Effect of substrate selection on indirect immunofluorescence testing of canine autoimmune subepidermal blistering diseases

Abstract

The detection by indirect immunofluorescence (IIF) of circulating antibodies in the serum of dogs with autoimmune subepidermal blistering diseases (AISBD) was regarded for a long time as an unrewarding tool. It was, however, demonstrated in humans that the sensitivity of IIF assays depended on the selection of the substrates used. The effects of substrate selection on IIF tests was thus studied by examining sera from 12 dogs with AISBD tested against 8 different substrates from 3 different normal dogs. Patients with AISBD suffered from bullous pemphigoid (n = 4 sera), mucous membrane pemphigoid (n = 4 sera), and epidermolysis bullosa acquisita (n = 4 sera). Substrates included canine tongue, canine lip, canine dorsal haired skin, and ventral haired skin. The same 4 substrates were also split with salt splitting technique (using 1 M sodium chloride), in order to cleave the basement membrane within the lamina lucida and to expose the targeted antigens. The strength of the specific fluorescence of each slide was scored after processing for IIF testing with anti-canine IgG polyclonal antibody. Other criteria, such as background fluorescence, easiness of the interpretation, and variations within a same substrate, were also assessed. Intact canine lip and canine salt-split lip demonstrated consistently stronger intensity of fluorescence and a better ease of interpretation. We concluded that the performance of IIF tests with such substrates was a reliable tool for the detection of circulating IgG autoantibodies of canine patients with AISBD.

Introduction

Autoimmune subepidermal blistering diseases (AISBD) consist of a group of mucosal and skin diseases that share common clinical, pathological, and immunological features (1). In humans, the classification of these diseases is based upon the clinical features and the demonstration of the antigens targeted by circulating antibodies. Animal patients with AISBD have been identified since 1978 and were given the generic diagnosis of bullous pemphigoid (BP) until 1995 (2). However, these patients exhibited different clinical and histopathological features. Moreover, the immunological findings and the prognosis were often different (1). Since 1995, some cases of AISBD have been investigated and reclassified using the clinical and immunological nomenclature currently established in humans (3,4,5,6). Mucous membrane pemphigoid (MMP), epidermolysis bullosa acquisita (EBA), and BP were subsequently individualized and are now considered the most frequently encountered diseases of this group (3,4,5).

Canine BP is now defined as a blistering dermatosis that affects mainly the face and the ventral aspect of the body and, less frequently, the mucous membranes. Antibody deposition occurs on the epidermal side of the dermal-epidermal junction and the antigen targeted is the NC16A extracellular domain of collagen XVII (BPAg2, BP180) (3).

Canine MMP is regarded as the mucous counterpart of BP. Antibody deposition usually, but not always, occurs at the same level of the dermoepidermal junction and the antigenic epitope usually targeted is the same extracellular domain of collagen XVII. However, in humans as well as in the dog, MMP appears as an immunologically heterogeneous disease with other antigens or epitopes sometimes being targeted by circulating antibodies (5).

Canine EBA is a more severe blistering disease that affects both skin and mucous membranes. Antibody deposition occurs on the dermal side of the dermal-epidermal junction. The antigen targeted is the amino-terminus NC1 segment of collagen VII, the latter being the main component of anchoring fibrils (4).

In veterinary dermatology, the detection of circulating antibodies was not deemed a rewarding diagnostic procedure for AISBD because of the low sensitivity of indirect immunofluorescence (IIF) testing (7,8,9). On the contrary, in human medicine, it has been considered for a long time that IIF can provide valuable information by allowing the detection of circulating antibasement membrane autoantibodies. However, studies have shown that the substrate used can greatly influence the results of IIF (10,11). Woodley (12) demonstrated that human skin, when incubated in 1 M sodium chloride, fractures cleanly through the lamina lucida zone of the epidermal basement membrane. This fracture places the NC16A fragment of the BP antigen on the epidermal side of the split and components of the lamina densa (including collagen VII) on the dermal side of the separation. Salt-split skin can thus be used to distinguish EBA and BP. Since 1995, veterinary studies have demonstrated that IIF can also be a valuable tool, provided good substrates were used (3,4). The effects of the substrate have also been studied for canine pemphigus and the study has demonstrated marked variation between the results obtained with different substrates (13). The same authors have also suggested that immunomapping of salt-split skin may be useful for the differential diagnosis of canine AISBD (14).

In the present study, we wish to determine whether the use of different substrates would influence the detection of circulating autoantibodies in dogs with AISBD. We will demonstrate that intact canine lip and canine salt-split lip are the substrates that offer the most consistent and easiest detection of circulating autoantibodies for dogs affected with this group of diseases.

Materials and methods

Specimen collection

The following substrates were used for IIF: canine tongue, canine lip, canine dorsal haired skin, and canine ventral haired skin. In order to evaluate the importance of individual variations within a same substrate, samples were taken by 6-mm punch biopsy on 3 different healthy dogs, provided by the local shelter immediately after euthanasia. One half of the samples (intact substrates) were snap-frozen in liquid nitrogen after OCT compound embedding (Tissue-tek; Sakura Finatek, Torrance, California, USA) and stored at −70°C until IIF was performed (3). The second half of the samples (salt-split substrates) were incubated in 1 M sodium chloride solution at room temperature for 24 h (as described elsewhere) (4). After incubation, the samples were washed in phosphate-buffered saline (PBS) and snap-frozen, as described above.

Patients

Sera from 12 dogs with AISBD (4 BP, 4 MMP, 4 EBA) were tested by IIF. The diagnosis of AIBSD had been established previously by collation of clinical, histopathological, and immunological findings. The criteria for selection used are listed in Table I (3,4,5). The sera from one normal dog and one dog with pemphigus vulgaris were used as negative and positive control, respectively.

Table I.

Indirect immunofluorescence

Indirect immunofluorenscence studies were performed as previously described (3). Briefly, 5-μm cryosections were obtained from each sample and air-dried for 15 min then fixed in acetone for 15 min, rinsed in PBS, and blocked in 1% normal rabbit serum solution for 20 min. Sections were then incubated with patient sera (1:50) for 60 min at room temperature in a moist chamber, washed with PBS, and then incubated for 30 min with fluorescein-isothiocyanate (FITC)-conjugated sheep anti-dog IgG (1:60) (ICN Biomedicals, Aurora, Ohio, USA). Sections were then washed with PBS, mounted with Vectashield-DAPI, cover slipped, and observed with an epifluorescence microscope.

Assessed criteria and evaluation of the fluorescence

Each section was scrutinized and the following points were assessed:

Pattern of specific fluorescence — The reactivity of the basement membrane zone (BMZ) was classified as linear or reticular. The fluorescence of the lateral aspects of the keratinocytes was also assessed. When salt-split substrates were used, the localization of the fluorescence (dermal or epidermal) was recorded.

Variations within the same substrates — The variation of the intensity of the fluorescence for each of the 8 substrates (intact and split lips, tongues, dorsal haired skin, ventral haired skin) was evaluated among each of the 3 donor dogs.

Intensity of specific fluorescence — The intensity of specific fluorescence was scored as follows: 0 (none), 1 (low), 2 (medium), 3 (strong), or 4 (very strong). Specific fluorescence was defined as a fluorescence that was stronger than or as strong as that of the dermis and of a pattern expected from the disease tested.

Intensity of the background fluorescence — The intensity of the dermal background fluorescence was evaluated and compared to that of the BMZ.

Quality of sectioning — Each section was evaluated based on morphologic criteria (aspect of the dermis and of the epidermis, integrity of the BMZ). Finally, the percentage of exploitable sections was evaluated for each substrate.

Results

Fluorescence pattern

On most sections it was possible to observe closely the fluorescence pattern. It appeared clearly that BP sera resulted in a reticulated fluorescence that extended to the lateral aspect of the basal keratinocytes (Figure 1). The EBA sera resulted in a strong, thick, linear fluorescence below the basal keratinocytes (Figures 2 and 4). These features were more easily observable on salt-split sections. On these sections, BP and MMP sera always resulted in fluorescence of the epidermal side of the clefts and EBA on the dermal side of this cleft (Figures 3 and 4).

Figure 1. Substrate: intact canine dorsal haired skin section (5 μm). Serum: bullous pemphigoid, 1:50. Antiserum: FITC-conjugated sheep anti-dog IgG, 1:60. Magnification 50X. The thinness of the epidermis renders the localization of the fluorescence uneasy.

Figure 2. Substrate: intact canine lip section (5 μm). Serum: epidermolysis bullosa acquisita, 1:50. Antiserum: FITC-conjugated sheep anti-dog IgG, 1:60. Magnification 50X. The thickness of the epidermis and the low background fluorescence of the dermis renders the interpretation easy.

Figure 4. Substrate: 5 μm salt-split lip section. Serum: epidermolysis bullosa acquisita, 1:50. Antiserum: FITC-conjugated sheep anti-dog IgG, 1:60. Magnification 50X. The fluorescence is assessed on the dermal side of the cleft.

Figure 3. Substrate: salt-split tongue section (5 μm). Serum: mucous membrane pemphigoid, 1:50. Antiserum: FITC-conjugated sheep anti-dog IgG, 1:60. Magnification 50X. The fluorescence is assessed on the epidermal side of the cleft. Note that the intensity of the background fluorescence is very intense with tongue substrates.

Variation within the same substrate

Individual variations within a same substrate were first evaluated. Samples collected from one dog demonstrated stronger fluorescence than those from the other 2 dogs. The average score for intact and split lips, tongues, dorsal haired skins, and ventral haired skins of this hypereactive dog was 2.3, while each of the other 2 dogs had an average score of 1.9. This discrepancy was present for each type of substrate (intact or split) and each disease. Moreover, 5 of 48 sections (10.4%) from this hypereactive dog demonstrated positive specific fluorescence even though sections or at least one section of another dog did not demonstrate any evidence of fluorescence.

Intensity of fluorescence

For intact substrates, the average score obtained for the lip was higher (2.1) that those of other substrates (Figure 1). This average score of the lip sections was higher than those of other sections for each disease (except EBA/ventral skin) (see Table II).

Table II.

With salt-split substrates, the same situation was evident. On the other hand, the average score of the salt-split substrates was higher than those of the intact substrates (see Table II). With regard to the individual diseases, the average score of the BP sera were higher than those of the other AIBSD sera (Table II).

Background fluorescence

The intensity of the dermal fluorescence background was very intense with the tongue substrate (Figure 3), making the interpretation of the specific fluorescence difficult. On the contrary, the significantly lower background fluorescence of the lip substrate (Figures 2 and 4) made for easier interpretation.

Quality of sectioning

With intact substrates, 100% of the tongue sections were usable (e.g. good enough quality to allow correct interpretation). With other substrates (lip, dorsal skin, ventral skin), percentages of usable sections were 90%, 64%, and 75%, respectively. With salt-split substrates, usable sections for tongue, lip, dorsal skin and ventral skin were 94%, 83%, 81%, and 75% respectively.

In conclusion, canine lip (intact or split) seems to be the best evaluated substrate for indirect immunofluorescence testing of canine AISBD because of a low dermal background fluorescence, the higher specific fluorescence, and the ease of sectioning.

Discussion

In human medical dermatology, IIF has been regarded as a rewarding diagnostic tool for many years. In 1985, Beutner noticed that 60% to 80% of the patients with bullous forms of EBA or BP were seropositive for circulatory anti-BMZ antibodies (10). On the other hand, lower percentages were found in non-bullous forms of EBA and BP and in all subtypes of MMP (cicatricial pemphigoid). In 1986, Bystryn demonstrated that the nature of substrates used influenced the results of 56% of assays for BMZ antibodies (11); 38% of the sera failed to react to one or more than one substrate and other substrates demonstrated different titers when different substrates were used. The best results were obtained using guinea pig esophagus (96% of positive results with this substrate) and it was possible to reach 100% when using guinea pig esophagus, monkey esophagus, and human skin, concomitantly.

It is now considered possible to detect circulating anti-BMZ autoantibodies in the serum of almost every human patient with EBA, BP, or MMP, provided proper substrates are used (15,16,17). Salt-split substrates, in which the epidermis is separated from the underlying dermis at the mid-level of the lamina lucida, appear particularly suitable for IIF detection of circulating autoantibodies.

In contrast, until 1995, it was considered that IIF testing was unreliable for the diagnosis of BP in dogs (7,8,9). However, Iwasaki et al performed IIF assays on the serum of a canine BP patient and demonstrated circulating anti-BMZ antibodies by using canine salt-split skin and bovine tongue (3). This study failed to detect these antibodies when intact canine skin was used.

In 1998, Olivry et al detected significant levels of circulating anti-BMZ specific antibodies (IgG and IgA) in a dog with EBA using intact canine tongue and lip as substrates (4). In another study, these investigators obtained 88% of positive tests with sera of dogs with MMP using canine salt-split lip (5).

The aim of our study was to determine the best substrates for IIF assays for detection of circulating IgG autoantibodies using sera of canine patients with AISBD. We obtained biopsies from 3 different dogs in order to evaluate the importance of individual variations in IIF testing. In 10.4% (5/48) of the tests, the immunofluorescence was negative with one particular substrate from one dog and present with the very same substrate from other dogs. Moreover, the substrates of one of the 3 dogs appeared to yield consistently better results than those of the 2 others. It seems wise, then, to use substrates from more than one dog when using such a diagnostic assay.

The best intact substrate used during our study was the canine lip. The main drawback of using canine tongue was the intensity of the background dermal fluorescence, which sometimes rendered interpretation of the assay difficult. The use of canine haired skin was even more difficult. For one, the thinness of the epidermis made the precise localization of the fluorescence impossible. Secondly, the presence of hair within the substrate made the cryosection more tedious and less suitable for easy interpretation.

The canine lip is characterized by a thick epithelium that allows easy detection of the location and pattern of specific fluorescence. In addition, its dermis often appears to exhibit less background reactivity than that of the tongue. Salt-split canine lip is also a suitable substrate. In fact, the average intensity of the fluorescence using salt-split lip is higher than that of the intact lip. In addition, the clefting allows one to determine on which side of the dermoepidermal junction the autoantibodies will bind. Indeed, with sera from dogs with BP and MMP, binding almost always occurs on the epidermal side of the clefts, while in dogs with EBA binding occurs on the dermal side of the splits.

We therefore conclude that using 2 intact canine lip sections from 2 different dogs and 2 salt-split lip sections from the same 2 dogs would constitute a suitable tool to detect specific circulating anti-BMZ antibodies in most sera from canine patients with AISBD.