Mineko Iwai

Endogenous type-C viral expression during lymphoma development in irradiated NFS mice.

The expression of the type-C retrovirus and the virus-related components in NFS mice were examined during preleukemic and leukemic phases after fractionated whole-body X irradiation. The NFS mice were highly susceptible to induction of thymoma by fractionated X irradiation. The leukemic tissues were negative for infectious type-C virus, as detected by both the XC-plaque test and mink S+ L- focus-inducing assays, but contained a substantially higher level of viral-specific RNA-dependent DNA polymerase activity and a major core protein p30 than the corresponding tissues from unirradiated age-control mice. In the preleukemic phase, the amount of p30-related antigen increased transiently in spleen. The leukemic cell lines established from radiation-induced lymphomas produced particulate entities with a buoyant density of about 1.15 g/ml. These virus-like particles lacked in vitro infectivity to mouse cells and mink lung cells and leukemogenicity in syngeneic mice. The p30-related antigens of these particles were immunologically similar to that of xenotropic virus derived from NZB mouse.

Lack of Evidence for the Involvement of Type-C and Type-B Retroviruses in Radiation Leukemogenesis of NFS Mice

Southern blot analysis revealed no difference between the DNA from radiation-induced thymic lymphomas and DNA from normal NFS mice. The probes used in the Southern blot analyses used a murine leukemia virus (MuLV) env DNA probe (pXenv), which specifically hybridizes with xenotropic and recombinant viral env genes, and mouse mammary tumor virus (MMTV) DNA probes (MMTV gag-pol, MMTV env, and MMTV LTR). This suggests that radiation leukemogenesis was not associated with gross alteration of the organization of these retroviral genomes. In DNA from radiation-induced thymic lymphoma, there was no indication of gross rearrangement in the common integration site of MuLV, pim-1, or in the common integration sites of MMTV, int-1 and int-2. Dot blot analysis of RNA from radiation-induced thymic lymphomas and normal thymuses demonstrated that there was no substantial difference between them in the expression of retroviral sequences, pim-1, pvt-1, int-1, or int-2, although transcripts that could be hybridized to the retroviral sequences were slightly elevated in some radiation-induced thymic lymphomas. These results show that radiation leukemogenesis does not appear to involve the activation of endogenous type-C and type-B retroviruses.

CLONING OF A CANCER CELL–PRODUCING HEPATOCYTE GROWTH FACTOR, VASCULAR ENDOTHELIAL GROWTH FACTOR, AND INTERLEUKIN-8 FROM GASTRIC CANCER CELLS

SummaryA cell colony (IM95m) that produces hepatocyte growth factor (HGF), vascular endothelial growth factor (VEGF), and interleukin-8 (IL-8) was cloned from gastric cancer cells (IM95 cell line). In culture medium, the highest levels of HGF, VEGF, and IL-8 were about 1.1, 0.9, and 0.17 ng/ml culture medium at 3 d from 105 cells. IM95m may be useful in elucidating the role of tumor cells in angiogenesis.

Mouse Mammary Tumor Virus Proviral Integration in the DD/Tbr Mice

The patterns of mouse mammary tumor virus (MMTV) integration in the DNA of spontaneousmammary tumors, salivary glands and livers of DD/Tbr mice were examined using MMTVenv,intlc andint-2c probes. The MMTVenv probe revealed 1 to 7 new proviral insertions in all mammary tumors. MMTV integration intoint-1 was observed in 10 of 18 mammary tumors, whereas that intoint-2 was seen in only 2 of 18 tumors. Of the 13 salivary glands examined, only 3 showed new MMTV proviral integrations, but rearrangement in int-1 orint-2 loci by MMTV was not observed. Immuno-colloidal gold electron microscopy revealed the presence of MMTV particles both in mammary tumors and in salivary glands, but no tumors were found to be developed in salivary glands. Taken together these results suggest that salivary glands support MMTV replication, but the virions thus produced may not lead to salivary gland tumorigenesis. It is suggested that the salivary gland is the source of horizontally transmitted MMTV in DD/Tbr mice.

DNA rearrangements of the int region in spontaneous mouse mammary tumors of SHN/S and SLN/S mice

SHN and SLN mice originating from the same Swiss albino stock are genetically very close to each other. The incidence and latent period of mammary tumor development in SHN mice were higher and shorter than those in SLN. To elucidate these differences in the behavior of mammary tumorigenesis, the frequency of insertion of mammary tumor viral genes within the int-1 and int-2 regions in spontaneous mammary tumors from their two substrains, SHN/S and SLN/S, were compared. The frequency of provirus integration into either int-1 or int-2 in DNAs from mammary tumors was 52% (11/21) in SHN/S and 45% (5/11) in SLN/S. The frequency of insertion within int-1 or int-2 could not account for the different susceptibilities of SHN/S and SLN/S.

Normal human salivary gland cells produce carcinoembryonic antigen-related antigen in collagen gels

Dear Editor: Recent studies on the functional characteristics of carcinoembryonic antigen (CEA), an antigen with oncofetal characteristics, have indicated that CEA is an intercellular adhesion molecule (1) which may control the functional activity of cellular collagen receptors (2). In this study we evaluated normal human salivary gland cells that were embedded in collagen gels cultured with serum free medium supplemented with mouse epidermal growth factor. The level of carcinoembryonic antigen-related antigens (CRAs) during culture was also determined. CRAs were detected by anti-human CEA monoclonal antibody that did not cross-react with nonspecific cross-reacting antigen (NCA) (3). Western blot analysis of the culture cells with this antibody demonstrated bands at approximately 140-180, 90, 70 and 30 kDa (data not shown). From the molecular weight, these bands were considered to correspond to CEA and normal fecal antigens (NFAs) (4). We are now performing a more detailed study on CRAs of normal human salivary gland cells. Normal cells were derived from fresh normal tissues of human submandibular (SGN-1, -2) and parotid glands (PGN-1, -2, -3) obtained at operation. Clusters containing about 30-50 cells from collagenase dissociated tissue were used for collagen gel culture. The culture medium consisted of Dulbeccos modified Eagles medium and Hams F12 in a 1:1 mixture. The medium was supplemented with insulin at 10 #g/ml, bovine serum albumin at 5 mg/ml, mouse epidermal growth factor at 5 ng/ml. Embedded cell culture within a collagen gel was performed as described previously (5). Within a few days of culture, normal submandibular gland cells produced branch-like outgrowths extending into the collagen gel matrix (Fig. 1). After about 5 days, the collagen gel containing the cells gradually detached from the plastic phase to become a floating collagen gel. The gel shrank from 3 cm 3 of primary volume to about 0.02 cm ~ during 3 weeks of culture. Histological sections of the gel

MORPHOLOGICAL STUDY OF NORMAL AND TUMOR CELLS IN THE HEAD AND NECK REGION USING IRRADIATED COLLAGEN GEL EMBEDDING TECHNIQUE: MORPHOLOGICAL STUDY USING COLLAGEN GEL

Fourteen specimens of normal, benign, or malignant tumor cells from the head and neck region were subjected for culture using 0.2% of irradiated collagen gel embedding technique. Re-differentiation of glands within the gel in serum-free medium was observed. There were marked differences in the growth patterns within the gel between normal or benign and malignant cells. Four normal glands and 5 out of 6 benign tumors showed branch-like growth patterns. On the other hand, malignant tumors showed no branch-like pattern, or could not grow at all. The results showed that the collagen gel technique could be useful for the differential diagnosis of malignancy in head and neck tumors.