Figures

Figure 1. Long-term expression of HBsAg in low, intermediate, and high viral loads

C57BL/6 (B6) mice were i.v. injected with low (10¹⁰ vg), intermediate (10¹¹ vg), or high (10¹²) viral loads of AAV/HBV. Sera were collected 6 months post-infection (p.i.) for the measurement of HBsAg by electrochemiluminescence. Error bars represent standard deviations (SD).

Figure 2. Functional impairment of HBV-specific CD8 T cells depended on viral loads

Mice were infected with different doses (10¹⁰, 10¹¹, or 10¹² vg) of AAV/HBV and treated with therapeutic vaccination. At week 2 p.i., mice were immunized by

intramuscular (i.m.) injection with 100 µg of pHBs plasmid DNA (encoding HBsAg) twice at a two-week interval. One week after the second DNA immunization (week 5 p.i.), mice were i.v. injected with 1 x 10⁸ PFU of adenoviral vector expressing pre-S2 envelope protein. (A) Sera were collected before (week 2 p.i.) and after (week 6 p.i.) the

treatment of therapeutic vaccination for the detection of HBsAg by

electrochemiluminescence. (B and C) Mice were sacrificed at week 6 p.i. and intrahepatic and splenic CD8 T cells were isolated, enriched with MACS beads, and subjected to interferon gamma (IFN-γ) ELISPOT assays, in which peptide-pulsed EL4 cells were co-cultured with CD8 T cells for stimulation. Data are representative of three independent experiments with 5 mice per group. Error bars represent SD. SFC,

spot-forming cells. HBs, HBs190-197 and HBs208-215 peptides. OVA, OVA257-264 peptide.

Med, RPMI 1640 medium.

Figure 3. PD-1 expression on intrahepatic CD8 T cells was upregulated in a viral load-dependent manner

Mice were infected with different doses of AAV/HBV and sacrificed 4 months after infection. Lymphocytes in liver and spleen were isolated and analyzed by flow cytometry. The expression of PD-1 was evaluated after gating on viable TCRβ⁺CD8⁺ cells. AAV/Empty served as a negative control. (A) The expression levels of PD-1 on intrahepatic and splenic CD8 T cells. (B) Percentages of PD-1⁺ cells in total intrahepatic

CD8 T cells. (C) Mean fluorescence intensity (MFI) of PD-1⁺ intrahepatic CD8 T cells.

Data were representative of three independent experiments with 3 to 5 mice per group.

Error bars represent SEM.

Figure 4. Expression of 2B4 and LAG-3 on intrahepatic CD8 T cells increased with greater extents during infection of higher viral loads

Mice were infected with different doses of AAV/HBV and sacrificed 4 months after infection. Lymphocytes in liver and spleen were isolated and analyzed by flow

cytometry. The expressions of 2B4, LAG-3, and Tim-3 were evaluated after gating on viable TCRβ⁺CD8⁺ cells. (A) Representative data of 2B4, LAG-3, and Tim-3

expression on CD8 T cells in the liver and the spleen of mice infected with different loads of AAV/HBV. (B and C) Summary results of 2B4, LAG-3, and Tim-3 expression on intrahepatic and splenic CD8 T cells. Data were representative of three independent experiments with 3 to 5 mice per group. Error bars represent SEM.

Figure 5. The levels of PD-1 upregulation on intrahepatic CD4 T cells depended on viral loads

Mice were infected with different doses of AAV/HBV and sacrificed 4 months after infection. Lymphocytes in liver and spleen were isolated and analyzed by flow cytometry. The expressions of PD-1, 2B4, LAG-3, and Tim-3 were evaluated after gating on viable TCRβ⁺CD4⁺ cells. (A) Representative data of PD-1, 2B4, LAG-3, and Tim-3 expression on CD4 T cells in the liver and the spleen of mice infected with different loads of AAV/HBV. (B and C) Summary results of PD-1, 2B4, LAG-3, and Tim-3 expression on intrahepatic and splenic CD4 T cells. Data were representative of three independent experiments with 3 to 5 mice per group. Error bars represent SEM.

Figure 6. HBV-specific immunity in early stage of infection was more potent in the absence of PD-1:PD-L1

Wild type (WT) and PD-L1 KO mice were infected with 10¹² vg of AAV/HBV and were sacrificed at week 2 p.i. (A) Sera were collected for the detection of HBsAg by electrochemiluminescence. (B to D) Intrahepatic and splenic CD8 T cells were isolated and subjected to IFN-γ ELISPOT assays. Representative photographs (B and C) and summary results (D and E) are shown. Data are representative of two independent experiments with 4 mice per group. Error bars represent SD.

Figure 7. Reduction of HBsAg in PD-L1 KO mice infected with low, intermediate, and high viral loads of AAV/HBV

WT and PD-L1 KO mice were infected with low (10¹⁰ vg), intermediate (10¹¹ vg), or high (10¹² vg) viral loads of AAV/HBV. Sera were collected at week 2 p.i. to determine the levels of HBsAg by electrochemiluminescence.

Figure 8. Higher viral loads elicited more intense functions of HBV-specific CD8 T cells in the early phase of infection in PD-L1 KO mice

PD-L1 KO mice were infected with different loads of AAV/HBV (n = 3 mice per group) and sacrificed two weeks p.i. IFN-γ ELISPOT was performed following isolation and enrichment of CD8 T cells from the liver and the spleen. (A and B) Representative data of IFN-γ production by HBV-specific CD8 T cells from the liver and the spleen of KO mice infected with different doses of AAV/HBV. (C and D) Summary results of IFN-γ spot-forming CD8 T cells from the liver and the spleen. Error bars represent SD.

Figure 9. Accelerated functional loss of HBV-specific CD8 T cells in PD-L1 KO mice infected with high viral load of AAV/HBV

Wild type and PD-L1 KO mice were infected with different doses of AAV/HBV. (A to C) Sera were collected at week 8 p.i. for the detection of HBsAg. (D to G) PD-L1 KO mice infected with different doses were immunized and sacrificed as described in

Figure 3. IFN-γ ELISPOT was performed with intrahepatic and splenic CD8 T cells.

Representative photographs (B and C) and summary results (D and E) are shown. Data are representative of two independent experiments with at least 3 mice per group. Error bars represent SD.

Figure 10. Blockade of LAG-3 accelerated the reduction of serum HBsAg in PD-L1 KO mice

WT and PD-L1 KO mice were i.p. injected with 200 µg α-LAG-3 monoclonal antibody (mAb) or Rat-IgG isotype control antibody on day -7, -1, 4, 8, 15, 21, 25, 39, 47, 56 (n

= 5 mice per group). Seven days after the first antibody injection (day 0), mice were infected with 10¹¹ vg of AAV/HBV by i.v. injection. Sera were collected at indicated time for the measurement of HBsAg by electrochemiluminescence. Error bars represent SD.

Figure 11. Identification of H-2Kb-restricted dominant epitopes of HBsAg and HBcAg

B6 mice were immunized twice at a two-week interval by intramuscular injection of pHBs and pHBc plasmid DNAs, which encode HBsAg and HBcAg, respectively, followed by electroporation. Splenic CD8 T cells isolated one week after the second immunization were subjected to IFN-γ ELISPOT assays using overlapping peptides shown in Table 1. (A and B) CD8 T cells from immunized mice were co-cultured with EL4 cells that were pulsed with peptide pools. (C and D) Individual peptides from positive pools (S31-S40, S61-S70, S71-S76, and C11-C20) in Figure 11A and B were used to identify the H-2Kb-restricted epitopes. N.D., not detected.

Figure 12. Detection of subdominant clones of HBV-specific CD8 T cells in chronically infected C57BL/6 mice

Intrahepatic CD8 T cells, which were pooled from 5 B6 mice infected with 10¹¹ vg AAV/HBV for 12 months, were co-cultured with EL4 cells pulsed with overlapping peptide pools. After 6 hours of culture in the presence of BFA, the productions of IFN-γ and TNF-α by total intrahepatic CD8 T cells (gated on viable TCRβ⁺CD8⁺ cells) were detected by flow cytometry following intracellular cytokine staining. Stimulations with EL4 cells pulsed with OVA peptide or RPMI medium alone served as negative control, while PMA plus ionomycin (P+I) stand for positive stimulation.

Figure 13. PD-1⁺ CD8 T cells from AAV/HBV-infected mice were irresponsive to stimulation using bone marrow-derived dendritic cells (BMDCs) infected with

adenoviral vectors expressing HBV antigens

Mice (n = 7) were infected with 10¹² vg AAV/HBV and sacrificed 6 weeks later. Pooled intrahepatic and splenic CD8 T cells were enriched by MACS beads and sorted by FACSAria following PD-1 staining. (A) PD-1 expression on CD8 T cells in the liver and the spleen of AAV/HBV infected mice. (B) PD-1 expression on CD8 T cells after sorting. (C) Bone marrow-derived dendritic cells (BMDCs) from PD-L1 KO mice were infected with 50 MOI of adenoviral vectors expressing luciferase (Ad-Luc), HBV (Ad-HBV), pre-S1 envelope protein (Ad-mS), or core protein (Ad-C). After 24 hours, infected BMDCs were washed and co-cultured with sorted CD8 T cells for

HBV-specific stimulation. Ad-Luc group served as a background control of responses against adenoviral vectors. Splenic CD8 T cells from mice immunized by twice DNA immunization and one adenoviral boost (DDA) were used as positive control for HBs- and HBc-specific responses. The production of IFN-γ and TNF-α by sorted CD8 T cells were detected by flow cytometry after intracellular cytokine staining.

Figure 14. PD-1⁺ intrahepatic CD8 T cells showed more intense function of TNF-α production against HBV peptides than the PD-1^- counterpart CD8 T cells after in vitro expansion

Mice were infected with 10¹² vg AAV/HBV and sacrificed 6 weeks later. Intrahepatic CD8 T cells were enriched by MACS beads and sorted by FACSAria following PD-1 staining. Sorted CD8 T cells were co-cultured with BMDCs and irradiated naïve splenocytes pulsed with overlapping peptide pools. On day 10, cells were restimulated with peptide-pulsed irradiated naïve splenocytes. Cells were subjected to TNF-α

ELISPOT assay on day 20 with the stimulation of peptide-pulsed EL4 cells for 24 hours.

OVA peptide served as a negative stimulation. (A) Representative data of TNF-α production by PD-1⁺ and PD-1^- intrahepatic CD8 T cells after in vitro expansion. (B) Summary results of TNF-α production by in vitro expanded PD-1⁺ and PD-1

-intrahepatic CD8 T cells of AAV/HBV-infected mice.

Figure 15. The frequencies of CD11a^hi CD8 T cells depended on viral loads of AAV/HBV

Mice were infected with different doses of AAV/HBV and sacrificed 4 months after infection. Intrahepatic lymphocytes were isolated and analyzed for surrogate activation markers by flow cytometry. TCRβ⁺CD8⁺ cells were gated for the analysis of CD11a^hi CD8 T cells. AAV/Empty served as a negative control. (A) Representative data of the frequencies of CD11a^hi cells among CD8 T cells in the liver of mice infected with different doses of AAV/HBV and age-matched naïve mice. (B) PD-1 expression on

CD11a^hi and CD11a^low intrahepatic CD8 T cells. (C) Summary results of the

percentages of CD11a^hi cells in total intrahepatic CD8 cells. Error bars represent SD.

Figure 16. Higher degree of oligoclonal T-cell expansion was observed in PD-1⁺ intrahepatic CD8 T cells than that of the PD-1^- counterpart CD8 T cells

Intrahepatic CD8 T cells from 5 B6 mice infected with 10¹¹ vg AAV/HBV for 12 months were pooled, enriched, and sorted by FACSAria following PD-1 staining. After RNA extraction from sorted cells, reverse transcription PCR was performed and cDNA products were subjected to PCR reactions using different forward TCRβ primers (Vβ1-20) with one reverse Cβ primer described in Table 2. The PCR products of TCRβ sequences were labeled with FAM fluorochrome by primer extension using Cβ-FAM primer shown in Table 2. The lengths of TCRβ cDNA fragments were analyzed by ABI3700 Analyzer and PeakScanner Software. Arrows indicate peaks differed from Gaussian distribution profile and represent the presences of oligoclonal T-cell expansions. Panels showed no signals resulted from inadequate amounts of PCR products. Mice immunized with plasmid DNA and adenoviral vector (DDA) served as positive control, and those infected with AAV/Empty (Empty) served as negative control.