Development of monoclonal antibodies specific to Marek disease virus-EcoRI-Q (Meq) for the immunohistochemical diagnosis of Marek disease using formalin-fixed, paraffin-embedded samples

Abstract

Marek disease (MD) is a viral disease characterized by the development of lymphoma in poultry. Although morphologic confirmation of lymphoma is used to diagnose MD, immunohistochemical detection of MD virus-EcoRI-Q (Meq), which is a viral protein that is expressed exclusively in MD tumor cells, would further improve the accuracy of diagnosis. We developed monoclonal antibodies (mAbs) that specifically detect Meq by immunohistochemistry (IHC) using formalin-fixed, paraffin-embedded (FFPE) sections. We evaluated the sensitivity and specificity of 14 mAbs that we produced, using FFPE samples of MDCC-MSB1 cells, MD tumor tissues, and tissues of uninfected chickens. Four different antigen retrieval conditions were investigated. Thirteen mAbs reacted with Meq in FFPE sections, but immunohistochemical reactivity and specificity varied depending on the mAb and antigen retrieval condition; heat-induced antigen retrieval (HIAR) was more effective at detecting Meq than the other tested conditions. HIAR pH 9 tended to increase immunoreactivity and decrease specificity. Of the 5 mAbs that immunoreacted strongly with Meq without nonspecific reactions under the optimal antigen retrieval conditions, 3 mAbs (1C1-121, 3A3-112, 5F7-82) did not produce background staining of tumor or non-tumor tissues; 2 mAbs (2C5-11, 4A5-54) produced background staining. The mAb 6B5-128 reacted moderately with Meq without nonspecific reactions and background staining. The remaining mAbs showed weak immunoreactivity or problematic nonspecific reactions. Our results suggest that some of our developed mAbs can be used in IHC to detect Meq in FFPE sections with high specificity, and that the use of IHC may greatly improve the diagnosis of MD.

Marek disease (MD), an infectious disease of poultry caused by the oncogenic Gallid alphaherpesvirus 2 (MD virus; MDV), is characterized by the formation of lymphoma.^6,20 Outbreaks of MD have been reported in various parts of the world, causing serious economic damage to the poultry industry. ²¹ Although the number of MD outbreaks has been reduced drastically by the widespread use of effective vaccines and the development of resistant chicken breeds, MD has not been eradicated.^{3,8,17,25,28,32}

Histologic examination and real-time PCR to quantify MDV genome copies are described as reliable methods for the definitive diagnosis of MD.^11,20,30 Given that most chicken flocks are exposed and persistently infected with MDV irrespective of lymphoma formation, testing including virus isolation, gene detection by conventional PCR, and serologic detection of antiviral antibodies has little value in the diagnosis of MD.^9,30 The advantage of histologic examination is the direct confirmation of lymphoma. Nevertheless, using histopathology alone is not always effective because it is difficult to differentiate MDV-induced lymphomas from other lymphoid tumors such as lymphoid leukosis, caused by avian leukosis virus (ALV), reticuloendotheliosis, caused by reticuloendotheliosis virus (REV), and possibly spontaneous lymphomas without viral etiology. Given that MD tumor cells are generally CD4+ T-cells, and that ALV or REV-induced lymphomas are usually of B-cell origin,^23,29 the detection of T-cell–specific antigens in tumor cells is helpful in the diagnosis of MD. However, this approach only shows the immunophenotype of tumor cells, but cannot demonstrate that tumor cells have been transformed by MDV. ³⁰ Therefore, the detection of an antigen exclusively expressed on MD tumor cells is considered to be a more effective approach for diagnosing MD.

MDV-EcoRI-Q (Meq) is a viral protein, specifically and highly expressed in the nuclei of tumor cells and tumor cell lines of MD.^13,18,22,26 Meq is encoded by the MDV-unique meq gene and is a basic leucine zipper protein, similar to oncoproteins known as the Fos/Jun family.^13,18 If the expression of Meq in lymphoma cells could be detected by immunohistochemistry (IHC), we could show a direct relationship between MDV infection and lymphomagenesis. The demonstration of Meq expression is ideal for the definitive diagnosis of MD and its differentiation from non-MD lymphomas. ¹¹

Several anti-Meq antibodies have been generated to detect Meq expression in MD tumor cells.^4,5,15,16 However, the previously produced anti-Meq antibodies have some limitations in their usability.^{1,4,5,10,14,27,29} We therefore aimed to generate an anti-Meq monoclonal antibody (mAb) that could detect Meq in MD tumor cells in formalin-fixed, paraffin-embedded (FFPE) sections by IHC with high specificity.

Materials and methods

Production of recombinant Meq

Recombinant Meq (rMeq), which contains 6 histidine (His) residues at its N-terminal, was prepared using the Bac-to-Bac baculovirus recombinant protein expression system (Thermo Fisher). The meq gene sequence (1,020 bp) of Md5, a very virulent MDV reference strain, was obtained from the National Center for Biotechnology Information database (GenBank AF243438.1). Codons of the meq gene were optimized using GeneOptimizer software (Thermo Fisher) to increase the efficiency of protein expression in Spodoptera frugiperda (Sf) cells without altering the amino acid sequence of the meq gene. The gene synthesized after codon optimization and incorporation of the His tag was cloned into a pFastbac1 vector. This vector was transfected into DH10Bac Escherichia coli competent cells to produce a recombinant bacmid. The recombinant baculovirus was generated by transfecting Sf9 cells with recombinant bacmid DNA.

rMeq was prepared from baculovirus-infected Sf21 cells 72 h after infection. The protein was purified from the cellular pellet by immobilized-metal affinity chromatography (Ni Sepharose 6 fast flow carrier, XK columns; GE Healthcare). The expression of rMeq was checked by western blotting (anti-6×-His-tagged mAb, clone 3D5; Thermo Fisher). The dispensed rMeq (0.3 mg/mL) was stored at −80°C until use.

Production of mouse mAbs against rMeq

Two 7-wk-old BALB/c mice (Japan SLC) were immunized intraperitoneally with rMeq antigen (10 μg/time) mixed with aluminum hydroxide–based adjuvant (2 mg in 100 μL/time) 3 times at 2-wk intervals. The final immunization was performed 3 d before cell fusion. After that, the mice were euthanized by the intraperitoneal injection of pentobarbital sodium (100 mg/kg body weight), and the spleens and serum were collected. Minced spleens were filtered through a 40-μm mesh (Thermo Fisher) and were fused with mouse myeloma P3U1 cells using dimethyl sulfoxide Hybri-Max reagent (MilliporeSigma).

The hybridomas secreting anti-rMeq IgG mAbs were screened by indirect ELISA as described below, and were cloned twice by single-cell sorting. The cloned mAbs were subtyped (rapid mouse antibody isotyping kit; Thermo Fisher).

Screening ELISA for detection of anti-rMeq mAb–producing hybridoma

We developed an indirect ELISA to screen hybridomas secreting mAbs that react with rMeq antigen. The rMeq (0.3 mg/mL) was mixed with the same amount of 0.1% Triton X-100 solution (Nacalai Tesque) and then applied to 96-well ELISA plates; 20% Block Ace solution (DS Pharma Biomedical) was used to reduce nonspecific reactions. The supernatant collected from wells of hybridomas was incubated on plates for 1 h at 37°C. Serum from a mouse immunized with rMeq was used as a positive control; the negative control was the serum from a non-immunized mouse. After washing, a secondary antibody, horseradish peroxidase–labeled rabbit anti-mouse IgG (gamma chain–specific; Zymed Laboratories), was applied for 1 h at 37°C. 2,2′-Azino-di-(3-ethylbenzthiazoline sulfonic acid) (MilliporeSigma) was used as a substrate for horseradish peroxidase. The enzymatic reaction was spectrophotometrically measured at 405 nm using a microplate reader (Multiskan FC basic; Thermo Fisher).

Preparation of FFPE sections

MDV-transformed cell line (MDCC-MSB1 [MSB1]) cells, ² MD tumor tissues, and normal tissues from MDV-uninfected chickens were prepared to evaluate the immunoreactivity of each mAb against Meq antigen in IHC. FFPE MD tumor tissues were derived from 3 specific pathogen–free (SPF) chickens (Nisseiken) that were infected experimentally with the MS1 strain of MDV, which is a very virulent MDV isolated in Japan. ¹² The chickens were inoculated intramuscularly with the virus at 2-d-old and euthanized at 57 d post-infection. Tumor tissues were collected at autopsy. Brain, spinal cord, heart, lung, proventriculus, gizzard, duodenum, cecum, pancreas, liver, kidney, thymus, spleen, bursa of Fabricius, and skin were collected from 3 uninfected SPF chickens as normal controls. Cultured MSB1 cell pellets after centrifugation at low speed, tumors, and normal tissues of uninfected chickens were fixed in 10% neutral-buffered formalin (Fujifilm Wako Pure Chemical) for 48 h and embedded in paraffin. The paraffin blocks were prepared in 2020 for MSB1 cells, in 2013 for MD tumor tissues, in 2018 for normal tissues of uninfected chickens.

IHC protocol

IHC was performed in 2020–2021 using newly sectioned histology slides. The deparaffinized sections were incubated in 0.3% H₂O₂ in methanol for 20 min at room temperature (RT) to block endogenous peroxidase activity. We tested 4 different antigen retrieval conditions in IHC: no treatment, heat-induced antigen retrieval (HIAR) with citrate buffer (pH 6) or EDTA buffer (pH 9), and enzyme-induced antigen retrieval (EIAR). HIAR was performed by incubating the sections in the buffer for 15 min in a microwave oven (500 W). For EIAR, the sections were treated with 0.1% actinase E (Kaken Pharmaceutical) for 10 min at 35°C. Then, 5% skim milk solution (Fujifilm Wako Pure Chemical) was applied for 20 min at RT to reduce nonspecific antibody binding. Supernatant fluid of cloned hybridoma cell culture containing anti-rMeq mAb was applied as the primary antibody. The sections were incubated with each anti-rMeq mAb diluted 1:3 in PBS for 20–24 h at 4°C. After rinsing with PBS, the sections were treated with a horseradish peroxidase polymer–based secondary antibody reagent (Histofine simple stain MAX PO [M] kit; Nichirei Bioscience), for 45 min at RT. The antigen–antibody reaction was visualized (DAB substrate kit; Nichirei Bioscience) under a microscope. The sections were counterstained with Mayer hematoxylin (Muto Pure Chemicals) and coverslipped after dehydration.

Evaluation by IHC of reactivity of anti-rMeq mAbs with Meq antigen on FFPE sections

The sensitivity and specificity of each mAb were assessed for MSB1 cells, MD tumors, and normal tissues of uninfected chickens. The lungs were selected as MD tumor tissues because the tumor cells were commonly found in all 3 birds and were well fixed by formalin. We defined specific immunoreactivity as a positive IHC reaction by the nuclei of tumor cells.^10,15,16 Reactivities of certain non-tumor cells were regarded as nonspecific reactions. Background staining was determined as staining of the entire section regardless of cell type. The intensity of immunostaining was scored into 4 categories: negative (–), weak (+), moderate (++), and strong (+++; Fig. 1). Evaluation was performed by observing the entire tissues or cells on the FFPE sections of the samples. For MSB1 cells, one specimen was evaluated for each mAb. For tumor or normal tissues, we averaged the scores of specimens from 3 birds of each.

Figure 1.

Four scoring categories of staining intensity of immunohistochemistry. A. – = negative. No staining of any tumor cells. B. + = weak. A small number of tumor cells are stained weakly. C. ++ = moderate. Many tumor cells are stained moderately. D. +++ = strong. Most tumor cells are stained strongly. Bars = 20 μm.

All experimental procedures involving mice and chickens were approved by the Ethics Committee of the National Institute of Animal Health, Japan (approvals 14-008, 19-070, 19-090).

Results

Establishment of hybridomas

The Codon Adaptation Index of the meq gene was improved from 0.75 to 0.95 by codon optimization. The successful production of rMeq was confirmed by western blotting using anti-6×-His-tagged mAb. Hybridomas were generated by the fusion of spleen cells from BALB/c mice immunized with rMeq with P3U1 myeloma cells. Fourteen clones of anti-rMeq IgG mAb–secreting hybridomas that were positive in ELISA using rMeq were established (Table 1). The growth of 1C1-121 was slower than that of the other hybridomas. The isotypes of mAbs were IgG2b only for 4A5-54, and IgG1 for the other 13 mAbs.

Table 1.

Results of immunohistochemical Marek disease virus-EcoRI-Q (Meq) detection in formalin-fixed, paraffin-embedded sections of MDCC-MSB1 (MSB1) cells, tumor tissues from chickens experimentally infected with Marek disease virus, and uninfected chickens using 14 anti-recombinant Meq monoclonal antibodies (mAbs) with different antigen retrieval conditions.

mAb	Isotype	Antigen retrieval condition
		No treatment				AE				HIAR pH 6				HIAR pH 9
		Meq detection				Meq detection				Meq detection		NS	Background	Meq detection		NS	Background
		MSB1	MD tumor	NS	Background	MSB1	MD tumor	NS	Background	MSB1	MD tumor			MSB1	MD tumor
1C1-121	IgG1	++	+	−	−	++	+++	−	−	+++	+++	−	−	+++	+++	−	−
1C2-3	IgG1	−	−	−	−	−	−	−	−	++	−	−	−	++	−	−	++
1C5-9	IgG1	−	−	−	−	−	−	−	−	+	+	−	−	+	+	−	−
1G6-62	IgG1	−	−	−	−	−	−	−	−	++	++	N	+	++	+++	N	++
2C5-11	IgG1	−	−	−	−	+	++	−	−	++	+++	−	+	++	+++	−	+++
3A3-112	IgG1	++	++	−	−	++	+++	−	−	+++	+++	−	−	+++	+++	−	+
4A5-54	IgG2b	+	−	−	−	+	++	−	−	++	+++	−	+	+++	+++	−	++
5C5-83	IgG1	++	+++	N	+++	+++	++	N	+++	++	+++	N	+++	+++	++	N	+++
5F7-82	IgG1	+	−	−	−	−	+	−	−	++	+++	−	−	+	+++	−	+
6B5-128	IgG1	+	−	−	−	++	−	−	−	++	+	−	−	+++	++	−	−
6H9-101	IgG1	++	++	N, G	+	++	+++	N, G	+	++	++	N, G	++	+++	+++	N, G	+++
9E5-11	IgG1	−	−	−	−	−	−	−	−	+	++	−	−	+	++	N	+
10D8-10	IgG1	++	++	−	−	+	++	N	+	++	+++	G	−	+++	+++	N, G	++
10G1-141	IgG1	++	++	−	+	+++	+++	N	++	+++	++	N	+++	+++	++	N	+++

AE = actinase E; Background = background staining; G = cytoplasm of globule leukocytes; HIAR = heat-induced antigen retrieval; MD tumor = tumor tissues of lungs from chickens experimentally infected with Marek disease virus; Meq = Marek disease virus-EcoRI-Q; MSB1 = MDCC-MSB1 cells; N = nuclei of non-tumor cells in systemic tissues of SPF chickens; NS = nonspecific reactions in tissues of uninfected SPF chickens. Scoring of immunostaining: – = negative; + = weak; ++ = moderate; +++ = strong.

Immunohistochemical reactivity of anti-rMeq mAbs against FFPE MSB1 cells

In IHC using FFPE MSB1 cells, all 14 established clones of anti-rMeq mAbs detected Meq expressed in the nuclei of MSB1 cells under HIAR conditions (Fig. 2A). Five mAbs (4A5-54, 5C5-83, 6B5-128, 6H9-101, 10D8-10) showed stronger staining intensity in HIAR pH 9 than in HIAR pH 6. On the other hand, many mAbs under untreated or EIAR conditions in IHC produced weaker staining than those with HIAR (pH 6 and pH 9; Table 1). Four mAbs (1C2-121, 1C5-9, 1G6-62, 9E5-11) failed to detect Meq in FFPE cells in IHC under untreated and EIAR conditions.

Figure 2.

Formalin-fixed, paraffin-embedded MDCC-MSB1 (MSB1) cells and tumor tissue of a lung from a Marek disease virus (MDV)-infected chicken. A. Immunohistochemistry (IHC) staining of MSB1 cells using anti-recombinant MDV-EcoRI-Q (rMeq) monoclonal antibody (mAb; 1C1-121) under heat-induced antigen retrieval (HIAR) pH 9 condition. The nucleus of each cell is labeled with the anti-rMeq mAb. Staining intensity is +++. Bar = 20 μm. B. Many MD tumor cells densely infiltrate pulmonary interstitial tissue and are scattered in capillary beds. H&E. Bar = 100 μm. C. Serial section of lung in image B. Strong IHC staining (+++) using anti-rMeq mAb (1C1-121) under HIAR pH 9 condition. Nuclei of most of the tumor cells are strongly positive for Meq. Bar = 100 μm. Inset: higher magnification of positive cells in the interstitial tissue. Bar = 20 μm. D. Serial section of lung in image B. Moderate IHC staining (++) using anti-rMeq mAb (6B5-128) under HIAR pH 9 condition. Staining intensity is weaker than in image C. Bar = 100 μm. Inset: higher magnification of positive cells in the interstitial tissue. Bar = 20 μm.

Immunohistochemical reactivity of anti-rMeq mAbs against FFPE tumor cells from chickens experimentally infected with MDV

With the exception of 1C2-3, 13 anti-rMeq mAbs immunohistochemically detected Meq antigen expressed in tumor cells of FFPE tissues from experimental cases of MD, depending on the antigen retrieval conditions. Similar to the IHC results of MSB1 cells, positive reactions with Meq antigen were observed in the nuclei of tumor cells.

Comparing the results of the 4 antigen retrieval conditions, the staining intensity of IHC tended to be higher in the order of HIAR pH9, HIAR pH6, EIAR, and no treatment (Table 1). 1C1-121, 2C5-11, 3A3-112, 4A5-54, 5C5-83, 5F7-82, 6H9-101, and 10D8-10 reacted strongly with Meq under HIAR conditions (Table 1; Fig. 2B, 2C). Of these, 1C1-121 and 3A3-112 were equally sensitive to Meq antigen by HIAR and EIAR. 10G1-141 showed a strong positive result under EIAR condition. 6B5-128 and 9E5-11 were moderately immunoreactive under HIAR, which was the best antigen retrieval condition for 2 mAbs (Fig. 2D). Given that the staining intensity of 1C5-9 at 1:3 dilution was consistently weak in all tested antigen retrieval methods in IHC, this mAb was judged as being unsuitable for the immunohistochemical detection of Meq antigen in FFPE samples.

The staining intensity of IHC varied from cell to cell (Fig. 2C, 2D). The nuclei of relatively large tumor cells tended to be stained strongly by most mAbs, for an unknown reason. IHC with 10D8-10 produced a strong positive reaction in mitotic chromosomes in addition to large tumor cells.

Specificity of anti-rMeq mAbs in IHC

Eight mAbs (1C1-121, 1C2-3, 1C5-9, 2C5-11, 3A3-112, 4A5-54, 5F7-82, 6B5-128) did not produce nonspecific reactions; 5 of these mAbs (1C2-3, 2C5-11, 3A3-112, 4A5-54, 5F7-82) produced background staining. IHC with 6 mAbs (1G6-62, 5C5-83, 6H9-101, 9E5-11, 10D8-10, 10G1-141) was associated with nonspecific reactions in the nuclei of cells from all tested tissues, as well as background staining. 6H9-101 and 10D8-10 also nonspecifically reacted with the cytoplasm of normal globule leukocytes, which are distributed mainly in the intestines.

Discussion

Of the 14 hybridomas secreting anti-rMeq IgG mAbs that we established, 6 anti-rMeq mAbs (1C1-121, 2C5-11, 3A3-112, 4A5-54, 5F7-82, 6B5-128) were particularly suitable for detecting Meq in FFPE sections by IHC given their immunoreactivity to Meq and specificity in not reacting with non-tumor cells of chickens. These 6 mAbs detected Meq expressed in both MSB1 cells and MD tumor cells without nonspecific reactions to normal tissues of uninfected chicken under certain antigen retrieval conditions, and would be promising for the definitive diagnosis of MD. On the other hand, 8 mAbs (1C2-3, 1C5-9, 1G6-62, 5C5-83, 6H9-101, 9E5-11, 10D8-10, 10G1-141) were unsuitable for diagnosing MD by IHC using FFPE samples. The IHC intensity, specificity, and suitable antigen retrieval conditions differed among the mAbs; HIAR was an appropriate condition for most mAbs.

Several antibodies have been produced to detect Meq by IHC.^5,15,16,27 To our knowledge, however, there are no previous reports of the adequate detection of Meq in FFPE sections using anti-Meq antibodies. For example, the use of anti-Meq mAb (clone 23B46) in IHC required a frozen tissue sample and a tyramide signal amplification system that is more sensitive and expensive than the polymer-based detection method.^10,11,14 Anti-Meq mAb (clone FD7) was used to detect Meq expressed only in paraformaldehyde-fixed cultured cells by an immunofluorescence assay.^5,31 Anti-Meq mAb (clone Lamba7) detected Meq only on frozen sections by an immunofluorescence assay. ⁴ Meanwhile, some related studies used monoclonal or polyclonal antibodies to detect Meq in FFPE sections, but these antibodies also reacted with cytoplasm of normal cells.^1,16,27 For instance, an anti-Meq mAb reacted immunohistochemically not only with MD tumor cells but also with the cytoplasm of proventricular gland cells. ²⁷ IHC using an anti-Meq polyclonal antibody showed occasional IHC reactions in the cytoplasm of ganglion cells in ALV-infected birds.^1,16 When IHC with anti-Meq antibodies is to be used for the definitive diagnosis of MD, antibodies that bind nonspecifically to the normal tissues of chickens may be inappropriate. In this regard, anti-rMeq mAbs produced in our study are promising in that they can detect Meq on FFPE sections without any nonspecific reactions to non-tumor cells of uninfected controls, by using the widely used and cost-effective polymer-based IHC method. We suggest the following steps for an accurate and definitive diagnosis of MD using anti-rMeq mAbs: first, check the presence of lymphoma tissue histologically, and then verify the expression of Meq and T-cell antigen in tumor cells by IHC.

Some limitations exist with our IHC to detect Meq. We did not examine the IHC response of Meq from more than one MDV strain. Slight diversity in the meq gene has been reported, such as insertion or deletion in the C-terminal proline-rich region, or point mutations.^7,19,24 It remains unknown whether the mAbs generated in our work can detect Meq expressed in MD tumor cells regardless of the variants of the meq gene or viral strains. Antibody cocktails containing ≥2 anti-rMeq mAbs in IHC may be one solution to increase the sensitivity of Meq detection in MD tumors. Further detailed analysis is required for the use and validation of our anti-Meq antibodies in field cases of MD from various chicken breeds. Another limitation includes the lack of evaluation using non-MD lymphomas such as ALV- or REV-induced lymphomas. Other factors that can influence the IHC for detecting Meq are fixatives, fixation time of the tissues, and variation of the IHC protocol, namely, the antigen retrieval method, dilution of mAb, or visualization method. These factors need to be evaluated in the future.