Contribution of the Lymphotoxin β Receptor to Liver Regeneration¹

The liver is a vital organ that performs many biological functions including synthesis, metabolism, and storage of carbohydrates, proteins, and lipids. The hepatocytes, which comprise the dominant cell population of the liver, effect these functions. Survival from insults such as viruses and toxins requires maintenance of hepatic mass. Therefore, in addition to the synthetic and metabolic function, the liver must continue to undergo self-renewal (1). Controlling the ability of the liver to regulate its mass has broad clinical implications. Organ shortages continue to press the need for liver transplantation alternatives, and the ability to increase the renewal of the liver would make split liver transplants and cellular transplants more feasible. In addition, flares of necroinflammatory activity seen in chronic hepatitis lead to increased hepatocyte turnover and fibrosis. Further understanding of liver growth control mechanisms offers a possible solution to both liver transplantation and chronic liver disease.

The partial hepatectomy is a particularly useful model for beginning to understand how the liver controls self-renewal (2). Partial hepatectomy, a procedure that removes 70% of the liver, is regarded as the preferred in vivo method to study liver growth due to its synchronized growth response in the liver (3, 4). The understanding of liver regeneration has focused on the cytokine and growth factor network involved in the process (5). These studies have defined TNF-α and IL-6 as important cytokines that prime the hepatocyte for response to growth factors such as hepatocyte growth factor, epidermal growth factor, and TGF-α (6, 7).

Besides release of cytokines and growth factors after partial hepatectomy, rapid changes in the cellular composition of the liver occur immediately after partial hepatectomy. For example, as early as 12 h after partial hepatectomy, there is an increase in the number of intrahepatic lymphocytes. This increase is due to an increase in the percentage of CD4-positive and NK T cells (8). However, the contribution of lymphocytes to the regenerative process has not been defined.

The lymphocyte-derived ligand, homologous to lymphotoxin (LT),³ exhibits inducible expression and competes with HSV glycoprotein D for herpesvirus entry mediator, a receptor expressed on T cells (LIGHT) is a TNF family member (TNF superfamily 14) that is expressed on activated T cells (9). LIGHT has two receptors, the LTβR, which is expressed on nonhemopoietic cells, and the herpes virus entry mediator receptor (HVEM), which is expressed on hemopoietic cells (9, 10). The expression of LIGHT on activated T cells can signal to lymphocytes via the HVEM receptor and to nonhemopoietic cells through the LTβR. Therefore, LIGHT performs a unique role in bridging hemopoietic and nonhemopoietic signaling.

In this study, we show that transgenic mice expressing LIGHT (Tg LIGHT) under the control of the T cell-restricted p56^lck promoter and CD2 enhancer display a massive increase in liver mass and markedly abnormal hepatocyte histology. The effect of LIGHT is dependent on the LTβR, which we demonstrate is expressed on hepatocytes. Immediately after partial hepatectomy, there is an increase in LIGHT and LTα from intrahepatic lymphocytes. Furthermore, LTβR-deficient mice (LTβR^−/−) show an impaired ability to regenerate their liver after partial hepatectomy, suggesting that the LTβR pathway is required for optimal liver regeneration. We demonstrate the importance of the LTβR signaling in supporting liver biology by demonstrating a deficiency in LTα also shows an impaired ability to undergo liver regeneration. Finally, we are able to show that treatment with a LTβR human fusion protein (LTβR-hFc) also blocks liver regeneration in wild-type mice. Therefore, this study establishes a key interaction between lymphocyte-derived ligands and hepatocytes and uncovers the LTβR and its ligands as a new pathway during liver regeneration.

Mice carrying a transgene for LIGHT have previously been reported (11). Male C57B/L6J, LTα^−/− mice (6–10 wk old) were a gift from D. Chaplin (Washington University, St. Louis, MO) and were backcrossed to a B6 background for 13 generations. LTβR^−/− mice were generated by K. Pfeffer (Institute of Medical Microbiology, Immunology, and Hygiene, Technical University of Munich, Munich, Germany) (12) and backcrossed to a B6 background for seven generations. HVEM^−/− mice were recently generated in K. Pfeffer’s laboratory and backcrossed to B6 for five generations with genetic targeting confirmed by Southern blot analysis and Ab staining (13). All mice were maintained at the University of Chicago and handled according to National Institutes of Health guideline and approved by the Institutional Animal Care and Use Committee. Bone marrow transplantation was performed by lethally irradiating recipients with 900 rad followed by i.v. injection of 3 million bone marrow cells which were isolated by flushing the femur and humerus of the donor with sterile PBS. Adoptive transfer was performed by harvesting the thymus from 6-wk-old mice and i.v. injection of 20 × 10⁶ cells into RAG1^−/− mice (The Jackson Laboratory). Treatment with LTβR-hFc (Biogen), was performed by i.v. injection of 200 μg of fusion protein 24 h before and immediately after partial hepatectomy. Anti-LTβR (ACH6) and negative control (Ha4/8) hamster Ab were kindly provided by J. Browning (Biogen).

The procedure was performed as originally described by Higgins and Anderson (4). The mice were anesthetized with ketamine and xylazine (100 mg/kg/10 mg/kg i.p.), the liver was exposed through a midline incision, and the median and left lobes were mobilized and delivered after ligature with a single silk 4-0 suture. Sham procedures consist of anesthetizing mice followed by opening of the skin and peritoneum and manipulating but not resecting the liver. Serum transaminase activity was determined by the method of Karmen using an Elan Diagnostics analyzer according to the manufacturer’s protocols in the University of Chicago Animal Research Core laboratory.

Intrahepatic lymphocytes were purified from the liver by pressing the liver through a steel mesh (no. 200) into PBS, centrifuged at 800 × g for 5 min with the resulting pellet suspended in a 35% Percoll-PBS-heparin (100 U/ml) solution and centrifuged at 800 × g for 20 min at room temperature. The pellet of mononuclear cells is cleared of RBC with a 5-min osmotic lysis (0.15 M NH₄Cl, 1 mM KHCO₃, 0.1 mM Na₂-EDTA, pH 7.3) and washed twice in PBS.

Lymphocyte RNA was extracted according to the manufacturer’s protocol and further purified using the RNeasy kit. cDNA was generated from 1 μg of total RNA after DNase treatment and combined with 200 U of Moloney murine leukemia virus reverse transcriptase (Promega), 750 nM dNTP, and 3 μg of random primers. LIGHT (forward, TCCGCGTGCCTGGAAA; reverse, AAGCTCCGAAATAGGACCTGG; probe, CGCCGGGTCAGACCACGTGACG), LTβ (forward, CACCTCATAGGCGCTTGGAT; reverse, ACGCTTCTTCTTGGCTCGC; probe, AGTGGGCAAGGGCTCAGCTGGG), LTα (forward, CCGACCTAGACCCACAAAAA; reverse, CTCCATCCTGACCGTTGTTT; probe, ATGCTCAGGGGACGACTCAAGCCTA), and GAPDH (forward, TCCACCACCATGGAGAAGGC; reverse, GGCATGGACTGTGGTCATGA; probe, TGCATCCTGCACCACCAACTGCTTAG) were detected by quantitative PCR using the TaqMan universal PCR master mixture (Applied Biosystems) combined with 300 nM either forward or reverse primer and 100 nM fluorescent oligonucleotide probe. The cycle threshold of amplification was determined using a Smartcycler and software (Cepheid). After generation of a standard curve for each gene, the expression levels for each gene were normalized to the expression of GADPH.

Frozen sections were cut from livers, allowed to air dry for 15 min, and incubated in PBS, 0.2% saponin, 5% normal goat serum for 15 min. Primary anti-LTβR Ab (clone ACH6) or control primary Ab (Ha 4/8), 10 μg/ml in 1% normal goat serum in PBS, was incubated on the tissue in a humid chamber for 2 h at room temperature. Secondary Ab consisted of goat anti-hamster IgG Alex 594 (Molecular Probes), 10 μg/ml, and Alexa 488 phalloidin (Molecular Probes) in 1% normal goat serum in PBS for 1 h at room temperature.

Tissue sections were fixed in 70% alcohol for 2 h followed by buffered formalin and processed either for routine H&E staining or for BrdU and immunohistochemical studies as per the manufacturer’s instructions (Zymed Laboratories). TUNEL staining was performed on paraffin-embedded, formalin-fixed tissue using the ApopTag Plus Peroxidase In Situ Apoptosis Detection kit (Chemicon International) according to the manufacturer’s directions. BrdU incorporation and TUNEL-positive nuclei were scored in 20 nonoverlapping high power fields (×40).

The liver regulates its mass through a strict homeostatic mechanism which maintains a 4–8% liver-body weight ratio (1). Much to our surprise, a TNF superfamily ligand, LIGHT (TNSF14), expressed under the proximal p56^lck promoter and CD2 enhancer, has a substantially increased liver size compared with littermate wild-type control mice (Fig. 1,A). When expressed as a percentage of total body weight, Tg LIGHT mice have significantly higher liver-body weigh ratios (Fig. 1,B). Liver histology reveals markedly enlarged hepatocytes and frequent mitotic figures in Tg LIGHT mice compared with age matched wild-type livers (Fig. 1, C and D). The hepatomegaly, abnormal histology, and frequent mitotic figures in Tg LIGHT mice raise the possibility that a T cell-derived cytokine disrupts the mechanisms that normally balance hepatocyte proliferation and death.

Next, we determine whether the effects of LIGHT on the liver could be transferred through bone marrow transplantation. The Tg LIGHT bone marrow transplanted into lethally irradiated wild-type mice resulted in massive hepatomegaly and abnormal liver histology. As expected, wild-type bone marrow transplant controls show normal liver size and histology (Fig. 2, A and B). To confirm that the changes in the liver were dependent on T cell-specific expression of LIGHT, adoptive transfer of T cells was performed. Twenty million thymocytes from 6-wk-old LIGHT Tg mice were transferred into RAG1^−/− mice. A similar substantial effect on the liver is seen in these mice with enlargement of hepatocytes compared with control RAG1^−/− mice that received 20 × 10⁶ wild-type thymocytes (Fig. 2, C and D). These findings support a novel idea that a T cell-derived ligand can have a profound effect on the liver homeostasis.

Given the effect of LIGHT on liver homeostasis, we wished to determine which of the two known putative receptors of LIGHT mediate this effect. Tg LIGHT mice were crossed to mice deficient in each of these receptors. The substantial effect of LIGHT on liver size and histology was lost when the Tg LIGHT mice were crossed with LTβR^−/− mice but not HVEM^−/− mice (Fig. 2, E and F). The LTβR-dependent effects of LIGHT on liver cells indicated the LTβR signaling pathway uniquely links lymphocyte-derived ligands to liver homeostasis. Therefore, it appears as though LTβR expression in the liver mediates the effect of LIGHT on the liver.

Next, we wished to determine whether the LTβR is expressed in the liver. An Ab to the LTβR and fluorescent secondary was used to detect LTβR expression in frozen sections of the mouse liver. Compared with LTβR^−/− liver, there was strong membranous expression of the LTβR on the hepatocytes (Fig. 3). Simultaneous staining with phalloidin was also performed to highlight the F-actin in the cytoskeleton of the hepatocyte. Merging the LTβR and phalloidin images confirms that the LTβR is expressed along the hepatocyte cellular membrane.

Because it appears as though signaling through the LTβR has a dominant effect on hepatocyte growth, we determined whether LIGHT and lymphotoxin expression was induced after a strong hepatocyte growth stimulus, partial hepatectomy. After the surgical removal of 70% of the liver mass in wild-type mice, lymphocytes were purified from the remnant liver, and RNA was isolated. There was an ∼3-fold induction of LIGHT mRNA expression in the intrahepatic lymphocytes, which is sustained from 4 to 12 h after partial hepatectomy (Fig. 4, □). LTα mRNA expression peaks at 4 h and begins to decline 12 h after partial hepatectomy (Fig. 4, ▦). As expected, the expression of LTβ mRNA remains at constant basal levels after partial hepatectomy with a slight increase at 12 h (Fig. 4, ▪). The induction of the LTβR ligand expression suggests that these ligands function in the inflammatory milieu that is rapidly induced after partial hepatectomy. These results indicate that hepatocytes express the LTβR and that the mRNA expression of LTβR ligands, LIGHT and LTα, are sharply induced in lymphocytes after partial hepatectomy.

Because the LTβR expression on the hepatocytes mediates liver proliferation in response to a T cell-derived ligand and LTβR ligands increase after partial hepatectomy, the relative contribution of LTβR to restoration of liver mass after partial hepatectomy was explored. Wild-type mice showed no significant increase in morbidity or mortality after partial hepatectomy, whereas there was a substantially increased morbidity and mortality of LTβR^−/− after partial hepatectomy (Fig. 5). We further examined the role of LTβR signaling after partial hepatectomy by determining whether mice deficient in another ligand known to signal through the LTβR, LTα, is also essential for liver regeneration. Mice deficient in LTα^−/− also displayed a significant decrease in survival after partial hepatectomy (Fig. 5). It suggests the essential role of LTβR signaling by LT during liver regeneration.

These data led us to consider whether decreased survival in the absence of LTβR signaling had a direct effect on the liver. LTβR^−/− and LTα^−/− mice showed evidence of liver damage 48 h after partial hepatectomy with significantly elevated serum aminotransferase levels compared with similarly treated wild-type mice (Fig. 6,A). Histological examination of the livers showed larger areas of necrosis in the LTβR^−/− and LTα^−/− mice 48 h after partial hepatectomy compared with wild-type mice (Fig. 6 B).

We also investigated whether there is an increase in the number of apoptotic hepatocytes using TUNEL staining. We could clearly demonstrate the number of TUNEL-positive nuclei were significantly increased in LTβR^−/− compared with wild-type mice at 24 and 48 h after partial hepatectomy (Fig. 6 C). Evidence that LTβR^−/− mice develop extensive areas of hepatocyte death when challenged with partial hepatectomy is consistent with an in vitro analysis that the LTβR-mediated microenvironment transduces viability signals in hepatocytes (14).

Next, we wanted to determine whether LTβR signaling was involved in initiating DNA synthesis after partial hepatectomy. Mice received a 2-h pulse with a thymidine analog, BrdU, 48 h after partial hepatectomy or a sham procedure. This allowed us to carefully examine the proliferation of the hepatocyte after partial hepatectomy by counting the number of BrdU-positive hepatocyte nuclei after immunohistochemical staining. There was significantly more DNA synthesis in the wild-type mice than in either LTβR^−/− or LTα^−/− mice (Fig. 7 A). As expected, sham procedures did not induce DNA synthesis.

The inability of the LTβR^−/− mice to synthesize DNA after partial hepatectomy was not due to a developmental lesion because short term treatment with LTβR-hFc also blocked the initiation of DNA synthesis in wild-type mice (Fig. 7 B). This indicated that signaling through the LTβR as well as a LTβR ligand(s) was important for the ability of the liver to regenerate hepatic mass. The combination of increased liver damage and a reduced regenerative response after partial hepatectomy induces significant mortality in mice that cannot activate the LTβR. Together, these data suggest that LTβR signaling through T cell-restricted ligands is critical for liver homeostasis.

In this study, we have revealed for the first time that signaling through the LTβR can have a substantial effect on the liver and that lymphocytes can play a novel role in liver regeneration. Our initial observation of proliferative and massively enlarged livers in Tg LIGHT mice led us to investigate two major questions. We wanted to know 1) whether a T cell-restricted ligand could have a dramatic liver phenotype and 2) what effect the ligand and receptor pathway has on the liver after partial hepatectomy.

We have previously demonstrated the expression of LIGHT on the T cells of Tg LIGHT mice; this agrees with what is known about the p56^lck promoter and CD2 enhancer that drives the LIGHT transgene (11, 15, 16, 17). Tg LIGHT mice have marked T cell infiltration of peripheral organs. Because there was a correlation between the presence of increased LIGHT expressing T cells and enlargement of the liver, we sought to determine whether this was a direct or indirect effect. In support of the direct action of T cell-derived LIGHT, we have found that either bone marrow transplantation or transfer of thymocytes from Tg LIGHT mice into wild-type or RAG1^−/− mice recipients resulted in the transfer of the abnormal liver phenotype. In addition, the direct effect of LIGHT on the liver was dependent on LTβR expressed on mature hepatocytes, but not on the HVEM receptor. This suggests that LIGHT was directly mediating effects on the liver. Therefore, based on our observations, it appears as though the LTβR functions as a bridge between LIGHT-expressing T cells and LTβR expression on hepatocytes.

In the Tg LIGHT mice, there exists a strong likelihood of a direct interaction between the T cell and the hepatocyte. The Tg LIGHT mice were engineered to retain LIGHT on the surface of the T cell through the introduction of a mutation at a suspected cleavage site. In the nontransgenic system, LIGHT is expressed on activated T cells, with few other cell types expressing LIGHT. It is possible that activated T cells deliver LIGHT to the liver and that LIGHT is then cleaved from the T cell membrane. Soluble LIGHT could then interact with the LTβR expressed on hepatocytes. The latter scenario is analogous to what occurs with other TNF superfamily ligands (18).

The physiological role of LIGHT and LTβR in liver has not been defined. The increased expression of LIGHT in our Tg mice was used to hint at LIGHT’s potential role during inflammation. There is in vitro evidence to suggest LIGHT does have a direct role on hepatocytes. Treatment of cultured human hepatocytes with LIGHT protects against TNF-mediated apoptosis (14). LIGHT and the LTβR does not appear to be essential for liver development during embryogenesis because mice deficient in LIGHT or LTβR appear to have normal liver histology (data not shown).

LTβR signaling by LIGHT appeared to mediate a proliferative environment in the liver, so we sought to define the role of the LTβR in liver growth. Partial hepatectomy is an in vivo method to deliver a strong growth stimulus to hepatocytes. First, we were able to show an increase in the mRNA of lymphocyte-derived LTβR ligands, LTα and LIGHT, immediately after partial hepatectomy. Second, a deficiency in the LTβR had a profound effect on the liver following partial hepatectomy. Not only was there a reduction in hepatocyte DNA synthesis but also there was significant liver damage and high mortality after partial hepatectomy. Third, the impaired liver-regenerative response could be replicated in wild-type mice treated with LTβR-Fc, suggesting a direct role for the LTβR. The LTβR appears to be uniquely situated to balance cellular death and DNA synthesis in hepatocytes after partial hepatectomy. LTβR signaling and the relative contribution of LIGHT and LTα to normal liver cell turnover will have to be explored.

With this study, we have been able to uncover an unexpected role of lymphocyte-restricted ligands and defined a new pathway in supporting liver regeneration. This system will allow us to 1) explore the environment created by activated lymphocytes during liver regeneration, 2) explore the events downstream of the LTβR that either protect hepatocytes from cell death, or 3) promote proliferation after partial hepatectomy.

The authors have no financial conflict of interest.