Demonstration of bovine herpesvirus type 1 and Mannheimia haemolytica antigens in specimens stored for up to 22 months in buffered formalin

Abstract

Bovine herpesvirus type 1 (BHV-1) and Mannheimia haemolytica antigens were demonstrated in lung tissues that were stored in 10% neutral phosphate buffered formalin for 1 to 22 months using the immunoperoxidase method. There were no differences observed in terms of labelling intensity and distribution of M. haemolytica antigens between specimens stored for 1 and 22 months. The labeling intensity in sections from 2-cm thick specimens was comparable to those from 0.2-cm thick specimens. There was no difference observed between pronase-treated and -untreated sections. However, for BHV-1, the labeling intensity in untreated sections was reduced in tissues that had been stored from 12 to 22 months. Sections from thin specimens stored in neutral buffered formalin for 22 months exhibited a stronger staining intensity than those from thick specimens.

Viral antigens, such as those of bovine herpesvirus type 1 (BHV-1) (1), classical swine fever virus (2), and porcine circovirus (3), have been demonstrated using the immunoperoxidase method in formalin-fixed, paraffin-embedded tissues kept at room temperature for 10 to 25 y. Prolonged storage of specimens in formalin has been reported to have adverse effects on the ability to demonstrate antigenic epitopes (4,5,6,7). Takimoto et al demonstrated herpes simplex virus type 1 antigen in mouse brain that had been stored in neutral buffered formalin for 6 mo (8).

The purpose of this study was to elucidate the relationship between extended formalin storage (from 1 to 22 mo) and the demonstration of bacterial and viral antigens. The immunoperoxidase method was used to detect BHV-1 and Mannheimia haemolytica antigens in lung tissues of cattle. Bulk (thick) specimens were also compared to thin ones in terms of antigen demonstration after extended storage in neutral buffered formalin.

The examination procedure and antigen distribution have been described previously (8). Archival lung tissues from 3 cattle (nos. 5 to 7) endobronchially infected with BHV-1, 2 cattle (nos. 8 and 9) infected with M. haemolytica and 2 co-infected cattle (nos. 12 and 13) were used. The tissues had been collected at post mortem and fixed in 10% neutral phosphate-buffered formalin for 7 d. Thereafter, smaller tissues from the right caudal lobe, 0.2 cm wide and 2 cm long, were cut, put into plastic unicassettes (Miles Laboratories, Naperville, Illinois, USA), and stored in 10% neutral buffered formalin at room temperature for varying periods of time prior to processing. Processing included routine methods used for preparing sections (4 μm) from paraffin-embedded tissues. Bulk specimens (10 cm × 10 cm × 2 cm) of the right caudal lobe from 7 cattle (nos. 5 to 9, 12, and 13) were also stored at room temperature in neutral buffered formalin until processing at 22 mo.

Both BHV-1 (nos. 5 to 7) and M. haemolytica (nos. 8, 9, 12, and 13) antigens in tissue sections were detected using the streptavidin-biotin (SAB) immunoperoxidase method (Histofine SAB reagent; Nichirei, Tokyo, Japan). After deparaffinization, endogenous peroxidase activity was blocked by treatment with 0.3% hydrogen peroxide in absolute methanol for 30 min. Sets of tissue sections were then treated with 0.1% pronase (Kaken Chemicals, Tokyo, Japan) in Tris-HCl buffer (0.05 M, pH 7.6) for 5, 10, 20, or 30 min at room temperature. One set of sections containing each antigen was not treated. The sections were incubated with normal goat (BHV-1) or rabbit serum (M. haemolytica), primary antibodies, SAB reagents, and 0.05% 3,3-diaminobenzidine tetrahydrochloride (Dohjin Chemical Laboratories, Kumamoto, Japan) in Tris-HCl buffer (0.05 M, pH 7.6). The primary antibodies used were goat anti-BHV-1 antibody (VMRD, Pullman, Washington, USA) diluted to 1:4096 and rabbit anti-M. haemolytica antibody (provided by Dr. K. Yamamoto, National Institute of Animal Health, Ibaraki, Japan) diluted to 1:16 384 both in Tris-HCl buffer. For the controls, non-immune goat and rabbit sera were used for BHV-1 and M. haemolytica, respectively. Counterstaining was done by using methylene green.

Sections were examined and scored subjectively as 1+ (weak), 2+ (moderate), or 3+ (strong) to indicate labeling intensity and the distribution and number of labeled cells (Table I). Both BHV-1 and M. haemolytica antigens were demonstrated in specimens stored for 22 mo in neutral buffered formalin. BHV-1 antigen was detected as brown granules in the cytoplasm and nucleus of degenerating bronchial and bronchiolar epithelial cells (Figures 1-3). Mannheimia haemolytica antigen was detected as dense, brown, diffuse immunolabeling in necrotic alveolar cells (Figures 4-6). There was no difference observed in terms of labeling intensity and distribution of BHV-1 antigens stored in neutral buffered formalin from 1 to 10 mo or between pronase-treated and -untreated sections during this period. From 12 to 22 mo, untreated sections exhibited a weaker labeling intensity compared to the treated ones. At 22 mo, the labeling intensity of untreated sections was very poor. The sections treated for 5 min exhibited a weaker intensity than those treated for 30 min. The labeling intensity of sections cut from 2-cm thick specimens was weak compared to those cut from 0.2-cm thick specimens.

Table I.

Figure 1. Lung section from a 0.2-cm thick specimen stored in 10% neutral buffered formalin for 22 mo. Immunoperoxidase SAB-staining, obtained with anti-BHV-1 serum, no pronase pretreatment. Bar = 100 μm.

Figure 2. Lung section from a 0.2-cm thick specimen stored in 10% neutral buffered formalin for 22 mo. Immunoperoxidase SAB-staining, obtained with anti-BHV-1 serum, pronase-pretreated for 30 min. Bar = 100 μm.

Figure 3. Lung section from a 2 cm-thick specimen stored in 10% neutral buffered formalin for 22 mo. Immunoperoxidase SAB-staining, obtained with anti-BHV-1 serum, pronase-pretreated for 30 min. Bar = 100 μm.

Figure 4. Lung section from a 0.2-cm thick specimen stored in 10% neutral buffered formalin for 22 mo. Immunoperoxidase SAB-staining, obtained with anti-M. haemolytica serum, no pronase pretreatment. Bar = 100 μm.

Figure 5. Lung section from a 0.2-cm thick specimen stored in 10% neutral buffered formalin for 22 mo. Immunoperoxidase SAB-staining, obtained with anti-M. haemolytica serum, pronase-pretreated for 30 min. Bar = 100 μm.

Figure 6. Lung section from a 2-cm thick specimen stored in 10% neutral buffered formalin for 22 mo. Immunoperoxidase SAB-staining, obtained with anti-M. haemolytica serum, pronase-pretreated for 30 min. Bar = 100 μm.

Mannheimia haemolytica antigen exhibited no difference in terms of the labeling intensity and distribution from 1 to 22 mo. Furthermore, no difference was observed in the demonstration of these antigens between treated and untreated sections. The labeling intensity of sections cut from 2-cm thick specimens and stored in neutral buffered formalin for 22 mo was comparable to sections from 0.2-cm thick specimens stored for the same period of time.

The results obtained in this study agree with the findings of Takimoto et al on the demonstration of herpes simplex virus in mouse brain stored in neutral buffered formalin for 6 mo (9). They are also consistent with earlier reports, that extended storage of specimens in neutral buffered formalin has no deleterious effect on antigens (10,11). Unbuffered formalin however, can cause loss in antigenicity after extended storage (9). It is not clear whether the loss of antigenicity is due to the formation of formic acid or improper fixation caused by the lowered pH (10,12,13).

The weak BHV-1 immunolabeling observed in tissues fixed for 12 to 22 mo appears to have resulted from antigen masking in tissue sections that were not treated with pronase (14). This effect became more pronounced at 22 mo when the sections treated with pronase for 5 min also exhibited a weaker labeling intensity compared with those treated for 30 min. This suggests that the process of fixation continued to take place throughout the storage period in formalin.

Neither the storage period nor the enzymatic digestion affected the demonstration of M. haemolytica antigens. This result suggests that the M. haemolytica antibodies used in this study recognized epitopes that were resistant to formalin fixation and subsequent tissue processing (13,15).

Storage of 0.2-cm thick specimens in neutral buffered formalin yielded better immunolabeling than 2-cm thick specimens after 22 mo for the BHV-1 antigens. It appears that the thickness of specimens stored in formalin for extended time periods is important. In conclusion, 10% neutral buffered formalin can be used as a preservative for specimens in retrospective histopathological and immunohistochemical studies in the diagnosis of bacterial and viral infectious diseases of domestic animals.