflammatory liver disease resulting in destruction of liver parenchyma is specific to the BALB/c background (7). Furthermore, the disease is 100% penetrant on the BALB/c background, as all BALB/c-TGF-1-/- mice predictably develop necroinflammatory liver disease and die before 18 days of age. Disease is associated with the accumulation of large numbers of activated hepatic CD4+ T cells strongly skewed to the Th1 effector cell phenotype, producing copious IFN- upon stimulation ex vivo. Experiments involving BALB/c-TGF-1-/-/IFN--/- double-knockout mice definitively demonstrate that IFN- is required for necroinflammatory liver disease (7). These observations suggest that the etiology of necroinflammatory liver disease in BALB/c-TGF-1-/- mice is autoimmune and implicate a role for lymphocytes of the adaptive immune system, in particular IFN--producing CD4+ Th1 cells. Here we directly test the hypothesis that necroinflammatory liver disease in BALB/c-TGF-1-/- mice is immune mediated by assessing whether lymphocytes of the adaptive immune system are required for disease development. In addition, we test the requirement for the CD4+ T cell subset specifically, through specific depletion of this subset from preclinical BALB/c-TGF-1-/- mice.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Mice were bred at the Dartmouth Medical School American Association for Accreditation of Laboratory Animal Care-accredited animal care facility and were treated humanely according to National Institutes of Health guidelines. BALB/c-TGF-1-/- pups were raised by crossing BALB/c(N6)-TGF-1+/- males and females. The derivation of BALB/c(N6)-TGF-1+/- breeders was previously described (7). Cages were monitored daily for the birth of new litters, defined as day 0. TGF-1 genotyping was performed on tail snips from 3- to 7-day-old pups by PCR using the following primer combination: 5'-TTGCTGTACTGTGTGTCCAG-3', 5'-CAGGACATAGCGTTGGCTAC-3', and 5'-TCCACAGAGAAGAACTGCTG-3'. This single PCR distinguishes the wild-type, heterozygous, and knockout genotypes. For the generation of BALB/c-TGF-1-/-/recombinase-activating gene 1 (RAG-1)-/- mice, BALB/c(N8)-TGF-1+/- mice were bred with BALB/c(N7)-RAG-1-/- mice (The Jackson Laboratory, Bar Harbor, ME). The RAG-1 genotype was determined by two PCR reactions using the following primers: for the wild-type Rag-1 gene, 5'-TCGTTTCAAGAGTGACGG-3' and 5'-ATGTCACAGGACGGTGTG-3'; and for the neo-insertion knockout RAG-1 gene, 5'-TCGTTTCAAGAGTGACGG-3' and 5'-ATGGAAGCCGGTCTTGTC-3'.
Aspartate aminotransferase (AST) analysis
Serum AST was determined using a Roche-Hitachi 917 Automatic Analyzer (Hitachi, Hialeah, FL), employing a UV kinetic enzymatic assay read at 340 nm.
Isolation of hepatic and splenic mononuclear cells (MNCs)
Under deep anesthesia, mice were decapitated and exsanguinated. Exsanguination before isolation of hepatic MNCs reduces by 
80% contamination by MNCs in circulation, leaving predominantly resident hepatic MNCs, and is as effective as portal vein perfusion in this regard (10). Portal vein perfusion was technically not possible in the very small 11- to 13-day-old BALB/c-TGF-1-/- mice analyzed. Spleens and livers from 11- to 13-day-old mice were dissected and weighed. Liver pieces were mechanically disrupted through a stainless steel mesh in RPMI/5% FCS. Liver homogenates were digested in 100 U/ml collagenase IV (Sigma-Aldrich, St. Louis, MO) and 10 U/ml DNase I (Roche, Indianapolis, IN) in serum-free RPMI at 37°C for 40 min. Cells were collected by centrifugation and resuspended in RPMI/5% FCS, and hepatic MNCs were isolated over Ficoll (Histopaque 1119; Sigma-Aldrich). Splenocytes were isolated by crushing the spleen between the frosted ends of two microscope slides, followed by isolation over Ficoll. Cells were carefully counted by hemocytometer, using trypan blue exclusion.
Flow cytometric analyses of hepatic and splenic MNCs
Abs and reagents purchased from BD PharMingen (San Diego, CA) included biotinylated anti-CD45, streptavidin-allophycocyanin, anti-CD3-PE, anti-CD19-PE, anti-CD11b-FITC, anti-CD8-FITC, and equivalently labeled isotype controls. Anti-CD4-FITC was prepared in the laboratory of K. Murphy (Washington University, St. Louis, MO). Stained cells were acquired using a FACSCalibur cytometer (BD Biosciences, San Jose, CA) and analyzed using CellQuest software (BD Biosciences). Positive staining was established using isotype control Abs in all instances. Liver or splenic MNCs were counted and stained sequentially with anti-CD45-biotin and streptavidin-allophycocyanin. Total recovered CD45+ cell number was determined by multiplying the total number of recovered cells by the percentage of CD45+. For splenic and hepatic MNC preparations, typically 95 and 50–95%, respectively, were CD45+. The density of CD45+ cells was determined by dividing the total number of recovered CD45+ cells by the weight (milligrams) of the organ. For subset analyses cells were stained for CD45 and either CD3 (T cells), CD19 (B cells), or CD11b (monocyte/macrophages). In some experiments CD3+ cells were further analyzed for CD4+ and CD8+ subsets.
Histology
After anesthesia and euthanasia of mice, livers were dissected out and fixed in buffered formalin, followed by paraffin embedding, sectioning, and staining with H&E using routine methods.
Cytolytic activity assays
Splenic and hepatic MNCs were isolated from 11- to 12-day-old BALB/c-TGF-1-/- and littermate control BALB/c-TGF-1+/+ mice and were used immediately as effector cells in a 6-h redirected lysis assay (11), similar to a published method (12). As shown in Fig. 6, freshly prepared hepatic BALB/c-TGF-1-/- MNCs were also fractionated before the cytolysis assay by depleting either CD4+ T cells or CD8+ T cells by negative selection using Ab-tagged magnetic beads (Miltenyi Biotec, Auburn, CA) and a column/magnet apparatus. In addition, enriched BALB/c-TGF-1-/- hepatic CD4+ or CD8+ cells were isolated from the column/magnet apparatus and washed. Positively selected aliquots were tested for subset enrichment by flow cytometry and were found to be >85% pure.
|
Lysis assays were performed in duplicate at five different E:T cell ratios (100:1, 20:1, 4:1, 0.8:1, and 0.16:1). Duplicate wells consistently showed <10% variation. The percent specific lysis for each well was calculated by standard methods using the amounts of 51Cr measured in the supernatant from target cells, as follows: % specific lysis = (measured cpm - spontaneously released cpm)/(total released cpm). Spontaneously released cpm refers to the amount of 51Cr released from labeled target cells cultured in the absence of added effector cells, and levels were consistently 15–20% of the total released cpm. Total released cpm refers to the amount of 51Cr measured from labeled target cells after several cycles of freezing-thawing. To account for potential variation in the proportion of T cells between MNC preparations, cell aliquots were stained for CD45 and CD3 and were analyzed by flow cytometry. This information was used to determine CD3-specific E:T ratios (not shown). The data were little changed by this additional analysis. The anti-CD3 mAb used (145-2C11) was in excess and therefore not rate limiting for this assay.
In vivo CD4+ T cell depletion
The CD4+ T cell-depleting mAb GK1.5 (13) was prepared from ascites fluid from hybridoma-injected SCID mice via ammonium sulfate precipitation and subsequent dialysis (10,000 kDa cutoff) in PBS. The presence of IgG was confirmed using SDS-PAGE and staining by Coomassie Blue. Protein concentration was determined with the Bio-Rad protein assay (Hercules, CA). In pilot assays the specificity and efficacy of CD4+ depletion were confirmed using GK1.5 injection in adult mice following standard protocols (14) with subsequent flow cytometric analysis of splenocyte populations. These analyses indicated specific depletion of the CD4+ T cell subset, with no effect on CD8+ T cells or (CD19+) B cells. For experimental analyses of BALB/c-TGF-1-/- mice, BALB/c-TGF-1-/- mice were injected on days 5, 6, and 7 of life with 0.025 mg/g purified GK1.5 mAb or with isotype control rat IgG2b mAb (KLH/G2b-1-2; eBioscience, San Diego, CA). Mice were euthanized on day 11 for transaminase analysis and histopathology and also to confirm the specificity and efficacy of CD4+ T cell depletion in BALB/c-TGF-1-/- mice via flow cytometric analyses of MNCs.
RT-PCR
Total RNA was isolated from homogenized livers using TRIzol (Life Technologies, Gaithersburg, MD) according to the manufacturer’s instructions. Two micrograms of total RNA was reverse transcribed to make cDNA (RNeasy Kit; Qiagen, Valencia, CA). PCR was performed for V14-J281 and for actin exactly as previously described (15). As a control for the RNA specificity of the PCR reaction the enzyme reverse transcriptase was omitted in the RT step. Subcloning and sequencing, using standard methodologies, confirmed the specific identity of the V14-J281 amplicon for a randomly selected wild-type sample.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-/- mice and in all seven 12- to 14-day-old BALB/c-TGF-1-/- mice assessed (7). To confirm and extend these results and to more precisely establish the kinetics of the onset of hepatocellular damage, transaminase levels were measured in sera from 27 BALB/c-TGF-1-/- mice of various ages. In BALB/c-TGF-1-/- mice younger than 8 days of age, serum AST levels were not different from littermate control levels (Fig. 1A). Corroborating the AST data, tissue sections of liver from BALB/c-TGF-1-/- mice younger than 8 days of age were histologically normal for age and indistinguishable from those from BALB/c-TGF-1+/+ littermate control mice (data not shown). At and after postnatal day 9 all BALB/c-TGF-1-/- mice exhibited serum AST levels above the normal mean (determined from littermate controls), and the vast majority (93%) mice exhibited levels >2 SD above the mean (Fig. 1A). After day 9 all BALB/c-TGF-1-/- livers appeared necrotic on gross examination (not shown) (7) and had histologic evidence of severe necroinflammatory liver disease, with subcapsular necrosis of hepatocytes and an associated mixed inflammatory response (Fig. 4B) similar to that described previously (7). In additional studies (n = 12), a strong correlation (r2 = 0.82) was observed between serum AST levels and the extent of TUNEL-positive areas of the liver (data not shown), confirming the utility of serum AST as a marker of hepatocellular damage in this mouse model. The temporal onset of hepatocellular damage (7–9 days) coincided with the failure of BALB/c-TGF-1-/- mice to gain weight appropriately, a cardinal sign of severe inflammation (Fig. 1B). All BALB/c-TGF-1-/- mice died before 18 days of age, with a mean survival of 13.8 ± 2.1 days (n = 20; data not shown) (7).
|
|
1-deficient BALB/c mice. For these analyses we purified intrahepatic MNCs from diseased 11- to 13-day old BALB/c-TGF-1-/- mice and from littermate control mice. In BALB/c-TGF-1-/- livers, the density (number of cells per milligram) of CD45+ cells was elevated 3.3-fold compared with that in littermate control livers (Fig. 2A; p = 0.067, -/- vs +/-; p = 0.012, -/- vs +/+). Subset analysis of the CD45+ cell population showed increases in cell densities for the T cell and macrophage lineages, but not for the B cell lineage (Fig. 2B). Within the T cell lineage, increases were seen for both the CD4+ and CD8+ cell subsets (Fig. 2C). The liver typically contains substantial numbers of NK cells and double-negative (CD4-CD8-) T cells (16, 17). However, the classic marker for the NK cell population (NK1.1) is not expressed in the BALB/c background (18), and we did not assess this population directly using flow cytometry. Additional functional studies of NK activity (see Fig. 5) showed no increase in NK function in BALB/c-TGF-1-/- livers. As for double-negative T cells, their numbers were indeed increased in BALB/c-TGF-1-/- livers (data not shown), but it is not clear whether these cells represent true resident hepatic double-negative T cells or activated single-positive T cells that have undergone CD4 or CD8 coreceptor down-regulation. Most hepatic T cells in BALB/c-TGF-1-/- mice have an activated cell surface phenotype (CD62Llow, CD44high, VLA4high; data not shown) (7).
|
|
1-/- livers were not paralleled in BALB/c-TGF-1-/- splenic cell populations. The densities of CD45+ cells were similar in the spleens of BALB/c-TGF-1-/-, BALB/c-TGF-1+/-, and BALB/c-TGF-1+/+ mice (Fig. 2D). T cells and macrophages were not expanded in BALB/c-TGF-1-/- spleens, although a relative paucity of B cells was observed (Fig. 2, E and F). These data indicate that a deficiency of TGF-1 results in an increase in the number of T cells and macrophages in liver (but not spleen) and a relative decrease in the number of B cells in spleen (but not liver).
These observations implicate lymphocytes of the adaptive immune system (T and/or B cells) in the pathogenesis of liver disease. To directly test this hypothesis we crossed BALB/c-TGF-1+/- mice with BALB/c-RAG-1-/- mice to produce BALB/c-TGF-1-/-/RAG-1-/- double-knockout mice. RAG-1 is necessary for the development of Ag receptors. Mice deficient in RAG-1 lack mature B and T cells, but not NK cells or cells of the innate immune system (19). BALB/c-TGF-1-/- mice that were wild-type or heterozygous for RAG-1 rapidly developed symptoms associated with inflammation (e.g., poor weight gain, hunched posture, and decreased mobility) and survived for 2 wk (Fig. 3A). By contrast, nine of 11 BALB/c-TGF-1-/-/RAG-1-/- mice generated in our facility showed extended survival and were in good health. One BALB/c-TGF-1-/-/RAG-1-/- mouse survived for 180 days of age before it was euthanized. Two of the 11 BALB/c-TGF-1-/-/RAG-1-/- mice died spontaneously of unknown causes at 39 and 70 days of age, respectively. Although smaller than littermate TGF-1+/-/RAG-1-/- mice (data not shown), BALB/c-TGF-1-/-/RAG-1-/- mice did not develop other symptoms associated with inflammation and were euthanized at various ages. Since most were euthanized, the mean survival shown (102 days) actually represents a minimum for BALB/c-TGF-1-/-/RAG-1-/- mice. To assess the presence of hepatocellular damage, serum AST levels were periodically analyzed in four BALB/c-TGF-1-/-/RAG-1-/- mice. For all four mice, serum AST levels were consistently in the normal range (Fig. 3B), indicating the absence of hepatocellular damage. In corroboration, there was an absence of pathology in histological sections of BALB/c-TGF-1-/-/RAG-1-/- livers (Fig. 4C), which were similar in appearance to BALB/c-TGF-1+/+ (Fig. 4A) and BALB/c-RAG-1-/- control livers (data not shown). These findings stand in stark contrast to the dramatic pathologic changes observed in livers from 11- to 13-day-old BALB/c-TGF-1-/- mice (Fig. 4B). Thus, necroinflammatory liver disease in BALB/c-TGF-1-/- mice is dependent upon the presence of a functioning adaptive immune system containing mature B and/or T cells. Moreover, since RAG-1-/- mice still generate functional NK cells, these data indicate that the presence of NK cells is not sufficient for the development of necroinflammatory liver disease in BALB/c-TGF-1-/- mice.
|
1-/- mice. The increase in numbers of hepatic T cells, but not B cells, suggests that T cell-mediated immunity plays a particularly important role. During immune responses T cells can be effectors of cytolytic damage on target cells. Therefore, we sought to determine whether T cells from BALB/c-TGF-1-/- livers exhibit abnormal cytolytic activity using a CD3-redirected lysis assay in vitro. In this assay FcR-expressing P815 mastocytoma cells are labeled with 51Cr and surface-coated with mAb to anti-CD3. During coculture cytolytic CD3+ T cells are activated by, and kill, the anti-CD3-coated target cells, permitting assessment of the total cytolytic capacity of a T cell population even when the relevant Ag is not known. NK cell cytolytic activity was determined using the YAC-1 NK target cell line.
Application of the anti-CD3-redirected lysis assay to hepatic and splenic MNC preparations from BALB/c-TGF-1-/- and wild-type mice indicated that BALB/c-TGF-1-/- hepatic MNC preparations exhibit abnormally high T cell cytolytic activity. Wild-type splenic MNCs exhibited some cytolytic activity against anti-CD3-coated 51Cr-labeled P815 target cells (Fig. 5A). By contrast, wild-type hepatic MNCs showed very little cytolytic activity (Fig. 5B). BALB/c-TGF-1-/- splenic MNCs exhibited cytolytic activity (Fig. 5C) similar to that of littermate control splenic MNCs (Fig. 5A). Strikingly, cytolytic activity in BALB/c-TGF-1-/- hepatic MNC preparations was greatly increased compared with that in hepatic wild-type littermate control hepatic MNC preparations (compare Fig. 5D with Fig. 5B). In all cell populations cytolytic activity was dependent upon T cell activation, as no activity was observed when P815 target cells were either left uncoated or coated with isotype control mAb. No NK cytolytic activity was observed in any MNC population, as indicated by the absence of lysis of YAC-1 target cells. The low CTL activity in wild-type liver is not due to a paucity of hepatic T cells, since CD3+ cells were abundant in wild-type liver MNC preparations (data not shown). Moreover, wild-type hepatic MNC cytolytic activity was low even after normalizing the data for CD3+ T cell numbers (determined by flow cytometry; data not shown). Thus, in BALB/c mice a deficiency of TGF-1 results in a strong up-regulation of CTL (but not NK) activity in liver MNCs.
To determine which T cell subset was responsible for cytolytic activity in this assay, BALB/c-TGF-1-/- hepatic MNC preparations were specifically depleted of either the CD4+ or CD8+ T cell subsets before the assay. Depletion of CD8+ T cells had little effect on cytolytic activity, whereas depletion of CD4+ T cells inhibited activity (Fig. 6). Results using isolated CD4+ cells and CD8+ cells corroborated these observations, as activity was very high in purified CD4+ T cells and was significantly lower in purified CD8+ T cells (Fig. 6). Thus, in hepatic BALB/c-TGF-1-/- MNC preparations, high cytolytic activity is preferentially associated with the CD4+ T cell subset.
BALB/c-TGF-1-/- livers accumulate large numbers of cytolytically active CD4+ T cells. This suggests that the CD4+ T cell subset plays an important role in the pathogenesis of hepatocellular damage in BALB/c-TGF-1-/- mice. To directly test the involvement of CD4+ T cells in vivo, presymptomatic BALB/c-TGF-1-/- mice were injected with the mAb GK1.5 to specifically deplete the CD4+ T cell subset before the onset of hepatocellular damage. The kinetic analysis of disease (Fig. 1) indicated that liver damage is not detectable (by either AST or histopathology) before 8 days of age, after which necroinflammatory liver disease becomes progressive and widespread. Therefore, BALB/c-TGF-1-/- pups were identified by PCR by 4 days of age and were injected with GK1.5 daily on days 5, 6, and 7; hepatocellular damage was assessed at 11 days of age. Injection of GK1.5 completely prevented the subsequent development of necroinflammatory liver disease. Serum AST levels for treated BALB/c-TGF-1-/- mice were not elevated and were similar to those in wild-type mice (Fig. 7). In addition, liver sections from GK1.5-treated BALB/c-TGF-1-/- mice (Fig. 4D) showed no histological evidence of the hepatocellular damage typically observed in liver sections from untreated BALB/c-TGF-1-/- mice. As a control, injection of isotype-matched irrelevant Ab into BALB/c-TGF-1-/- mice had no effect, with necroinflammatory disease developing (Figs. 4E and 7) as in untreated BALB/c-TGF-1-/- mice. Additional analyses confirmed the efficacy of the GK1.5 mAb treatment in vivo, showing that the CD4+ T cell subset, but not other lymphocyte subsets, was specifically depleted (data not shown). Thus, CD4+ T cells are required for the development of necroinflammatory liver disease in BALB/c-TGF-1-/- mice.
|
-chain, V14-J281. Previously, NKT cells have been shown to contribute to hepatocellular damage in mice injected with the T cell mitogen Con A (20). Therefore, we considered whether the expansion of hepatic CD4+ T cells observed in BALB/c-TGF-1-/- livers might represent an expansion of the NKT cell subset. RT-PCR analysis (Fig. 8) indicated that V14-J281 TCR expression was not enhanced in BALB/c-TGF-1-/- livers. To the contrary, V14-J281 expression was significantly lower in BALB/c-TGF-1-/- livers than in littermate control livers. Thus, the expansion of CD4+ T cells in BALB/c-TGF-1-/- livers is not due to an expansion of the V14-J281-expressing NKT cell subset.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1 spontaneously develop aggressive necroinflammatory liver disease and succumb rapidly, typically by 2 wk of age. While our previous report (7) demonstrated a requirement for the type 1 cytokine IFN- in hepatocellular damage in BALB/c-TGF-1-/- mice, we had no direct evidence that liver disease in these mice requires cells of the adaptive immune system. Several conclusions can be drawn from the current work. First, it definitively establishes that liver damage in BALB/c-TGF-1-/- mice has an immune pathogenesis, requiring cells of the adaptive immune system. Moreover, in vivo neutralization studies using the GK1.5 mAb demonstrate a requirement for the CD4+ T cell subset in hepatocellular damage. The logical converse of this observation is that non-CD4+ T cells (e.g., CD8+ T cells, NK cells, B cells, etc.) are not sufficient (in the absence of CD4+ T cells) for the development of liver disease. Less prominent roles for CD8+ T cells and for NK cells are also indirectly supported by results from the ex vivo cytolysis assays using BALB/c-TGF-1-/- hepatic MNCs, which detected no NK-mediated activity, comparatively little CD8+ T cell activity, but robust CD4+ T cell activity. The individual contributions of other lymphoid subsets will have to be determined experimentally in vivo. Nevertheless, demonstration of the requirement for adaptive lymphocytes in general and for CD4+ T cells specifically is strong evidence that necroinflammatory liver disease in BALB/c-TGF-1-/- mice results from an autoimmune attack upon liver parenchyma.
The liver is an abundant source of NKT cells, a CD4+ T cell subset restricted by the nonclassical class I MHC molecule CD1d that can be activated by lipid Ags presented by CD1d. The observations that CD4+ T cells with high cytolytic activity are greatly expanded in BALB/c-TGF-1-/- livers and that hepatocellular damage requires CD4+ T cells suggested a possible pathogenic involvement of NKT cells. Due to the absence of expression on the BALB/c background of the polymorphic NKT marker NK1.1, we did not directly measure the number of NKT cells. Nevertheless, the observation of suppressed hepatic expression of V14-J281 mRNA in BALB/c-TGF-1-/- liver indicates that the NKT population is significantly reduced in number compared with that in wild-type liver. This is even more remarkable given the much larger total number of CD4+ T cells in BALB/c-TGF-1-/- liver. Thus, NKT cells probably compose only a very small minority of hepatic CD4+ T cells in BALB/c-TGF-1-/- liver, arguing against a pathogenic role for this subset. The data actually support the opposite, that NKT cells might act as suppressor cells, and a deficiency in this cell subset might contribute to pathogenesis. This is consistent with an accumulating body of evidence that NKT cells inhibit the development of autoimmunity (21, 22, 23) and with recent observations that NKT cells are deficient in several human autoimmune diseases, among them AIH (24).
Is necroinflammatory liver disease in BALB/c-TGF-1-/- mice Ag-specific? The requirement for lymphocytes of the adaptive immune system and for CD4+ T cells more specifically implicates an Ag-specific mechanism. However, evidence from a variety of experimental systems suggests that activated T cells can mediate hepatocellular damage through non-Ag-specific bystander effects. In mice, Con A injection causes Ag-nonspecific activation of hepatic T cells, inducing hepatocellular damage mediated through the release of soluble hepatotoxic molecules such as TNF- or IFN- (25, 26, 27). Moreover, following injection of OVA peptide, activated OVA peptide-specific TCR transgenic CD8+ T cells traffic to the liver, where they induce hepatocellular damage through Fas ligand (FasL)-Fas-dependent mechanisms (28). Mehal et al. (29) have also demonstrated that intrahepatic accumulation of activated OVA-specific TCR-transgenic CD8+ T cells causes hepatocellular damage. Damage ensues even when hepatocytes are rendered (through genetic manipulation of their MHC) incapable of presentation of the OVA peptide, indicating that liver damage can occur through mechanisms that do not involve cognate T cell recognition of hepatocytes (29). Similar conclusions were reached by Bowen et al. (30), who showed that TCR-transgenic CD8+ T cells induce Ag-nonspecific hepatocellular damage through a mechanism that requires the inflammatory cytokines IFN- and TNF-. The extent to which the hepatic BALB/c-TGF-1-/- CD4+ T cell response is Ag specific remains an important unanswered question.
How might CD4+ T cells be triggered? If the response is Ag specific, CD4+ T cells might be triggered by hepatocytes directly, through Ag:TCR interactions. Under normal conditions, however, hepatocytes typically express very little, if any, surface class II MHC (31, 32), which would appear to argue against such a mechanism. Then again, it is likely that hepatocytes in BALB/c-TGF-1-/- mice express substantial class II, since TGF-1-/- mice show widespread aberrant constitutive expression of class II mRNA and protein (33). Class II expression is strongly up-regulated by IFN- (34), suggesting a possible mechanism for the involvement of IFN- (7). Through up-regulation of class II MHC complexes on the surface of hepatocytes, IFN- could prime hepatocytes for subsequent destruction by CD4+ CTLs. TGF-1 and IFN- have opposite effects on the regulation of class II MHC (35). High expression of IFN- mRNA is observed in BALB/c-TGF-1-/- livers (J. T. Lin and J. D. Gorham, unpublished observations), and the up-regulatory activity of IFN- on class II expression might be potentiated in TGF-1-/- mice, which lack the down-regulatory activity of TGF-1. Alternatively, CD4+ T cells might be triggered not by hepatocytes, but, more conventionally, by class II-expressing hepatic APCs, which include Kupffer cells, liver dendritic cells, and liver sinusoidal endothelial cells. A third possibility is that CD4+ T cells are triggered extrahepatically, then migrate to the liver, which selectively retains activated T cells (36).
How might CD4+ T cells bring about the demise of hepatocytes? It is intriguing to speculate that CD4+ T cells might themselves be proximal direct effectors of hepatocyte death. Indeed, the demonstration of robust CTL activity within the CD4+ T cell subset increases the plausibility of this scenario. CD4+ T cells might mediate direct cytotoxicity through the actions of FasL-Fas or granzyme/perforin mechanisms or perhaps indirectly, through the release of hepatotoxic molecules such as TNF- or IFN-. FasL, TNF-, and IFN- mRNA are all up-regulated in BALB/c-TGF-1-/- livers (7) (data not shown). Alternatively, CD4+ T cells might not be the proximal direct effectors of hepatocellular death, but might participate instead by providing more typical help, engaging other cell types to be the final death effectors. These pathways are not mutually exclusive, and determining the relative contribution of each to liver damage remains an important research goal.
Finally, this work provides evidence that TGF-1 plays a critical role in homeostatic regulation of the hepatic immune system. In the absence of TGF-1, there is substantial expansion in the numbers of T cells and macrophages in liver. Moreover, TGF-1 also inhibits the development or accumulation of hepatic T cells with potent cytolytic activity, particularly within the CD4+ cell subset. In additional experiments (J. T. Lin and J. D. Gorham, unpublished observations), we have observed similar abnormal hepatic expansions in T cells and macrophages in TGF-1-/- mice on a 129-based background, which do not develop necroinflammatory liver disease. This indicates, first, that hepatic expansions in T cells and macrophages are not sufficient to initiate hepatocellular damage (i.e., there must be other triggers as well), and, second, that TGF-1’s inhibition of hepatic immune expansion is not limited to the BALB/c genetic background. That is, a basic function of TGF-1 appears to be to regulate the number, composition, and activity of immune cells in the liver. That equivalent abnormalities were not observed in splenic MNC preparations may indicate that the immune functions of TGF-1 in this regard are somewhat specific to the liver. It remains to be determined, however, whether the increases observed in hepatic T cell and macrophage numbers represents a true dysregulation of resident intrahepatic immune cells or, rather, reflects an increase in trafficking to the liver of immune cells activated at extrahepatic sites.
The liver has a complex immune system, populated with a large variety of classical and nonclassical T cells and NK cells (16). Delineating the effects of TGF-1 on these various hepatic immune cell subsets remains important work for future investigations. In vitro, TGF-1 has been shown to potently inhibit the development and/or activity of both T cells and NK cells (for review, see Ref. 37). The current work shows that, at least with respect to maintaining hepatic immune homeostasis and preventing immune-mediated hepatocellular damage in BALB/c mice, TGF-1’s inhibition of T cells (and CD4+ T cells in particular) is more relevant than its inhibition of NK cells. These various observations have potentially important implications for understanding hepatic immune function, not only in autoimmune liver diseases, but also in immunity to liver-specific pathogens, such as HCV, HBV, malaria, schistosomiasis, and Listeria, and in liver-specific immunity observed in graft-vs-host disease.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 L.A.R. and J.T.L. contributed equally to this work.
3 Address correspondence and reprint requests to Dr. James D. Gorham, Department of Pathology, Dartmouth Medical School, One Medical Center Drive, Lebanon, NH 03756. E-mail address: james.d.gorham{at}dartmouth.edu
4 Abbreviations used in this paper: AIH, autoimmune hepatitis; AST, aspartate aminotransferase; FasL, Fas ligand; MNC, mononuclear cell; RAG, recombinase-activating gene.
Received for publication September 30, 2002. Accepted for publication February 18, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and interferon- in the liver biopsies of children with chronic liver disease. J. Hepatol. 21:816.[Medline]
1-deficient mice develop necroinflammatory IFN--dependent hepatitis. J. Immunol. 166:6413.1 gene results in multifocal inflammatory disease. Nature 359:693.[Medline]
1 null mutation in mice causes excessive inflammatory response and early death. Proc. Natl. Acad. Sci. USA 90:770.14 NKT cell-mediated cytotoxicity by interleukin 4 in an autocrine mechanism resulting in the development of concanavalin A-induced hepatitis. J. Exp. Med. 191:105.-galactosylceramide prevents autoimmune diabetes in non-obese diabetic mice. Nat. Med. 7:1052.[Medline]
14-J281 NKT cells naturally regulate experimental autoimmune encephalomyelitis in nonobese diabetic mice. J. Immunol. 168:6007.24+) V11+ NKT cell numbers are decreased in a wide variety of diseases that are characterized by autoreactive tissue damage. Clin. Immunol. 100:144.[Medline]
-/- mice, but not in TNF--/- mice: role for IFN- in activating apoptosis of hepatocytes. J. Immunol. 159:1418.[Abstract]
plays a critical role in T cell-dependent liver injury in mice initiated by concanavalin A. Gastroenterology 111:462.[Medline]
1 (TGF-1) controls expression of major histocompatibility genes in the postnatal mouse: aberrant histocompatibility antigen expression in the pathogenesis of the TGF-1 null mouse phenotype. Proc. Natl. Acad. Sci. USA 90:9944. suppresses IFN- induction of class II MHC gene expression by inhibiting class II transactivator messenger RNA expression. J. Immunol. 158:2065.[Abstract]
. Annu. Rev. Immunol. 16:137.[Medline]Related articles in The JI:
This article has been cited by other articles:
|
|
 |
|
  R. T. Robinson, J. Wang, J. G. Cripps, M. W. Milks, K. A. English, T. A. Pearson, and J. D. Gorham End-Organ Damage in a Mouse Model of Fulminant Liver Inflammation Requires CD4+ T Cell Production of IFN-{gamma} but Is Independent of Fas J. Immunol., March 1, 2009; 182(5): 3278 - 3284. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Ghannad, D. Nica, M. I. Garcia Fulle, D. Grenier, E. E. Putnins, S. Johnston, A. Eslami, L. Koivisto, G. Jiang, M. D. McKee, et al. Absence of {alpha}v{beta}6 Integrin Is Linked to Initiation and Progression of Periodontal Disease Am. J. Pathol., May 1, 2008; 172(5): 1271 - 1286. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. H. Wrzesinski, Y. Y. Wan, and R. A. Flavell Transforming Growth Factor-{beta} and the Immune Response: Implications for Anticancer Therapy Clin. Cancer Res., September 15, 2007; 13(18): 5262 - 5270. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. T. Robinson and J. D. Gorham TGF-beta1 Regulates Antigen-Specific CD4+ T Cell Responses in the Periphery J. Immunol., July 1, 2007; 179(1): 71 - 79. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I.-K. Park, L. D. Shultz, J. J. Letterio, and J. D. Gorham TGF-{beta}1 Inhibits T-bet Induction by IFN-{gamma} in Murine CD4+ T Cells through the Protein Tyrosine Phosphatase Src Homology Region 2 Domain-Containing Phosphatase-1 J. Immunol., November 1, 2005; 175(9): 5666 - 5674. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. T. Lin, S. L. Martin, L. Xia, and J. D. Gorham TGF-{beta}1 Uses Distinct Mechanisms to Inhibit IFN-{gamma} Expression in CD4+ T Cells at Priming and at Recall: Differential Involvement of Stat4 and T-bet J. Immunol., May 15, 2005; 174(10): 5950 - 5958. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Devitt, K. G. Parker, C. A. Ogden, C. Oldreive, M. F. Clay, L. A. Melville, C. O. Bellamy, A. Lacy-Hulbert, S. C. Gangloff, S. M. Goyert, et al. Persistence of apoptotic cells without autoimmune disease or inflammation in CD14-/- mice J. Cell Biol., December 20, 2004; 167(6): 1161 - 1170. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), heterozygous (
), or null (
and 
), for RAG-1. 
