Chromosomal localization of human leukocyte, fibroblast, and immune interferon genes by means of in situ hybridization

Recombinant plasmids containing cDNA inserts from human leukocyte interferon (IFN-α), fibroblast interferon (IFN-β), or immune interferon (INF-γ) genes were radiolabeled and hybridized in situ to human metaphase chromosome preparations. The results localized the human IFN-α and IFN-β genes to the short arm of chromosome 9(p21→pter) and localized the IFN-γ gene to the long arm of chromosome 12(q24.1).     

Two antigenically distinct species of human interferon.

Rabbit antisera prepared against interferon produced in human fibroblast cell cultures stimulated with poly(1).poly(C) neutralized the activity of interferon preparations produced in various human fibroblast cultures timulated either with poly(1)poly)C) or with viruses. However, these antisera showed no detectable neutralizing activity against interferon produced in cultures of human leukocytes. On the other hand, most rabbit antisera against the human leukocyte interferon were active in neutralizing both homologous interferon and fibroblast interferons. A preparation of antiserum against leukocyte interferon, active against both leukocyte and fibroblast interferons, was shown by affinity chromatography to have two distinct antibody populations, one of which was specific for the fibroblast interferon. We conclude that the heterologous neutralizing activity of sera from rabbits immunized with leukocyte interferon is liekly to be due to the presence of two antigenic species of interferon. The major antigenic species of leukocyte interferon preparations (designated "Le") is distinct from huamn fibroblast interferon. The minor species of leukocyte interferon ("F") is either identical with, or closely related to, interferon produced in human fibroblast cultures.     

Spontaneous production of human interferon.

Several established lines of human lymphoblastoid cells were evaluated for abilities to produce interferons. Some cell lines were able to produce interferon when induced with either Newcastle disease virus or Sendai virus, whereas others failed to produce detectable interferon when so induced. However, several cell lines were able to spontaneously produce interferon without induction. Spontaneously produced interferon was liberated by cells only during logarithmic growth phase, reaching levels ranging from about 10 reference units/ml of growth medium for some cell lines to 1000 reference units/ml for others. The interferons produced by induced lymphoblastoid cells and the spontaneously produced interferons were all characterized as type I human leukocyte interferon by high levels of cross-species antiviral activities on bovine cells and by neutralizations by antiserum to human leukocyte interferon but not by antiserum to human fibroblast interferon. However, analysis by electrophoresis in sodium dodecyl sulfate/polyacrylamide gels revealed that spontaneously produced interferon was less size heterogeneous than human leukocyte interferon, migrating as a single band of activity with a peak at 20,000 daltons, whereas human leukocyte interferon contained peaks of major activity at 23,000 and 18,000 daltons and virus-induced Namalva lymphoblastoid cell interferon migrated predominantly as the 18,000-dalton form. Also, although neither virus-induced primary leukocyte interferon nor any of the virus-induced lymphoblastoid cell interferons were neutralized by antiserum to mouse interferon, all of the spontaneously produced interferons were neutralized by antiserum to mouse interferons. These data suggest significant structural similarities between the active cores of certain interferons from phylogenetically diverse animal species.     

Microcell-mediated transfer of a single human chromosome complements xeroderma pigmentosum group A fibroblasts.

Chromosomes from an immortalized aneuploid human fibroblast cell line were randomly tagged with the selectable marker neo by transfection with the plasmid pSV2neo. Somatic cell fusions between transfected human cells and mouse A9 cells generated pools of G418-resistant human-mouse hybrid clones containing various numbers of human chromosomes. Microcell-mediated chromosome transfer from the hybrid pools to xeroderma pigmentosum complementation group A (XP-A) cells in culture and selection for G418-resistant colonies resulted in the identification of XP cells with enhanced resistance to ultraviolet radiation. Screening of subclones from selected pools of human-mouse hybrids facilitated the identification of hybrids containing a single neo-tagged human chromosome. Transfer of this chromosome to XP-A cells (but not to XP-F or XP-C cells) results in enhanced resistance to ultraviolet light and enhanced excision repair capacity. The identification of a single human chromosome that complements the phenotype of XP-A cells in culture provides the potential for genetic mapping of the complementing gene and for its isolation by molecular cloning.     

Human leukocyte interferon: structural and biological relatedness to adrenocorticotropic hormone and endorphins.

Anti-alpha-corticotropin [anti-ACTH alpha (1-13)](also alpha-melanotropin) and anti-gamma-endorphin antisera neutralized human leukocyte interferon activity but not fibroblast interferon activity. Human leukocyte interferon was not neutralized by anti-human lutenizing hormone (lutropin) or follicle-stimulating hormone (follitropin) antisra. Conversely, antisera to human leukocyte interferon neutralized ACTH activity. The neturalization of human leukocyte interferon by anti-human leukocyte interferon serum was partially blocked by ACTH. These studies show strong antigenic relatedness among human leukocyte interferon, ACTH, and endorphins, implying that there are underlying structural similarities. Structural relatedness is shown by pepsin cleavage of ACTH activity from human leukocyte interferon. The implication for the natural functions of human leukocyte interferon are discussed.     

Expression of human and mouse nonhistone chromosomal proteins in hybrid mouse erythroleukemia cells containing a single human chromosome.

The nonhistone chromosomal proteins of a series of hybrid mouse erythroleukemia cell lines containing human chromosome 16 were investigated by two-dimensional gel electrophoresis to determine if such cells contained nonhistone chromosomal proteins of both human and mouse origin. Comparison of the two-dimensional gel electrophoretograms of the nonhistone chromosomal proteins of mouse and human cell lines showed 400 and 280 chromosomal proteins, respectively, of which about 75% were electrophoretically identical. The two-dimensional gel electrophoretogram of a cloned hybrid mouse erythroleukemia cell line that retained a tetraploid complement of mouse chromosomes and human chromosome 16 (as the only human chromosome) displayed a nonhistone chromosomal protein of pI 6.2 and Mr 65,000. This protein, which comigrates with a nonhistone chromosomal protein present in the human cell line used to produce this hybrid cell and which is also present in two additional human cells lines studied, could not be detected in the mouse erythroleukemia parent before fusion. This polypeptide also was shown by similar techniques to be associated with the presence of human chromosome 16 in four out of five other independently derived hybrid mouse erythroleukemia cell lines that contained a near tetraploid complement of mouse erythroleukemia chromosomes.     

Serial transfer of a human gene to rodent cells by sequential chromosome-mediated gene transfer.

The human hypoxanthine phosphoribosyl-transferase (IMP:pyrophosphate phosphoribosyltransferase, EC 2.4.2.8) gene (hprt) has been serially transferred to mouse cells and then to Chinese hamster fibroblasts by two cycles of metaphase chromosome isolation and incubation with recipient cells. Human metaphase chromosomes were incubated with mouse A9 cells deficient in hypoxanthine phosphoribosyltransferase, and independent colonies expressing the human species form of this gene were isolated in a selective medium. Metaphase chromosomes isolated from two of these clonal lines were incubated with Chinese hamster fibroblasts deficient in hypoxanthine phosphoribosyltransferase; five resulting independent colonies again expressed the human species of this gene. The transfer frequencies in the two cycles of chromosome-mediated gene transfer were similar (about 10(-7)). These results indicate that the transferred human chromosome fragment is closely associated with the chromosomes of the mouse A9 cells and it is probably integrated into the chromosomal DNA of the recipient cell.     

Interferon enhancement of the invasive capacity of Ewing sarcoma cells in vitro

The ability of interferons to reduce cell proliferation in vitro and in vivo is a well-studied phenomenon. To extend such observations, the effect of interferons on the invasiveness in vitro of human malignant cells derived from a Ewing sarcoma was evaluated. Two related parameters were examined: (i) production of type IV (basement membrane) collagenase and (ii) penetration of human amnion basement membrane and collagenous stroma. After 6 days of treatment with crude fibroblast, leukocyte, or lymphoblastoid interferon at 100 units/ml in serum-free medium, type IV collagenase levels increased 2- to 4-fold per cell relative to those of untreated controls. With homogeneous fibroblast and lymphoblastoid interferons, a 2-fold elevation in type IV collagenase was detected after 2 days, with further increases, occasionally dramatic, occurring on the 4th and 6th day of treatment. The ability of Ewing sarcoma cells to invade human amnion connective tissue was measured after 6 days of treatment with various interferons. Relative to the behavior of untreated controls, crude leukocyte interferon, homogeneous lymphoblastoid interferon, and homogeneous fibroblast interferon at 100 units/ml augmented invasiveness 3-, 17- and 22-fold, respectively, when cells were allowed 4 days in which to traverse the amnion. When untreated cells were exposed simultaneously to the amnion and to homogeneous lymphoblastoid or fibroblast interferon, a 4- to 5-fold increase in invasiveness above control levels was observed in 2 days. These data emphasize the complexity of interferon-induced phenomena. In any overview, the effects of interferon on both the tumor cell and the host must be considered.     

Species-specific suppression of histone H1 and H2B production in human/mouse hybrids.

Ten human/mouse hybrid cell lines that segregate either human or mouse chromosomes were examined for the expression of human- and mouse-specific histones H1 and H2B. Results of this study indicate that the human and mouse chromosomes in hybrid cells that segregate human chromosomes (M greater than H hybrids) contain only mouse histone H1 and H2B. Chromosomes in hybrid cells that segregate mouse chromosomes (H greater than M hybrids) contain only human H1 and H2B histones. Loss of the ability to produce either human or mouse histones does not seem to be due to the loss of specific human or mouse chromosomes because M greater than H hybrids retaining at least one copy of each human chromosome contain only mouse H1 and H2B and H greater than M hybrids retaining at least one copy of each mouse chromosome contain only human H1 and H2B histones. These results, together with those concerning histone H4 acetylation levels and ratios of variants of histones H3 and H2A that are like those in the dominant parent cell type, indicate that the control mechanisms affecting H1 and H2B expression in H greater than M and in M greater than H hybrid cells affect expression of histones H2A, H3, and H4 genes as well. The present data thus support the hypothesis that none of the histone genes that are active in the recessive parent cell type is expressed in hybrid lines that segregate recessive cell chromosomes.     

Loss of mouse chromosomes in somatic cell hybrids between HT-1080 human fibrosarcoma cells and mouse peritioneal macrophages.

Somatic cell hybrids between mouse peritioneal macrophages and HT-1080 human fibrosarcoma cells lose mouse chromosomes and retain the entire complement of human chromosomes. In contrast, somatic cell hybrids between cells derived from two different mouse continuous cell lines and HT-1080 human cells were found to lose human chromosomes preferentially. Loss of mouse chromosomes is not a general property of hybrids between mouse macrophages and transformed human cells; the hybridization of mouse macrophages with cells derived from five different human fibroblast lines transformed by simian virus 40 resulted in the production of hybrid clones that preferentially lost human chromosomes.     

Use of Monkey-Mouse Hybrid Cells for the Study of the Cellular Regulation of Interferon Production and Action

Four clones of a somatic monkey-mouse hybrid cell line were studied. One of these clones produced both mouse and monkey interferon, while the three others produced only mouse interferon. All four were, however, sensitive to both mouse and monkey (or human) interferon. The karyotype analysis of these four hybrid clones and the parental cell lines enabled us to locate the possible genetic site governing the synthesis of monkey interferon on a small subtelocentric monkey chromosome. The genetic site responsible for the synthesis of the antiviral protein is located on a different (monkey or mouse) chromosome.     

Human interferons protect plants from virus infection.

Unfractionated human leukocyte interferon, as well as highly purified subspecies of this interferon, and a purified recombinant of human leukocyte interferon produced in bacteria are active in suppressing multiplicability of tobacco mosaic virus in tobacco leaf discs. Human fibroblast interferon exhibits diverse levels of antiviral activity against tobacco mosaic virus but becomes as active as human leukocyte interferon upon incubation with glycosidases. The effect of interferon is reversible; normal multiplication of tobacco mosaic virus resumes upon removal of interferon.     

Sialidosis and galactosialidosis: chromosomal assignment of two genes associated with neuraminidase-deficiency disorders.

The inherited human disorders sialidosis and galactosialidosis are the result of deficiencies of glycoprotein-specific alpha-neuraminidase (acylneuraminyl hydrolase, EC 3.2.1.18; sialidase) activity. Two genes were determined to be necessary for expression of neuraminidase by using human-mouse somatic cell hybrids segregating human chromosomes. A panel of mouse RAG-human hybrid cells demonstrated a single-gene requirement for human neuraminidase and allowed assignment of this gene to the (pter----q23) region of chromosome 10. A second panel of mouse thymidine kinase (TK)-deficient LM/TK- -human hybrid cells demonstrated that human neuraminidase activity required both chromosomes 10 and 20 to be present. Analysis of human neuraminidase expression in interspecific hybrid cells or polykaryocytes formed from fusion of mouse RAG (hypoxanthine/guanine phosphoribosyltransferase deficient) or LM/TK- cell lines with human sialidosis or galactosialidosis fibroblasts indicated that the RAG cell line complemented the galactosialidosis defect, but the LM/TK- cell line did not. This eliminates the requirement for this gene in RAG-human hybrid cells and explains the different chromosome requirements of these two hybrid panels. Fusion of LM/TK- cell hybrids lacking chromosome 10 or 20 (phenotype 10+,20- and 10-,20+) and neuraminidase-deficient fibroblasts confirmed by complementation analysis that the sialidosis disorder results from a mutation on chromosome 10, presumably encoding the neuraminidase structural gene. Galactosialidosis is caused by a mutation in a second gene required for neuraminidase expression located on chromosome 20.     

Assignment of the gene for methylthioadenosine phosphorylase to human chromosome 9 by mouse-human somatic cell hybridization.

The purine and polyamine metabolic enzyme methylthioadenosine (MeSAdo) phosphorylase is abundant in normal cells and tissues but is lacking from many human and murine malignant cell lines and from cells of some human leukemias in vivo. To explore the genetic control of MeSAdo phosphorylase expression, we measured levels of the enzyme in somatic cell hybrids prepared by fusing MeSAdo phosphorylase-deficient mouse L cell lines with human fibroblasts. In the hybrid clones, MeSAdo phosphorylase activity segregated concordantly with adenylate kinase 1, a marker for human chromosome 9, but not with enzyme markers for any other human chromosome. In hybrid clones derived from human fibroblasts with a reciprocal translocation between chromosomes 9 and 17, MeSAdo phosphorylase activity was confined to cells containing the 9pter----9q12 region. In every case, the enzyme-positive hybrid clones displayed bands of MeSAdo phosphorylase activity with isoelectric points characteristic of both the human and murine enzymes. These results indicate that the structural gene for human MeSAdo phosphorylase, designated MTAP, can be assigned to the 9pter----9q12 region of human chromosome 9. Furthermore, these studies with interspecies somatic cell hybrids show that the MeSAdo phosphorylase-deficient state is recessive in mouse L cell lines.     

Multiple active sites on human interferons.

Human interferons stimulated in peripheral leukocytes and foreskin fibroblasts are active in cultures of human and rabbit cells. The dominant factors in leukocyte and fibroblast interferons responsible for antiviral activity in rabbit cells were shown to be antigenically distinct from each other as well as from rabbit interferon. In addition, leukocyte interferon contained also a minor component with antigenic determinants characteristic of fibroblast specificity, which could be isolated by affinity chromatography on Sepharose-bound antibodies directed against firboblast interferon. Neutralization tests with selected anti-interferon sera suggested that the antiviral activities of leukocyte and fibroblast interferons in human and rabbit cells were associated with single molecules. A model is proposed where molecules of human interferon contain multiple reactive sites each of which is capable of interaction with cells of a different species. The number and distribution of these determinant sites may vary with the source of the human interferon and account for the differential in antiviral protection expressed in homologous and phylogenetically unrelated host cells.     

Identification of a yeast artificial chromosome clone encoding an accessory factor for the human interferon gamma receptor: evidence for multiple accessory factors.

Human chromosomes 6 and 21 are both necessary to confer sensitivity to human interferon gamma (Hu-IFN-gamma), as measured by the induction of human HLA class I antigen. Human chromosome 6 encodes the receptor for Hu-IFN-gamma, and human chromosome 21 encodes accessory factors for generating biological activity through the Hu-IFN-gamma receptor. A small region of human chromosome 21 that is responsible for encoding such factors was localized with hamster-human somatic cell hybrids carrying an irradiation-reduced fragment of human chromosome 21. The cell line with the minimum chromosome 21-specific DNA is Chinese hamster ovary 3x1S. To localize the genes further, 10 different yeast artificial chromosome clones from six different loci in the vicinity of the 3x1S region were fused to a human-hamster hybrid cell line (designated 16-9) that contains human chromosome 6q (supplying the Hu-IFN-gamma receptor) and the human HLA-B7 gene. These transformed 16-9 cells were assayed for induction of class I HLA antigens upon treatment with Hu-IFN-gamma. Here we report that a 540-kb yeast artificial chromosome encodes the necessary species-specific factor(s) and can substitute for human chromosome 21 to reconstitute the Hu-IFN-gamma-receptor-mediated induction of class I HLA antigens. However, the factor encoded on the yeast artificial chromosome does not confer antiviral protection against encephalomyocarditis virus, demonstrating that an additional factor encoded on human chromosome 21 is required for the antiviral activity.     

Genes coding for sensitivity to interferon (IfRec) and soluble superoxide dismutase (SOD-1) are linked in mouse and man and map to mouse chromosome 16.

By using 12 hamster-mouse hybrids segregating a mouse T(16;17)Bnr Robertsonian translocation chromosome in conjunction with 10 similar hybrids segregating normal mouse chromosomes, we have shown that the loci that control cellular sensitivity to interferon (IfRec) and code for the soluble enzyme superoxide dismutase (SOD-1) (superoxide:superoxide oxidoreductase; EC 1.15.1.1) are syntenic in the mouse and map to mouse chromosome 16. IfRec and SOD-1 are also syntenic in man. They have previously been assigned to the distal segment of the long arm of human chromosome 21, trisomy for which causes Down syndrome. Because both IfRec and SOD-1 map to mouse chromosome 16, it will now be possible to use mice trisomic for this chromosome to determine whether certain aspects of the Down syndrome phenotype in man are caused by an altered dosage of IfRec and SOD-1.     

Human chromosome 7 carries the beta 2 interferon gene.

A cDNA clone (pAE20-4) corresponding to the 1.3-kilobase human beta 2 interferon mRNA was used as a probe in blot-hybridization experiments of DNA from a panel of human-rodent somatic cell hybrids containing overlapping subsets of human chromosomes. The DNA hybridization experiments showed that the human beta 2 interferon gene is located on human chromosome 7. This assignment is consistent with previous experimental data in which the expression of the translationally active 1.3-kilobase beta 2 interferon mRNA was assayed in various somatic cell hybrids. Blot-hybridization experiments using DNA from different human cell strains and cell lines reveal distinct EcoRI restriction fragment length polymorphisms of the human beta 2 interferon gene.     

Entrapment of metaphase chromosomes into phospholipid vesicles (lipochromosomes): Carrier potential in gene transfer

Transfer of genes from one type of cultured mammalian cell to another by using isolated metaphase chromosomes has been reported with a frequency of one per 10(6)-10(8) cells. Very recently a rate of 16/10(6) has been reported with Chinese hamster ovary cells [Spandidos, D. A. & Siminovitch, L. (1977) Proc. Natl. Acad. Sci. USA 74, 3480-3484]. To increase the frequency of gene transfer, we isolated metaphase chromosomes from hypoxanthine guanine phosphoribosyltransferase (HGPRT) positive cells, entrapped them in liposomes, and fused the lipochromosomes with HGPRT-negative cells. Lipochromosomes were prepared with cholesterol and egg lecithin, using isolated metaphase chromosomes from a mouse-human somatic hybrid cell line (A9/HRBC2); the entire X chromosome, including the HGPRT, glucose-6-phosphate dehydrogenase, and phosphoglycerate kinase genes, is the only recognizable human genetic material retained by the hybrids. Enclosure of the chromosomes in the lipid envelope was confirmed by electron and fluorescence microscopy and differential centrifugation. These lipochromosomes were fused with HGPRT(-) mouse cells (A9) in the presence or absence of polyethylene glycol and transferents were selected in hypoxanthine/aminopterin/thymidine (HAT) medium. The frequency of transfer was at least once per 10(5) cells, a minimum 10-fold improvement over previous methods. The selected cells contained HGPRT activity similar to the amount found in the A9/HRBC2 cells. Starch gel electrophoresis verified that the observed HGPRT activity in the transferents is due to the human enzyme. Human glucose-6-phosphate dehydrogenase and phosphoglycerate kinase were also identified electrophoretically in the transferents. Karyotyping with C and Q banding did not reveal the presence of the whole human X chromosome or a visible extra fragment of a human chromosome associated with the mouse genome. The biochemical data strongly suggest, however, that transfer of a portion of the human X chromosome has occurred in these transferents. Thus, at least three X-linked genes have been transferred from one cell to another with high frequency, using metaphase chromosomes.     

Expression of human and suppression of mouse nucleolus organizer activity in mouse-human somatic cell hybrids.

Most mouse-human somatic cell hybrids show preferential loss of human chromosomes, absence of human 28S ribosomal RNA, and suppression of human nucleolus organizer activity, as visualized by the Ag-AS silver histochemical stain. In contrast, the mouse-human hybrids studied here show preferential loss of mouse chromosomes. The hybrids were made by fusion of HT-1080-6TG human fibrosarcoma cells with BALB/c mouse peritoneal macrophages or strain 129 mouse teratocarcinoma cells. The Ag-AS staining method shows nucleolus organizer activity of chromosomes 13, 14, 15, 21 (rarely), and 22 in the human parent and chromosomes 12, 15, 16 (rarely), and 18 in the BALB/c mouse parent. In the hybrid cells the human nucleolus organizer regions are active, as shown by Ag-AS staining and involvement in "satellite association." The mouse nucleolus organizer regions are not stained by the Ag-AS method even though mouse chromosomes 12, 15, and 18 are present in the BALB/c hybrids and at least one copy of each mouse chromosome is present in the teratocarcinoma-derived hybrids. Thus, in these mouse-human hybrids, unlike those that lose human chromosomes, only human nucleolus organizer activity is expressed, and mouse nucleolus organizer activity is suppressed.