Molecular structure of human fibroblast and leukocyte interferons: probe by lectin and hydrophobic chromatography.

Structural differences between human leukocyte virus-induced interferon and human fibroblast polyinosinic-polycytidylic acid (rIn-rCn)-induced interferon have been noted in previous studies. This study reports the behavior of human leukocyte and fibroblast interferon, induced by virus and by rIn-rCn, in several lectin and hydrophobic chromatographic systems. Differences in both glycosylation and in hydrophobicity of human leukocyte and fibroblast interferons are documented. Human fibroblast interferon is a glycoprotein, whereas our evidence suggests that human leukocyte interferon probably is not. Also, fibroblast interferon is more hydrophobic than leukocyte interferon, as probed on several hydrophobic adsorbents. The possible relationships of these differences to each other and to antigenic variations are discussed. Generally, the differences appear to be attributable to the cell type in which the interferon was induced. However, our results suggest that at least subtle differences in the processing of the induction signal (virus or rIn-rCn) within the same cell type may occur, slightly altering some structural features.     

Nucleotide sequence of the chromosomal gene for human fibroblast (beta 1) interferon and of the flanking regions

The nucleotide sequence of the human fibroblast (beta 1) interferon chromosomal gene and its flanking regions was determined. These results confirm the absence of intervening sequences in the gene. The presence of some sequences in the upstream flanking region homologous to similar features for other eukaryotic genes was revealed: these include not only the TATAAAT sequence and the consensus sequence (reported by Benoist et al., 1980) but also two additional motifs, one of which is so far present only in inducible genes. Furthermore, a striking similarity between the upstream flanking regions of the human beta 1 and alpha 1 interferon genes is observed.     

Gene reorganization during serial divisions of normal human cells

We have followed during serial divisions of human fibroblasts the presence in chromosomal and extrachromosomal DNA, of two genes that are expressed in fibroblasts, actin and interferon, and of one that is not expressed, globin. The intensity of the blot hybridization of the actin and globin probes with chromosomal DNA diminished during serial divisions of diploid fibroblasts. The interferon gene remained constant throughout the human fibroblast life span. Chromosomal DNA sequences were present in extrachromosomal circular DNA which appeared at the end of the fibroblast life span. The results could explain some functional changes that occur in these cell populations when their division potential declines.     

Molecular weight of the functional unit of human leukocyte, fibroblast, and immune interferons

There is good agreement between the target molecular weight and the known molecular weight of human leukocyte interferons (about 20,000). The target molecular weight of fibroblast interferon, 31,000 to 42,000, is significantly larger than the monomer molecular weight of 21,000 to 24,000, suggesting that the dimer may be the predominant active functional unit in solution. A range from 63,000 to 73,000 for the target molecular weight of several different fractions of immune interferon (including natural crude as well as the recombinant form) indicates that the functional form of the immune interferon may be a trimer or tetramer. Thus, these studies indicate that the functional unit of leukocyte interferon is the monomer, that of fibroblast interferon is a dimer, and that of immune interferon is probably a tetramer (or trimer).     

Fibroblast interferon in man is coded by two loci on separate chromosomes.

We have examined viral and poly(rl):poly(rC) induction of interferon synthesis in several human, mouse and Chinese hamster cell lines, and in hybrids derived from the fusion of such cells. We observed species and cell-type differences in inducer effectiveness and in the kinetics of interferon production. In some cases, parental characteristics are preserved in somatic cell hybrids, and in other cases, the expression of the donor phenotype is modulated by the epigenetic state of the recipient cell. Mapping studies in human/mouse and human/Chinese hamster hybrids indicate that there are at least two structural genes for human fibroblast interferon. Chromosomes 2 and 5 each contain genetic information for the synthesis of fibroblast interferon. Gene dosage experiments indicate that one gene is on the long arm of chromosome 2 and another is on the short arm of chromosome 5. Leukocyte interferon genes could not be mapped to these chromosomes, but this negative result could be influenced by the epigenetic state of the hybrid cells.     

Human interferon enhances gunaylate cyclase activity

Partially purified human leukocyte interferons, partially purified human lymphoblastoid interferon, and human fibroblast interferon enhanced rat liver, kidney, and splenic guanylate cyclase {E.C.4.6.1.2.} activity 2–4 fold at 5 μIU concentration. Dose-response relationships revealed that the human leukocyte interferons enhanced splenic guanylate cyclase activity at concentrations as low as 0.01 μIU while a concentration of 1 μIU for partially purified human lymphoblastoid interferon and 10 μIU concentration for human fibroblast interferon were necessary to see any effect on guanylate cyclase activity.     

Nucleotide sequence of a portion of human chromosome 9 containing a leukocyte interferon gene cluster

The nucleotide sequence of a 9937 base-pair portion of human chromosome 9, which contains two complete leukocyte interferon genes (LeIF-L and J), the complete intergenic region, and part of a third related possible pseudogene (LeIF-M), has been determined. The coding regions of the L and J genes are separated by 4363 nucleotides. The coding regions for the putative L and J interferons are 96% homologous and are each surrounded by about 3500 nucleotides of flanking sequences, which are also highly homologous. The L and J genes and their respective flanking sequences comprise a 4000 nucleotide leukocyte interferon gene repeat unit; the L gene repeat unit contains two major insertions not present in the J gene repeat unit. The J gene repeat unit is flanked by sequence features reminiscent of those found surrounding transposable elements. Both the L and J gene repeat units are embedded within sequences that are highly repeated in the human genome. Structural features identified within this portion of chromosome 9 may have been important for the generation of this interferon gene cluster.     

Enzymatic modifications of human fibroblast and leukocyte interferons

The sensitivity of highly purified human fibroblast interferon and partially purified human leukocyte interferon to several proteolytic and glycolytic enzymes was determined with respect to antiviral activity, isoelectric point, molecular weight, and thermal stability. Leucine aminopeptidase altered the distribution of isoelectric points for both interferons but produced little change in molecular weights; this enzyme somewhat reduced the activity of only leukocyte interferon. Treatment of fibroblast interferon with carboxypeptidases A and B did not greatly decrease antiviral activity, but it did slightly reduce the molecular weight of the interferon and substantially altered the distribution of isoelectric point values; similar treatment of leukocyte interferon caused some loss in activity, especially of the 17,000-molecular-weight species. Both interferons were inactivated rapidly by treatment with the endoproteases trypsin, pepsin, bromelain, and subtilisin. Chymotrypsin shifted the isoelectric points of both interferons, but only leukocyte interferon was significantly inactivated. Treatment with neuraminidase and beta-galactosidase changed the isoelectric point distribution but did not affect the activity or thermal stability of either interferon; such a treatment reduced the molecular weight of fibroblast interferon and the size heterogeneity of leukocyte interferon. Treatment with neuraminidase and then leucine aminopeptidase greatly reduced the activity of both interferons, especially leukocyte interferon. The data indicate that biologically active forms of fibroblast and leukocyte interferons can be distinguished by their relative sensitivity to certain proteases.     

Cloning of eukaryotic genes in single-strand phage vectors: the human interferon genes

Using oligonucleotide probes with defined sequences, we have selected clones from a human lymphocyte cDNA library which represent human leukocyte (HuIFN-α) and fibroblast (HuIFN-β) interferon gene sequences. Double-stranded f1 phage DNA was used as the vector for initial cloning of cDNA. Clones carrying interferon gene sequences were identified by hybridization with the oligonucleotide probes. The same oligonucleotide probes were used as primers for dideoxy chain termination sequencing of the clones. One HuIFN-α clone, 201, has a nucleotide sequence different from published HuIFN-α sequences. Under control of the lacUV5 promoter, the 201 gene has been used to express biologically active HuIFN-α in Escherichia coli.     

Increased nitroblue tetrazolium dye reduction by human neutrophils stimulated by interferon in vitro

Human leukocyte interferon enhanced nitroblue tetrazolium dye (NBT) reduction by human neutrophils (PMNs). Increase in NBT reduction paralleled increase in interferon dose. When human leukocyte interferon was heated to 60 C or 80 C for 30 min, both the antiviral activity and the effect on NBT reduction decreased. Human leukocyte interferon neutralized with anti-human leukocyte interferon serum showed no effect on NBT reduction. A human fibroblast interferon preparation also enhanced NBT reduction. The species dependency of interferon was shown in NBT reduction as well as in antiviral activity.     

Enhancement of T cell cytotoxic responses by purified human fibroblast interferon.

Purified polyribonucleotide-induced human fibroblast interferon (HFIF) was tested for its effects on proliferative and cytotoxic human T cell responses to alloantigens. The addition of HFIF (100 to 400 IFU/ml) to mixed leukocyte cultures decreased alloantigen-induced lymphocyte proliferative responses as determined by both recovery of responding cells and by 3H-thymidine incorporation into responding cells. However, HFIF, but not the mock interferon preparation, increased the cytotoxic response of T cells to allogeneic cells by 4- to 5-fold when expressed in terms of lytic units. Although fibroblast and leukocyte interferons have different physicochemical and biologic properties, the results reported here are in concert with previous findings concerning the effects of virus-induced leukocyte interferon on human T cell functions.     

Chemical characterization of recombinant human leukocyte interferon A using fast atom bombardment mass spectrometry

Proteolytic digests of biologically active fractions of recombinant human leukocyte interferon A expressed in large quantities in Escherichia coli were analyzed by fast atom bombardment mass spectrometry and high-performance liquid chromatography. The values observed in the mass spectra of digests of the major fraction of recombinant human leukocyte interferon A with trypsin and Staphylococcus aureus protease V8 accounted for 93% of the amino acid sequences of human leukocyte interferon A predicted from the nucleotide sequence of the gene encoding the protein, indicating that the major fraction of recombinant human leukocyte interferon A was expressed with the same amino acid sequence as that translated from the nucleotide sequence of the gene encoding the protein. Mass spectrometry of proteolytic digests of two minor fractions of recombinant human leukocyte interferon A and mass and amino acid analyses of their high-performance liquid chromatography fractions showed that the amino group of the N-terminal amino acid residue of interferon was in part acetylated, and the Cys-1 and Cys-98 residues were oxidized to cysteic acid or linked to glutathione. These findings suggest that amino acid residues in recombinant proteins prepared in large quantities in E. coli are modified post-translationally.     

Human leukocyte interferon (HuIFN-alpha): potent endorphin-like opioid activity

Human leukocyte interferon, but not fibroblast or immune interferons, binds to opiate receptor $\text{in}\text{,}\text{vitro}$. When injected intracerebrally into mice, human leukocyte, but not fibroblast or immune interferon, caused potent endorphin-like opioid effects. These effects include analgesia, lack of spontaneous locomotion and catalepsy. All of these actions of human leukocyte interferon were preventable and reversible by the opiate antagonist naloxone. The findings suggest that some of the side effects of leukocyte interferon therapy may be mediated by opiate receptor binding. They also provide evidence for a regulatory circuit between the immune and neuroendocrine system. This putative circuit could be an etiologic site for certain psychopathological states.     

The gene for human fibroblast interferon (IFB) maps to 9p21

Summary The interferons have been classified into alpha, beta, and gamma (leukocyte, fibroblast, and immune). We used a human genomic clone for 1 interferon IFB to determine the gene copy number in two patients with unbalanced rearrangements of 9p. Our results provide evidence for regional assignment of this gene to 9p21.     

Interferon effect on ribosomal ribonucleic acid related to chromosome 21 ploidy.

Antiviral and cell-growth-inhibitory activities of human interferon were shown to be related to the activity of a gene or genes present on chromosome 21. The 18s rRNA is vital to cell growth; it is capable of a viral-mRNA-recognition function and it is coded for by genes a portion of which are present on chromosome-21. A previously reported ability of human interferon to affect rRNA metabolism is characterized by a decrease in the sucrose-gradient-peak ratio of radiolabelled 28S to 18S rRNA in extracts from the cytoplasm of interferon-treated human fibroblasts. In the present report, interferon dose-response curves are presented demonstrating a direct relationship between a decrease in this ratio and interferon concentrations in the media. By using this virus-independent cytoplasmic rRNA assay, eight human fibroblast lines, differing in chromosome 21 ploidy, were tested for sensitivity to human interferon. Two monosomy-21, two euploid-21 and four trisomy-21 cell lines were tested. The monosomy-21 cell populations were significantly less sensitive to interferon than the other six cell types tested. Of the cell lines tested, the most sensitive, by a wide margin, was a trisomy-21 line. Trisomy-21 cell monolayer sensitivity, however, varied widely within the range from normal to supersensitive. These observations suggest that interferon's ability to affect rRNA metabolism is related to the activity of a gene or genes present on chromosome 21.     

Alveolar macrophage lipoxygenase products of arachidonic acid: isolation and recognition as the predominant constituents of the neutrophil chemotactic activity elaborated by alveolar macrophages

Human immune interferon preparations have anticellular activity on human cell lines (WISH and HEp-2). This anticellular activity copurified with the human immune interferon and appears to be a function of the immune interferon molecule. On the basis of a unit of antiviral activity, purified human immune interferon had about 20 and 100 times more anticellular activity than purified fibroblast or leukocyte interferon, respectively. The possible implications of this finding in the treatment of human neoplasia are discussed.     

Expression of human leukocyte interferon type A in Saccharomyces cerevisiae under the control of regulatory elements of the yeast gene URA3

Expression of a chromosomal gene for human leucocyte interferon A was obtained due to interaction of 5'-nontranslating region of a cloned interferon gene with the regulatory sequences from UNA3 yeast gene. The sequence of 5'-nontranslating region from interferon gene essential for its expression in yeast is localized within 130 b p. from the initiating codon. Due to increasing of plasmid stability and copy number a 60-fold increase. in interferon yield was obtained in yeast transformants reaching the level of 6 X 10(7) u X 1(-1). The data are presented supposing the existence of functional polycistronic mRNA in yeast.     

Binding of human fibroblast interferon to concanavalin A-agarose. Involvement of carbohydrate recognition and hydrophobic interaction.

Human fibroblast interferon binds to a concanavalin A-agarose (Con A-Sepharose) equilibrated with methyl alpha-D-mannopyranoside, or levan; in contrast, it is only partially retarded on a similar column equilibrated with ethylene glycol. Interferon does not bind, however, to a lectin column equilibrated with both methyl alpha-D-mannopyranoside and ethylene glycol. Thus, a hydrophobic interaction between fibroblast interferon and the immobilized lectin seems to account for a large portion of the binding forces involved. Other hydrophobic solutes, such as dioxane, 1, 2-propanediol, and tetraethylammonium chloride, were found equally or more efficient than ethylene glycol in displacing interferon from the lectin column. The elution pattern of interferon from a concanavalin A-agarose (Con A-Sepharose) column, at a constant ehtylene glycol concentration and with an increasing mannoside concentration, reveals the existence of four distinct interferon components. The selective adsorption to and elution from a concanavalin A-agarose (Con A-Sepharose) column resulted in about a 3000-fold purification of human fibroblast interferon and complete recovery of activity. The specific activity of the partially purified interferon preparation is about 5 X 10(7) units per mg of protein. The chromatographic behavior of human leukocyte interferon is remarkable in that it does not bind to concanavalin A-agarose at all indicating the absence of carbohydrate moieties recognizable by the lectin, or if present, their masked status. When concanavalin A was coupled to an agarose matrix (cyanogen bromide activated) at pH 8.0 and 6.0 human fibroblast interferon bound to both lectin-agarose adsorbents and could be recovered with methyl alpha-D-mannopyranoside. Concanavalin A, immobilized directly on agarose matrix at pH 8.0 and 6.0, thus displays only carbohydrate recognition toward interferon. By contrast, unless a hydrophobic solute was included in the solvent containing methyl mannoside, human fibroblast interferon could not be recovered from concanavalin A-agarose coupled at pH 9.0. When concanavalin A was immobilized via molecular arms, in tetrameric as well as dimeric forms, the binding of interferon again occurred exclusively through carbohydrate recognition. Thus, the hydrophobic interaction can be eliminated by appropriate immobilization of the lectin, and then adsorbed glycoproteins, as exemplified here by interferon, can be recovered readily with methyl mannoside alone.