Failure of CD4 T-Cells to Respond to Liver-Derived Antigen and to Provide Help to CD8 T-Cells

CD4 T-cell help is required for the induction of efficient CD8 T-cells responses and the generation of memory cells. Lack of CD4 T-cell help may contribute to an exhausted CD8 phenotype and viral persistence. Little is known about priming of CD4 T-cells by liver-derived antigen. We used TF-OVA mice expressing ovalbumin in hepatocytes to investigate CD4 T-cell priming by liver-derived antigen and the impact of CD4 T-cell help on CD8 T-cell function. Naïve and effector CD4 T-cells specific for ovalbumin were transferred into TF-OVA mice alone or together with naïve ovalbumin-specific CD8 T-cells. T-cell activation and function were analyzed. CD4 T-cells ignored antigen presented by liver antigen-presenting cells (APCs) in vitro and in vivo but were primed in the liver-draining lymph node and the spleen. No priming occurred in the absence of bone-marrow derived APCs capable of presenting ovalbumin in vivo. CD4 T-cells primed in TF-OVA mice displayed defective Th1-effector function and caused no liver damage. CD4 T-cells were not required for the induction of hepatitis by CD8 T-cells. Th1-effector but not naïve CD4 T-cells augmented the severity of liver injury caused by CD8 T-cells. Our data demonstrate that CD4 T-cells fail to respond to liver-derived antigen presented by liver APCs and develop defective effector function after priming in lymph nodes and spleen. The lack of CD4 T-cell help may be responsible for insufficient CD8 T-cell function against hepatic antigens.


Introduction
While activation of CD8 T-cells proceeds in the absence of CD4 T-cells, the latter's help is crucial for the development of long lasting memory. Although the primary response of CD8 T-cells activated in the absence or presence of CD4 T-cells is comparable, CD8 T-cells generated in the absence of CD4 T-cell help display poor recall responses and produce little cytokines upon their second encounter with antigen [1,2]. CD8 T-cells primed in the absence of CD4 T-cell help display an exhausted phenotype characterized by high levels of PD-1 [3]. This exhausted phenotype and the lack of CD4 T-cell help may contribute to the persistence of virus in chronic hepatitis. Indeed, lack of CD4 T-cells help may explain why CD8 T-cells fail to eradicate hepatitis C virus [4].
The liver represents a unique environment for antigen presentation because several populations of professional and non-professional antigen-presenting cells (APCs) reside there, and its environment in general skews T-cell responses towards tolerance (reviewed in [5]). In addition to the possibility of antigen presentation within the liver by resident dendritic cells (DCs) [6], Kupffer cells (KCs) [7], liver sinusoidal endothelial cells (LSECs) [8], and hepatocytes [9], the liver is also sampled by dendritic cells, which subsequently migrate to secondary lymphatic organs and present antigen there [10]. While CD4 T-cells recognize antigen obtained by phagocytosis and presented by APCs, CD8 T-cell responses are restricted to antigens derived from brake-down products within the target cell. However, through a mechanism termed cross-presentation, CD8 T-cells may also respond to exogenous or endogenous antigens presented by professional APCs after uptake and processing [11].
CD4 T-cells need to have access to the same antigen that is recognized by CD8 T-cells to be able to provide help. We and others have developed mouse models, in which a neoantigen is expressed in the liver and which allow the investigation of both CD8 and CD4 T-cell activation by this antigen [12,13,14,15]. Using these models, it has become evident that CD8 T-cells primed in the liver by APCs acquire effector function rather than become tolerant. The ability of liver APCs to prime CD4 T-cells in an antigen specific manner is less well characterized. Expression of MHC-II molecules on hepatocytes triggers CD4 T-cell activation [16], but T-cells primed under these conditions are defective [17]. LSECs constitutively express MHC-II but it is still a matter of debate whether they are capable of stimulating CD4 T-cells [18,19,20,21]. In vitro experiments demonstrate that CD4 T-cells primed by LSECs lack effector function [18] or display a regulatory phenotype [22]. CD4 T-cells primed by liver DCs in vitro proliferate, yet little Th1 effector cytokines are produced [23].
We have analyzed CD4 T-cell activation by liver-derived antigen in vitro and in vivo and investigated the role of CD4 T-cell help in a CD8 T-cell dependent model of immune-mediated liver injury.

Failure of naïve CD4 T-cells to respond to liver-derived antigen in vitro
To determine the capability of specific populations of liver APC to activate CD4 T-cells, we purified DCs, KCs, or LSECs from the livers of TF-OVA mice, which express ovalbumin in hepatocytes. The endogenous antigen should be present in the different types of APCs, which acquire the antigen through phagocytosis of hepatocyte remnants. Therefore, incubation of OT-II CD4 or OT-I CD8 T-cells with specific APCs will test the APCs' ability to stimulate naïve T-cells with an antigen that is derived from an endogenous source and presented or cross-presented after processing.
OT-I T-cell proliferation was observed after incubation with DCs, KCs, or LSECs, indicating that liver-derived antigen had indeed been taken up and processed by the APCs (Fig. 1). Nevertheless, OT-II T-cells did not proliferate when incubated with either one of the APCs presenting liver-derived antigen. When liver APCs were incubated with exogenous ovalbumin beforehand, proliferation of CD4 T-cells was restored, indicating that APCs are in principle capable of processing antigen and activating CD4 T-cells (data not shown). DCs isolated from the spleens of TF-OVA mice induced proliferation in both OT-I and OT-II T-cells. These data indicate that liver APCs are incapable of stimulating naïve CD4 T-cells with liver-derived antigen in vitro although they do stimulate CD8 T-cells.
Failure of naïve CD4 T-cells to respond to liver-derived antigen in vivo Similar to the observed effects in vitro, no priming of naïve OT-II T-cells in the liver was observed in vivo [13]. Rather, activation occurred in spleen and liver-draining lymph node ( Fig. 2A). Since the majority of CD4 T-cells were activated in the spleen, we transferred naïve OT-II T-cells into splenectomized TF-OVA mice. Priming of OT-II T-cells was observed in the liver draining lymph node but not in the liver in splenectomized TF-OVA mice. Our results imply that the spleen is dispensable for the priming of naïve CD4 T-cells and that APCs within the liver are insufficient to induce their priming in vivo, leaving the liver draining lymph node as the key location for priming of CD4 T-cells by liverderived antigen.
To investigate more thoroughly the requirements for the priming of naïve CD4 T-cells by liver-derived antigen in vivo, we generated bone-marrow chimeras, in which professional APCs are incapable of presenting the antigen to OT-II T-cells. In TF-OVA mice reconstituted with bone marrow from MHC-II 2/2 mice, no priming of OT-II T-cells was observed, even after a prolonged observation period of 68 h (Fig. 2B), indicating that professional bone-marrow derived APCs are required for the priming of naïve CD4 T-cells by liver-derived antigen. At the 68 h time point, proliferating CD4 T-cells were retrieved from the liver. The lack of proliferation at the earlier time point and the fact that only cells were retrieved that had divided multiple times suggest that these cells had migrated to the liver after activation in spleen and lymph  nodes. CD4 T-cells retrieved from the liver exclusively belonged to a population that had undergone several rounds of proliferation in consistence with the liver's preference to recruit activated rather than naïve T-cells [24].
Inflammation of the liver leads to up-regulation of MHC-II molecules on hepatocytes as well as to maturation of professional APCs. Both mechanisms could influence antigen presentation to CD4 T-cells and potentially allow their priming within the liver. To test whether this is the case, we induced hepatitis by transfer of OT-I T-cells prior to transfer of OT-II T-cells. At the peak of inflammation (i.e. 6 days after transfer of OT-I T-cells), we confirmed the presence of inflammation by measuring ALT-levels ( Fig. 3A) before transferring carboxy-fluorescein succinimidyl-ester (CFSE)-labeled OT-II T-cells. Migration of OT-II T-cells to the liver and spleen was evaluated at 20 h, before proliferation of CD4 T-cells contributes to increased cell numbers. Significantly increased numbers of CFSE + cells were observed in liver and spleen of mice suffering from hepatitis as compared to controls, demonstrating enhanced recruitment of CD4 T-cells to these organs (Fig. 3B). However, even under inflammatory conditions, we observed virtually no proliferation of transferred OT-II T-cells in the liver, as compared to the spleen (Fig. 3C). Thus, an inflammatory response does not promote antigen presentation sufficiently to allow priming of naïve CD4 T-cells by liver-derived antigen in the liver.
CD4 T-cells primed by liver-derived antigen display deficient Th1-effector function Next we investigated whether naïve CD4 T-cells primed in TF-OVA mice exert effector function. To this end, CFSE-labeled OT-II T-cells were transferred into TF-OVA mice. To prove that activation is antigen-dependent, controls were carried out in B6 mice. Sixty-eight hours after transfer, lymphocytes were isolated from liver and spleen and re-stimulated in vitro. While restimulation induced considerable production of IL-2, very little Interferon-c was produced by OT-II T-cells retrieved from liver or spleen (Fig. 4A). To test whether OT-II cells primed in TF-OVA mice acquired a regulatory phenotype, we stained cells retrieved from liver and spleen for CD25 and FoxP3. Less than one percent of CFSE + CD4 + T-cells expressed CD25 and FoxP3 simultaneously (Fig. 4B), illustrating that induction of a regulatory phenotype in OT-II T-cells primed in TF-OVA mice is a rare event. Thus, activation of naïve CD4 T-cells by liver-derived antigen resulted in induction of cells capable of producing IL-2 but with a defective Th1-response, rather than in induction of complete anergy or a regulatory phenotype.

Effector CD4 T-cells accumulate in the liver of TF-OVA mice
Given the fact that naïve CD4 T-cells are not primed in the liver of TF-OVA mice, we tested whether effector CD4 T-cells migrate to the liver of TF-OVA mice and recognize their cognate antigen there. Effector OT-II cells with a Th1-phenotype were generated in vitro. Their ability to produce Interferon-c was demonstrated after in vitro re-challenge with PMA/ionomycin (Fig. 5A). Twenty hours after transfer into TF-OVA or B6 control mice, effector OT-II T-cells were present in the liver in larger quantities than naïve T-cells, independently of antigen-expression (Fig. 5B, 20 h). Proliferation of these cells was observed 68 h after transfer in liver, spleen, and draining lymph nodes of TF-OVA mice (Fig. 5B, 68 h). A higher degree of CFSE-dilution was observed in TF-OVA mice than in B6 controls, indicating antigen-specific activation of the transferred effector OT-II T-cells. In contrast, no proliferation occurred after transfer of naïve OT-II T-cells into B6 mice.
Next we tested whether OT-II T-cells accumulate in the liver of TF-OVA mice after antigen recognition. To this end, 4 million naïve or effector OT-II T-cells were transferred into TF-OVA mice, and lymphoid organs and the liver were removed to enumerate OT-II T-cells at different time points. Since CFSEdilution is complete after 3-4 days, the Va2-chain of the OT-II Tcell receptor was used to estimate the number of OT-II T-cells. The increase of CD4 + Va2 + cells in livers of TF-OVA mice was much more pronounced than in livers of B6 control mice after transfer of effector T-cells (Fig. 5C). In the spleen, numbers of CD4 + Va2 + cells increased on days 2 and 3 but decreased to background levels at day six. In contrast, no accumulation of CD4 T-cells was observed in the liver after transfer of naïve CD4 Tcells, which accumulated solely in the spleen. In TF-OVA mice injected with effector OT-II T-cells, a dense lymphocytic infiltrate in the portal and periportal areas was observed 6 days after transfer (Fig. 5D).
Our results confirm that activated CD4 T-cells accumulate in the liver of TF-OVA mice and respond to antigen.

Immune-mediated liver injury is amplified by effector but not naïve CD4 T-cells
We investigated the capacity of naïve and effector OT-II T-cells to induce liver damage. No significant increase in ALT-levels was observed after transfer of 4 million naïve or effector OT-II T-cells into TF-OVA mice (Fig. 6A), demonstrating that neither naïve nor effector CD4 T-cells are sufficient to induce hepatitis.
Since provision of help to CD8 T-cells is one of the main tasks of CD4 T-cells, we examined whether hepatitis induced by CD8 T-cells is augmented by addition of CD4 T-cells. The combination of naive OT-II and OT-I T-cells did not increase the severity of hepatitis compared to OT-I T-cells alone, even when transfer of CD4 T-cells preceded the transfer of CD8 T-cells by three days to ensure presence of CD4 T-cells in the liver at the time of CD8 Tcell priming. In contrast, the addition of effector OT-II T-cells to OT-I T-cells increased the severity of hepatitis, as judged from ALT-release. Three animals with very high ALT-levels died during the course of the experiment due to fulminant liver damage, while no animal died in any of the other groups. Extensive necrosis was present in the livers (data not shown). Therefore, we reduced the number of OT-I T-cells in the cotransfer experiment to such an extent that transfer of CD8 T-cells alone did not cause ALT-release (Fig. 6B). Co-transfer of effector OT-II T-cells with low numbers of OT-I T-cells lead to ALTrelease, thus converting a sub-threshold damage elicited by CD8 T-cells alone to overt hepatitis when CD4 T-cell help is provided.
To prove that amplification of CD8 T-cell function by CD4 Tcell help is responsible for this observation, we determined the degree of cytolysis exhibited by OT-I T-cells in different settings. In vivo cytolytic activity was increased by the addition of effector OT-II T-cells (Fig. 6 C, D), suggesting that augmentation of CD8 T-cell activity is indeed responsible for the observed effect on the severity of liver damage.

Discussion
The advent of mouse models of autoimmune hepatitis that do not rely on alloreactivity of T-cells [25,26] provides experimental systems to study the role of T-cells in a model, in which professional as well as non-professional APCs contribute to T-cell priming [12,13,14]. These models have challenged the paradigm that CD8 T-cells primed by liver-derived antigen are rendered tolerant. Rather, effector function and hepatitis result from CD8 T-cell activation by liver-derived antigen when professional APCs are presenting the antigen. However, the effect of CD4 T-cells in such a setting remains elusive. We report here on the role of CD4 T-cells in an antigen specific model of immune-mediated liver injury.
We employed exogenous TCR-transgenic CD4 T-cells reactive to the neoantigen expressed in the liver and investigated their activation by different types of APCs in vivo and in vitro. We demonstrate that priming is achieved in the liver-draining lymph node and spleen by professional APCs but not by APCs within the liver. This finding is in sharp contrast to our findings on CD8 Tcells, which are efficiently primed within the liver. No proliferation was observed of CD4 T-cells primed in vitro on LSECs or in vivo in bone-marrow chimeras, in which only non-professional APCs are capable of presenting the antigen. Thus, LSECs are insufficient to prime CD4 T-cells by liver-derived antigen, at least under noninflammatory conditions. Likewise, KCs and DCs isolated from the liver did not prime CD4 T-cells in vitro, and the absence of proliferating CD4 T-cells in the liver demonstrates that they do not activate CD4 T-cells by liver-derived antigen in vivo, either. Wuensch et al. also observed a lack of proliferation of CD4 T-cells in response to antigen introduced into hepatocytes by viral infection, albeit in a model in which the antigen is not crosspresented by professional APCs [27]. Therefore, direct comparison of the priming ability of specific APC populations on CD8 and CD4 T-cells is not possible. Our data indicate that the level of liver-derived antigen acquired by liver APCs in vivo suffices to stimulate CD8 T-cells. It is unclear whether the amount of antigen presented by the various APCs is too little to allow activation of CD4 T-cells or whether the low expression levels of MHC-II as well as of co-stimulatory molecules on liver APCs (reviewed in [28]), particularly KCs and LSECs, prevent activation. Observations obtained after addition of large amounts of exogenous antigen in vitro probably do not reflect the physiological situation in vivo and should therefore be interpreted with caution.
In a study by Lüth et al. [29], in vivo activation of CD4 T-cells by antigen expressed in the liver was studied. MBP-specific CD4 Tcells isolated from spleen and liver of mice transgenic for MBP expressed in hepatocytes proliferated, and a regulatory phenotype resulted, leading to suppression of experimental autoimmune encephalitis caused by autoreactive CD4 T-cells. The authors did not determine the type of APC and the site of activation leading to induction of the regulatory phenotype. In our experiments, only a small minority (,1%) of OT-II T-cells retrieved from liver or spleen displayed a regulatory phenotype as determined by expression of CD25 and FoxP3. However, we cannot exclude a regulatory function in the absence of FoxP3 expression, and even the small population of cells with a regulatory phenotype could suffice to suppress effector T-cells, as demonstrated by Lüth et al. Varying levels of antigen expression under the transgenic promoters or differences in the uptake or processing of the antigen by APCs may also contribute to the observed discrepancies.
Our findings have implications for the understanding of T-cell reactivity towards liver antigen. CD4 T-cells are not activated by liver-derived antigen in contrast to CD8 T-cells, possibly explaining why CD4 T-cell help is found infrequently in patients with viral hepatitis and exhaustion of the CD8 T-cell response results, leading to viral persistence. On the other hand, our data suggest that autoreactivity against liver antigen is most likely triggered by CD8 and not CD4 T-cells. Patients with autoimmune hepatitis display a mixed infiltrate of inflammatory cells in the liver, including CD8 and CD4 T-cells. In initial reports, CD4 T-cells were suggested as the main effector population causing liver damage in autoimmune hepatitis [30,31,32]. More recently it has become evident that CD8 T-cells play a dominant role, especially in the early phase of the disease [33,34]. Our data support this notion.
In summary, our data add to existing evidence for a pivotal role of CD8 T-cells in the development of acute immune-mediated liver injury. Without CD8 T-cells, no liver damage is observed. Naïve CD4 T-cells are not primed within the liver by liver-derived antigen, neither by professional nor by non-professional APCs, but require activation in spleen or lymph nodes before re-locating to the liver. Effector CD4 T-cells are capable of infiltrating the liver but do not cause liver damage by themselves. While help provided by effector CD4 T-cells augments the severity of liver damage, CD4 T-cell help is dispensable for the induction of liver injury in the acute setting. CD4 T-cells likely play a more crucial role in maintaining the prolonged activation of CD8 T-cells observed in chronic hepatitis and in the formation of memory CD8 T-cells.

Animals and cells
All animals received humane care according to institutional criteria. All animal procedures were approved by the Landesamt für Gesundheit und Soziales, Berlin (registrations G0020/04 and G0191/09).
TF-OVA mice were described before [13]. Bone-marrow chimeras were generated by lethal irradiation of TF-OVA mice, followed by supplementation with C57BL/6 I-Ab 2/2 bonemarrow (Taconic, Hudson, NY). Mice were used for experiments after 6 weeks. Splenectomy was carried out under anesthesia with xylazin/ketamin, and mice were used for experiments after 2 weeks.
Naïve OT-I CD8 and OT-II CD4 T-cells were isolated from lymph nodes and spleen of OT-I [36] or OT-II [37] mice, respectively, using magnetic sorting (Miltenyi Biotec, Bergisch- Gladbach, Germany). Purity of preparations was above 90% for OT-I T-cells and above 80% for OT-II T-cells. For in vitro coculture experiments and generation of CD4 effector cells, CD4 + Tcells were further selected for CD62L expression using magnetic beads (Miltenyi Biotec).
Liver APCs were isolated from TF-OVA mice by perfusion of the liver with 0.5 mg/ml collagenase in RPMI (Sigma, Taufkirchen, Germany). Fragmented livers were incubated for 20 min in digestion media at 37uC in an agitating incubator, passed through a nylon mesh, and briefly centrifuged at 300 rpm. Non-parenchymal cells were purified from the supernatant on a 42% Percoll gradient (Sigma, 2000 rpm, 20 min). Cells were separated into LSEC + and LSEC 2 fractions by magnetic separation with anti-CD146 beads (Miltenyi Biotec). The LSEC + fraction was further sorted by FACS for CD31 + CD45 2 LSEC (both antibodies from ebioscience, San Diego, CA), resulting in .99% purity. From the LSEC 2 fraction, KCs or DCs were isolated by FACS-sorting, resulting in .90% purity of F4/80 + CD11b + KCs or in .95% purity of CD11c + DCs, respectively (antibodies from ebioscience and BD Biosciences, Heidelberg, Germany). Spleen DCs were obtained by FACS sorting of CD11c + cells, resulting in .95% purity.
For in vitro co-culture experiments, 2610 5 APCs were cultured in 96 well flat bottom wells in 200 ml RPMI supplemented with 10% FCS, 1% penicillin/streptomycin, 1% gentamycin (Biochrom, Berlin, Germany), 1% L-Glutamin (Invitrogen, Karlsruhe, Germany), and 0.05 mM b-mercaptoethanol (Sigma), and allowed to settle for 24 h before addition of 1610 5 carboxy-fluorescein succinimidyl-ester (CFSE, Invitrogen, Karlsruhe, Germany) labeled T-cells [38]. Proliferation of OT-II or OT-I T-cells was analyzed after three days. OT-II and OT-I T-cells incubated alone were used as negative control. OT-II or OT-I T-cells incubated with irradiated spleen cells or specific APC populations and ovalbumin (100 mg/ml) served as positive control.

Histology
For histology, livers were perfused with PBS and fixed for 24 h in 4% paraformaldehyd, followed by embedding in paraffin. 4 mm sections were stained with hematoxylin and eosin.

FACS analysis
To isolate intrahepatic lymphocytes, livers were perfused with PBS/0.5% BSA, fragmented, passed through a 70 mm nylon mesh, and then treated as described above, before purification on a discontinuous 40/70% Percoll gradient. Antibodies against CD4, CD25, and Va2 were from BD-Biosciences. For intracellular staining, cells were fixed and permeabilized using BD-Cytofix/CytopermTM (BD-Biosciences). Staining was performed with antibodies anti-Interferon-c, anti-IL-2 (eBioscience), and anti-IL-4 (BD-Biosciences). Intracellular staining for FoxP3 was performed using the APC-FoxP3 staining set (eBioscience), according to the manufacturer's instructions. Cells were analyzed on a Becton Dickinson FacsCalibur using the CellQuest software.

T-cell restimulation and in vivo cytolysis assay
For restimulation experiments, 4 million CFSE-labeled OT-II T-cells were transferred into TF-OVA mice. Mice were sacrificed at day 3, non-parenchymal cells were isolated from liver and spleen, cultured in complete RPMI, and activated in 20 nM phorbol 12-myristate 13-acetate/1 mM ionomycin (Sigma). After 1 hour, 2 mg/ml brefeldin-A (Sigma) was added, and cells were analyzed for intracellular cytokines after an additional 3 hours. Restimulation of in vitro generated effector cells was performed likewise.
For in vivo cytolysis assays, 4 million naïve or effector OT-II Tcells, 1 or 4 million naïve OT-I T-cells, or a combination of 1 or 4 million OT-I T-cells and 4 million naïve or effector OT-II T-cells were transferred into TF-OVA mice. Six days later, splenocytes from C57/BL6 mice were labeled in 7,5 mM or 0,75 mM CFSE and pulsed with 1 mg/ml SIINFEKL or left untreated for 1 hour, respectively. SIINFEKL-pulsed splenocytes and control splenocytes were mixed in equal numbers, and 8 million splenocytes were injected intravenously. As a control, splenocytes were injected into mice that had not received T-cells. After 5 hours, cells were isolated from lymphatic organs and liver and analyzed for CFSE staining. Specific lysis was calculated as follows: 100x[1-(%CFSE lo (control)/%CFSE high (control))/(%CFSE lo (OT-I)/ %CFSE high (OT-I))].

Alanine aminotransferase measurement
Blood was collected into separation tubes and sera were stored at 220uC before automatic analysis on a Roche modular analyzer (Grenzach-Wyhlen, Germany).

Statistical analysis
Statistical analyses were performed with GraphPad Prism 5 software (GraphPad Software, San Diego, CA). The Student' s t test and the Mann-Whitney test were used as applicable.