Secondary Logo

Journal Logo

Original Article

MHC Class I Assembly Function and Intracellular Transport Routes for Hepatitis B Virus Antigen Cross-presentation by Heat Shock Protein gp96

Qin, Lijuan1,2; Liu, Yongai1,2; Xu, Yuxiu1; Li, Yang1; Hu, Jun1; Ju, Ying1; Zhang, Yu1; Wang, Shuo1; Li, Zihai3; Li, Changfei1; Li, Xin1,∗; Meng, Songdong1,2,∗

Editor(s): Wang, Haijuan

Author Information
Infectious Diseases & Immunity: July 2022 - Volume 2 - Issue 3 - p 183-192
doi: 10.1097/ID9.0000000000000058
  • Open



Heat shock protein (HSP) gp96, an endoplasmic reticulum (ER)-resident chaperone of the HSP90 family, is uniquely associated with a wide range of antigenic peptides from tumors and viruses. Studies on rodent models and clinical trials have demonstrated that gp96–peptide complexes purified from tumors or cell-based gp96-Ig-secreting vaccines generate antigen-specific T-cell immunity against tumors.[1–3] The gp96-based prophylactic and therapeutic vaccines against various pathogens, including influenza virus, human papillomavirus (HPV), herpes virus, simian immunodeficiency virus (SIV), and hepatitis B virus (HBV), have been tested in mouse models for antiviral cytotoxic T-cell (CTL) responses. Although conventional inactivated influenza vaccines have low cross-protective efficiency, adding gp96 as an adjuvant effectively improves cross-protection against challenge by a heterologous virus by boosting the T cell response against conserved viral epitopes.[4,5] The recombinant HPV16 oncoprotein E7-gp96 fusion protein was shown to induce a higher IFN-γ+ T-cell response than E7 protein alone and significantly delayed tumor occurrence and growth in mice.[6] Herpesvirus epitopes complexed in vitro with gp96 elicit specific CTLs in immunized mice.[7] Vaccination with cells secreting gp96-Ig carrying viral peptides protects macaques from highly pathogenic SIV infections.[8] In addition, immunization with a combination of HBV surface and core antigen formulated with gp96 led to a marked enhancement in specific cellular responses toward the virus and significantly increased the antiviral effect of vaccination.[9] As T-cell–mediated immunity plays a major role in viral control and clearance, these studies provide valuable insights into the use of gp96 as an adjuvant in treating viral infections. Many roles have been proposed for gp96 in T-cell activation. It is now apparent that all of these are direct or indirect results of one of the two basic functions: its role as a chaperone for antigen presentation to MHC molecules and its role in the induction of innate immunity. In response to intradermal or subcutaneous immunization, gp96–antigen complexes access draining lymph nodes and are predominantly internalized by CD11b+ macrophages.[10] Chaperoning by gp96 substantially increased the efficiency of antigen uptake through the gp96 receptor CD91 via an antigen scavenger receptor-mediated mechanism.[1,11,12] Endocytosed gp96 may translocate to the cytosol and cross-present the associated peptides to MHC class I (MHC-I) molecules, facilitating prime effector function in cognate CD8+ T-cells.[2,13] Gp96 and calreticulin in the ER constitute a relay line for cellular peptide transfer to MHC-I molecules in a concerted and regulated manner.[14] In addition, gp96 interacts with CD91 or Toll-like receptors, leading to the maturation of antigen-presenting cells (APCs), secretion of cytokines, and priming of specific Th cell subsets.[15–17] More recently, gp96 was found to selectively activate cytokine production in NK cells via APCs, which is required for T-cell function.[18]

Although gp96 has been shown to possess a unique ability for antigen binding and presentation to generate CTL immunity against viruses and cancer, the precise cellular compartment of antigen delivery in this process and its regulation of the assembly of MHC-I molecules with antigenic peptides are largely unclear. In addition, alternative pathways for exogenous antigen presentation and CD8+ T-cell activation by gp96 have been proposed because no reduction in the generation of MHC-I–peptide complexes was observed under gp96 depletion in mouse fibroblasts.[19] Consequently, the entire route of antigen presentation by gp96 requires further investigation, and it remains largely unexplored whether endogenous gp96 directly interact with MHC-I assembly complexes.

Antigen presentation and MHC-I peptide loading are critical for generating antiviral and antitumor cytotoxic CD8+ T-cells. The assembly of MHC-I molecules with antigenic peptides in the ER is a coordinated and regulated process that requires the concerted action of the cytosolic proteasome, transporter associated with antigen processing (TAP) for peptide delivery into the ER, and accessory chaperones for MHC-I loading in the ER, including calnexin, calreticulin, tapasin, and ERp57.[20–22] However, the overall architecture of the peptide-loading complex (PLC) and its underlying molecular mechanisms remain only partially understood.

This study investigated the direct role of gp96 in the regulation of the assembly of MHC-I molecules with HBV-derived peptides. Furthermore, we explored the mechanisms underlying the antigen-presenting function of recombinant gp96 as an adjuvant to activate specific T cells.

Material and methods


Hsp90b1flox/flox C57BL/6 mice were a gift from Professor Zihai Li.[23] Hepatocyte-specific gp96 KO mice were generated by crossing albumin-Cre mice with Hsp90b1flox/flox mice to obtain Albumin-Cre-Hsp90b1flox/flox mice. HLA-A2.1 C57BL/6 mice were purchased from the Jackson Laboratory. Animal studies were performed according to the guidelines of the Institute of Microbiology, Chinese Academy of Sciences of Research Ethics Committee (approval number PZIMCAS2011001).

Cell culture

Human hepatoma cell lines HepG2, 293T, and RAW264.7 were obtained from the American Type Culture Collection (Manassas, VA, USA) and cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (FBS) and 100 U/mL streptomycin/penicillin. The mouse embryonic H-2b fibroblast cell line K41 and the T-cell hybridoma cell line B3Z were maintained in Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with 10% FBS and 100 U/mL penicillin/streptomycin.

Antibodies and reagents

Conformation-sensitive mouse anti-H-Kb monoclonal antibody, mouse anti-gp96 monoclonal antibody, anti-EEA1 antibody, anti-LAMP-1 antibody, and anti-calreticulin antibody were obtained from Santa Cruz Biotechnology (Dallas, TX, USA). Mouse anti-HSP70 and conformation-sensitive anti-human leukocyte antigen HLA Class 1 antibodies were obtained from Abcam (Cambridge, UK). Conformation-sensitive H-2Kd antibody was obtained from BD Biosciences (Franklin Lakes, NJ, USA). The H-Kb-binding epitope OVA8 (SIINFEKL), extended peptide OVA13 (SIINFEKLTEWTS), control peptide HBc18–27 from the HBV core protein, and gp96 inhibitory peptide p37 fused with the cell-penetrating peptide PTD (protein transduction domain of transactivator of transcription; PTD-p37)[24,25] were chemically synthesized by Jier Biological Company (Shanghai, China). Primaquine, chloroquine, MG132, brefeldin A (BFA), and cycloheximide (CHX) were purchased from Sigma-Aldrich (St. Louis, MO, USA).

Construction of stable cell lines

Stable gp96 knockdown was performed in HepG2 cells using an shRNA lentiviral expression system. Oligonucleotides encoding short hairpin transcripts directed against the gp96 mRNA were synthesized. Luciferase shRNA was used as a negative control. The shRNA sequences for gp96 and luciferase were 5′-GCAAGGACATCTCTACAAA-3′ and 5′-CTTACGCTGAGTACTTCGA-3′, respectively. The procedures used to generate stable knockdown cell lines have been described previously.[26] Lentiviral plasmids containing cDNA of the herpes simplex virus (HSV) ICP47 protein, which inhibits the TAP-mediated translocation of antigen-derived peptides across the ER membrane,[27] or HSV UL49.5 protein, which inhibits TAP through the conformational arrest of the transporter and degradation of TAP by proteasomes,[28] were co-transfected with lentiviral packaging helper plasmid pSPAX2 and vesicular stomatitis virus glycoprotein G expressing plasmid pMD2G into 293T cells. Supernatants were harvested 48 hours after transfection and used to infect K41 cells for 24 hours. Three days after infection, GFP-positive cells were sorted by flow cytometry using a BD FACS Aria III.

Expression of recombinant gp96 proteins

The Bac-to-Bac Baculovirus Expression System was used to express recombinant gp96 and GST-gp96 fusion proteins as previously described.[25]

GST pull-down

A total of 20 μg of GST or GST-gp96 fusion protein was incubated with 50 μL glutathione sepharose 4B for 1 hour at 4°C, and cell lysates were then added and incubated for 4 hours at 4°C. The agarose beads were washed twice with PBS and resuspended in an SDS loading buffer. After boiling, the supernatant was subjected to western blotting analysis.

Co-immunoprecipitation (co-IP)

The cell lysates were incubated with 2 μg of the relevant antibodies overnight at 4°C. The resulting immunocomplex was precipitated using Protein G Sepharose beads (GE Healthcare, IL, USA) for 4 hours at 4 °C. Beads were collected by brief centrifugation and washed three times with cold PBS. The washed beads were separated using SDS-PAGE for western blot analysis.

Confocal laser scanning microscopy

Confocal microscopy was performed on non-permeabilized cells as previously described.[29] Images were obtained using a TCS SP2 confocal laser-scanning microscope (Leica Microsystems).

Flow cytometry

The cells were then disassociated with 10 mM EDTA and washed with PBS. After blocking, cells were sequentially stained with anti-H-Kb or anti-HLA Class 1 (Clone:W6/32) monoclonal antibody and donkey-anti-mouse-AF488 secondary antibody at 4°C for 1 hour. Cells were washed twice and resuspended in PBS. The fluorescence intensity was detected using a FACSCalibur flow cytometer (BD Biosciences). The results were analyzed, and the mean fluorescence intensity was calculated using FlowJo software.

Immunization of mice

Female HLA-A2.1/Kb transgenic mice (6–8 weeks old) were immunized subcutaneously with 10 μg of HBV surface (HBsAg) and core (HBcAg) proteins with 10 μg of recombinant gp96 as an adjuvant at weeks 1, 2, and 4, respectively, as described previously.[30] Mice were sacrificed 1 week after the last immunization. CD8+ T-cells were purified using MACS (Miltenyi Biotec, Bergisch-Gladbach, Germany) according to the manufacturer's instructions.

Cytotoxicity assay

HepG2 cells transfected with pcDNA3.1-HBV1.3 or pcDNA3.1 as control were labeled with 2 μM carboxyfluorescein succinimidyl ester and seeded into a 96-well plate as target cells. Isolated CTLs were added at different effector/target (E/T) ratios. Cells were incubated at 37°C for 4 hours, and the target cells were stained with propidium iodide (PI) using the Vybrant apoptosis assay kit (Invitrogen, Waltham, MA, USA).

Immunocytological localization of the gp96-OVA13 peptide complex

Recombinant gp96 was conjugated with Alexa Fluor 488 according to the manufacturer's instructions. Recombinant gp96-OVA13 peptide complexes were generated by incubating mixtures of 20 μg of protein and 20 μg of peptide at 50°C for 10 minutes, followed by incubation for 30 minutes at 22°C. The unbound peptide was removed by ultrafiltration. RAW246.7 cells were incubated with gp96-OVA13 complex at 37°C for 1 to 3 hours, washed, and fixed with 4% paraformaldehyde for 30 minutes. Cells were stained with anti-EEA1 Ab for early endosomes, anti-LAMP-1 Ab for late endosomes, anti-calreticulin Ab for ER, and anti-HSP70 Ab for the cytoplasm followed by Alexa 594-conjugated secondary antibody, and then visualized with a Leica SP8 confocal scanning laser microscope system.

B3Z assay

K41 cells were harvested and resuspended in RPMI 1640 medium supplemented with 10% FBS. Cells were transfected with gp96 or luciferase siRNA or treated with TAT-p37 peptide or geldanamycin.[31] For peptide presentation inhibition assays, K41 cells were pre-incubated with chloroquine, primaquine, MG132, or BFA at 37°C for 2 hours. Thereafter, the cells were plated in a 96-well cell plate and pulsed with OVA8 (100 μg/mL) peptide or gp96-OVA13 complex (400 μg/mL) for 2 hours. Cells were fixed with 1% PFA for 20 minutes. Subsequently, 1 × 105 B3Z cells were added to each well and co-cultured for an additional 16 to 18 hours at 37°C. Cells were washed once with PBS and lysed using Z-buffer (0.12 mM chlorophenol red-beta-D-galactopyranoside (CPRG, Calbiochem, USA), 100 mM 2-mercaptoethanol, 9 mM MgCl2, and 0.125% NP-40 in PBS). After 4 hours of incubation at 37°C, the reaction was stopped by adding stop buffer (300 mM glycine and 15 mM EDTA in water) and the absorbance was read at 630 nm using an ELISA reader.

Statistical analysis

Statistical analyses were performed using two-tailed Student's t-test or two-way ANOVA using GraphPad Prism 5.01. Statistical significance was set at P < 0.05.


Reduced cell surface levels of MHC-I in gp96 knockdown cells

To determine the possible role of gp96 in MHC-I levels and stability, we generated a stable cell line with gp96 knockdown using shRNA HepG2-gp96i, in which gp96 expression was almost completely depleted [Figure 1A]. HepG2-luci cells stably transfected with a luciferase siRNA were used as controls. As shown in Figure 1B, compared with the control, gp96 knockdown significantly reduced the cell surface levels of MHC-I by approximately 80% (P < 0.01), as determined by flow cytometry using a conformation-sensitive monoclonal antibody. MHC-I in HepG2-gp96i cells showed significantly lower stability than HepG2-luci cells upon CHX treatment [Figure 1C]. In addition, a pronounced increase in MHC-I ubiquitination levels in HepG2-gp96i cells was observed compared to that in the control cells [Figure 1D]. Conversely, overexpression of gp96 in 293T cells led to elevated cell surface expression of MHC-I [Figure 1E]. The levels of total and cell surface MHC-I were decreased by 55% and 35%, respectively. In gp96 knockdown 293T cells [Figure 1F]. These results indicate that cellular gp96 promotes the accumulation of MHC-I by inhibiting its degradation via the ubiquitin-proteasome pathway.

Figure 1:
Effect of cellular gp96 on MHC-I levels. (A) Western blot analysis of gp96 expression in HepG2 cells that were stably transfected with gp96 siRNA or luciferase siRNA (mock). (B) Flow cytometry analysis of MHC class I expression in HepG2-luc (dotted line) and HepG2-gp96i (solid line) cells. Cells stained with control IgG served as control (grey). (C) HepG2-luc and HepG2-gp96i cells were treated with 50 μg/mL CHX for the indicated hours and cell lysates were subjected to western blotting. MHC-I levels were analyzed using Image J software and normalized to GAPDH. The ratio of MHC-I and GAPDH levels at 0 hour was set as 1.0. (D) HepG2-luc and HepG2-gp96i cells were treated with 10 μM MG132 for 4 hours. Cell lysates were immunoprecipitated with the anti-MHC-I antibody, and then subjected to western blotting with a anti-ubiquitin antibody. (E) Western blot analysis of gp96 expression in 293T cells transfected with the plasmid of pcDNA3.1-gp96 or pcDNA3.1 as control (left panel), and flow cytometric analysis of MHC-I expression in 293T cells transfected with pcDNA3.1-gp96 (solid line) or pcDNA3.1 (dotted line) as control (right panel). Cells stained with control IgG served as negative control (NC) (grey). (F) Western blot analysis of MHC-I expression in 293T and 293T-gp96i cells (left panel). Flow cytometry analysis of MHC-I expression in 293T and 293T-gp96i cells (right panel), cells stained with control IgG served as a negative control.

Endogenous gp96 interacts with MHC-I and is involved in antigen presentation

Furthermore, we explored the mechanism of gp96-regulated MHC-I levels and stability. As shown in Figure 2A and B, endogenous gp96 was associated with MHC-I and TAP in co-immunoprecipitation assays with either an anti-MHC-I or anti-TAP antibody. The interaction between gp96 and MHC-I was further confirmed by the GST pull-down assay [Figure 2C]. Furthermore, confocal microscopy analysis showed that gp96 partially colocalized with MHC-I in HepG2 cells [Figure 2D]. Knockdown of the antigen presentation molecules calreticulin [Figure 2E] or TAP [Figure 2F] with RNAi resulted in a similar reduction in cell surface MHC-I as gp96 siRNA, and simultaneous knockdown of calreticulin or TAP along with gp96 caused only a modest decrease in cell surface MHC-I compared with that by each alone. These results indicated that endogenous gp96 affects MHC-I levels via the antigen presentation pathway.

Figure 2:
Cellular gp96 binds to MHC-I and participates in antigen presentation. (A and B) Co-immunoprecipitation analysis of the interaction of MHC-I or TAP with endogenous gp96. (C) GST pull down for the interaction between purified GST-gp96 and MHC-I. (D) Confocal microscopy of gp96 and MHC-I in HepG2 cells. (E and F) Flow cytometry of MHC-I expression in HepG2 and HepG2-gp96i cells transfected with CAR siRNA (E) or TAP siRNA (F). Cells incubated with fluorescence-labeled secondary antibodies alone served as background controls (grey). Dotted line marks the MFI (mean fluorescence intensity, calculated by FlowJo). (G) Flow cytometric analysis of MHC-I in HepG2-luci and HepG2-gp96i cells transfected with pHBV1.3 or pcDNA3.1 as control. (H) Cytotoxicity assay. HepG2-luci (control) and HepG2-gp96i cells were transfected with pHBV1.3, labeled with CFSE, and mixed with splenocytes from HBsAg/HBcAg-immunized HLA-A2.1/Kb transgenic mice at different ratios for 4 hours. The mixed samples were stained with PI and analyzed by flow cytometry. E:T, effector-to-target cell ratio. Results are presented as mean ± SD of five mice. P < 0.05; ∗∗ P < 0.01 compared to control.

To further demonstrate that gp96 presents cellular viral peptides to MHC-I molecules, HepG2-luci and HepG2-gp96i cells were transfected with HBV expression plasmids. HBV expression in HepG2-luci cells caused an increase in MHC-I levels, and the effect of HBV on MHC-I was completely abolished by gp96 knockdown [Figure 2G]. In addition, a higher cytotoxic effect of CTLs from HBV antigen-immunized HLA-A2 mice was observed in control cells than in gp96-knockdown cells [Figure 2H].

We examined whether gp96 knockdown could impair peptide cross-presentation using the well-characterized model antigen ovalbumin (OVA). T cell hybridoma B3Z synthesizes β-galactosidase when its T cell receptor engages the H-2Kb-restricted epitope OVA8/Kb complex. As shown in Figure 3A, B3Z activation observed in gp96 knockdown K41 cells was significantly lower than in the control cells. Similar results were obtained in cells treated with the gp96-specific inhibitory peptide PTD-p37 [Figure 3B] and the chemical inhibitor geldanamycin [Figure 3C].

Figure 3:
Knocking down of gp96 impairs antigen cross-presentation. K41 cells were transfected with gp96 siRNA or control siRNA for 48 hours (A), or treated with gp96 inhibitor peptide PTD-p37 or control peptide (B), or with geldanamycin or DMSO as control (C). OVA13 (SL8C) presentation was then tested using the B3Z assay. The response of B3Z was measured as the specific absorbance at 630 nm using an ELISA reader. The experiments were performed in triplicates. Results are presented as the mean ± SD. P < 0.05; ∗∗ P < 0.01 compared to control.

In summary, these data suggest that endogenous gp96 and other ER members, including TAP and calreticulin, constitute a relay line for loading degraded cellular peptides onto MHC-I molecules.

Hepatocyte-specific deletion of gp96 results in impaired antiviral CTL response

Thereafter, we determined the effect of gp96 on the HBV-specific hepatic CTL response in an AAV/HBV-infected mouse model that can produce HBV antigens and complete virions from plasmid-bearing hepatocytes.[32,33] To better determine virus-specific CTL response and clearance of HBV, 15 μg pAAV-HBV1.2 plasmid/mice were used to generate an acute mouse model.[32] Hepatocyte-specific gp96 knockout mice and control mice were injected with AAV/HBV 1.2 plasmid at the age of 6 weeks. As shown in Figure 4, gp96-KO mice exhibited significantly higher levels of serum HBsAg than wild-type control mice 7 days after HBV infection (by 1.6 fold at day 7, by 6.5 fold at day 10 and by 2.9 fold at day 13 compared to control mice). In addition, wild-type mice exhibited higher hepatic HBV-specific CTL responses than gp96-KO mice by 34% [Figure 4]. We further conducted a naïve CD3+ T-cell adoptive transfer experiment to exclude the possibility of the participation of immune cells other than T-cells in gp96-mediated HBV clearance. gp96-KO mice showed higher serum HbsAg levels than wild-type mice (by 2.3 fold at day 12, by 2.2 fold at day 18 and by 2.6 fold at day 25 compared to control mice) [Figure 4]. These results indicated that hepatic gp96 plays a major role in the activation of T-cell–mediated clearance of HBV.

Figure 4:
gp96 knock-out in hepatocytes leads to impaired antiviral CTL response. (A) Six to eight weeks old wild type and hepatic gp96-KO mice were hydrodynamically injected with 15 μg pAAV-HBV1.2 plasmid. Serum levels of HBsAg were examined with ELISA at the indicated time points (n=6). (B) Six to eight weeks old wild type and gp96-KO mice were treated with 10-Gy γ-ray at –1 days and hydrodynamically injected with 15 μg pAAV-HBV1.2 plasmid at day 0. Mice were then transferred with 3 × 106 naïve CD3+ T-cells/mice at day 2. Serum levels of HBsAg were examined using ELISA at the indicated time points. Results are presented as mean ± SD of three mice. (C) Hepatic lymphocytes were isolated from mice 2 weeks after hydrodymanic injection. Thereafter, cells were stimulated with 20 μg/mL HBV core peptide (aa 93–100) and analyzed using flow cytometry for percentage of IFN-γ+ CD8+ T-cells. (D) Results are presented as mean ± SD of six mice. P < 0.05; ∗∗ P < 0.01.

gp96 directs exogenous antigen to ER

We investigated the distribution pattern of exogenous gp96 uptake by the macrophages. We performed a kinetic experiment by incubating the fluorescence-labeled gp96-OVA13 complex with RAW264.7 cells. Cells were fixed at 1, 2, or 4 hours post-incubation and stained with specific antibodies against markers of organelle structures, including EEA1 [Figure 5A], LAMP1 [Figure 5B], HSP70 [Figure 5C], and calreticulin [Figure 5D]. Colocalization of gp96 with the early endosome marker EEA1 was observed 1 and 2 hours post-incubation. Notably, fluorescence-labeled gp96-OVA13 was detected in the ER as early as 2 hours after incubation and was most evident at 4 hours. In contrast, gp96-OVA13 seldom colocalizes with the lysosome marker LAMP-1 or the plasma marker HSP70. These results indicate that the exogenous gp96–peptide complex was endocytosed and delivered to the ER by macrophages via the static endosomal pathway.

Figure 5:
Intracellular localization of the gp96-OVA13 peptide complex in RAW264.7 cells. Confocal microscopy analysis of the localization of exogenous gp96-OVA13 labeled with Alexa Fluor® 488 (green) in RAW264.7 cells. The cells were incubated with the gp96-OVA13 complex for 0, 1, 2, and 4 hours. The cells were then fixed with paraformaldehyde. EEA1 (A), LAMP-1 (B), HSP70 (C), and calreticulin (D) were stained with the relevant antibodies, followed by TRTIC-labeled secondary antibody (red). Scale bar, 100 μm. Arrows indicate co-localized areas.

Presentation of exogenous gp96-chaperoned peptide is endosome- and proteasome-dependent and TAP-independent

Finally, we examined the cross-presentation efficiency of the OVA peptide OVA13 with exogenous gp96. K41 cells were pulsed with OVA8 (SL8), OVA13 (SL8C), or gp96-OVA13 (SL8C) complexes for different times and then co-cultured with B3Z T-cells. As shown in Figure 6A, an efficient OVA13 peptide representation was observed in the presence of gp96. Compared to gp96-OVA13, B3Z cell stimulation with OVA13 peptide alone was significantly lower. The cross-presentation kinetics of OVA13 with gp96 showed that the equilibration time for peptide presentation was approximately 2 hours. OVA13 cross-presentation by gp96 was largely blocked by the endosomal acidification inhibitor chloroquine [Figure 6B] and the proteasome inhibitor MG132 [Figure 6C], indicating that endosomal uptake and proteasome degradation were essential for gp96-mediated peptide cross-presentation. In contrast, the TAP transporter inhibitors ICP47 or UL49.5 [Figure 6D] and the membrane recycling inhibitor primaquine [Figure 6E] had only minor effects on gp96-mediated OVA13 cross-presentation, indicating that gp96-mediated peptide cross-presentation was independent of the MHC-I recycling and TAP pathways. In addition, the Golgi secretion inhibitor BFA strongly inhibited OVA13 peptide cross-presentation [Figure 6F], suggesting that the cell surface MHC-OVA8 for B3Z recognition was mainly derived from the ER-to-Golgi pathway.

Figure 6:
Effect of antigen presentation inhibitors on gp96-mediated cross-presentation of OVA13. (A) Kinetics of cross-presentation of OVA13 peptide by gp96 in K41 cells. The cells were incubated with gp96, OVA13 peptide, or gp96-OVA13 peptide complex at the indicated time points and then washed and fixed. The efficiency of cross-presentation of OVA13 peptide were analyzed by B3Z assay. The OVA8 peptide served as positive control. K41 cells were preincubated with chloroquine (B), MG132 (C), primaquine (E), or BFA (F) at 37°C for 2 hours, then pulsed with gp96-OVA13 complexes. The cells were transfected with ICP47 or UL49.5 expressing plasmid before being subjected to the assay (D). P < 0.05; ∗∗ P < 0.01. NS, not significant. Results are presented as mean ± SD of three replicates.


This study evaluated the contribution of gp96 to the presentation of antigenic peptides in CTL recognition. Our study on gp96 knockdown in HepG2 cells demonstrated an important role for gp96 in facilitating the loading of MHC-I molecules with HBV-specific peptides. Under hepatic gp96 depletion, the efficiency of MHC I assembly was reduced by 75%, and impaired CTL were observed for HBV antigens [Figure 1B and Figure 2H]. Moreover, we characterized the cellular compartment of gp96 as an adjuvant-mediated T-cell immune response. Endocytosed gp96 may target chaperone peptides to endosomes, then to the cytoplasm for peptide processing by the proteasome, and eventually to the ER for MHC-I peptide loading [Figure 6B, C]. Therefore, this study provides further evidence for the gp96-mediated peptide-loading function of MHC-I molecules in T-cell recognition.

Although ER-resident gp96 has long been proposed to be involved in antigen processing and presentation for T-cell recognition, several studies have focused on its immunological function in T-cell activation and the detailed antigen delivery route, and its regulation of MHC-I assembly are largely unexplored. As contradictory results have been reported in this area, innate immunity has been proposed to be the main pathway for gp96-induced CTL activation.[19] This study provides direct evidence that cellular gp96 interacts with MHC-I and TAP, identifying endogenous gp96 as a new member of the PLC. Both in vitro and in vivo experiments showed that gp96 affects the formation of MHC-I-HBV peptide complexes in hepatocytes. Using the classic OVA antigen presentation model, we demonstrated that, similar to cellular gp96, exogenous gp96 could efficiently cross-present antigens associated with MHC-I molecules for T-cell recognition in APCs. Therefore, our study elucidated the underlying mechanism of gp96-mediated T-cell activation and revealed the universal function of gp96 in antigen presentation.

Previous studies have indicated that gp96 knockdown or knockout has no effect on the total number of cell surface MHC-I molecules in murine fibroblasts[19] or bone marrow-derived macrophages.[16] However, in our model system, gp96 depletion decreased the levels of cell surface MHC-I and antiviral CD8+ T-cell–mediated cytotoxicity in human HepG2 cells [Figure 1B, Figure 2H]. This discrepancy may be attributable to the gp96-p53-MHC-I regulatory pathway. Depleting gp96 leads to increased p53 by affecting the activity of its E3 ligase Mdm2,[34] and elevated p53 may increase MHC class I expression by upregulating ER aminopeptidase 1 (ERAP1).[35] It is possible that p53-induced MHC-I expression counteracted the reduction in MHC-I assembly in gp96-depleted cells. However, this hypothesis requires further investigation. These data highlight that the regulation of MHC-I assembly by gp96 can be cell type-specific, resulting from different expression profiles and reciprocal regulatory interactions of cellular gp96, p53, ERAP1, and MHC-I. In addition, we could not completely determine the possibility that decreased HBV replication in the liver of gp96-depleted mice may also affect T-cell activity, as suppression of gp96 expression decreases HBV production,[36] and gp96-KO in the liver leads to derangement of sphingolipid metabolism and promotes the accumulation of long-chain ceramides.[37] Given that gp96 knockdown had a moderate effect on HBV replication, and gp96-KO livers exhibited limited liver steatosis at the age of 6 weeks when the mice were injected with AAV-HBV, we considered that the impaired virus-specific CTL response in the liver-specific gp96-KO mice was mainly due to compromised viral antigen presentation by hepatic gp96 depletion.

This study observed reduced MHC-I and decreased cytotoxic effect of gp96-knockdown in HBV-transfected cells [Figure 2G and H]. Further studies are needed to reveal the role of gp96 in MHC I presentation of viral antigens and the virus-specific cytotoxic effects of CTLs.

Approximately 250 million people worldwide are chronically infected with HBV. Chronic hepatitis B (CHB) is associated with a much higher risk of progressive cirrhosis, liver failure, and hepatocellular carcinoma.[38] In HBV infections, virus-specific CTL-mediated immune responses most likely play a dominant role in determining the viral clearance, disease progression, and outcome. Hepatocytes are not only targets of CTLs during viral infection of the liver but also regulate T-cell responses. In addition to dendritic cells, Kupffer cells, and liver sinusoidal endothelial cells, virus-harboring hepatocytes directly contribute to the presentation of viral antigens and influence the de novo priming of T-cells within the liver microenvironment.[39–41] Our previous studies have shown that gp96 binds to the HLA-A11-restricted HBV epitope in virus-infected liver tissues.[36,42] This study demonstrated that gp96 knockdown hepatoma cells expressed reduced cell-surface MHC-I and exhibited increased MHC-I ubiquitylation and degradation, indicating that gp96 depletion interfered with MHC-I trafficking and peptide assembly. We further provide evidence that endogenous gp96 in hepatocytes is directly associated with PLC, which may facilitate peptide loading. Moreover, we performed hepatocyte-specific deletion of gp96 by crossing Hsp90b1flox/flox mice with albumin-Cre mice. A higher virus-specific CTL response was observed in AAV/HBV-infected wild-type mice than in gp96-KO mice [Figure 4B]. In addition, gp96-KO mice displayed significantly higher serum HBsAg levels in the liver than in wild-type mice [Figure 4A]. This acute HBV-infected model was generated by hydrodynamic injection of 15 μg pAAV-HBV1.2 plasmid into wild-type and hepatic gp96-KO mice. In contrast to acute HBV infection, CHB virus-specific T-cell responses are generally rather low due to long-term extensive stimulation with viral antigens, expression of co-inhibitory immune receptors (e.g., PD-1/PD-L1, TIM-3), and direct inhibition of Tregs. Hepatic gp96-mediated CTL immunity may be partly compromised in the immunosuppressive liver environment of patients with CHB. It is important to investigate the potential association between hepatic gp96 levels and virus-specific T-cell immunity in HBV-infected patients, as studies have shown elevated gp96 expression in the liver tissues of certain CHB populations.[43] Previous studies demonstrated that HBV-specific CTLs can be primed and differentiated efficiently in the liver, independent of secondary lymphoid tissues. In addition, hepatic presentation of HBV antigens requires the cooperation of hepatocytes and specialized antigen-presenting Kupffer cells.[44,45] Therefore, our findings suggest that gp96 in hepatocytes may be involved in viral peptide presentation in the ER and may promote the incorporation of newly assembled class I molecules into the loading complex.

Extracellular gp96, as an adjuvant, elicits efficient CTL responses through cross-presentation via unknown mechanisms. Messmer et al. recently showed that following immunization, gp96 is endocytosed predominantly by murine macrophages through CD91 and may then be dissociated from the chaperoned peptide and subsequently translocated to the cytosol.[10] Moreover, gp96 and calreticulin form a relay line in the ER to sequentially transfer cellular peptides to MHC-I, indicating an indispensable role for gp96 in MHC-I peptide processing and presentation.[14] Using immunofluorescence colocalization and endosomal acidification inhibition analysis, we observed that the internalized gp96–peptide complex was sorted to the static endosome and subsequently released into the cytosol. The peptide disassociated from gp96 may be subjected to proteasomal cleavage via the inhibition of proteasome-blocked peptide presentation by gp96. TAP inhibition only had a minor effect on exogenous gp96-mediated antigen presentation [Figure 6D]. We hypothesized that cytosolic gp96 may re-chaperone the processed peptide from the proteasome and relocate it to the ER, bypassing the peptide-transport pathway of TAP. Exogenous gp96 was observed to translocate to the ER following incubation with RAW264.7 cells by immunofluorescence colocalization and immunoprecipitation assays [Figure 5]. The C-terminal KDEL sequence of gp96 may facilitate gp96 translocation via interactions with the KDEL receptor in the ER.[46,47] In contrast to its cytosolic counterpart HSP90, whose chaperone peptide was cross-presented through an endosome-recycling pathway,[46] we also observed that the gp96 associated peptide was presented to MHC-I molecules via the ER–Golgi secretory pathway.

In summary, our data demonstrated that cellular gp96 promotes the assembly and antigen-presenting function of MHC class I molecules via HBV antigen cross-presentation. Our study aimed to address the mechanism underlying gp96 as an adjuvant-induced CTL response in a concerted and regulated manner within different cellular compartments. Given that gp96 is involved in diverse aspects of innate and adaptive immunomodulation, a further understanding of gp96 function in the antigen-binding and cross-presentation complex that orchestrates T-cell immune networks will allow the identification of novel combination strategies for designing a more effective gp96-based vaccine against pathogen infections.

Limitation of this study

Our study examined the role and underlying mechanism of HSP gp96 in the presentation of antigenic peptides for recognition by CTLs. However, the effects of gp96 on antigen-specific MHC class I molecules require further investigation. Furthermore, the correlation between hepatic gp96 expression and specific T-cell responses in chronic HBV-infected patients needs to be verified.


This work was supported by the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB29040000), the Industrial innovation team grant from Foshan Industrial Technology Research Institute, Chinese Academy of Sciences, and the National Natural Science Foundation of China (32070163, 81761128002, 81871297).

Author Contributions

Songdong Meng and Xin Li conceived the project. Zihai Li supervised the project. Lijuan Qin, Yongai Liu, Yuxiu Xu, and Yang Li performed the experiments and analyzed the data. Jun Hu, Ying Ju, Yu Zhang, Shuo Wang, and Changfei Li assisted with the performance of some experiments. Lijuan Qin, Yongai Liu, and Yuxiu Xu edited the original draft. Xin Li, Lijuan Qin, and Songdong Meng critically revised the manuscript. All authors revised the manuscript and approved the final manuscript.

Conflicts of Interest



[1]. Randazzo M, Terness P, Opelz G, et al. Active-specific immunotherapy of human cancers with the heat shock protein Gp96-revisited. Int J Cancer 2012;130(10):2219–2231. doi:10.1002/ijc.27332.
[2]. Strbo N, Garcia-Soto A, Schreiber TH, et al. Secreted heat shock protein gp96-Ig: next-generation vaccines for cancer and infectious diseases. Immunol Res 2013;57(1–3):311–325. doi:10.1007/s12026-013-8468-x.
[3]. Fromm G, de Silva S, Giffin L, et al. Gp96-Ig/costimulator (OX40L, ICOSL, or 4-1BBL) combination vaccine improves T-cell priming and enhances immunity, memory, and tumor elimination. Cancer Immunol Res 2016;4(9):766–778. doi:10.1158/2326-6066.CIR-15-0228.
[4]. Ju Y, Fan H, Liu J, et al. Heat shock protein gp96 adjuvant induces T cell responses and cross-protection to a split influenza vaccine. Vaccine 2014;32(23):2703–2711. doi:10.1016/j.vaccine.2014.03.045.
[5]. Zhang H, Zheng HG, Guo P, et al. Broadly protective CD8(+) T cell immunity to highly conserved epitopes elicited by heat shock protein gp96-adjuvanted influenza monovalent split vaccine. J Virol 2021;95(12). doi:ARTN e00507-2110.1128/JVI.00507-21.
[6]. Mohit E, Bolhassani A, Zahedifard F, et al. The contribution of NT-gp96 as an adjuvant for increasing HPV16 E7-specific immunity in C57BL/6 mouse model. Scand J Immunol 2012;75(1):27–37. doi:10.1111/j.1365-3083.2011.02620.x.
[7]. Navaratnam M, Deshpande MS, Hariharan MJ, et al. Heat shock protein-peptide complexes elicit cytotoxic T-lymphocyte and antibody responses specific for bovine herpesvirus 1. Vaccine 2001;19(11–12):1425–1434. doi:10.1016/S0264-410x(00)00381-9.
[8]. Strbo N, Vaccari M, Pahwa S, et al. Cutting edge: novel vaccination modality provides significant protection against mucosal infection by highly pathogenic simian immunodeficiency virus. J Immunol 2013;190(6):2495–2499. doi:10.4049/jimmunol.1202655.
[9]. Zhao B, Wang YZ, Zhang Y, et al. TAT-mediated gp96 transduction to APCs enhances gp96-induced antiviral and antitumor T cell responses. Vaccine 2013;31(3):545–552. doi:10.1016/j.vaccine.2012.11.011.
[10]. Messmer MN, Pasmowitz J, Kropp LE, et al. Identification of the cellular sentinels for native immunogenic heat shock proteins in vivo. J Immunol 2013;191(8):4456–4465. doi:10.4049/jimmunol.1300827.
[11]. Matsutake T, Sawamura T, Srivastava PK. High efficiency CD91- and LOX-1-mediated re-presentation of gp96-chaperoned peptides by MHC II molecules. Cancer Immun 2010;10:7.
[12]. Binder RJ. Immunosurveillance of cancer and the heat shock protein-CD91 pathway. Cell Immunol 2019;343:103814. doi:10.1016/j.cellimm.2018.05.007.
[13]. Doody ADH, Kovalchin JT, Mihalyo MA, et al. Glycoprotein 96 can chaperone both MHC class I- and class II-restricted epitopes for in vivo presentation, but selectively primes CD8(+) T cell effector function. J Immunol 2004;172(10):6087–6092. doi:10.4049/jimmunol.172.10.6087.
[14]. Kropp LE, Garg M, Binder RJ. Ovalbumin-derived precursor peptides are transferred sequentially from gp96 and calreticulin to MHC Class I in the endoplasmic reticulum. J Immunol 2010;184(10):5619–5627. doi:10.4049/jimmunol.0902368.
[15]. Pawaria S, Binder RJ. CD91-dependent programming of T-helper cell responses following heat shock protein immunization. Nat Commun 2011;2. doi:ARTN 52110.1038/ncomms1524.
[16]. Yang Y, Liu B, Dai J, et al. Heat shock protein gp96 is a master chaperone for toll-like receptors and is important in the innate function of macrophages. Immunity 2007;26(2):215–226. doi:10.1016/j.immuni.2006.12.005.
[17]. Ansa-Addo EA, Thaxton J, Hong F, et al. Clients and oncogenic roles of molecular chaperone gp96/grp94. Curr Top Med Chem 2016;16(25):2765–2778. doi:10.2174/1568026616666160413141613.
[18]. Sedlacek AL, Kinner-Bibeau LB, Binder RJ. Phenotypically distinct helper NK cells are required for gp96-mediated anti-tumor immunity. Sci Rep-Uk 2016;6. doi:ARTN 2988910.1038/srep29889.
[19]. Lev A, Dimberu P, Das SR, et al. Efficient cross-priming of antiviral CD8(+) T cells by antigen donor cells is GRP94 independent. J Immunol 2009;183(7):4205–4210. doi:10.4049/jimmunol.0901828.
[20]. Hulpke S, Tampé R. The MHC I loading complex: a multitasking machinery in adaptive immunity. Trends Biochem Sci 2013;38(8):412–420. doi:10.1016/j.tibs.2013.06.003.
[21]. Adiko AC, Babdor J, Gutiérrez-Martínez E, et al. Intracellular transport routes for MHC I and their relevance for antigen cross-presentation. Front Immunol 2015;6. doi:ARTN 33510.3389/fimmu.2015.00335.
[22]. Gao B, Adhikari R, Howarth M, et al. Assembly and antigen-presenting function of MHC class I molecules in cells lacking the ER chaperone calreticulin. Immunity 2002;16(1):99–109. doi:10.1016/s1074-7613(01)00260-6.
[23]. Rachidi S, Sun S, Wu BX, et al. Endoplasmic reticulum heat shock protein gp96 maintains liver homeostasis and promotes hepatocellular carcinogenesis. J Hepatol 2015;62(4):879–888. doi:10.1016/j.jhep.2014.11.010.
[24]. Qian L, Fan H, Ju Y, et al. A peptide-based inhibitor of gp96 suppresses HBsAg expression and HBV replication by upregulation of p53. J Gen Virol 2019;100(8):1241–1252. doi:10.1099/jgv.0.001289.
[25]. Liu W, Chen M, Li X, et al. Interaction of Toll-like receptors with the molecular chaperone gp96 is essential for its activation of cytotoxic T lymphocyte response. Plos One 2016;11(5):e0155202. doi:10.1371/journal.pone.0155202.
[26]. Li CF, Wang YZ, Wang SF, et al. Hepatitis B virus mRNA-mediated miR-122 inhibition upregulates PTTG1-binding protein, which promotes hepatocellular carcinoma tumor growth and cell invasion. J Virol 2013;87(4):2193–2205. doi:10.1128/Jvi.02831-12.
[27]. Ahn K, Meyer TH, Uebel S, et al. Molecular mechanism and species specificity of TAP inhibition by herpes simplex virus protein ICP47. EMBO J 1996;15(13):3247–3255. doi:10.1002/j.1460-2075.1996.tb00689.x.
[28]. Koppers-Lalic D, Reits EA, Ressing ME, et al. Varicelloviruses avoid T cell recognition by UL49.5-mediated inactivation of the transporter associated with antigen processing. Proc Natl Acad Sci USA 2005;102(14):5144–5149. doi:10.1073/pnas.0501463102.
[29]. Li X, Sun L, Hou J, et al. Cell membrane gp96 facilitates HER2 dimerization and serves as a novel target in breast cancer. Int J Cancer 2015;137(3):512–524. doi:10.1002/ijc.29405.
[30]. Wang S, Qiu L, Liu G, et al. Heat shock protein gp96 enhances humoral and T cell responses, decreases Treg frequency and potentiates the anti-HBV activity in BALB/c and transgenic mice. Vaccine 2011;29(37):6342–6351. doi:10.1016/j.vaccine.2011.05.008.
[31]. Koo BH, Apte SS. Cell-surface processing of the metalloprotease Pro-ADAMTS9 is influenced by the chaperone GRP94/gp96. J Biol Chem 2010;285(1):197–205. doi:10.1074/jbc.M109.039677.
[32]. Dembek C, Protzer U. Mouse models for therapeutic vaccination against hepatitis B virus. Med Microbiol Immun 2015;204(1):95–102. doi:10.1007/s00430-014-0378-6.
[33]. Huang LR, Wu HL, Chen PJ, et al. An immunocompetent mouse model for the tolerance of human chronic hepatitis B virus infection. Proc Natl Acad Sci USA 2006;103(47):17862–17867. doi:10.1073/pnas.0608578103.
[34]. Wu B, Chu XY, Feng C, et al. Heat shock protein gp96 decreases p53 stability by regulating Mdm2 E3 ligase activity in liver cancer. Cancer Lett 2015;359(2):325–334. doi:10.1016/j.canlet.2015.01.034.
[35]. Wang B, Niu D, Lai L, et al. p53 increases MHC class I expression by upregulating the endoplasmic reticulum aminopeptidase ERAP1. Nat Commun 2013;4:2359. doi:10.1038/ncomms3359.
[36]. Fan H, Yan X, Zhang Y, et al. Increased expression of Gp96 by HBx-induced NF-kappaB activation feedback enhances hepatitis B virus production. Plos One 2013;8(6):e65588. doi:10.1371/journal.pone.0065588.
[37]. Rachidi S, Sun SL, Wu BX, et al. Endoplasmic reticulum heat shock protein gp96 maintains liver homeostasis and promotes hepatocellular carcinogenesis. J Hepatol 2015;62(4):879–888. doi:10.1016/j.jhep.2014.11.010.
[38]. Seto WK, Lo YR, Pawlotsky JM, et al. Chronic hepatitis B virus infection. Lancet 2018;392(10161):2313–2324. doi:10.1016/s0140-6736(18)31865-8.
[39]. Dolina JS, Cechova S, Rudy CK, et al. Cross-presentation of soluble and cell-associated antigen by murine hepatocytes is enhanced by collectrin expression. J Immunol 2017;198(6):2341–2351. doi:10.4049/jimmunol.1502234.
[40]. Knolle PA. Staying local-antigen presentation in the liver. Curr Opin Immunol 2016;40:36–42. doi:10.1016/j.coi.2016.02.009.
[41]. Riedl P, Reiser M, Stifter K, et al. Differential presentation of endogenous and exogenous hepatitis B surface antigens influences priming of CD8(+) T cells in an epitope-specific manner. Eur J Immunol 2014;44(7):1981–1991. doi:10.1002/eji.201343933.
[42]. Meng SD, Gao T, Gao GF, et al. HBV-specific peptide associated with heat-shock protein gp96. Lancet 2001;357(9255):528–529. doi:10.1016/S0140-6736(00)04050-2.
[43]. Zhu XD, Li CL, Lang ZW, et al. Significant correlation between expression level of HSP gp96 and progression of hepatitis B virus induced diseases. World J Gastroenterol 2004;10(8):1141–1145. doi:10.3748/wjg.v10.i8.1141.
[44]. Murata Y, Kawashima K, Sheikh K, et al. Intrahepatic cross-presentation and hepatocellular antigen presentation play distinct roles in the induction of hepatitis B virus-specific CD8(+) T cell responses. J Virol 2018;92(21). doi:10.1128/JVI.00920-18.
[45]. Benechet AP, De Simone G, Di Lucia P, et al. Dynamics and genomic landscape of CD8(+) T cells undergoing hepatic priming. Nature 2019;574(7777):200–205. doi:10.1038/s41586-019-1620-6.
[46]. Han JM, Park SG, Liu B, et al. Aminoacyl-tRNA synthetase-interacting multifunctional protein 1/p43 controls endoplasmic reticulum retention of heat shock protein gp96 – its pathological implications in lupus-like autoimmune diseases. Am J Pathol 2007;170(6):2042–2054. doi:10.2353/ajpath.2007.061266.
[47]. Hou JW, Li X, Li CF, et al. Plasma membrane gp96 enhances invasion and metastatic potential of liver cancer via regulation of uPAR. Mol Oncol 2015;9(7):1312–1323. doi:10.1016/j.molonc.2015.03.004.

Hepatitis B virus; Antigen cross-presentation; Cytotoxic T-cell; gp96

Copyright © 2022 The Chinese Medical Association, published by Wolters Kluwer Health, Inc.