Development and characterization of mouse anti-canine PD-L1 monoclonal antibodies and their expression in canine tumors by immunohistochemistry in vitro

Abstract

Immune escape is the hallmark of carcinogenesis. This widely known mechanism is the overexpression of immune checkpoint ligands, such as programmed cell death protein 1 and programmed death-ligand 1 (PD-1/PD-L1), leading to T cell anergy. Therefore, cancer immunotherapy with specific binding to these receptors has been developed to treat human cancers. Due to the lack of cross-reactivity of these antibodies in dogs, a specific canine PD-1/PD-L1 antibody is required. The aim of this study is to develop mouse anti-canine PD-L1 (cPD-L1) monoclonal antibodies and characterize their in vitro properties. Six mice were immunized with recombinant cPD-L1 with a fusion of human Fc tag. The hybridoma clones that successfully generated anti-cPD-L1 antibodies and had neutralizing activity were selected for monoclonal antibody production. Antibody properties were tested by immunosorbent assay, surface plasmon resonance, and immunohistochemistry. Four hybridomas were effectively bound and blocked to recombinant cPD-L1 and cPD-1-His-protein, respectively. Candidate mouse monoclonal antibodies worked efficiently on formalin-fixed paraffin-embedded tissues of canine cancers, including cutaneous T-cell lymphomas, mammary carcinomas, soft tissue sarcomas, squamous cell carcinomas, and malignant melanomas. However, functional assays of these anti-cPD-L1 antibodies need further investigation to prove their abilities as therapeutic drugs in dogs as well as their applications as prognostic markers.

Keywords:

• Dog
• canine PD-L1
• anti-canine PD-L1 antibody
• immune checkpoint inhibitor

1. Introduction

Cancer immune evasion is one part of a critical process of carcinogenesis in humans and animals. There are multiple mechanisms that malignant cells use to escape functional host immunity, such as the inhibition of cytotoxic T cells or natural killer cells, the downregulation of antigen-presenting molecules (major histocompatibility class I) in cancer cells, the secretion of immune suppressive cytokines (TGF-β), and the recruitment of regulatory T cells or tumor-promoting macrophages (Vinay et al. 2015). In addition, the expression of immune checkpoint molecules (i.e. cytotoxic T lymphocyte antigen 4 (CTLA4), programmed cell death protein 1 and its ligand (PD-1/PD-L1), etc.) limits T-cell activation and function. PD-1 is a co-inhibitory receptor, expressed on the transmembrane of activated T cells, that interacts with two ligands, PD-L1 and PD-L2 (Latchman et al. 2001; Yamazaki et al. 2002). PD-L2 expression has been observed in mast cells, dendritic cells (DCs), and macrophages, whereas PD-L1 has been noted in hematopoietic (T/B lymphocytes, DCs, macrophages, and mast cells) and non-hematopoietic cells (vascular endothelial cells, hepatocytes, epithelial cells, myocytes, pancreatic islet cells, and astrocytes; Jubel et al. 2020). Both PD-L1 and PD-L2 have also been detected in tumor cells and tumor stroma in humans, such as non-small cell lung carcinoma (NSCLC), head and neck squamous cell carcinoma (HNSCC), melanoma, Hodgkin’s lymphoma (HL), and breast cancer (Mittendorf et al. 2014; Prat et al. 2017; Vari et al. 2018). However, PD-L1 is a dominant inhibitory ligand of PD-1 in the tumor microenvironment (Sun et al. 2018).

The expression of the PD-1/PD-L1 axis has been investigated in various canine malignant neoplasms, including cell lines and clinical samples. PD-L1 mRNA expression was observed in canine B-cell lymphoma, urothelial cell carcinoma, mast cell tumor (MCT), melanoma, histiocytic sarcoma, and osteosarcoma cell lines (Maekawa et al. 2014; Hartley et al. 2017; Kumar et al. 2017; Pinard et al. 2022). In addition, the overexpression of PD-L1 in dogs has been mentioned in many types of cancer, including squamous cell carcinoma (SCC), malignant melanoma (MM), mammary adenocarcinoma, hepatocellular carcinoma, renal cell carcinoma (RCC), MCT, diffuse large B-cell lymphoma (DLBCL), and soft tissue sarcoma (STS) (Maekawa et al. 2014, 2016; Shosu et al. 2016; Ambrosius et al. 2018; Stevenson et al. 2021). The level of PD-L1 expression has also been correlated with prognosis in particular tumors in dogs. A higher expression of PD-L1 was associated with aggressive behavior and decreased survival time in DLBCLs, malignant mammary tumors, and osteosarcomas (Ariyarathna et al. 2020; Aresu et al. 2021; Cascio et al. 2021).

Using anti-PD-L1 antibodies, which are known as immune checkpoint inhibitors (ICIs) for blocking the cascade of the PD-1 and PD-L1 pathways, has shown a promising effect for neoplastic treatment in humans and dogs. This immunotherapy has been approved to treat human melanoma, NSCLC, HNSCC, HL, and RCC. Due to the increased expression of PD-L1 in canine malignant cells/tumor-infiltrating lymphocytes and the effective treatment of PD-L1 blockade in human neoplasms, monoclonal antibodies (mAbs) against canine PD-L1 (cPD-L1) have been developed in recent studies (Maekawa et al. 2017; Nemoto et al. 2018; Choi et al. 2020). A generation of anti-cPD-L1 mAbs—H7-9 and G11-6 (Nemoto et al. 2018), c4G12 (Maekawa et al. 2017), and JC071, 173, 194, and 205 (Choi et al. 2020)—efficiently inhibited the binding of canine PD-1 (cPD-1) to cPD-L1, significantly enhanced interferon-γ (IFN-γ) production by dog peripheral blood mononuclear cells (PBMCs), and increased production of T lymphocytes. The effects of c4G12 were further investigated to treat oral MMs with lung metastasis in 29 dogs (Maekawa et al. 2021). The overall response rate of this immunotherapy was 7.7% with one dog having complete remission. Even though the response rate was quite low, the survival time of dogs receiving immunotherapy was prolonged compared to the control group. Therefore, the production of anti-cPD-L1 mAb concurrent with its diagnostic and prognostic ability and effectiveness in cancer immunotherapy will be helpful to increase its highest utility in various canine neoplasms. The objectives of this study were to generate mouse mAbs against cPD-L1, to investigate their characterization, and to determine their application on immunohistochemistry as a pilot retrospective study in specific canine tumor subtypes.

2. Materials and methods

2.1. Study animals and cell lines

Three BALB/c and three ICR female mice, 6–8 weeks old, were purchased from Siam Nomura International (Bangkok, Thailand) and were handled according to the Institute Animal Care and Use Committee (IACUC) practices at Chulalongkorn University (IACUC no. 0202560). P3/NSI/1-Ag4-1 (NS-1) and Sp2/0-Ag14 (SP2) myeloma cell lines were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA). Myeloma cells were maintained in complete medium (DMEM supplemented with 10% fetal bovine serum, 1 mM sodium pyruvate, 1% non-essential amino acids, 1% penicillin/streptomycin, 50 μg/ml gentamycin, and 1% beta-mercaptoethanol) and cultured at 37 °C in a 5% CO₂ humidified incubator.

2.2. Mouse immunization and hybridoma generation

BALB/c and ICR mice were immunized with cPD-L1 human Fc (70110-D02H, Sino Biological Inc., Beijing, China) as follows: a single bolus intramuscular and subcutaneous injection of 50 µg protein in Complete Freund’s Adjuvant (CFA; Pacific Immunology, Ramona, CA, USA, 1:1, v/v), a monthly interval of 2 boosters with 25 µg protein in Incomplete Freund’s Adjuvant (Pacific Immunology, 1:1, v/v), and final booster via intraperitoneal injection with 25 µg protein in sterile normal saline solution. Blood from immunized mice was collected via a facial vein for detecting antibody titers by enzyme-linked immunosorbent assay (ELISA). Two mice that provided the highest IgG-specific titer against cPD-L1 were selected for splenectomy and underwent hybridoma generation four days after the final immunization. Mice were euthanized by overdose inhalation of isoflurane and cervical dislocation for aseptic splenectomy.

Electrofusion was used for hybridoma generation, as previously described by Phakham et al. (2022). In brief, splenocytes were separated, filtered through a 100 µm cell strainer, and counted using a Countess 3 FL automated cell counter (Invitrogen, Waltham, MA, USA). Splenocytes and myeloma cells were mixed at a ratio of 2:1 and washed several times with sterile DPBS buffer (Gibco, Grand Island, NY, USA) and BTXpress cytofusion (BTX) buffer. The cell suspension was electroporated by the BTX ECM 2001 Electrofusion (Harvard Bioscience, Inc., Holliston, MA, USA). Fused cells were recovered in ClonaCell-HY medium C (Stemcell Technologies, Vancouver, Canada) at 37 °C in a 5% CO₂ humidified incubator overnight.

2.3. Hybridoma cloning and selection

The culture supernatants from the hybridoma minipools were determined for cPD-L1-specific IgG levels using ELISA. For primary screening, minipools providing ELISA signals (cPD-L1+, Human-IgG–) were considered positive hybridoma clones that produced mouse anti-cPD-L1 mAbs. Then, positive minipools were further screened for cPD-1/cPD-L1 neutralizing activity by ELISA. Minipools producing antibodies inhibiting cPD-1/cPD-L1 interaction were sub-cloned to obtain monoclonality by limiting dilution technique. Selected hybridoma monoclones were allowed to grow for 7–10 days before binding and neutralizing activities were determined. Hybridoma clones producing mAbs confirmed that both cPD-L1 binding and cPD-1/cPD-L1 neutralizing activities by ELISA were cultured and expanded in a complete medium and then transferred to a serum-free medium (SFM4MAb, Cytiva, South Logan, UT, USA) for mAb production. The culture supernatant was harvested and purified using protein A column chromatography. Antibody isotyping was performed to identify the isotype of the purified mAbs.

2.4. Canine PD-L1 binding activity by ELISA

Culture supernatant or purified monoclonal antibody (100 µl/well) was added to a cPD-L1 hu Fc-coated ELISA plate (10 ng/well) and incubated at 37 °C for 1 h and then washed three times with 0.05% Tween-20 in PBS (PBST) buffer. For the human PD-L1 cross-reactivity profile, human PD-L1 hu Fc (10 ng/well) was used as an antigen for the ELISA coating. Goat anti-mouse IgG Fcγ-HRP antibody (Jackson ImmunoResearch Inc., West Grove, PA, USA) dilution at 1:8000 in PBST (100 µl/well) was added to the assay plate and incubated at 37 °C for 1 h. The SIGMAFAST OPD (o-phenylenediamine dihydrochloride, Sigma Aldrich, Saint Louis, MO, USA) substrate solution (100 µl/well) was added to the plate and incubated at room temperature in the dark for 20 min. The reaction was stopped by adding 2 N H₂SO₄ (50 µl/well). The absorbance was measured at 492 nm using a Cytation 5-cell imaging multi-mode reader (BioTek, Winooski, VT, USA).

2.5. Canine PD-1/canine PD-L1 neutralizing activity by ELISA

Culture supernatant or purified monoclonal antibody (60 µl/well) and cPD-L1 hu Fc (3 µg/ml, 60 µl/well) were co-incubated with shaking (600 rpm) in a 96-well V-shaped plate at 37 °C for 30 min. A 100 µl of co-incubated reaction was added to an ELISA plate coated with cPD-1-His protein (70109-D08H, Sino Biological Inc., Beijing, China) at 100 ng/well, followed by incubation with goat anti-human IgG Fcγ-HRP antibody dilution at 1:5000 in PBST, and finally with the SIGMAFAST OPD substrate solution (Sigma Aldrich, Saint Louis, MO, USA). The reaction was stopped by adding 2 N H₂SO₄. The absorbance was measured at 492 nm using a Cytation 5-cell imaging multi-mode reader (BioTek, Winooski, VT, USA). The percentage of neutralization was calculated according to the following formulation: 1$-$ (A492_Sample/A492_{Negative control}) × 100.

2.6. Protein purification using ÄKTA pure protein purification system

Cultured supernatants containing anti-cPD-L1 immunoglobulin were harvested by centrifugation (4500×g, at room temperature for 10 min) and then filtrated through a 0.45 µm PES syringe filter before being applied to the protein A FF column (Cytiva, South Logan, UT, USA), as described previously (Phakham et al. 2022). The purified antibody was quantified by spectrophotometry (A280) using a Cytation 5-cell imaging multi-mode reader (BioTek) and stored at 4 °C (short-term) or −80 °C (long-term).

2.7. Canine PD-L1 binding kinetics by surface plasmon resonance (SPR)

The Biacore T200 (Cytiva) was used to determine the binding kinetics between the antibodies and the cPD-L1 protein. Purified mouse mAb at 1 µg/ml was injected into an individual flow cell of a CM5 equipped with goat anti-mouse IgG Fc-γ specific antibody at 30 µl/min for 90 s. A single-cycle binding kinetics analysis was performed by sequentially injecting mouse mAbs at various concentrations (0.8018, 1.604, 3.208, 6.415, and 12.83 nM) with interval cycles of association time for 60 s and dissociation time for 120 s. The uncoated reference cells and buffer blank signals were subtracted from the sensorgrams. Biacore T200 evaluation software (version 3.1) was used to calculate the association constant (k_a), dissociation constant (k_d), and equilibrium constant (K_D).

2.8. Immunohistochemistry to detect PD-L1 in canine cancers

The archives of formalin-fixed paraffin-embedded (FFPE) tissues were selected from the Department of Pathology, Faculty of Veterinary Science, Chulalongkorn University. Five blocks from each cutaneous T-cell lymphoma (CTL), soft tissue sarcoma (STS), and tubular mammary carcinoma (TMC) were cut into 3-µm thickness, deparaffinized, rehydrated, and blocked endogenous peroxidase by 0.3% (v/v) H₂O₂ for 30 min. The tissues were then retrieved antigen by citrate buffer pH 6 at 95 °C for 30 min, followed by blocking non-specific binding protein with 2.5% (w/v) bovine serum albumin for 20 min. The antibody dilutions from each clone (5D2, 15D12, 8F2, and 21A1) varied between 1:25, 1:50, 1:100, and 1:1500 and were incubated at 4 °C for at least 12 h. The anti-rabbit/mouse Ig-HRP (Envision, Dako, Glostrup, Denmark) was used to detect a signal from the antigen–antibody binding complex, and diaminobenzidine was used as a chromogenic substrate. Normal lymph nodes from a dog were used as a positive control. After a well-developed protocol with a suitable antibody dilution was achieved, a preliminary study of five MMs, five SCCs, and one peripheral T-cell lymphoma (PTCL) was further investigated for PD-L1 expression in dogs.

3. Results

3.1. Hybridoma screening and selection

Among the three BALB/c and three ICR mice immunized with cPD-L1 Fc protein, ICR mouse number 2 and BALB/c mouse number 3 provided the highest cPD-L1 IgG titers. After electrofusion, hybridoma minipools were selected in a complete medium containing hypoxanthine, aminopterin, and thymidine (HAT). An ELISA binding signal to cPD-1/cPD-L1 was provided by 112 hybridoma minipools obtained from the ICR mouse; however, only four hybridomas postulated > 20% neutralizing activity, which was selected for subcloning. For the BALB/c mouse, 98 hybridoma minipools showed specific binding against cPD-L1, while only four hybridomas exhibited neutralizing activity. In summary, eight candidate hybridoma minipools were then subcloned by limiting dilution to achieve a hybridoma monoclone. Finally, we successfully isolated four hybridoma monoclones producing mAbs with cPD-1/cPD-L1 neutralizing activity, including 5D2, 15D12, 18F2, and 21A1.

3.2. Neutralizing activity of mouse anti-canine PD-L1 monoclonal antibodies

A competitive ELISA was developed to evaluate the cPD-1/cPD-L1 neutralizing activity of purified mouse mAbs. The results demonstrated that all purified mAbs were able to inhibit the cPD-1/cPD-L1 interaction. Among these candidates, 21A1 displayed the highest neutralizing activity, with a half-maximal inhibition concentration (IC₅₀) value of 0.918 µg/ml (Figure 1A, green line). In comparison, the others inhibited the cPD-1/cPD-L1 interaction with IC₅₀ values ranging from 0.932 to 2.466 µg/ml (Table 1). In addition, epitope binning data of four anti-cPD-L1 mAbs demonstrated that they can be classified into two groups; group 1 (5D2 mAb and 18F2 mAb) and group 2 (15D12 mAb and 21A1 mAb).

Figure 1. (A) Binding profile and (B) cPD-1/cPD-L1 neutralizing profile of four mouse anti-cPD-L1 mAbs. Data are presented as mean ± SD.

Table 1. Summary characteristics of four purified mouse anti-cPD-L1 mAbs.

Antibody	Myeloma partner	Antibody isotype	Binding activity, EC50 (ng/ml)	Neutralizing activity, IC50 (µg/ml)
5D2	NS-1	IgG1, kappa	8.63	1.818
15D12	SP2	IgG1, kappa	6.55	0.932
18F2	SP2	IgG1, kappa	8.62	2.466
21A1	SP2	IgG1, kappa	6.08	0.918

3.3. Binding activity of mouse anti-canine PD-L1 monoclonal antibodies

To determine the cPD-L1 binding activity, ELISA was used, and the results revealed that these antibodies strongly bound to recombinant cPD-L1 proteins with half-maximal effective concentration (EC₅₀) values around 6.08–8.63 ng/ml (Table 1). Interestingly, 21A1 displayed the highest binding activity, with an EC₅₀ value of 6.08 ng/ml (Figure 1B, green line). Moreover, all anti-cPD-L1 mAbs could not bind to recombinant human PD-L1 protein confirming their species specificity using the same immunoassay platform.

3.4. Binding kinetics of mouse anti-canine PD-L1 monoclonal antibodies

The SPR technique was used to evaluate the binding kinetics of the purified mAbs against the cPD-L protein. A single cycle of kinetics was performed, and the results showed that all antibodies exhibited high cPD-1 binding affinity in a sub-nanomolar range (K_D = 1.37 × 10⁻¹⁰ to 6.72 × 10⁻¹¹ M), as shown in Table 2. The 5D2 mAb demonstrated the highest binding affinity against cPD-L1 protein with an equilibrium dissociation constant (K_D) of 21.49 pM. Figure 2 shows the SPR sensorgrams of each purified mouse anti-cPD-L1 antibody.

Figure 2. SPR sensorgrams of purified mouse anti-cPD-L1 mAb clone (A) 5D2, (B) 15D12, (C) 18F2, and (D) 21A1.

Table 2. Binding kinetics data of mouse anti-cPD-L1 mAbs against its target.

Antibody	Canine PD-L1 binding kinetics
	ka (1/M.s)	kd (1/s)	KD (M)
5D2	1.864E + 6	4.007E − 5	2.149E − 11
15D12	1.493E + 6	2.049E − 4	1.373E − 10
18F2	1.910E + 6	1.284E − 4	6.723E − 11
21A1	1.757E + 6	3.947E − 4	2.247E − 10

k_a = association rate constant; k_d= disassociation rate constant; K_D = equilibrium disassociation constant.

3.5. Immunohistochemistry against PD-L1 in canine cancers

Four dilutions were compared to the reaction in normal canine lymph nodes to evaluate the lowest concentration of each antibody clone that showed a strong immunohistochemical signal intensity. Heat-induced epitope retrieval in a low-pH buffer was successful in recovering PD-L1 expression in the FFPE samples. The appropriate dilution for 5D2 and 15D12 was 1:100 (final concentrations at 0.04 and 0.11 mg/ml, respectively), while 18F2 and 21A1 had to be used at dilution 1:25 (final concentrations at 0.1 and 0.14 mg/ml, respectively) with comparable results. Then, all four anti-cPD-L1 mAbs were investigated for PD-L1 expression in canine CTLs, TMCs, and STSs. More than 90% of the neoplastic population exhibited PD-L1 in all positive FFPE cases. PD-L1 was expressed in all CTLs and TMCs; however, only four STSs (4/5, 80%) showed PD-L1 immunoreactivity. In addition, PD-L1 clone 5D2 showed the highest performance with intermediate to strong signal intensities in all samples (Table 3; Figure 3). For the preliminary study, five MMs, five SCCs, and one PTCL illustrated strong reactivity to PD-L1 clone 5D2 (Figure 4). To determine the cross-reaction with human PD-L1, four unstained slides of human oral SCC were immunohistochemically labelled with cPD-L1 mAbs. For 5D2 and 15D12, the neoplastic cells showed nonspecific reactions in the cytoplasmic areas, while for 18F2 and 21A1, none of the neoplastic cells exhibited immunoreactivity.

Figure 3. Canine PD-L1 immunohistochemical staining in cutaneous T-cell lymphoma. (A) Clone 5D2. (B) Clone 15D12. (C) Clone 18F2. (D) Clone 21A1. 400×. IHC. Bar = 10 µm.

Figure 4. Immunohistochemistry staining of cPD-L1-positive tissues of clone 5D2. (A) 100% of neoplastic cells in squamous cell carcinoma, (B) oral malignant melanoma, (C) soft tissue sarcoma, and (D) tubular mammary carcinoma presents strong membranous and cytoplasmic staining for cPD-L1 (400×). IHC. Bar = 10 µm.

Table 3. Immunohistochemical results of mouse anti-cPD-L1 mAbs in five cases of cutaneous T-cell lymphoma, tubular mammary carcinoma, and soft tissue sarcoma.

	FFPE samples	Cutaneous T-cell lymphoma (N = 5)				Tubular mammary carcinoma (N = 5)				Soft tissue sarcoma (N = 5)
	Clone	5D2	15D12	18F2	21A1	5D2	15D12	18F2	21A1	5D2	15D12	18F2	21A1
Signal intensity	Strong	4/5 (80%)	2/5 (40%)	2/5 (40%)	2/5 (40%)	4/5 (80%)	1/5 (20%)	–	1/5 (20%)	4/5 (80%)	2/5 (40%)	–	–
	Intermediate	1/5 (20%)	3/5 (60%)	2/5 (40%)	2/5 (40%)	1/5 (20%)	4/5 (80%)	3/5 (60%)	2/5 (40%)	–	2/5 (40%)	–	1/5 (20%)
	Weak	–	–	1/5 (20%)	1/5 (20%)	–	–	2/5 (40%)	2/5 (40%)	–	–	4/5 (80%)	3/5 (60%)
	Negative	–	–	–	–	–	–	–	–	1/5 (20%)	1/5 (20%)	1/5 (20%)	1/5 (20%)

4. Discussion

A number of ICIs against PD-L1 are commercially available in human medicine, such as avelumab, durvalumab, and atezolizumab, and they have been approved for treating human solid cancers (Murciano-Goroff et al. 2020). However, no anti-PD-L1 monoclonal antibody specific to dogs exists in the companion animal industry. One study investigated the cross-reactivity and functionality of seven proven human ICIs, including atezolizumab, avelumab, and durvalumab, in vitro and ex vivo in dogs. Only atezolizumab was bound to recombinant and natural canine PD-1 protein, blocked the interaction between PD-1 and PD-L1, and increased the production of IFN-γ (Pantelyushin et al. 2021). Therefore, anti-cPD-L1 mAbs are required for clinical treatment. Several research groups have been developed and established cPD-L1-specific mAbs (Maekawa et al. 2017; Nemoto et al. 2018; Choi et al. 2020). Maekawa et al. (2017) established a rat-dog chimeric anti-PD-L1 mAb, consisting of constant region Ig from a dog with the variable region from a rat to reduce immunogenicity and avoid neutralization from antibody response in dogs. According to their results, this chimeric mAb worked efficiently in vitro because it enhanced the production of IL-2 and IFN-γ and the proliferation of T lymphocytes. Dogs with cancer (seven malignant melanomas and two undifferentiated sarcomas) were repeatedly received chimeric cPD-L1 mAb. The patients were well tolerated and tumor burdens were reduced in some dogs (Maekawa et al. 2017). The administration of this antibody was further investigated in 29 dogs with advanced-stage oral MMs. Adverse event effects of any grade were observed in 50% of patients, and treatment dogs had prolonged survival times compared to the control group (Maekawa et al. 2021). Nemoto et al. (2018) also developed rat anti-cPD-L1 mAbs by immunization with NRK/cPDL1 transduced cells. Two rat hybridomas secreting monoclonal antibodies for cPD-L1 were cloned, and both mAbs had binding and inhibiting efficiency against NRK/cPDL1 cells detected by western blot. However, IFN-γ production in concanavalin A-stimulated PBMCs from nine dogs after incubating with cPD-L1 mAb was notably increased in six dogs, suggesting that some factors, such as the number of activated T cells, the level of PD-L1 expression, and the antibody affinity to T cells, had an impact on treatment response in each patient (Nemoto et al. 2018). Nonetheless, only the cPD-1 antibody was conducted in a clinical study to treat canine cancers reported by Nemoto’s colleagues (Igase et al. 2020). Another study generated mouse mAb specific for cPD-L1 using two immunization methods with recombinant cPD-L1 Ig (Choi et al. 2020). Four mAbs of anti-cPD-L1 exhibited specificity for either recombinant or natural canine ligands. Moreover, these mAbs had blockade ability over canine PBMCs that produced 3–6 folds IFN-γ comparing to the control group. In addition, our work successfully invented four mouse mAbs of anti-cPD-L1 using a traditional prime-boost strategy using CFA. These antibodies showed specific binding to both recombinant and native molecules in dogs. They also had the diagnostic potential to detect the expression of PD-L1 in canine tumor cells using immunohistochemistry. Nevertheless, functional assays of four candidate-specific cPD-L1 antibodies and clinical applications as therapeutic targets are required in future investigations.

In this study, we evaluated the diagnostic application of these anti-cPD-L1 mAbs to evaluate PD-L1 expression in canine FFPE samples. All four mouse cPD-L1 mAbs worked well and bound with the natural ligand of cPD-L1 in the FFPE tissues. High expression of PD-L1 was exhibited in 100% of neoplastic cells in CTL, TMC, MM, SCC, PTCL, and STS similar to previous studies (Maekawa et al. 2016; Shosu et al. 2016; Ariyarathna et al. 2020; Maekawa et al. 2021). According to the ELISA and binding kinetic results, the 5D2 and 18F2 mAbs showed high binding affinity to recombinant cPD-L1. In addition, the 5D2 mAb had the greatest diagnostic application on the immunohistochemical platform. Other applications, such as western blot, flow cytometry, and immunocytochemistry, are needed to confirm their performance and utility for diagnosis and prognosis.

Recently, immunotherapy against ICIs has effectively shown beneficial treatment in various types of human cancers. In veterinary oncology, immunotherapy using ICIs is still lacking in a large-scale commercial industry, including ICI targeting PD-L1. A few publications have reported on the therapeutic effects of PD1/PD-L1 checkpoint inhibitors in clinical studies. Some dogs with clinical stage IV spontaneous oral MM showed partial or complete responses after treatment with PD1/PD-L1 blockage antibodies (Igase et al. 2020; Maekawa et al. 2021). Furthermore, the upregulation of PD-L1 in cancer dogs was correlated with prognosis. A high risk of relapse/metastasis and short survival time was observed in cPD-L1-overexpressed cancers, for example, DLBCL, mammary carcinoma, and anal sac adenocarcinoma (Ariyarathna et al. 2020; Aresu et al. 2021; Minoli et al. 2022). Therefore, a prospective clinical study with an increased sample size of various canine cancers will be helpful in confirming their treatment efficacy and the prognostic value of PD-1/PD-L1 immunotherapy.

5. Conclusion

We successfully generated mouse monoclonal antibodies against the canine PD-L1 molecule. The 5D2, 15D12, 18F2, and 21A1 mAbs effectively bound to both recombinant and native forms of canine PD-L1 protein, as confirmed by ELISA, SPR, and immunohistochemistry, respectively. Nevertheless, the potential application of these antibodies as immune checkpoint inhibitors for cancer immunotherapy in veterinary applications necessitates additional in vitro and clinical investigations.

Acknowledgments

We would like to thank Dr.Sirinun Pisamai, Department of Veterinary Surgery, Faculty of Veterinary Science, Chulalongkorn University for providing unstained slides of human oral squamous cell carcinoma. SS was supported by the Postdoctoral Fellowship, Ratchadapisek Somphot Fund, Chulalongkorn University.

Disclosure Statement

The authors report there are no competing interests to declare.

Additional information

Funding

This research was funded by the Thailand Science Research and Innovation Fund, Chulalongkorn University (CU_FRB65_food(21)_185_31_04) and the Center of Excellence for Companion Animal Cancer, Faculty of Veterinary Science, Chulalongkorn University.