HCC tissue specimens from 77 cases (64 males and 13 females), collected between January and December 2014 from the Pathology Department of the Daping Hospital Third Military Medical University (Chongqing, China), were fixed in formalin and embedded in paraffin. All patients included in the analysis were diagnosed with HCC following surgery and had received no adjuvant radiotherapy, chemotherapy, biological therapy or other treatments. HCC was staged according to the Union for International Cancer Control/American Joint Committee On Cancer 2010 tumor, node and metastasis (TNM) staging criteria. This project was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of Daping Hospital and Research Institute of Surgery, Third Military Medical University. Written informed consent was obtained from the subjects for the entire research process in this study. All specimens were anonymized and handled according to the aforementioned ethical and legal standards.

The mouse myeloma cell line SP2/0-Ag14 was obtained from the American Type Culture Collection (cat. no. CRL-1581; Manassas, VA, USA). Specific pathogen-free female BALB/c mice (weight, 18–20 g; age, 6–7 weeks) were purchased from Charles River Laboratories (Saint Constant, QC, Canada). Procedures involving animals and their care were conducted at Yes Biotech Laboratories registered animal facility (license no. 0107-01; Yes Biotech Laboratories Ltd., Mississauga, ON, Canada) under the Animals for Research Act and at the supervision of Ontario Ministry of Agriculture, Food and Rural Affairs (Guelph, ON, Canada). The mice were maintained at 20–23°C, with a humidity of 40–60% and exposed to 12 h/12 h light-dark cycles. Mice were fed on Laboratory Rodent Diet 5001 purchased from Ren’s Pets Depot (Oakville, ON, Canada). The procedures and pain level assessments were reviewed and approved by Yes Biotech Laboratories Animal Care Committee and complied with the Animals for Research Act.

Freund’s complete adjuvant and Freund’s incomplete adjuvant solutions were purchased from Sigma-Aldrich; Merck KGaA (Darmstadt, Germany). Incomplete medium used for hybridoma development consisted of base medium and supplements purchased from Gibco; Thermo Fisher Scientific, Inc. (Waltham, MA, USA). The incomplete medium contained 50% RPMI-1640 (cat. no. 11875-093), 50% Dulbecco’s modified Eagle’s medium/F12 (cat. no. 11320-033), 1% L-glutamine (200 mM; cat. no. 15039-027), 1% minimum essential medium-non essential amino acids (cat. no. 11140-050), 0.1% mercapotoethanol (cat. no. 21985-023), and 1% Antibiotic-Antimycotic (cat. no. 15240-062). Complete medium consisted of incomplete medium supplemented with 10% fetal bovine serum (cat. no. 10091-148; Gibco, Thermo Fisher Scientific, Inc.). Polyethylene glycol 1500 [PEG 1500; 50% (w/v)] was purchased from Roche Diagnostics (Lavel, QC, Canada). Recombinant Human PD-1/PDCD1 protein was purchased from ACROBiosystems (Newark, DE, USA). Control mouse anti-PD-1 monoclonal antibody (clone ID: UMAB199), horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (cat. no. ZB-2305) and diaminobenzidine (DAB) were purchased from OriGene Technologies, Inc. (Beijing, China). Normal goat serum and neutral gum were purchased from Wuhan Boster Biological Technology, Ltd. (Wuhan, China). 4-chloro-1-naphthol substrate (cat. no. C8890) was purchased from Sigma-Aldrich; Merck KGaA. The Mouse Typer Sub-Isotyping kit was purchased from Bio-Rad Laboratories, Inc. (Hercules, CA, USA).

The mAbs were prepared by Yes Biotech Laboratories Ltd. A total of 4 BALB/c mice were used for this project. The mice were divided into two equal groups, and were immunized using the same procedure. The first immunization was performed subcutaneously at multiple sites on the back of each mouse with 200 μl emulsion consisting of phosphate-buffered saline (PBS) containing 25 μg recombinant PD-1 protein (cat. no. PD1-H5221; ACROBiosystems) and 100 μl complete Freund’s adjuvant solution. The immunization was boosted twice intraperitoneally (IP) with an emulsion containing 25 μg PD-1 protein and 100 μl incomplete Freund’s adjuvant at 3-week intervals. At 10 days following the second booster injection, 4 μl blood was collected from the tip of the mouse tail. The antibody titers in the serum were assessed by indirect ELISA method using the recombinant PD-1 (immunogen)-coated ELISA plates (developed in-house by Yes Biotech Laboratories Ltd.). The negative control in the indirect ELISA was PBS, which was the diluent for the mouse serum samples. The blank control was substrate reading. A final booster injection was performed with 25 μg PD-1 protein in 50 μl PBS (IP). At 4 days following the final booster injection, the mice with the desired antibody titer (>10⁶) were sacrificed by cervical dislocation by a trained animal technician. The spleen was removed and spleen cells were isolated. A total of 10⁸ cells were obtained from each spleen of the immunized mice. The spleen cells were mixed with the Sp2/0-Ag14 myeloma cells, which were pre-selected with 8-azaguanine (cat. no. A5284; Sigma-Aldrich; Merck KGaA) for fusion in a sterile centrifuge tube in the presence of 50% PEG 1500 (cat. no. 783641; Roche Applied Science), as previously described (31,32). At 24 h following fusion, the hybridomas were selected by adding 100 μl of complete medium supplemented with 1% hypoxanthine-aminopterin-thymidine (cat. no. 31062-011; Gibco; Thermo Fisher Scientific, Inc.) to each well. At 8 days following fusion, the supernatants were tested for PD-1 antibody by indirect ELISA method using the recombinant PD-1 (immunogen)-coated plates. Cells whose supernatants exhibiting strong reactivity with immunogen in the indirect ELISA were subcloned by a limiting dilution procedure in complete medium supplemented with 1% hypoxanthine-thymidine (cat. no. 11067-030; Gibco; Thermo Fisher Scientific, Inc.).

The mAb isotypes secreted by the hybridomas were determined using the Mouse Typer Sub-Isotyping kit (cat. no. 1722051; Bio-Rad Laboratories, Inc.) according to the manufacturer’s instructions. This ELISA-based kit uses a panel of ultrapure reagents to determine the mouse sub-isotypes IgG1, IgG2a, IgG2b, IgG3, IgM, IgA, κ and λ chains. In the experiment, PD-1 was coated onto an ELISA microplate, and then each specific anti-PD-1 monoclonal antibody was added to the respective columns. The isotype antibodies were added to respective horizontal rows. Immunoglobulin isotypes were then determined using goat anti-rabbit (H+L) HRP conjugate (dilution, 1:2,000; cat. no. 172-1019; Bio-Rad Laboratories, Inc.) and 3,3′,5,5′-tetramethylbenzidine substrate (cat. no. 00-4201-56; Thermo Fisher Scientific, Inc.). Subsequently, the mAbs were purified using GammaBind Plus sepharose resin (cat. no. 17088602; GE Healthcare Life Sciences, Little Chalfont, UK).

A 15% PAGE gel was prepared and SDS-PAGE was performed as described previously (33) using recombinant PD-1 protein (cat. no. PD1-H5221; ACROBiosystems) following denaturation for 10 min at 100°C. The PD-1 protein was loaded at 40 ng/lane. Following electrophoresis, the protein was transferred to polyvinylidene difluoride membranes (Merck KGaA) at 150 mA for 2 h in transfer buffer at 4°C. The membranes were washed five times with PBS containing 0.05% Tween -20 (PBST), and blocked with 3% gelatin in PBS at room temperature for 2 h. The membranes were subsequently cut into strips along the lane dividers. Each of the 14 prepared PD-1 antibodies were quantified with a spectrophotometer by measuring the absorbance of the antibody samples at 280 nm. The antibodies were then diluted to 1 μg/ml in PBST, and incubated with individual strips at room temperature for 2 h. Following five washes with PBST, the strips were incubated with HRP-conjugated goat anti-mouse IgG (Yes Biotech Laboratories Ltd.) diluted to 1:1,000 in PBST with 1% bovine serum albumin (cat. no. 81-003-07; Merck KGaA) for 1 h at room temperature. The strips were then washed five times with PBST. The 40-kD PD-1 bands were visualized by incubating the membranes with 4-chloro-1-naphthol substrate (cat. no. C8890; Sigma-Aldrich; Merck KGaA).

The paraffin-embedded tissues specimens collected from HCC patients were prepared and stored by the Pathology Department of the Daping Hospital Third Military Medical University. The 3-μm-thick tissue sections were dewaxed in xylene, rehydrated in alcohol and immersed in 3% hydrogen peroxide for 10 min at room temperature to suppress endogenous peroxidase activity. Antigen retrieval was performed by heating each section for 30 min in 0.01 mol/l sodium citrate buffer (pH 6.0) at 100°C. Sections were then washed three times in PBS (each for 5 min). After blocking any nonspecific binding with ready-to-use normal goat serum for 30 min at 37°C, the sections were incubated in a humidified chamber overnight at 4°C with each of the 14 prepared PD-1 mAbs (diluted, 1:400). The commercial anti-PD-1 mAb (dilution, 1:400; cat. no. UMAB199) was used as the positive control, while PBS was used as the negative control. The sections were then incubated with HRP-conjugated goat anti-mouse IgG (diluted, 1:500; cat. no. ZB-2305; OriGene Technologies, Inc.) for 30 min at room temperature. Following three additional washes, peroxidase activity was developed with DAB for 5 min at room temperature. The sections were counterstained with 0.2% hematoxylin for 1 min at room temperature. The optical density of the sections was quantitatively analyzed using Image-Pro Plus 6.0 software (Media Cybernetics, Inc., Rockville, MD, USA).

The paraffin-embedded HCC tissues were stained with 0.2% hematoxylin for 5 min and counterstained with 0.5% eosin for 1 min at room temperature to identify the cancer tissues and the adjacent normal regions. According to the results of the mAb analyses, the highest-specificity PD-1 mAb (9E11) was selected to detect the expression of PD-1 in HCC tissues, in order to assess the correlation between the expression of PD-1 in the tumor tissues and the clinicopathological features of patients, and to evaluate the feasibility of utilizing an anti-PD-1 mAb in the treatment of HCC. The IHC methods used were the same as those described above.

Two experienced pathologists evaluated tissue staining intensity and positive cell percentage scoring in a double-blind manner (34,35). A total of five T lymphocyte infiltration-dense regions were selected for counting the number of PD-1⁺ T lymphocytes in each tissue section. The expression levels of PD-1 were defined by a ‘quickscore’, which was calculated according to the methods described previously (34,35). Briefly, the proportion of positively stained cells throughout the section was termed category A, and was assigned a score ranging from 0 to 3 (0, 0–4%; 1, 5–24%; 2, 25–49%; 3, 50–100%). The entire section was scanned at low microscopic power in order to gauge the levels of staining intensity. The average intensity defined as negative (colorless), weak (pale yellow), intermediate (pale brown), or strong staining (dark brown) was assigned a score ranging from 0 to 3, and termed category B. The product (A + B) was recorded as the total score. Total scores <3 were considered to indicate negative PD-1 expression, while total scores ≥3 were considered to indicate positive PD-1 expression.

In addition to the aforementioned methods, the number of positive cells per field of view was evaluated under a high-powered microscope by two experienced pathologists as described previously (36). A total of five T lymphocyte infiltration-dense regions (at ×200 magnification) were selected for counting the number of PD-1 positive cells in each tissue section. PD-1 expression levels in HCC tissues were determined by calculating the number of PD-1⁺ T lymphocytes per field of view.

IBM SPSS Statistics software (version 23.0; IBM Corp., Armonk, NY, USA) was used for statistical analysis. The results were expressed as the mean ± standard deviation. Differences among multiple groups were compared by one-way analysis of variance followed by the Games-Howell post hoc test. The χ² and Fisher’s exact test were used for the comparison of categorical variables. The Mann-Whitney U test was applied for the comparison of ranked data. The evaluation of multiple factors was performed by stepwise binary logistic regression. P<0.05 was considered to indicate statistically significant difference.

The recombinant human PD-1/PDCD1 protein was used as an antigen to immunize BALB/c mice. Blood was collected from the mice tail tips at 10 days following the final booster injection, and the antibody titers in the serum were assessed by indirect ELISA. As shown in Table I, serum antibody titers reached 1:10⁷, and the desired immunization effect was therefore achieved for subsequent cell fusion experiments.

Splenocytes from the immunized BALB/c mice were fused with the mouse myeloma cell line Sp2/0-Ag14. PD-1 mAbs were generated and assigned the following names: 1C4, 1H8, 2C4, 2H7, 3E5, 5B2, 5H7, 7C9, 7E5, 7G12, 9A5, 9B4, 9E11 and B1C4. Following repeated serial culture in vitro, cryopreservation and revival, the hybridomas maintained a normal morphology and continuously secreted PD-1 mAbs (data not shown).

The isotypes of the 14 mAbs were determined using a Mouse Typer Sub-Isotyping kit. As shown in Table II, the heavy chains of 13 mAbs were IgG1, while clone 9A5 was observed to be IgG2a. All 14 mAbs were observed to contain κ light chains.

The 14 mAbs were used as primary antibodies in western blot analysis to detect PD-1 protein. The molecular weight of recombinant human PD-1/PDCD1 protein is 40 kDa. As shown in Fig. 1, the 2C4, 5B2, 5H7, 7E5, 7G12, 9A5, 9B4, 9E11 and B1C4 mAbs demonstrated a high specificity for PD-1, indicating that these antibodies may be used as detection antibodies in western blotting. By contrast, a weak reaction was observed for 1C4 and 2H7 mAbs, and no reaction was observed for the 7C9, 1H8 and 3E5 mAbs. This indicates that these five mAbs are unsuitable for the detection of PD-1 by western blot analysis. However, it remains unclear whether they would be useful for the additional immunostaining experiments, such as IHC analysis.

The 14 mouse anti-human PD-1 mAbs were used for IHC analysis of HCC tissues as primary antibodies. As shown in Fig. 2, the 9E11 and 2H7 mAbs exhibited strong positive staining, particularly the 9E11 clone, which was similar to the commercial antibody UMAB199 in detecting PD-1 expression (Fig. 2B). Using the 9E11 mAb, the staining was dark, the background of tissues was clear and there was non-specific staining. By contrast, the B1C4 mAb exhibited lighter staining when compared with the UMAB199 positive control. The remaining 11 mAbs exhibited more non-specific staining, as demonstrated by the staining pattern observed when the 7E5 mAb was used (Fig. 2); the antibody bound to additional non-specific proteins on tumor cell membranes. The results indicated that these mAbs demonstrate poor specificity for PD-1 in HCC tissues. Therefore, the 9E11 mAb, which exhibited the strongest specificity for PD-1, was used to investigate the clinical significance of PD-1 expression in HCC tissues for subsequent experiments.

A total of 77 HCC tumor specimens were examined by IHC. The distribution of PD-1 expression in the HCC tissue samples is presented in Figs. 3 and 4. The images in Fig. 3 display HE and IHC staining results where the same HCC tissue sections were stained in tandem. The tumor tissues (black arrows) and peritumoral tissues (red arrows) were clearly distinguished by comparing the HE and IHC images of the same HCC tissue section. A large number of T lymphocytes were identified in the peritumoral tissues at the tumor-normal adjacent tissue junction. The distribution of PD-1⁺ T lymphocytes at different HCC tissue locations, as detected by IHC, is presented in Fig. 4. An accumulation of a small number of PD-1⁺ T lymphocytes was observed in the nodular margin of the HCC tissue (Fig. 4A). A large number of T lymphocytes were present in the peritumoral tissues at tumor-normal adjacent tissue junction (Fig. 4B and C) and primarily in the liver portal region (Fig. 4B). In addition, due to the low degree of T lymphocyte infiltration in the tumor interior, fewer PD-1⁺ T lymphocytes were observed in this region (Fig. 4D). Therefore, peritumoral tissues at the tumor-normal adjacent tissue junction were selected to analyze the correlation between PD-1 expression in HCC and the clinical and pathological characteristics of patients.

Under a double-blind setting, the results of IHC analysis for the 77 HCC cases (Table III) were evaluated by two pathologists according to the staining intensity of PD-1 combined with the ratio of PD-1-positive cells. Using these results, the correlation between PD-1 expression in the HCC tissues and the clinical and pathological characteristics of patients were statistically evaluated. As demonstrated in Table IV, PD-1 expression in HCC tissues was significantly associated with tumor size, venous thrombosis and the degree of tumor differentiation. The rate of positive PD-1 expression in HCC cases with a tumor size ≤5 cm was higher than that in tumors >5 cm in size (P<0.01). The proportion of HCC tissue samples derived from patients with venous thrombosis that were positive for PD-1 expression was higher than in the proportion of patients positive for PD-1 expression without venous thrombosis (P<0.05). In addition, the positive rate of PD-1 in poorly differentiated (highly malignant) HCC was significantly higher than in moderately and highly differentiated HCC (P<0.01). The expression of PD-1 in HCC tissues was not correlated with patient sex, median age, hepatic cirrhosis, hepatitis B infection, TNM stage or tumor location (Table IV).

The number of PD-1 positive cells in multiple fields of view (×200 magnification) for each tissue section was then determined (Fig. 5). The number of PD-1⁺ T lymphocytes was closely correlated with tumor size and differentiation (P<0.01; Fig. 5A and B). PD-1 expression was significantly increased in tumors ≤5 cm in size or in poorly differentiated tumors, which is consistent with the aforementioned IHC results. By contrast, no correlation between the number of PD-1⁺ T lymphocytes in HCC tissues and the presence of venous thrombosis was observed (Fig. 5C).

The results indicate that PD-1 expression in HCC tissues may be closely correlated with tumor size and the degree of differentiation and moderately correlated with venous thrombosis. However, PD-1 expression does not appear to be correlated with additional clinical characteristics, including patient sex, age, cirrhosis, hepatitis B infection, TNM stage or tumor location.