Retrieval of leukocyte antigens in paraffin-embedded rat tissues.

We compared the immunohistochemical reactivity of various rat leukocyte antigens in frozen and paraffin-embedded thymus, spleen, abdominal lymph node, liver, and brain tissues of healthy Sprague-Dawley rats, fixed in various fixatives. Immune reactivity after fixation in Methacarn was superior to that of 4% neutral buffered formalin, a mercury-based fixative (B-5), or Carnoy. Microwave (MW) antigen retrieval (AR) enhanced antigen reactivity. Ten of the 11 leukocyte antigens studied could be retrieved in Methacarn-fixed, paraffin-embedded sections with a reactivity comparable to that obtained on frozen sections.

Normal colon of Sprague-Dawley rats. An immunohistochemical study.

The patterns of expression of the human-tumor-associated antigens, CO17-1A, GA73-3, BR55-2, GICA19-9, CA50 and carcino-embryonic antigen (CEA) were studied in the normal colonic mucosa (the last three also in the serum) of Sprague-Dawley rats. Four immunohistochemically different segments were identified: caecum, ascending colon, transverse colon and descending colon. The immunohistochemical reactions of the cells at the lower part of the crypt were essential for the distinction of the four segments. In the caecum, the MAbs 17-1A, 73-3 and 19-9 stained the glycocalyx of the cells of the lower part of the crypts and the Golgi apparatus of the intercalated cells (IC). MAb55-2 stained very weakly the goblet-like cells (GLC) of the lower part of the crypt of transverse colon, in addition to a nearly complete lack of reaction in the upper part of the crypts. In the ascending colon, the lower part of the crypts showed a characteristic diffuse staining of the intercalated cells with MAb55-2. The perinuclear and mucosal staining observed in the GLC of the transverse colon with MAbs 17-1A, 73-3 and 19-9 as against the supranuclear and Golgi zone staining observed in the GLC/goblet cells (GC)/columnar cells (CC) of the lower part of crypts of the descending colon with the same MAbs, distinguished the former segment from the latter. The IC demonstrated by immunohistochemistry in the lower parts of the crypts of caecum and ascending colon appear to correspond to the replicating cells of the colonic crypts.

Tumor-associated antigens common to humans and chemically induced colonic tumors of the rat.

The expression of human tumor-associated antigens CO17-1A, GA73-3, BR55-2, GICA 19-9, and CA50 and of carcinoembryonic antigen was immunohistochemically studied in the colonic mucosa of 70 Sprague-Dawley rats. Fifty were treated with 1,2-dimethylhydrazine (DMH) (with EDTA as a vehicle), ten were treated with EDTA only, and ten were untreated normal rats. The tumors were histogenetically divided as: (a) adenocarcinomas arising from villous adenomas; (b) adenocarcinomas arising from lymphoid-associated mucosa (LAM); and (c) adenocarcinomas arising in flat mucosa. Of 44 colonic adenocarcinomas, BR55-2 was expressed in 41 tumors, CO17-1A in 40 tumors, GA73-3 in 38 tumors, and GICA 19-9 in 38 tumors. CA50 and carcinoembryonic antigen were not expressed in the tumors. The highest antigenic expression (number of cells) was observed in adenocarcinomas arising in villous adenomas and the lowest in those arising in flat mucosa. Adenocarcinomas arising in LAM had an intermediate expression. The expression of these antigens had no correlation to the localization of the tumor and to the differentiation. The expression of these antigens was similar in the non-lymphoid-associated normal colonic mucosa of the untreated, EDTA-treated, and DMH-treated rats. In DMH-treated rats, LAM demonstrated increased expression (number of cells) and increased staining intensity of these tumor-associated antigens. In six of the 50 DMH-treated rats, only LAM expressed carcinoembryonic antigen. CA50 was not expressed in the normal colon of untreated, of EDTA-treated, and of DMH-treated rats, nor was it in DMH-induced tumors. None of the tumor-associated antigens (GICA 19-9 and CA50 and carcinoembryonic antigen) was detected in serum. It is concluded that this animal model would be of value in the preclinical evaluations of monoclonal antibodies for therapy in humans.