key: cord-0766019-ajfarayh
authors: Dörfelt, Stefanie; Matiasek, Lara A.; Felten, Sandra; Sangl, Laura; Hartmann, Katrin; Matiasek, Kaspar
title: Antigens under cover—The preservation and demasking of selected antigens for successful poststaining immunocytochemistry of effusion, brain smears, and lymph node aspirates
date: 2019-02-25
journal: Vet Clin Pathol
DOI: 10.1111/vcp.12702
sha: 2ab7c15271a9a80bd7ac4617b211afbf832e6a1a
doc_id: 766019
cord_uid: ajfarayh

BACKGROUND: In clinical cytology, the applicability of an ancillary test such as immunocytochemistry is too often limited by low sample volume, poor cell representation, and sample preservation. Diagnosticians often read Romanowsky‐stained cytology, although specific techniques such as immunocytochemistry are often essential for a definitive diagnosis. OBJECTIVES: The goal of the present study aimed to investigate if immunocytochemistry on previously‐stained cytologic specimens was possible. Different pretreatments were examined to determine which treatment preserved antigenicity best. METHODS: One hundred and twenty‐two impression smears and 64 fine‐needle aspirate preparations of brain and lymph nodes were processed and evaluated microscopically. The impact of staining cytologic preparations with a modified Wright's stain, using a destaining method, performing a coverslipping and decoverslipping process, and subjecting smears to a microwave treatment (MWT) were examined for the immunolabeling of selected nuclear, cytoplasmic, and plasmalemmal antigens, as well as intracellular feline coronavirus (FCoV). Biotinylated secondary antibodies were used, and the bound primary antibody was visualized using an ABC amplification kit. RESULTS: Cellular antigens were reliably detected with immunocytochemistry after smears were stained with a Romansky stain and were coverslipped early after staining and stayed coverslipped until immediately before immunolabeling. The staining intensity reached the same levels as that of the controls if the films underwent MWT in citrate buffer. In contrast, FCoV antigen detection was abolished after any physicochemical interference. CONCLUSIONS: Poststaining immunocytochemistry represents a practical tool for additional investigations on prestained cytologic specimens when searching for cellular antigens. Paired untreated samples should be kept in case the workup requires testing for more vulnerable viral antigens.

Investigation of cytologic specimens obtained by fine-needle aspiration (FNA) is the least invasive approach to obtain a diagnosis of disease. [1] [2] [3] Cytomorphology, however, can provide only limited insights into the biology of a neoplastic disorder or the etiology of an infectious process. It may be necessary to combine cytomorphology with specific disease markers to elucidate the underlying disease and collect predictive data. 2 In contrast to histologic investigations, which allow for multiple tissue sections that can examine individual cells, repetitive impression smears and FNA of the same tissue can by no means reproduce the same composition of cells and tissue components. 2 In addition to these technologic limitations, the type and distribution of a lesion can pose essential limitations to its holistic cytologic characterization. Thus, detectability is compromised if the lesion is (a) poorly exfoliative and of low cellularity; (b) affecting tissues that are difficult to access (eg, the central nervous system (c) has atypical cells and pathogens that are poorly dense or scantly distributed; and (d) has cells that are highly vulnerable to shearing stress.

To detect changes in samples with low cellularity, multiple investigations should be carried out on the same slide rather than single investigations on multiple slides with variable cell numbers.

The simplest secondary procedures use conventional stains to highlight microorganisms, chemical compounds, matrix components, and subcellular structures with histochemistry after a destaining procedure. With the advancement of panoptic stains, special stains have mainly focused on the assessment of microbial and fungal organisms (eg, Ziehl-Neelsen, Fite-Faraco, Gram, Fontana-Masson), mucins and polysaccharides (eg, Alcian blue, Periodic acid-Schiff), iron pigments, and melanin (eg, Prussian blue reaction, Nile blue A, Fontana-Masson). 2 As cellular identification is rarely needed in these situations, ancillary staining often can be conducted on spare slides. 2 Other infectious diseases and tumor cell antigens, however, could require preselection and cytomorphologic characterization of the cells to be stained. 2, 3 This holds true, in particular, for advanced tumor diagnoses. 3, 4 Immunocytochemistry (ICC) on Romansky-stained (prestained) slides (poststaining ICC, PSICC) could help identify the histogenetic origin of tumors or cell types, characterize cells that cannot be identified with conventional stains, and obtain further prognostic and predictive information. 3, [5] [6] [7] [8] For infectious diseases, PSICC is a promising tool for the intracellular detection of viral proteins. One of the most common situations is in the detection of feline coronavirus (FCoV) for the diagnosis of feline infectious peritonitis (FIP), in which the presence of intracellular viral antigens in macrophages must be demonstrated. 9, 10 This approach requires preservation of both the target antigen and host cell morphology.

Historically, PSICC methodology was tailored to individual cases. Therefore, systematic studies on the reproducibility, efficacy, and reliability of immunostaining are rare. 1, [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] Hence, the present study aimed to fill this gap by establishing reproducible guidelines for PSICC labeling of selected cellular and viral antigens in tissue imprints, FNA, and cytocentrifuged fluids with the focus specifically on antigen preservation and retrieval. Target antigen selection was driven by our daily neurocytologic specimen caseload various mammalian species with possible inflammatory and infectious CNS diseases.

This study investigated (a) the ability to apply ICC to prestained cytologic specimens, (b) whether coverslipping could be used to preserve antigenicity, and (c) whether microwave treatment (MWT) enhanced immunodetection in prestained smears. Therefore, ICC staining quality was evaluated on smears after prestaining with a modified Wright's stain (modWS), destaining with hydrochloric acid (HCl), and then comparing the now unstained samples with and without coverslipping and MWT. The effects were evaluated for a set of cellular antigens with a distinct subcellular distribution (study A) and intracellular FCoV proteins (study B 

We tested the impact of individual pretreatments on the immunostaining of nuclear, cytoplasmic, and membrane antigens. Therefore, this systematic evaluation employed distinct methodologic trials, as illustrated in Figure 1 and Table 1. A series of 122 impression smears and 64 FNA preparations were taken from the brain and lymph nodes of a cat and a pig during postmortem examination (<1/2 hour after euthanasia) for reasons unrelated to the study's purpose. CNS cells were obtained from the postcruciate gyrus of the brain after extensive craniectomy, encephalectomy, and trimming, as described elsewhere. 25, 26 Lymphocyte-enriched samples were collected from the superficial cervical and para-aortic lymph nodes via FNA using a 22-gauge syringe. All preparations were air-dried immediately. 3) plus the additive effects of additional coverslipping-decoverslipping cycles (SDS-CDC, 36 slides) (trial 4) were investigated with another 60 slides. These trials were compared to the 36 CDC-treated slides (trial 5) and to a series of 24 slides not subjected to any treatment, which served as the gold standard (trial 6) ( Table 1 ). All samples were stored for up to 3 months.

All slides except those from trial 6 were subjected to the laboratory treatments mentioned above, which preceded ICC. Coverslipped samples from trials 2, 4, and 5 were immersed in xylene to achieve liquefaction of the mounting medium until the coverslips slipped off.

Thereafter, the slides were put through a graded ethanol series (2 × 100%, 1 × 96%, 1 × 70%; 5 minutes each) and immersed in distilled water.

Prestained and decoverslipped slides (trials 2 and 4) and those left uncovered after the modWS method (trial 3) were immersed in 1% HCl until the color had completely faded as identified with microscopy. Freshly destained slides (trials 2, 3, and 4) and nonstained decoverslipped slides from trial 5 were immersed in distilled water until ICC was performed ( Figure 1 , Table 1 ).

Before performing ICC, one set of slides from each trial (trials 2, 3, and 6: n = 12 each; trials 4 and 5: n = 18 each; Figure 1 and Table 1 ) was subjected to the MWT-based antigen retrieval. For this F I G U R E 1 Schematic illustration of the staining trials. Trial 1: Slides were stained and coverslipped. These slides were used for cytomorphologic evaluations. Trial 2: The procedure included staining, coverslipping, and microwave treatments of the slides. After storage and directly before immunocytochemistry, these specimens were decoverslipped and destained, and half of the slides were microwave-treated again. Trial 3: Slides were only stained and stored before immunocytochemistry, at which time the slides were destained, and half were microwave treated. Trial 4: Slides were stained and coverslipped. After storage and directly before immunocytochemistry, they were decoverslipped and destained, and half of them were microwave treated. Trial 5: This procedure included only the coverslipping of the slides. Before immunocytochemistry, the slides were decoverslipped, and half were microwave treated. Trial 6: No pretreatment was carried out with these specimens, and they were stored frozen. Only half of the slides were pretreated in the microwave before immunocytochemistry. These unstained and untreated slides served as the gold standard for the study procedure, the slides were transferred to a cuvette containing 0.01 mol/L of a citrate buffer solution (pH 6). Based on our immunohistochemical protocol, MWT was carried out for 5 minutes/800 W followed by 20 minutes/250 W. The slides were left in citrate buffer for 30 minutes to cool down to room temperature. All further steps except for the primary antibody (pAB) incubations were carried out at room temperature.

The MWT slides and those kept in distilled water (destained and unstained slides without MWT, Table 1 Upon removal of the nonimmune serum, lymph node FNAs were incubated with an anti-CD3 antibody for detection of the corresponding T cell antigen. Other markers, specifically, NeuN and GFAP, were used on separate brain smears ( Table 2) . Each run contained site-matched slides in which a pAB was replaced by an antibody diluent, serving as a negative control (altogether n = 12). Incubation was carried out for 18 hours at 4°C (39.2°F).

After repeated wash steps with PBS, the slides were mounted with biotinylated goat anti-rabbit/mouse IgG antibodies (both DAKO, Glostrup, Denmark) for 1 hour. Bound pAB was subsequently visualized using an ABC amplification kit (Vectastain; Vector Laboratory Inc, Burlingame, CA, USA) using diaminobenzidine tetrahydrochloride (DAB) as the chromagen. The enzyme reaction was blocked with PBS rinses. Then, the slides were counterstained with Mayer's hematoxylin (AppliChem GmbH, Darmstadt, Germany), underwent an ascending ethanol series, and were coverslipped as described above.

Cell yields and preservation were assessed with bright field microscopy by two different clinical pathologists. In separate sessions, these observers, who were blinded to the smear origins, the pretreatment types, and the pAB type, evaluated the ICC outcomes.

ICC quality determinants were (a) cell type specificity, (b) intensity of cellular staining, (c) the presence/absence of acellular background activity, and (d) nonspecific staining. These parameters were semiquantitatively scored by using a 4-point system (0-3). In the event of intraobserver disagreement, the slides were reviewed on a T A B L E 1 Different pretreatments applied before performing the immunocytochemistry trials (Trials 1-6) Germany), and mean values were newly assigned.

For statistical evaluations, the final scores were compared between the different procedures using nonparametric statistical algorithms. A P of ≤ 0.05 indicated statistical significance.

This study evaluated the impact of pretreatment on FCoV antigen immunolabeling.

Study B was conducted on pleural (n = 6) and abdominal (n = 16) effusions of FIP cases confirmed with histopathology and FCoV antigen-positive immunohistochemistry. 9 These effusions were collected using conventional thoracentesis or abdominocentesis procedures.

Harvested fluids were mounted onto uncoated standard slides (Langenbrinck, Emmendingen, Germany) using a cytocentrifuge (Hettich Universal 16, Adelsried, Germany). According to our in-house protocol for effusions, centrifugation chambers were spun for 5 minutes at 250g, after which the supernatant was removed, and the coated slides were centrifuged for 1 minute at 416g. Then, the slides were air-dried and stored at −20°C (−4°F) until further processing.

To evaluate the preservation and antigenicity of viral antigens, the samples underwent trials 3, 4, 5, and 6 (with and without MWT), as described in study A. ICC was performed using a mouse monoclonal antibody directed against the coronavirus nucleocapsid (clone FIPV3-70; Table 2 ). Incubation and subsequent steps were identical to those applied for NeuN labeling (see above) using the ABC enhancer and DAB.

The ICC signal was re-evaluated by both observers concerning cellular and subcellular localization, intensity, specificity, and background, as described above.

To evaluate cell preservation, the slides were microscopically assessed. ModWS-stained slides revealed a representative cell yield, with fewer than 5% of cells suffering crush artifact and fewer than F I G U R E 2 The outcome of immunocytochemical staining for cellular antigens (blue arrows: specific positive staining; black arrows: immunonegative cells). Counterstaining with Mayer's hematoxylin. A, Immunocytochemistry on unstained and untreated films served as the gold standard, and which illustrates optimal staining results. The CD3 signal was strongest in the periphery of the lymphocyte cytoplasm close to the membrane. NeuN immunoreactivity was restricted to neuronal nuclei, while GFAP diffusely stained the soma and fragmented processes of astrocytes (procedure depicted in trial 6 without microwave treatment). B, The impact of xylene-based coverslipping and decoverslipping appeared negligible. Hence, microwave treatment was not necessary (procedure depicted in trial 5 without microwave treatment). C, Prestaining with Wright's stain and then bleaching with hydrogen chloride abolished the specific immunostaining of all markers (blue arrow), if the slides were left uncoverslipped (procedure depicted in trial 3 without microwave treatment). D, Coverslipping after prestaining with a Wright's stain appeared superior in being able to preserve antigenicities and specific staining intensities (procedure depicted in trial 4 without microwave treatment). E, Microwave treatment completely rescued the compromising effects of staining/destaining and coverslipping/ decoverslipping (procedure depicted in trial 4 with microwave treatment). NeuN = neuronal nuclei, GFAP = glial fibrillary acidic protein, CD3 = T cell marker. The wave image depicts microwave treatment, the upside down y image depict antibody treatment, the coverslip indicates the coverslipping/decoverslipping technique DÖRFELT ET AL.

| 103 (GFAP: score 1.5 ± 0.58) (Figure 2A ; Table 3 ). For GFAP only, a background with brownish, cloudy, sometimes flocculent low-intensity staining was evident in all specimens (4/4) ( Table 3 ). However, this background did not interfere with the distinctive staining of astrocyte somata. Regarding the subcellular distribution of ICC on these plain slides, the CD3 signal was strongest in the periphery of the lymphocyte cytoplasm close to the membrane. NeuN immunoreactivity was restricted to neuronal nuclei, while GFAP diffusely stained the soma and fragmented processes of astrocytes.

Even after coverslipping and decoverslipping (CDC only, trial 5), ICC permitted the identification of all three cell types ( Figure 2B ).

However, both the staining intensity and cell specificity of the CD3

and NeuN specimens were significantly lower than those observed in trial 6 (P ≤ 0.04), while GFAP background staining was stronger (P = 0.02) (Tables 3 and 4 ).

The SDS (trial 3) did not interfere with immunolabeling if ICC was performed within a few days after the Romanowsky staining.

However, a delay of 1-2 weeks abolished immunolabeling significantly ( Figure 2C ). Hence, the detectability of all markers was severely compromised (cell specificity: P = 0.013; intensity: Tables 3 and 5 ). Weak cellular staining was seen in isolated lymphocytes and astrocytes only, while the majority of cells in all specimens remained immunonegative.

Early coverslipping of modWS-stained specimens (trials 2 and 4)

rescued the antigenicity of all three antigens compared with the prestained, noncoverslipped slides in trial 3 (intensity and cell specificity: Figure 2D ; Tables 5 and 6 ). A specific signal was obtained on these slides, with a median signal intensity (SI) of 1.0-1.5 (Table 6 ). Nonspecific background staining was restricted mostly to GFAP slides (10/10) but remained distinct from cellular chromagen enrichment, as seen in Figure 2D ( Table 6) . Notably, SDS/CDC affected the subcellular distribution of immunostaining; the CD3 signal extended throughout the lymphocyte cytoplasm rather than accentuating the periphery, while GFAP staining in astrocytes leaked into the nucleus. NeuN staining was still mainly confined to neuronal nuclei; however, the signal was patchy and did not encompass the entire nuclear area ( Figure 2D) . The positive effect of MWT was restricted to coverslipped slides.

The omission of coverslipping in trial 3 failed to rescue antigenicity.

All types of chemical pretreatments (trials 3, 4, and 5) completely abolished immunostaining of the FCoV nucleocapsid ( Figure 3A and B; Table 7 ). Therefore, in contrast to the immunodetection of cellular antigens, coverslipping did not protect viral antigenicity on stained slides (Table 7) . Likewise, MWT precluded immunostaining on plain slides from trial 6 and, hence, failed to rescue antigen labeling on the pretreated slides. contamination, and differences in yield and cell composition among consecutive samples. 2, 29, 30 Traditionally, MWT is used in immunohistochemistry with formalin-fixed tissues for antigen recovery from the cross-linking of aldehyde bridges. 37, 38 Similarly, MWT has proven useful for antigen demasking in cytologic specimens after fixation with ethanol and formalin 37 and after cytologic staining. 1 However, MWT failed to recover viral antigen detectability in this study. Instead, MWT alone effectively abolished the immunolabeling of FCoV in untreated and unstained films; furthermore, it was not possible to reestablish immunostaining after antecedent cytologic staining and coverslipping. Thus, this FCoV epitope appears to be much more vulnerable to changes induced by physicochemical factors than the other cellular proteins investigated in this study.

Whether viral antigen preservation can be achieved by prior fixation of a smear, as described for other microbial antigens, such as Chlamydia, 39 is currently under investigation.

Other studies have proposed that fixation by formalin or ethanol can improve ICC results. 1, 11, 13, 14, 26, 36, 37, 39 for FCoV detection in air-dried prestained films.

We are grateful to the head technician of our Neuropathology & Cytology Lab, Mrs. Karin Stingl, for excellent technical assistance.

The authors have indicated that they have no affiliations or financial involvement with any organization or entity with a financial interest in, or in financial competition with, the subject matter or materials discussed in this article.

Establishing a protocol for immunocytochemical staining and chromogenic in situ hybridization of Giemsa and Diff-Quick prestained cytological smears

Use of destained cytology slides for the application of routine special stains

Immunocytochemistry and cutaneous neoplasia ('something will turn up

Immunocytochemistry: It's role in diagnosis of undifferentiated neoplasms by fine needle aspiration cytology

Immunocytochemistry in diagnostic cytology

Immunocytochemical differentiation of neoplastic and hyperplastic canine epithelial lesions in cytologic imprint preparations

The role of immunocytochemistry in diagnostic pathology

High-grade canine T-cell lymphoma/leukemia with plasmacytoid morphology: a clinical pathological study of nine cases

Diagnostic utility of cerebrospinal fluid immunocytochemistry for diagnosis of feline infectious peritonitis manifesting in the central nervous system

Immunocytochemical demonstration of feline infectious peritonitis virus within cerebrospinal fluid macrophages

Immunocytochemical staining of c-erb B-2 oncogene in fine-needle aspirates of breast carcinoma: a comparison with tissue sections and other breast cancer prognostic factors

Immunocytochemical demonstration of glial fibrillary acidic protein in imprint smears of human brain tumors

Ki-67 antigen detection in unstained and destained cytological samples

p63 immunostaining in destained bronchoscopic cytological specimens

Immunocytochemical detection of KI-67 in Diff-Quick-stained cytological smears of canine mammary gland tumours

MIB-1 immunostaining on cytological samples: a protocol without antigen retrieval

Immunocytochemical staining of unstained versus previously stained cytologic preparations

Mammary analogue secretory carcinoma of parotid: Is preoperative cytological diagnosis possible?

Role of fine needle aspiration cytology, cell block preparation and CD63, P63 and CD56 immunostaining in classifying the specific tumor type of the lung

Immunocytochemical analysis of prostate stem cell antigen as adjunct marker for detection of urothelial transitional cell carcinoma in voided urine specimens

Immunoreactivity of p16 in anal cytology specimens: histologic correlation

A pilot study of Galectin-3, HBME-1, and p27 triple immunostaining pattern for diagnosis of indeterminate thyroid nodules in cytology with correlation to histology

Monoclonal antibody 44-3A6 as an adjunct in cytodiagnosis of adenocarcinomas in body fluids

Immunocytochemistry on cytologic specimens of limited quantity

International Veterinary Epilepsy Task Force recommendations for systematic sampling and processing of brains from epileptic dogs and cats

Diagnostic yield and accuracy of postmortem cytological sampling from the brain surface of animals with neurological abnormalities

Classification of feline intraocular neoplasms based on morphology, histochemical staining, and immunohistochemical labelling

Color Atlas of Cytology of the Dog and Cat, 1st edn

Sample collection and preparation in cytology: increasing diagnostic yield

Diagnostic Cytology and Hematology of the Dog and Cat, 2nd edn

Inhibitors of the polymerase chain reaction in Papanicolaou stain. Removal with a simple destaining procedure

Rapid staining techniques in cytopathology: a review and comparison of modified protocols for hematoxylin and eosin, Papanicolaou and Romanowsky stains

Tissue fixation and the effect of molecular fixatives on downstream staining procedures

A study on the appropriate fixation for the procedures for the better preservation of cellular antigenicity and morphology of the blood smear in immunocytochemistry: an improvement of the immunostain technique using alkaline-phosphatase (ALP) as a labeling enzyme

Immunocytochemistry: an indispensable technique in routine cytology

Immunostaining of cell preparations: a comparative evaluation of common fixatives and protocols

When tissue and antibodies get along: revisiting the technical aspects of immunohistochemistry -the red, brown, and blue technique

Antigen retrieval techniques: current perspectives

Immunocytochemical detection of Chlamydia trachomatis on cytological specimens

Antigens under cover-The preservation and demasking of selected antigens for successful poststaining immunocytochemistry of effusion, brain smears, and lymph node aspirates