Optimization of plasmid vectors for high-level expression in lung epithelial cells

Nonviral gene therapy approaches use a plasmid vector to express the desired transgene. We have systematically examined several regulatory elements within plasmid vectors that govern gene expression, e.g., the promoter, enhancer, intron, and polyadenylation signal, by constructing a series of plasmids that differed only in the particular sequence element being evaluated. Of the several promoters and polyadenylation signal sequences that were tested, the human cytomegalovirus (CMV) immediate early gene promoter and the addition of polyadenylation signal sequences from the bovine growth hormone (BGH) gene or rabbit beta-globin gene produced the highest levels of expression in vitro. The inclusion of a hybrid intron 3 to the promoter further increased expression 1.6-fold. The addition of a region of the CMV enhancer 5' to several weak promoters increased expression 8- to 67-fold, and co-transfection with a second plasmid encoding a chimeric transcription factor also enhanced expression. On the basis of these results, the CMV promoter, the hybrid intron, and the BGH polyadenylation signal were selected for consistent high level expression in vitro and in the mouse lung. However, expression was transient, with greater than 60% loss of activity in the first 7 days. This transient expression was not specific to CMV promoter-containing plasmids, because plasmids containing other heterologous promoters showed a similar profile of transient expression in vivo. These comparative analyses begin to provide a basis for the development of optimized expression plasmids for gene therapy of lung diseases.

Mapping of the X-breakpoint involved in a balanced X;12 translocation in a female with mild mental retardation

Balanced chromosomal abnormalities such as translocations and inversions have been identified in many genetic diseases. Cloning of the breakpoints involved in these abnormalities has led to the identification of the disease-related genes. Recent reports suggest the presence of a mental retardation locus at Xq11-12. We have identified a female patient with a balanced translocation t (X;12) (q11;q15) associated with mild mental retardation. We identified a yeast artificial chromosome spanning the X-chromosome breakpoint by using fluorescent in situ hybridization techniques. A cosmid library of this YAC has been constructed and the search for candidate genes is in progress.

Prp (proline-rich protein) genes linked to markers Es-12 (esterase-12), Ea-10 (erythrocyte alloantigen), and loci on distal mouse chromosome 6

The closely linked proline-rich protein (Prp) genes, coding for abundant salivary proteins, are located on distal mouse chromosome 6. They are part of a conserved linkage group that is represented on human chromosome 12p. Two other markers, Ea-10 and Es-12, that were previously unassigned to a chromosome are closely linked to Prp genes in the mouse.

Recombination between kappa chain genetic markers and the Lyt-3 locus

Recombination has been detected for the first time between chromosome 6 loci controlling kappa chain expression in normal mouse serum immunoglobulin and the Lyt-3 locus. The recombination event occurred at the 26th or 27th backcross generation during the derivation of the Lyt-2a, Lyt-3a-congenic line B6.PL(85NS). The line is now homozygous for the Lyt-2a, Lyt-3a allele(s) at N30F13 and homozygous animals express the Igk-Ef1b allele derived from C57BL/6. The frequency of recombination has been estimated to be 0.30% based on the present results and previous studies in which no recombination was detected. The results rule out the hypothesis that the Lyt-3 locus itself controls the light chain phenotype observed in normal serum immunoglobulin.

Molecular heterogeneity of D-end products detected by anti-H-2.28 sera. III. Reactivity of certain anti-H-2.28 alloantisera with Qa-2 antigen

In capping experiments with peripheral T lymphocytes, two anti-H-2.28 sera (AKR anti-AKR.L, anti-Kb, and C3H anti-C3H.B10, k anti-b) that do not contain any Qa-2-specific antibodies are able to redistribute not only the H-2.28-positive H-2 molecules, but also Qa-2 molecules. This is due to the capacity of these sera to react with Qa-2 molecules because on cells where all known molecules of the H-2d haplotype were capped (K1d, K2d, Dd, Md, Ld, L2d), both antisera still reacted when the cells came from a Qa-2 positive Dd strain (B10.A) but not when the cells were of Qa-2 negative strain (BALB/cByA). The reaction with Ia and non-H-2 antigens was excluded in these experiments. These data show that Qa-2 and H-2 antigens share some specificities of the H-2.28 family. Other anti-private and anti-public anti-H-2 sera failed to react with the Qa-2 molecules.

Genetic and endocrine control of renin activity in the submaxillary gland of the mouse

Basal activity of submaxillary gland (SMG) renin is high in female mice that carry the Rnrs allele and is induced to higher levels by treatment with dihydrotestosterone (DHT). To determine whether the difference in basal activity between high (Rnrs/Rnrs) and low (Rnrb/Rnrb) strains is due to enhanced sensitivity of Rnrs/Rnrs strains to endogenous androgen, we first studied the effect of several types of endocrine ablation on SMG renin in young female mice, and second, we removed normal androgen receptor protein by introducing the X-linked Tfm gene. Adrenalectomy with or without castration had no effect on basal SMG renin; hypophysectomy decreased basal renin activity 400-fold but did not abolish responsiveness to DHT. Loss of androgen receptor did not affect basal renin activity but did prevent enhancement by DHT. Basal and induced renin activities in L.AKR(Alll)/Cy, a congenic strain homozygous for Rnrs introduced from AKR/J into the background of C57L/J, an Rnrb/Rnrb type strain, are intermediate between levels observed in the original strains. We conclude that (1) the basal level of SMG renin is regulated directly or indirectly by some pituitary hormone(s) but not by androgen, (2) androgen induction of renin activity requires a normal androgen receptor, and (3) major gene(s) that regulate basal as well as induced SMG renin are in a circumscribed region of chromosome 1.

Induction in C57BL/KsJ mice of complement-dependent antibody cytotoxic to cultured beta cells

Despite widespread evidence that autoimmune mechanisms may contribute to the beta cell necrosis associated with type I insulin-dependent diabetes mellitus (IDDM), it has not heretofore been demonstrated that islet cell antibodies (ICAs), directed primarily against cytoplasmic antigens, are capable of specific lysis of beta cells. We utilized a readily accessible source of mouse pancreatic islets [CBA/J mice lacking exocrine pancreas (exocrine pancreatic insufficiency syndrome)] to experimentally induce ICAs inbred mice. Homogenates of these islets were injected weekly for four weeks into syngeneic (CBA/J) and allogeneic (A/J, C57BL/6J, C57BL/KsJ) recipients. Only C57BL/KsJ inbred mice showed the induction of a high titer (greater than or equal to 160) antiserum cytotoxic to 51Cr-labeled CBA/J lymph node target cells. Neither the immunized C57BL/sJ mice with circulating ICAs nor any of the other immunized strains showed any decrease in glucose tolerance as compared with vehicle controls. Moreover, no morphologic evidence of islet necrosis or atrophy was apparent. Thus the ICAs induced were reactive with alloantigenic determinants of the donor and unreactive with antigenic determinants of the recipient strain. The C57BL/KsJ antiserum was further screened for anti-islet cell cytotoxic activity using both a 51Cr release assay from CBA/J islet cell monolayer cultures, and immunocytochemical staining of sections of Bouin's fixed, paraffin-embedded pancreas. This antiserum was cytotoxic to CBA/J beta cells in monolayer culture, but not the other non-beta islet cell types. Immune lysis of the beta cell required rabbit complement. At a concentration of 1% antiserum and 4% complement, beta cell lysis was evident between 3 and 4 h at 37 degrees C. Ultrastructural examination of beta cells exhibiting cytopathic changes revealed cytoplasmic disarray rather than any obvious lytic events at the plasma membrane. Grossly distended, rough, endoplasmic reticulum containing intracisternal type A retrovirus was the most prominent feature distinguishing antiserum and control serum-treated beta cells. This model system suggests that ICAs which recognize beta cell cytoplasmic antigens are capable of specifically lysing beta cells via a complement-dependent mechanism. Immunocytochemical staining revealed that, in addition to islet beta-cells,, the antiserum (1/500 dilution) stained a macrophage-like cell in the spleen and lymph nodes, as well as an epithelial-like cell in the thymus. The possibility is discussed that this multiple specificity may have been due to passenger leukocytes present in the islet homogenates used to immunize.

Leukemogenesis, immune responsiveness, and murine leukemia virus expression in congenic AKR/J mice differing at H-2

In this study we examined the leukemia incidence, ectropic and xenotropic murine leukemia virus expression, and immune responsiveness of congenic AKR/J mice differing at the H locus. Congenic AKR.L-H-2b/1 mice, bearing the H-2b haplotype derived from C57L/J, were found to have a significant delay in time of death due to leukemia relative to that of AKR/J (H-2k) mice. The expression of ecotropic murine leukemia virus was found to be identical in both strains. The expression of xenotropic murine leukemia virus did vary, however, with the AKR.L-H-2b/1 mice showing a significantly reduced level of virus expression relative to AKR/J mice. In addition to these observations, we found that the AKR.L-H-2b/1 mice have an enhanced blastogenic responsiveness to phytohemagglutinin and to specific antigen to which they had previously been sensitized. Concomitant enhanced antibody response was not found. We suggest that the stronger cellular response, relative to AKR/J, may contribute to the delay in leukemia onset and to reduced xenotropic virus expression observed with the congenic mice.

Murine lymphocyte alloantigens. I. The Ly-6 locus

The Ly-6 locus and the Ly-6.2 specificity have been previously described, and we now further define this locus and the allelic specificities, Ly-6.1 and Ly-6.2. Back-cross studies and examination of several recombinant inbred (RI) lines demonstrate that Ly-6 is distinct from other loci determining CMAD, except for ALA-1 and evidence is presented for the identity of Ly-6 and ALA-1. The Ly-6 congenic strain C3H.B6-Ly-6b is described and was used to prepare antisera in combinations congenic for Ly-6. The allelism of the Ly-6.1 and Ly-6.2 specificities was confirmed by the reactivity of these antisera with a segregating F2 generation. Antisera can be prepared between the Ly-6 congenic strains, although the magnitude of the response is under the control of one or more genes in A strain, not linked to H-2 or Ly-4 loci. Absorption analysis using separated T and B lymphocyte populations demonstrate that Ly-6.2 does not have exclusive peripheral T cell representation as originally reported, but is also present on B cells in amounts one-fourth to one-eighth of that found on T cells and is found in greater quantities on T blast cells.

Murine lymphocyte alloantigens. I. The Ly-6 locus

The Ly-6 locus and the Ly-6.2 specificity have been previously described, and we now further define this locus and the allelic specificities, Ly-6.1 and Ly-6.2. Back-cross studies and examination of several recombinant inbred (RI) lines demonstrate that Ly-6 is distinct from other loci determining CMAD, except for ALA-1 and evidence is presented for the identity of Ly-6 and ALA-1. The Ly-6 congenic strain C3H.B6-Ly-6b is described and was used to prepare antisera in combinations congenic for Ly-6. The allelism of the Ly-6.1 and Ly-6.2 specificities was confirmed by the reactivity of these antisera with a segregating F2 generation. Antisera can be prepared between the Ly-6 congenic strains, although the magnitude of the response is under the control of one or more genes in A strain, not linked to H-2 or Ly-4 loci. Absorption analysis using separated T and B lymphocyte populations demonstrate that Ly-6.2 does not have exclusive peripheral T cell representation as originally reported, but is also present on B cells in amounts one-fourth to one-eighth of that found on T cells and is found in greater quantities on T blast cells.

Effect of rabbit strain on activity level and cytotoxicity of serum complement. III. Comparison of four tumor target cells

Serum samples from about 10 males and 10 females from each of 15 genetically defined strains of rabbits and from one hybrid were tested as sources of complement for the microtiter lymphocytotoxicity test using tumor cells from SaI-A, 6C3HED-A, BW5147, and L-cells as target cells. The tumor cells were tested with appropriate H-2 antisera. Results indicated strain differences particularly evident when the target cells were eithter 6C3HED-A or L-cells. BW5147 worked well with most all strains and SaI-A was extremely demanding of all strains of rabbits. However, strains IIIC/J, IIIVO/J and the F1 hybrid between them were still among the six best strains when tested against the tumor cells as they were when tested against the lymph nodes in previously published data. Strain WH/J which carries the gene, ha, for hereditary lymphosarcoma, proved to be a good complement source with SaI-A and 6C3HED-A, both sarcoma in nature. However, it was not a good complement source with BW5147, a lymphatic leukemia, or L-cells. This is possible evidence for a common tumor antigen associated with lymphosarcoma.

Effect of rabbit strain on activity level and cytotoxicity of serum complement. II. Comparison of five murine target cells

Serum samples from about 10 males and 10 females from each of 15 genetically defined strains of rabbits and from one hybrid were tested as sources of complement for the microtiter lymphocytotoxicity test using as target cells lymph node cells from C3H/HeJ, B10/Sn, BALB/cJ, and DBA/2J strain mice. The lynph node cells were tested with appropriate H-2 alloantisera. Results indicated marked strain differences. Correlation analysis of these data with data using primarily aliquots from the same serum samples tested against lymph node cells from B10.A/Sn mice showed clearly that a genetically defined population of rabbits that provided serum of high quality for one of these test systems would in general work reasonably well in any one of the other four. The correlation coefficients in all possible combinations ranged from +0.49 to +0.92. Of the five strains tested, the results obtained from BALB/cJ appear to be the best predictors of the results in the other four strains (r values were +0.74, +0.57, "0.76, and +0.92). Studies are in progress to test rabbit serum complement samples against the more demanding tumor cells.

Effect of rabbit strain on activity level and cytotoxicity of serum complement

Serum samples from about 10 males and 10 females from each of 16 genetically defined strains of rabbits and from one hybrid were tested as sources of complement for the microtiter lymphocytotoxicity test. Lymph nodes from mice of strain B10.A/Sn were used as the target cells with a strong H-2 alloantiserum. Results indicate marked strain differences. Rabbit serum from a hybrid between the two best strains IIIC/J and IIIVO/J was almost as satisfactory as the parental strains and better than all other strains tested in this system, indicating a genetic basis for the response. The results of testing aliquots of these serum samples against other test systems including human cell lines will have to be in hand before a general statement can be made, but preliminary unpublished data against lymph nodes from A/HeJ and C3H/HeJ strain mice and tumor cells from an ascites form of a sarcoma Sal-A are encouraging and suggest a consistent superiority of the two strains and the hybrid between them referred to above.

Autosomal phosphoglycerate kinase linked to mouse major histocompatibility complex

The mouse autosomal locus that determines the form of phosphoglycerate kinase found only in testes is shown here to be closely linked to but not included within the major histocompatibility complex on chromosome 17. Data are presented that strongly favor the location of this locus, designated Pgk-2, distal to H-2, Qa-1, and Qa-2, and closely associated with T1a. The Pgk-2 strain distribution pattern for 103 inbred and congenic strains of mice is given. Because Pgk-2 is polymorphic among inbred strains, it should be of value in linkage studies.

Host-gene control of type-C RNA tumor virus expression and tumorigenesis in inbred mice

We analyzed the relationship of genetic factors determining the expression of endogenous type-C RNA tumor viruses and other host-gene markers to tumorigenesis. A hybridization experiment was performed with mice of strains AKR/J and C57L, the first filial (F(1)) generation hybrids, the second filial (F(2)) generation hybrids, and the backcrosses to the two parental strains. The results demonstrated a highly significant and predictable association between the expression of complete infectious virus or the viral group-specific (gs) antigen in spleens of young mice and tumorigenesis later in life. Most of the tumors were thymic leukemia and reticulum sarcoma, but other mesenchymal, as well as epithelial, tumors were also observed. Tumors occurred preferentially in gs-antigen- or virus-positive mice of all crosses; in the C57L-backcross and F(2) mice segregating for gs-antigen and virus expression, a few gs-antigen-negative mice developed reticulum cell sarcomas. At the time of their occurrence, the mice were all gs-antigen-positive, and most had virus as well.A minor effect of the major histocompatibility locus, H-2, on leukemogenesis was found in the F(2) mice. Several tumor types were also found that we have never observed in the two parental strains. Our data provide the most direct biological evidence in favor of the viral oncogene theory. Thus, from the presence or absence of expression in early life of splenic gs antigen or virus, we can predict whether or not a tumor is likely to develop later in life. These findings suggest that the genome of endogenous type-C RNA viruses is the major determinant for tumorigenesis although they provide no clues about the factors responsible for the various histological types.

Ly-4, a new locus determining a lymphocyte cell-surface alloantigen in mice

A new locus is described that determines an alloantigen on the surface of lymphocytes. It differs from loci previously described in that the corresponding antibody, at least in most antisera, reacts almost exclusively with node lymphocytes, and weakly or not at all with thymuslymphocytes. Positive strains include C57BL/6, C57BL/10, C57L, C57BR/cd, and RF; negative strains include BALB/c, C3H, and SJL. The symbol Ly-4 is assigned, with the C57BL/6 allele being Ly-4(b), and the BALB/c allele Ly-4(a).

The molecular basis of codominant expression of the histocompatibility-2 genetic region

H-2 alloantigens in an intact, undegraded form can be solubilized by the nonionic detergent NP-40 and isolated by indirect immunoprecipitation and electrophoresis in sodium dodecyl sulfate on polyacrylamide gels. With the use of an immunological precipitation method, the question about the number of cellular H-2 gene products in heterozygous cells has been partially resolved. At least four separable gene products were detected in cells heterozygous at the H-2 genetic region-one for each of the H-2D genes of the two parental haplotypes, and one for each of the H-2K genes of the two parental haplotypes. This is in contrast to the two H-2 gene products reported previously for cells homozygous at the H-2 region. The findings establish that the two parental H-2 haplotypes on homologous chromosomes in heterozygous cells are ultimately expressed as different glycoprotein molecules.