Genetic aspects of tumour metastasis.

The application of techniques for the transfer of genetic material between cells when used with recombinant DNA technologies and gene cloning procedures hold exciting possibilities for the genetic analysis of the metastatic behaviour of tumour cells. So far these genetic studies have produced confusing results, but some of the problems have been defined more clearly. For a more complete understanding of the metastatic phenotype a more critical analysis of the models used to mimic the process and a greater appreciation of the deficiencies in the techniques applied to study it are necessary.

BioBreeding/Worcester (BB/Wor) rats are deficient in the generation of functional cytotoxic T cells.

The BioBreeding/Worcester (BB/Wor) rat provides a good model of spontaneous autoimmune diabetes. There are several sublines of the BB/Wor rat. The diabetes prone (DP) sublines develop diabetes at a frequency of 50 to 80% from 60 to 120 days of age. The DP rats are lymphopenic, have a severe deficit in phenotypic OX 19+ OX 8+ cytotoxic T cells (Tc), and lack RT 6.1 T cells. These rats have a relative increase in OX 19- OX 8+ natural killer (NK) cells and in NK activity as compared with the diabetes resistant (DR) sublines. The DR sublines have a normal complement of phenotypic Tc and RT 6.1 T cells, fewer NK cells, and lower NK activity than the DP rat. The ability to elicit functional Tc in the BB/Wor rat has not been well studied. In these experiments, by using a model of lymphocytic choriomeningitis virus (LCMV) infection in DP and DR rats, we have studied the functional activity of Tc in these lines. Seven days after infection with LCMV, DR rats develop lymphocytes which are cytotoxic for LCMV-infected syngeneic fibroblasts. These cytotoxic lymphocytes are phenotypic Tc (OX 19+ OX 8+), and do not kill Pichinde virus-infected syngeneic fibroblasts or LCMV-infected allogeneic fibroblasts. This cytotoxic activity is accompanied by an increase in phenotypic Tc from 17 to 33%. DP rats produced neither functional nor phenotypic Tc. These studies confirm that NK cells are the predominant cytotoxic lymphocyte in the BB/Wor rat and suggest that these rats may not utilize a Tc mechanism in islet destruction or another immunologic process such as graft rejection.

Molecular analysis of the murine lupus-associated anti-self response: involvement of a large number of heavy and light chain variable region genes.

The mRNA encoding heavy and light chains of a hybridoma-derived monoclonal IgM, kappa anti-immunoglobulin (rheumatoid factor) and an IgG3, kappa anti-histone autoantibody from systemic lupus erythematosus and arthritis-prone MRL/Mp-lpr/lpr mice have been molecularly cloned, and the nucleotide sequences corresponding to their variable regions have been determined. To investigate whether autoantibodies with specificities frequently observed in lupus disease might share common structural components, the sequences obtained in this study have been compared with those of a monoclonal MRL/Mp-lpr/lpr IgM, kappa anti-DNA autoantibody previously analyzed in our laboratory (J. Exp. Med. 1985. 161: 805). The 3 immunoglobulins employed different heavy chain variable region (VH) genes belonging to the large J588 VH gene family, kappa light chain variable region (V kappa) genes from 3 different V kappa groups, and different diversity and joining segments. Our findings suggest that murine lupus-associated autoantibodies of different specificities do not have genetic components in common to signal their self-reactive nature and are encoded by a large number of immunoglobulin gene elements.

Immunogenetics of collagen-induced arthritis.

CIA is a unique experimental model of disease, which may be particularly relevant to the study of RA, since it represents a true autoimmune reaction against a major joint component. Since it represents a model of the association of MHC genes and disease, it has a broad relevance to the study of disease, particularly the contribution of class II MHC genes to autoimmunity. This disease model has been demonstrated in three species, and a marked influence of immunogenetic control is apparent. Native type II collagen is a large protein with a genetically conserved structure, giving rise to a number of cross-reactive antigen epitopes between species. Susceptibility to CIA correlates with an Ir directed against certain epitopes, and this autoimmune response appears to be controlled by MHC-linked genes. In the mouse, susceptibility has been specifically associated with variations in the I-A beta chain. The immunogenetic control may be effected through the T-cell recognition of the collagen molecule, resulting in antibodies and reactive cells with a specificity for native autologous type II collagen. Although the specificity of the anti-type II collagen response is the critical element in the generation of the autoimmune reaction, production of a high level of antibody is important, as well as a complement-fixing isotype. Cell-mediated immunity to type II collagen contributes to the disease chronicity, either through the generation of inappropriate T help, or a suppressor defect, and the production of factors. However, the precise sequence of events, from antigen recognition to joint damage, has yet to be elucidated fully. Some aspects of this experimental model bear a striking resemblence to RA. Hopefully, this experimental model will provide valuable insights into the significance of collagen autoimmunity and the contribution of class II genes to the pathology of arthritic disease.

Ceroid lipofuscinosis in sheep. I. Bis(monoacylglycero)phosphate, dolichol, ubiquinone, phospholipids, fatty acids, and fluorescence in liver lipopigment lipids.

The ceroid lipofuscinoses are inherited lysosomal diseases of children characterized by a fluorescent lipopigment stored in a variety of tissues. Defects in lipid metabolism or the control of lipid peroxidation have been postulated to explain their pathogenesis. In the present study, lipopigment was isolated from the liver of sheep affected with ceroid lipofuscinosis. It was 70% protein, the rest being mainly lipids. These were only one-sixth as fluorescent as total liver lipids, but contained a number of fluorophors. None were major components of the lipopigment or the postulated fluorescent product of lipid peroxidation. Lipopigment lipids included the lysosomal marker bis(monoacylglycero)phosphate that contained 42.9% linoleate and 16.5% linolenate. Lipopigment neutral lipids were dolichol, dolichyl esters, ubiquinone, free fatty acids, and cholesterol, indicative of a lysosomal origin of the lipopigment. Phosphatidylcholine, phosphatidylinositol, phosphatidylserine, and phosphatidylethanolamine were present in proportions and with fatty acid profiles typical of lysosomes. No differences were found between the lipids of total control and affected livers, nor the fatty acid profiles of their phosphatidylcholine, phosphatidylethanolamine, or triglycerides. It is concluded that ovine ceroid lipofuscinosis is not a lipidosis, nor does the lipopigment arise from the abnormal peroxidation of lipids. Strong similarities between the lipopigment and the age pigment lipofuscin were noted.

Ceroid lipofuscinosis in sheep. II. The major component of the lipopigment in liver, kidney, pancreas, and brain is low molecular weight protein.

Previous studies on the lipopigment from the livers of sheep affected with ceroid lipofuscinosis showed that the disease does not involve a defect in lipid metabolism or abnormal lipid peroxidation and that most of the lipopigment was proteinaceous. In this study, lipopigment was isolated from liver, kidney, pancreas, and brain of affected sheep without the use of proteolytic enzymes. Lipopigment from all tissues was two-thirds protein. Modified silver staining after sodium dodecyl sulfate-polyacrylamide gel electrophoresis showed a major band of Mr = 14,800, heterogeneous material between Mr = 5,000 and 9,000, and a major band of Mr = 3,500. These compounds did not stain for RNA or carbohydrate and were digested by a nuclease-free protease as expected for protein. They are not normal lysosomal proteins. Lipopigment levels of dolichol, ubiquinone, and cholesterol were consistent with the lipopigment being protein-enriched lysosome-derived cytosomes. The presence of the Mr = 3,500 proteins in whole affected tissue homogenates distinguished them from homogenates of normal tissues. It was concluded that low Mr proteins are specifically stored in ovine ceroid lipofuscinosis and that the ceroid lipofuscinoses may result from inherited defects in lysosomal protein catabolism.

Origins of pathogenic anti-DNA idiotypes in the NZB X SWR model of lupus nephritis.

The investigations with the NZB X SWR model show that the development of systemic autoimmune disease is a multistep, multigene process. Severe lupus nephritis in the NZB X SWR hybrids results from the interaction of genes inherited from both the autoimmune NZB and the normal SWR parents. A similar genetic interaction occurs in the NZB X NZW hybrids, but in this model, both the parental strains are abnormal and the nature of the gene products or their mechanism of action is unknown. In the NZB X SWR model, we have been able to identify a restricted subpopulation of nephritogenic anti-DNA antibody idiotypes that are encoded by genes of the normal SWR parents. Thus, these are one set of genes that determine the development of severe lupus nephritis in the F1 hybrids. In addition, another set of genes allows for the expansion of B cells that produce such pathogenic anti-DNA idiotypes in the F1 hybrids since such B-cell clones remain dormant in the normal SWR parents. The latter category of genes, presumably specifying defects in immunoregulation, are probably inherited from the NZB parents or may be the result of complementation of genes inherited from both parents. Further investigations with the NZB X SWR model will help us define the immunoregulatory defects in SLE that are specific for the T and B cells involved in pathogenic autoantibody production.

Studies on an animal model of alcoholism.

Past and ongoing studies indicate that the selectively bred P line of rats satisfies virtually all the suggested criteria for an animal model of alcoholism. They attain pharmacologically active levels of BAC and develop tolerance and physical dependence with voluntary oral ethanol ingestion, while in the free-feeding state. Ethanol is positively reinforcing to the P rats and consumption appears to be directed by the post-ingestive, pharmacological effects of ethanol, as revealed by the intragastric self-administration studies. Some interesting differences between the P and the NP lines have been uncovered. They differ in the content of serotonin in several brain regions and they respond differently to ethanol. The P rats develop acute tolerance to sedative-hypnotic doses of ethanol more rapidly than do the NP rats, and they exhibit stimulation with low doses of ethanol. These differences suggest hypotheses on mechanisms underlying alcohol-seeking behavior which can now be tested experimentally. It should be emphasized, however, that the described findings are the product of but a single genetic experiment. Clearly, replication is needed, and we are currently doing this, using a better defined, heterogeneous stock of rats. This one experiment, however, has demonstrated the feasibility of developing animal models of alcoholism and offers hope that the genetic and biological basis of alcohol-seeking behavior can be explored in the laboratory. The screening and testing of pharmacological agents able to deter alcohol-seeking behavior is an obvious practical application of this model.

Structural comparisons between mouse and human prealbumin.

In an attempt to construct model systems for familial amyloidotic polyneuropathy, prealbumin cDNA was cloned from a mouse liver cDNA library, using previously cloned human prealbumin cDNA as a hybridization probe. The primary structure of mouse prealbumin deduced from the cDNA sequence shows that it consists of 147 amino acids, including a whole prealbumin sequence (127 amino acids) and a putative signal sequence (20 amino acids). These numbers are in complete agreement with those determined for the human prealbumin. Among the 127 amino acid residues of the mature human prealbumin, 25 are replaced by different amino acids in the mouse prealbumin. Interestingly, 24 out of the 25 substituted amino acids are located at the outer surface of the protein, and the regions corresponding to the core and central channel of the protein are almost completely conserved. The cloned cDNA provided essential information for manipulating amyloidosis in mice.

Mast cells and their degranulation in the Tsk mouse model of scleroderma.

The Tsk mouse is a genetically transmitted example of cutaneous fibrosis which has been compared with human scleroderma. During a systematic histopathological study of the Tsk mouse, both an increased number and an increased proportion of degranulated mast cells were observed. The consistent association of mast cells and fibrosis in scleroderma, graft-vs-host reactions (GVHR), and now the Tsk mouse raises the question of a pathogenetic role for mast cells in fibrotic disorders in general.

In vivo prevention of thyroid and pancreatic autoimmunity in the BB rat by antibody to class II major histocompatibility complex gene products.

Evidence is accumulating that the development of insulin-dependent diabetes mellitus involves autoimmune phenomena, both in the human and in the BB rat model. A strong association is observed in both cases with alleles of the class II major histocompatibility complex (MHC). Results of the present study show that autoimmune phenomena, as assessed by the presence of clinical diabetes or histological thyroiditis, are prevented by the injection of monoclonal antibodies to class II gene products in the BB rat. Immunosuppression was specifically obtained with a monoclonal antibody to the murine I-E equivalent, as opposed to the murine I-A equivalent, of the rat major histocompatibility complex. This represents indirect evidence for I-E subregion control of immune responses to islet cell and thyroid antigens in the BB rat model. The frequent occurrence of anaphylactic type deaths in young (1 month old) animals receiving more than six weekly injections of partially purified homologous (rat) monoclonal antibodies to rat class II gene products underscores the potential risks of this type of immunotherapy. The presumed immunologic mechanism (IgE antibody) and its specificity (anti-allotype, anti-idiotype, or anti-impurity) must be clarified to assess the risks and feasibility of this type of therapy.