PD-L1 expression in tissues of cats with naturally occurring feline infectious peritonitis

Abstract

Among feline coronaviruses (FCoVs), feline enteric coronavirus (FECV) is a ubiquitous enteric pathogen in feline populations and responsible for asymptomatic to mild enteric disease. However, in a small number of cats, viral mutations within the host combined with impaired host immune responses may lead to feline infectious peritonitis (FIP), a fatal, systemic infection. Programmed death-1 (PD-1) is a major immune-checkpoint protein expressed on activated T cells, and its engagement by the ligands PD-L1 and PD-L2 on target cells leads to suppression of T T-cell activity. Previously, we found that infection with FIPV, but not FECV, up-regulated the expression of PD-L1 in feline cell lines, which attenuated T-cell activities. We also demonstrated that feline interferons (IFNs), particularly IFN-γ, are potent inducers of PD-L1 in cells. In this study, we assess the expression of PD-L1 in association with FIPV in the tissues of four cats with naturally occurring FIP using histopathology, real-time RT-qPCR, and confocal microscopy. By confocal microscopy, we found that numerous cells co-localized with FCoV and PD-L1 in various tissues of all animals, with about 50% of FCoV-positive cells also stained with PD-L1. Interestingly, about 10% of PD-L1 positive cells were also positive for FCoV, which is likely due to severe inflammation and cytokines such as IFN-γ. These findings support a tissue-level association between PD-L1 expression and FCoV infection in some tissues from cats with FIP, while also indicating substantial variation among tissues and animals.

Introduction

There are two biotypes of feline coronavirus (FCoV), feline infectious peritonitis virus (FIPV) and feline enteric coronavirus (FECV). FECV is a common enteric pathogen in feline populations and commonly presents as an enteric infection (Pedersen et al., 2008). However, in a small population of cats, this enteric infection can give rise to a fatal systemic disease known as feline infectious peritonitis (FIP). Clinically, cats with FIP exhibit various signs such as pyrexia, lethargy, weight loss, fluid accumulation in body cavities, and neurologic or ocular signs, depending on the form of the disease (Pedersen, 2014a; Pedersen et al., 2015; Ritz et al., 2007). Without antiviral treatment, FIP is invariably fatal, and the exact mechanism underlying its development remains unknown.

It is widely accepted that FIP develops in individual cats as a consequence of viral mutations combined with an inadequate cellular immune response following FECV infection. The hallmark distinction between FECV and FIPV is the shift in viral tropism from enterocytes to macrophages (Pedersen et al., 2008). Currently, mutations in the ORF 3c gene and the spike (S) gene have been associated with this tropism change and FIP development (Chang et al., 2012; Hoffmann et al., 2020; Licitra et al., 2013; Papa et al., 2021; Pedersen et al., 2009; Pedersen et al., 2012; Wu and Zhao, 2021). T-cell immunity plays a vital role in the control of virus-infected cells, and resistance to FIP development has been associated with a robust FCoV-specific T-cell response (de Groot-Mijnes et al., 2002; Paltrinieri et al., 2003). The importance of effective T-cell immunity in the course of coronavirus infections has also been demonstrated in studies of murine hepatitis virus (MHV), a murine coronavirus (Compton et al., 2004; Homberger, 1997; Homberger et al., 1998; Pedersen, 2014b). In these studies, T-cell-deficient mice infected with an enterotropic strain of MHV were shown to develop a fatal systemic condition resembling FIP (Compton et al., 2004; Homberger, 1997; Homberger et al., 1998; Pedersen, 2014b). Many viruses have mechanisms to circumvent T-cell immunity, and the programmed cell death protein 1 (PD-1) axis is a well-established target for viral evasion strategies. PD-1 is a major checkpoint protein expressed on activated T cells, B cells, and myeloid cells. It regulates overactive T-cell responses by inhibiting downstream signal cascades (Chamoto et al., 2023; Latchman et al., 2001; Patsoukis et al., 2020; Yokosuka et al., 2012). PD-1 has two ligands: PD-L1, which is expressed by a wide variety of cell types, and PD-L2, which is primarily expressed on hematopoietic cells (Dong et al., 2002; Ishida et al., 2002; Iwai et al., 2003; Liang et al., 2003; Loke and Allison, 2003; Sharpe et al., 2007; Zhong et al., 2007). Macrophages, the target cells of FIPV, express both PD-L1 and PD-L2, as well as PD-1 (Dong et al., 2002; Ishida et al., 2002; Iwai et al., 2003; Liang et al., 2003; Loke and Allison, 2003; Sharpe et al., 2007; Zhong et al., 2007). Inflammatory cytokines, such as interferon (IFN)-γ, are well-known inducers of PD-L1 and PD-L2 expression (Garcia-Diaz et al., 2017). Many viruses have also been shown to up-regulate PD-L1 expression as an immune evasion strategy (Channappanavar et al., 2012; Huang et al., 2023; Lafon et al., 2008; Maier et al., 2007; Planès et al., 2014; Raftery et al., 2018; S et al., 2022; Sabbatino et al., 2023; Schönrich and Raftery, 2019; Shaabani et al., 2016; Telcian et al., 2011; Valero-Pacheco et al., 2013; Xu et al., 2010; Yuan et al., 2024).

Previously, we reported that PD-L1 is up-regulated in Crandell–Rees feline kidney (CRFK) cells and Felis catus whole fetus (Fcwf-4) cells infected with serotype II FIPV-1146 strain and serotype I FIPV Black strain, whereas infection with serotype II FECV-1683 strain did not induce PD-L1 expression (Zabiegala et al., 2025). In addition, PD-L1 expression induced by FIPV infection suppressed T-cell activation in a co-culture assay with Jurkat cells. Furthermore, we found that feline IFNs, particularly IFN-γ, are potent inducers of PD-L1 expression in these cells. These findings suggest that FIPV exploits host immune pathways to suppress T-cell activity, thereby facilitating survival of FIPV.

Given these findings, we sought to investigate PD-L1 expression in the tissues of cats with naturally occurring FIP. Using post-mortem tissue samples, we examined the co-localization of PD-L1 and FCoV using confocal microscopy and assessed the RNA levels of PD-L1 and FCoV using real-time RT-qPCR. The results revealed the clustered regions of co-localized PD-L1 and FCoV in various tissues of all animals with about 50% of FIPV-positive cells were also stained with PD-L1. We also found only small fractions of PD-L1 positive cells were positive for FIPV. The relative levels of PD-L1 RNA and viral RNA varied among different tissues and individual cats, which might be due to inflammatory cytokines such as IFN-γ (Berg et al., 2005). These findings support our previous findings and warrant further studies on the role of PD-L1 and T cells in FIP development in cats.

Materials and Methods

Sample collection and preparation

Frozen and fixed tissue samples were obtained from four deceased cats (designated as Cats A, B, C and D) diagnosed with FIP by immunohistochemistry (IHC) at Kansas State Veterinary Diagnostic Lab (Table 1). Frozen tissues were initially stored at −20°C before being transferred to −80°C for long-term storage. For histology, IHC, and confocal microscopy, tissue samples were fixed in 10% formalin (StatLab, McKinney, Texas) for a minimum of 24 h prior to paraffin embedding. Tissues were dehydrated six times in increasing concentrations of ethanol (70–100%) for total 240 min at 40°C in a Sakura Tissue-Tek VIP 6 Vacuum Infiltration Processor (Sakura Finetek USA, Inc., Torrance, California). Tissues were embedded in paraffin (Leica Biosystems, Nussloch Germany), heated to 61°C, and sectioned into 4-μm-thick layers using a Leica RM2235 rotary microtome (Leica Biosystems, Nussloch, Germany). Tissues were transferred to positively charged histological slides until further preparation.

Table 1:

Summary of clinical signs and histopathological lesions of four cats with FIP

	Cat A	Cat B	Cat C	Cat D
Age, Sex, Breed	3-year-old MN DSH	11-year-old FS DSH	8-month-old FS Bengal	4-month-old MN Siamese
Clinical Signs/Given History	Lethargy Pyrexia Inappetence Ascites Leukocytosis Lymphadenopathy	Unable to thermoregulate Thrombocytopenia Lymphocytosis Monocytosis Hyperphosphatemia Hypocalcemia Hypochloremia Hypoalbuminemia Hyperglobulinemia Hyperbilirubinemia	Self-resolving hematochezia (1 month prior) Found deceased	Euthanized
Form	Wet	Mixed	Mixed	Mixed
Effusion	Abdominal	Abdominal, thoracic	Abdominal, Pericardial	Pericardial

Notes: MN = male neutered, DSH = domestic shorthair, FS = female spayed.

Histology and immunohistochemistry

Slides were processed in a Leica Autostainer ST5010 (Leica Microsystems, Nussloch Germany) for hematoxylin and eosin staining. Slides were baked at 60°C for 10 min. Paraffin-embedded tissue slides were treated three times with XS-3 for 3 min each at room temperature. Slides were rehydrated using decreasing concentrations of ethanol (100%, 100%, 95%, and 85%) for 5 min each and then washed for 1 min with tap water at room temperature. Slides were then stained with hematoxylin for 12 min at room temperature, followed by a 2-min wash in tap water. Select Acid Rinse was then used for 1 min at room temperature, and then slides were washed for 1 min with tap water. Select Bluing Solution was then used for 1 min at room temperature, followed by a 1-min tap water wash. The slides were then treated with 95% ethanol for 1 min and then counterstained with eosin for 30 s. This was followed by three washes in 100% ethanol for 1 min each at room temperature, followed by three washes in XS-3 for 3 min each at room temperature. The slides were then automatically transferred to the CV5030 automatic coverslipper, where coverslips were mounted.

For diagnosis of FIP using immunohistochemistry, slides were processed in a Leica BOND-MAX automatic stainer (Leica Microsystems, Nussloch Germany) with the Bond Polymer Refine Detection kit – DAB chromogen (Leica Biosystems, Newcastle, United Kingdom). Paraffin was removed with BOND Dewax Solution (Leica Microsystems) for 30 s at 72°C, three washes in alcohol at room temperature, and three washes in BOND Wash Solution (Leica Microsystems) at room temperature, with the final wash lasting 5 min. Pretreatment antigen retrieval was performed with BOND Epitope Retrieval Solution 2 (Leica Biosystems, Newcastle, United Kingdom), an EDTA-based pH 9.0 solution, for 20 min at 100°C, followed by three washes with BOND Wash Solution at 35°C, with a final wash for 3 min at room temperature. Primary incubation of the mAb FIPV3–70 clonal antibody (Custom Monoclonals International, Sacramento, California) at a dilution of 1:1,000 was performed for 15 min at room temperature and was followed by three washes with BOND Wash Solution at room temperature. Secondary incubation with the PowerVision Poly-HRP α-mouse IgG polymer (Leica Biosystems, Newcastle, United Kingdom) was then performed for 25 min at room temperature. Final washes with BOND Wash Solution were then performed at room temperature for 3 min, 5 min, and then 2 min, followed by a fourth and final wash for 5 min. The slides were washed twice with deionized water at room temperature and then washed with Mixed DAB Refine (Leica Microsystems), followed by a 10 min incubation at room temperature. Slides were washed four more times with deionized water at room temperature. The slides were then counterstained for 5 min with hematoxylin from the Bond polymer Refine Detection kit at room temperature. Final washes were performed with deionized water, BOND Wash Solution, and then deionized water at room temperature.

RNA extraction from tissues

For RNA extraction, 60 mg of three sections of each tissue was homogenized with an Omni Bead Ruptor 24 (Kennesaw, GA) homogenizer using 1.4-mm ceramic beads. Homogenates were centrifuged at 18,373 ×g for 3 min prior to aliquoting for RNA extraction. Total RNA of cell lysates was extracted using Qiagen RNeasy Mini purification kit (Germantown, MD).

Serotyping of FCoVs and partial sequence analysis of S protein and 3C-like protease (3CLpro) of FCoVs from the cats

For 3CLpro, cDNA was generated using the Promega GoTaq 2-step RT-qPCR system (Madison, WI) with a reverse primer targeting the FCoV 3CLpro on samples with the highest viral load from three cats (the liver from Cat A, spleen from Cat B, and small intestine from Cat C), following the manufacturer’s protocol. PCR amplification was performed with an initial denaturation step at 94°C for 2 min, followed by 40 cycles of denaturation (94°C for 30 s), annealing (55°C for 30 s), and extension (72°C for 1.5 min). For geno(sero)typing, primers targeting the 3′ end of S gene listed in Table 2 (lubs, lffs, lcfs, nlubs, nlffles, and nlcfs) were used as previously described by Addie et al (2003). Types I and II FCoV were distinguished by different sizes of PCR products. The amplified 3CLpro and S genes were sequenced by Genewiz (South Plainfield, NJ). Multiple sequence alignment was performed using Clustal Omega for phylogenetic analysis as previously described (Madeira et al., 2024). The reference serotype I and II FCoVs included FIP-I-C1Je (GenBank No: DQ848678), FIP-I-black (EU186072), FIP-I-26M (KP143512), FIP-I-28O (KP143508), FIP-I-27C (KP143507), FIP-I-HLJ/HRB (KY566209), FIP-I-UU3 (FJ938061), FIP-I-UU2 (FJ938060), FIP-I-HLJ-DQ (KY292377), FIP-II-1146 (DQ010921), FECV-II-1683 (JN634064), FIP-II-DF2 (DQ286389), FIP-II-NTU156 (GQ152141). FECV-I-RM (FJ938051), and FECV-I-UG-FH8 (KX722529).

Table 2.

Primers and probe used in this study

Primer/probe		Sequences
fPD-L1 Forward		5' - CTATGGCGGTGCTGACTATAAG - 3'
fPD-L1 Reverse		5' - CCTCAGCCTGACACATTAGTT - 3'
fPD-L1 Probe		5' - /56-FAM/TCTGTGGAT/ZEN/CCTGTCACCTCTGAACA/3IABkFQ - 3'
β-actin Forward		5' - GGCATCCACGAAACTACCTT - 3'
β-actin Reverse		5'-AGCACTGTGTTGGCGTACAG-3'
β-actin Probe		5'-/56-FAM/ATCATGAAG/ZEN/TGTGACGTGGACATCCG/3IABkFQ-3’
FCoV 3’ UTR Forward		5’ - GATCCAGACGTTAGCTCTTCC - 3'
FCoV 3’ UTR Reverse		5’ - GGAGGTACAAGCAACCCTATT - 3'
FCoV 3CLpro Forward		5’ - AATGAGTTGTCTAGTGTGCGTTTA - 3’
FCoV 3CLpro Reverse		5’ - TAATACTAACAAATGTCGGGTTAAT - 3’
FCoV S	lubs	5' - CCACACATACCAAGGCC - 3'
	lffs	5' - GTTTCAACCTAGAAAGCCTCAGAT - 3'
	lcfs	5' - GCCTAGTATTATACCTGACTA - 3'
	nlubs	5' - CCTAGAAAGCCTCAGATGAGTG - 3'
	nlffles	5' - CCTAGAAAGCCTCAGATGAGTG - 3'
	nlcfs	5' - CAGACCAAACTGGACTGTAC - 3'

Real-time RT-qPCR for fPD-L1 and FCoV

RT-qPCR was performed using the SuperScript One-Step RT-PCR system with Taq DNA Polymerase (Thermo Fisher) for feline PD-L1 (fPD-L1), FCoV (3’ UTR), and β-actin as a control. Table 2 contains the sequences of primers and probes used for real-time RT-qPCR. For quantitative analysis, samples were incubated at 50°C for 30 min, and then pre-denatured at 95°C for 2 min using random hexamers. PCR amplification was conducted over 40 cycles, with denaturation at 94°C for 15 s, annealing at 55°C for 30 s, and extension at 72°C for 30 s. Relative mRNA levels of PD-L1 and FCoV RNA, normalized to β-actin, were calculated using the 2^−Δct method. To evaluate the relationship between levels of PD-L1 and FCoV RNA in each cat, log–log line non-linear regression analysis was performed to calculate R² using GraphPad Prism software version 9.5 (San Diego, CA). The data points for the analysis were 8, 9, 12, and 9 for Cat A, B, C, and D, respectively.

Immunofluorescence confocal microscopy

Sections of the various tissues from Cats A–D were examined using confocal microscopy. Slides for immunofluorescence confocal microscopy were prepared manually and processed for immunofluorescence staining. Paraffin-embedded tissue slides were treated three times with xylenes for 3 min each. Slides were rehydrated using decreasing concentrations of ethanol (100%, 95%, 70%, and 50%) for 5 min each. Slides were rinsed with deionized distilled water before treatment with phosphate-buffered saline (PBS, pH 7.4) with 0.05% Tween20 for 10 min. Antigens were retrieved by heat-induced antigen retrieval using a Tris-EDTA solution (pH 9.0) for 20 min at 100°C, and washed with deionized distilled water once cooled. Slides were permeabilized with PBS with 0.5% Tween20 (Fisher Scientific, Pittsburgh, PA) in PBS 10 min at room temperature. After three washes with PBS, the slides were incubated with a blocking buffer of 5% bovine serum albumin in PBS with 0.05% Tween20 for 1 h at room temperature. Slides were then treated with a mixture of rabbit anti-PD-L1 antibodies 28–8 and RM1012 (Abcam, Waltham, MA), E1L3N rabbit mAb (Cell Signaling Technology, Danvers, MA), and mouse anti-FCoV antibodies (BioRad, Hercules, CA) diluted in blocking buffer at 4°C overnight. Anti-FCoV antibodies were diluted at a 1:200 ratio, anti-PD-L1 clone 28–8 was diluted at a 1:100 ratio, and anti-PD-L1 RM1012 clone and E1L3N clone were diluted at a 1:200 ratio. These PD-L1 antibodies were anti-human clones used because no commercial feline PD-L1 antibody was available, and their use in feline formalin-fixed paraffin-embedded is a limitation. Tissues were washed three times with PBS 0.05% Tween20 for 5 min before incubation with an anti-rabbit secondary antibody conjugated with AlexaFluor 555 (1:500) or an anti-mouse secondary antibody conjugated with AlexaFluor 647 (1:1,000) for 2 h at room temperature. Tissues were washed three times with PBS 0.05% Tween20 before coverslips were mounted in Prolong Gold Antifade Mountant with DNA Stain DAPI (Invitrogen, Waltham, MA). Samples were scanned with a Zeiss LSM 700 confocal microscope (Zeiss, Oberkochen, Germany). An unstained slide was used to calibrate the scope and remove autofluorescence signals, which also served as a control. Z slices were acquired at 40× magnification and Fiji software was used to process images, generate Costes’ masks from Z-projections, and calculate the Manders coefficients. Fcwf-4 cells infected with or without FIPV-1146 were used for controls for confocal staining.

Results

Gross and histological lesions

The primary gross findings in Cat A (3-year-old, male neutered, domestic shorthair) included ascites, fibrin deposits on the serosa, and multifocal to coalescing hepatic nodules (Table 1). The spleen contained mixed inflammatory cells and fibrin accumulation on the serosal surface. Mesenteric lymph nodes were effaced by pyogranulomatous inflammatory aggregates and fibrin, and the capsules were expanded severely by fibrin and inflammatory cells (chronic serositis). Histologically, the liver had severe, multifocal pyogranulomatous to lymphocytic, histiocytic, and neutrophilic necrotizing hepatitis (Fig. 1A). Mild perivascular inflammation was present in the kidney. Diagnostic IHC for FCoV performed on the liver and spleen showed strong immunopositive staining within inflammatory foci in the liver and along the serosal surface of the liver and spleen (Fig. 1B; only liver is shown).

Figure 1:

Histological lesions of Cats A and B. (A) Cat A, liver, H&E stain, 10× magnification. Severe multifocal pyogranulomatous to lymphocytic histiocytic, and neutrophilic necrotizing hepatitis. (B) Cat A, liver, FCoV IHC, 10× magnification. Strong immunopositive in inflammatory foci in the liver and throughout the capsule. (C) Cat B, liver, H&E, 10× magnification. Overlying the capsular surface of the liver is a thick band of fibrin with low numbers of intermixed neutrophils and macrophages. Deeper within the hepatic parenchyma is a nodular aggregate of epithelioid macrophages with fewer lymphocytes and neutrophils surrounding a core of necrotic debris (pyogranuloma). (D) Cat B, liver, FCoV IHC, 10× magnification. Several macrophages within areas of inflammation and fibrin along the hepatic capsule and within the hepatic parenchyma exhibit strong cytoplasmic immunolabeling for feline coronavirus, consistent with FIP.

Bicavitary effusion was noted on Cat B (11-year-old, female spayed, domestic shorthair) with fibrinous and granulomatous peritonitis and pleuritis (Table 1). Limited histopathology was performed on the liver and various lymph nodes. Histologically, the parenchyma of the liver had multifocal to coalescing granulomatous aggregates and the capsule was expanded by fibrin, granulomatous inflammation, and karyorrhectic debris (Fig. 1C). Examined lymph nodes were effaced by pyogranulomatous inflammatory aggregates with capsules moderately expanded by fibrin and lymphoplasmacytic inflammation. Throughout the lymph nodes and liver (Fig. 1D), there were multifocal, strongly immunopositive areas for FCoV IHC.

Gross necropsy of Cat C (8-month-old, female spayed, Bengal) showed severe ascites, along with adhesions of the omentum to the large intestine and peritoneum (Table 1). The small intestine had severe, acute fibrinosuppurative serositis/peritonitis with extension into the tunica muscularis (Fig. 2A). Lymph nodes were effaced by severe, acute fibrinosuppurative and necrotizing lymphadenitis with lymphoid depletion (Fig. 2C). Diagnostic IHC for FCoV showed strong immunopositive staining multifocally throughout the intestinal serosa (Fig. 2B) and the lumen of the intestine, and throughout the lymph nodes (Fig. 2D), with mild immunopositive staining in the liver.

Figure 2:

Histological lesions of Cats C and D. (A) Small intestine serosal surface, H&E stain, 10× magnification. Severe, acute, fibrinosuppurative serositis/peritonitis. (B) Cat C, small intestine, FCoV IHC staining, 20×. Multifocal regions of strong immunopositive staining in inflammatory foci in the serosa and tunica muscularis. (C) Cat C, lymph node, H&E, 20×. Marked lymphoid depletion with poorly developed follicles. Large numbers of degenerate neutrophils mixed with bacterial colonies. (D) Cat C, lymph node, FCoV IHC staining, 20×. Multifocal areas of strong FCoV immunopositive staining were seen throughout the node. (E) Cat D, kidney, H&E, 10×. Marked interstitial granulomatous inflammation with central coagulative to lytic necrosis and vascular infiltration. (F) Cat D, kidney, FCoV IHC staining, 10×. Focal area of strong immunopositive staining within inflammatory foci in the corticomedullary region.

Initial gross examination of Cat D (4-month-old, male neutered, Siamese) showed severe depletion of adipose tissue, moderate muscle atrophy, and moderate to severely pale mucous membranes. In the thorax, mild to moderate serofibrinous effusion was found in the pericardial sac and mild to moderate diffuse pulmonary edema was noted. Histologically, the leptomeninges and choroid plexi within the cerebrum, cerebellum, and brainstem were expanded by moderate, multifocal, granulomatous to pyogranulomatous inflammatory aggregates with phlebitis, lymphangitis, as well as multifocal periventricular and perivascular encephalitis. The kidneys contained multifocal to coalescing granulomatous interstitial nephritis with phlebitis, lymphangitis, parenchymal necrosis, interstitial edema, periglomerular fibrosis, tubular ectasia, proteinosis, and hemorrhage (Fig. 2E). Mild to moderate multifocal granulomatous to pyogranulomatous hepatitis with hepatic necrosis was observed in the liver and mild to moderate multifocal lymphohistiocytic serositis was observed in the spleen. Moderate, locally extensive, pyogranulomatous lymphadenitis was observed in the lymph nodes. Throughout the liver parenchyma, and within the kidney interstitium and urothelium, there were occasional strongly immunopositive areas of FCoV staining (Fig. 2F).

Genetic Analysis of 3CLpro and S

Genotyping RT-PCR confirmed all FCoVs from the cats as serotype I FIPV based on the size of the amplified products (~376 bp) in gels. The amino acid homology of 3CLpro (280 out of 302 aa) and partial sequence of S protein (region surrounding heptad repeat 2 in the fusion domain) from Cats A, B, and C showed 98.35–98.76% and 90.68–94.17% homology among themselves, respectively. We were unable to generate DNA sequencing data from Cat D tissues; therefore, further sequence analysis was not included in this study. The phylogenetic tree based on the partial S-gene sequence from the conserved heptad repeat 2 region showed A, B, and C forming a closely related cluster with short internal branch lengths. The phylogenetic tree derived from the 3CLpro likewise displayed short branch lengths and tight clustering of A, B, and C. Taken together, both analyses demonstrated the same overall pattern of close relatedness among Cats A, B, and C (Figs. 3A and 3B).

Figure 3:

Phylogenetic analysis of FCoV S (A) and 3CLpro (B) from Cats A (A-liver), B (B-spleen), and C (C-SI) and reference serotype I and II FECV and FIPV strains.

Viral load and PD-L1 mRNA expression

Relative PD-L1 expression in relation to viral load was variable among cats (Figs. 4A–D). Cat A demonstrated elevated PD-L1 expression alongside increasing FCoV load (R² = 0.90, Fig. 4A), with the liver showing the highest values (Fig. 4A), which is consistent with the severe necrotizing hepatitis and strong FCoV immunostaining observed in this tissue (Figs. 1A and 1B). The PD-L1 and FCoV RNA levels were moderate in the spleen and lowest in the kidney (Fig. 4A), which aligns with the relative severity of the histopathological changes observed in these tissues. In Cat B, two out of three sections of both liver and spleen showed increased PD-L1 expression at higher viral levels (Fig. 4B). Similarly to Cat A, the liver and spleen showed the highest viral load and PD-L1 expression, whereas the kidney generally showed lower viral and PD-L1 RNA expression levels (Fig. 4B). An overall moderate association was observed between PD-L1 and FCoV RNA levels in Cat B (R² = 0.64, Fig. 4B). However, Cats C and D showed no apparent correlation between viral load and PD-L1 expression (Figs. 4C and 4D).

Figure 4:

The relative levels of PD-L1 mRNA and FIPV RNA. Relative PD-L1 mRNA (Y-axis) and FCoV RNA (X-axis), normalized to β actin, were plotted as 2–ΔCt values for Cats A (A), B (B), C (C), and D (D).

Confocal microscopy of PD-L1 and FCoV expression

The expression of PD-L1 in tissues infected with FIP was assessed using confocal microscopy. Overall, the highest PD-L1 expression in tissues was commonly clustered with or around virus-infected cells in liver, kidney, lymph nodes, and serosa of the small intestine (Figs. 5–7). Colocalization assessments were performed on the Z-projection of the tissues shown in Figs. 5–7 at 40× magnification and quantified using the Manders coefficient. PD-L1 immunostaining is shown in yellow, FCoV immunostaining is shown in red, and Costes’ mask colocalization assessment is shown in white. Confocal images are presented in Figs. 5–7, with cells with apparent co-localizations shown in Fig. 5 and 6, and cells without co-localizations in Fig 7.

Figure 5:

Confocal microscopy of PD-L1 expression (yellow), FCoV infection (red), and colocalization (white) in the liver of Cats A (A and B), B (C and D), and C (E and F). (A) Liver capsule of Cat A, 20× magnification. The white square demarcates the region shown in B. (B) Costes mask of Z-projection of the liver of Cat A at 40× magnification. (C and D) Costes mask of Z-projection of the liver of Cat B at 40× magnification. (E) Liver of Cat C at 20× magnification with the white square demarcating the region seen in F. (F) Costes mask of Z-projection of the liver of Cat C at 40× magnification.

Figure 7:

Confocal microscopy of PD-L1 expression (yellow), FCoV infection (red), and colocalization (white) in the small intestine (A and B) and the lymph node (C and D) of Cat C. (A) Serosa of the small intestine of Cat C at 20× magnification with a white square demarcating the section seen in B. (B) Costes’ mask of the Z-projection of the serosa of the small intestine of Cat C at 40× magnification. (C) Lymph node of Cat C at 20× magnification with a white square demarcating the section seen in D. (D) Costes mask of the Z-projection of lymph node of Cat C at 40× magnification.

Figure 6:

Confocal microscopy of PD-L1 expression (yellow), FCoV infection (red), and colocalization (white) in the lymph nodes of Cat B and the kidney of Cat D. (A) Lymph node capsule of Cat B at 20× magnification. The white square demarcates the 40× magnification seen in B. (B) Costes mask of Z-projection of the lymph node of Cat B at 40× magnification. (C) Kidney of Cat D at 40× magnification. (D) Costes mask of the Z-projection of the kidney of Cat D at 40× magnification.

The livers of Cats A (Figs. 1B and 5A), B (Figs. 1D and Fig. 5C), and C (Figs. 2B and 5E) were strongly positive for FCoV. In Cat A, approximately 97% of FCoV expression overlapped with the PD-L1 signal, while 17% of PD-L1 expression overlapped with FCoV infection (Fig. 5B). Approximately 44% (Fig. 5C) and 68% (Fig. 5D) of FCoV expression overlapped with the PD-L1 signal in the sections assessed in Cat B, while approximately 12% (Fig. 5C) and 4% (Fig. 5D) of PD-L1 expression overlapped with FCoV infection. Approximately 43% of FCoV in the liver of Cat C overlapped with the PD-L1 signal, while approximately 19% of PD-L1 expression overlapped with FCoV infection (Fig. 5F).

The lymph node of Cat B (Fig. 6A) and the kidney of Cat D (Figs. 2F and 6C) were strongly positive for FCoV. In the lymph node of Cat B, approximately 70% of FCoV expression overlapped with the PD-L1 signal, while approximately 10% of PD-L1 expression overlapped with FCoV infection (Fig. 6B). In the kidney of Cat D, approximately 43% of FCoV expression overlapped with the PD-L1 expression, while only 8% of PD-L1 expression overlapped with FCoV (Fig. 6D). However, this overlap was not observed in the tissues assessed in Cat C. The serosal and muscularis layers of the small intestine of Cat C were strongly positive for FCoV (Fig. 2B and 7A) and PD-L1 (Fig. 7B), and while expression was clustered together in the tissue, there was little to no overlap observed between the two. Similarly, the lymph node of Cat C showed strong staining for both FCoV (Fig. 2D and 7C) and PD-L1 (Fig. 7C), though little expression overlap was observed.

Discussion

Cytotoxic T cells and natural killer (NK) cells play a crucial role in eliminating virus-infected cells, and inhibition of these cells can promote viral survival. Viruses employ diverse strategies to manipulate cell-signaling pathways, thereby attenuating host immune responses. Some viruses such as severe acute respiratory syndrome Coronavirus 2 (SARS-CoV-2) and human cytomegalovirus (HCMV) inhibit T-cell activation by preventing the surface expression of the MHC-I-peptide complex (Ahn et al., 1996; Ahn et al., 1997; Chevalier and Johnson, 2003; Jones et al., 1996; Lehner et al., 1997; Park et al., 2004; Reyburn et al., 1997; Stagg et al., 2009; Zhang et al., 2021). Another mechanism by which viruses evade cell-mediated cytotoxicity is through the upregulation of immune checkpoint proteins. The PD-1/PD-L1 axis is a key checkpoint for T cells, functioning to suppress activation signaling when the T-cell receptor is engaged by an antigen. Many cancers and viruses have been shown to exploit the PD-1/PD-L1 signaling pathway as a mechanism of immune evasion. Viruses including lymphocytic choriomeningitis virus (LCMV), human immunodeficiency virus (HIV), simian immunodeficiency virus (SIV), influenza virus, herpes virus, respiratory syncytial virus, hepatitis B virus, human cytomegalovirus (HCMV), Hantavirus and SARS-CoV-2 have been found to up-regulate the expression of PD-L1 (Huang et al., 2023; Sabbatino et al., 2023; Schönrich and Raftery, 2019).

Additionally, PD-L1 expression can be induced via inflammatory signals such as IFNs. In particular, IFN-γ is a potent inducer of PD-L1, as demonstrated by us (Zabiegala et al., 2025) and other groups (Eppihimer et al., 2002; Garcia-Diaz et al., 2017). IFN-γ, primarily secreted by T cells and NK cells, is produced in response to various inflammatory cytokines such as IL-12, IL-15, IL-18, and type I IFNs (Castro et al., 2018). Studies of feline PD-L1 expression are very limited, with no reports on the typical expression levels in various tissues. One study assessed the expression of various proteins including PD-L1 on PBMCs cats with varying types of FCoV infection using an aptamer-based proteomics approach, but no difference was noted in the PBMCs among the groups (Curtis et al., 2024). Maekawa et al. assessed the PD-L1 expression in different tumor types using a PD-L1 antibody and showed positive staining on squamous cell carcinoma, mammary adenocarcinoma, fibrosarcoma, and renal cell carcinoma (Maekawa et al., 2023). Previously, it was reported that FIPV-1146 infection up-regulates PD-L1 mRNA expression in CRFK cells, and increased PD-L1 mRNA expression is seen in PBMCs from cats with FIP, implicating the PD-1/PD-L1 axis in immune evasion (Harun et al., 2013). However, to date, tissue-level association of PD-L1 with FIPV has not been reported, highlighting a gap in understanding of checkpoint regulation and FIP. In cats with FIP, increased levels of IFN-γ have been reported in the tissues, including necrotizing/fibrinous and granulomatous lesions (Berg et al., 2005; Dean et al., 2003; Gunn-Moore et al., 1998; Kiss et al., 2004). Therefore, in this study, we evaluated the expression of PD-L1 in various tissues from four cats with naturally occurring FIP (Fig. 4). The tissue samples used in this study were obtained from the Kansas State Veterinary Diagnostic Lab, and they were recent cases from the past 1–2 years. These cats were infected with serotype I FIPV, as determined by the S-protein N-terminal region amplified from their tissues, and the phylogenetic analyses show that the viruses from Cats A–C are overall closely related.

When the RNA levels of PD-L1 and FCoV were evaluated, some tissues and animals showed a correlation between PD-L1 and FCoV levels, while others did not (Fig. 4). The co-localization results showed that while a high percentage (40–97%) of FCoV-positive cells were also positive for PD-L1, only about 10% PD-L1-positive cells co-localized with FCoV antigen (Figs. 5–7). However, overall viral load did not correlate well with PD-L1 expression by RT-qPCR or confocal analysis, despite the presence of cells co-expressing both markers (Fig. 5E). The findings indicate that the relationship between PD-L1 expression and viral load is variable across tissues and animals. The reason for this variable correlation remains uncertain, although differences in the timing of PD-L1 induction following viral infection, disease stage or local inflammation levels may contribute to this variability. We previously demonstrated that both FIPV infection as well as feline IFN-γ are potent inducers for PD-L1 in both CRFK and Fcwf-4 cells (Zabiegala et al., 2025), and inflammatory cytokines are likely abundant in tissues with necrosis, fibrosis, and granulomatous inflammation (Figs. 1 and 2). Taken together, these findings suggest that PD-L1 up-regulation during FIP reflects both direct viral infection and cytokine-mediated responses within inflamed microenvironments. The relative influence of these factors appears to vary across tissues and among individual cats, consistent with the heterogeneous RNA expression patterns, lesion severity, and confocal co-localization profiles observed. There is strong evidence that T cells are largely responsible for the control of FCoV infection and the prevention of FIP (de Groot-Mijnes et al., 2002; Paltrinieri et al., 2003). Thus, increased PD-L1 expression would be most advantageous to FIPV during the early stages of tropism change and systemic spread and establishment of FIP infection. As FIP progresses, however, lymphopenia often arises, rendering PD-L1 expression largely less efficacious in suppressing T-cell responses. In end-stage FIP, as examined here, the relationship of FIPV and the PD-1/PD-L1 axis may be less apparent, making it difficult to discern how much PD-L1 contributes to immune dysregulation at this late stage. This highlights the importance of investigating earlier phases of infection, when T-cell populations are more intact and PD-L1-driven immunomodulation may play a more active role in shaping disease progression. Additionally, understanding the role of PD-1/PD-L1 axis in cats that receive antiviral treatment and subsequently relapse will be critical, as effective clearance of virus-infected macrophages may depend on overcoming PD-L1–mediated immune suppression. The main limitations of the study include a small number of cases tested and lack of control of non-FIP cats. Without the controls, it is not possible to determine whether PD-L1 expression is specific to FIP or reflects generalized inflammation. Currently we are continuing the studies on PD-L1 and interferons in FIP development with additional FIP as well as non-FIP cases.

Conclusion

The results support a tissue-level association between PD-L1 expression and FCoV infection in some tissues from cats with FIP with substantial variation among tissues and animals, which warrant for further analysis for the role of elevated PD-L1 levels by FCoV in the development of the FIP clinical disease.

• PD-L1 expression was assessed in tissues from four cats with naturally occurring FIP.

• Co-localization of FCoV and PD-L1 was confirmed by confocal microscopy.

• Virus infection and inflammation play roles in PD-L1 induction and FIP development.

Availability of data and material

The data that support the findings of this study are available on request from the corresponding author (KOC).