Regulation by interferon alpha of immunoglobulin isotype selection and lymphokine production in mice.

Antigens and infectious agents that stimulate interferon alpha(IFN-alpha) production in mice induce antibody responses that are predominantly of the immunoglobulin (Ig)G2a isotype and contain little or no IgE. This suggested the possibility that IFN-alpha might have a role in directing Ig isotype selection. Consistent with this possibility, we have found that injection of mice with recombinant mouse IFN-alpha suppresses IgE secretion, enhances IgG2a secretion, and has no independent effect on IgG1 secretion in mice stimulated with a foreign anti-IgD antibody. Injection of mice with polyinosinic acid.polycytidylic acid (poly I.C), an inducer of macrophage IFN-alpha production, also suppresses the anti-IgD antibody-induced IgE response and stimulates the IgG2a response; these effects are blocked by a sheep antibody that neutralizes mouse IFN-alpha/beta. Both recombinant IFN-alpha and poly I.C have maximum IgE suppressive and IgG2a stimulatory effects when injected early in the anti-IgD antibody-induced immune response. Addition of IFN-alpha to mouse B cells cultured with lipopolysaccharide (LPS) + interleukin 4 (IL-4) suppresses both IgG1 and IgE production, but much less potently than IFN-gamma. IFN-alpha suppresses anti-IgD antibody-induced increases in the level of splenic IL-4 mRNA, but enhances the anti-IgD antibody-induced increase in the splenic level of IFN-gamma mRNA. These results are consistent with the effect of IFN-alpha on Ig isotype expression in mice, as IL-4 stimulates IgE and suppresses IgG2a secretion while IFN-gamma exerts opposite effects. These observations suggest that antigen presenting cells, by secreting IFN-alpha early in the course of an immune response, can influence the nature of that response both through direct effects on B cells and by influencing the differentiation of T cells.

Three-dimensional structure of recombinant human interferon-gamma.

The x-ray crystal structure of recombinant human interferon-gamma has been determined with the use of multiple-isomorphous-replacement techniques. Interferon-gamma, which is dimeric in solution, crystallizes with two dimers related by a noncrystallographic twofold axis in the asymmetric unit. The protein is primarily alpha helical, with six helices in each subunit that comprise approximately 62 percent of the structure; there is no beta sheet. The dimeric structure of human interferon-gamma is stabilized by the intertwining of helices across the subunit interface with multiple intersubunit interactions.

Zinc mediated dimer of human interferon-alpha 2b revealed by X-ray crystallography.

BACKGROUND: The human alpha-interferon (huIFN-alpha) family displays broad spectrum antiviral, antiproliferative and immunomodulatory activities on a variety of cell types. The diverse biological activities of the IFN-alpha's are conveyed to cells through specific interactions with cell-surface receptors. Despite considerable effort, no crystal structure of a member of this family has yet been reported, because the quality of the protein crystals have been unsuitable for crystallographic studies. Until now, structural models of the IFN-alpha's have been based on the structure of murine IFN-beta (muIFN-beta). These models are likely to be inaccurate, as the amino acid sequence of muIFN-beta differs significantly from the IFN-alpha's at proposed receptor-binding sites. Structural information on a huIFN-alpha subtype would provide an improved basis for modeling the structures of the entire IFN-alpha family. RESULTS: The crystal structure of recombinant human interferon-alpha 2b (huIFN-alpha 2b) has been determined at 2.9 A resolution. HuIFN-alpha 2b exists in the crystal as a noncovalent dimer, which associates in a novel manner. Unlike other structurally characterized cytokines, extensive interactions in the dimer interface are mediated by a zinc ion (Zn2+). The overall fold of huIFN-alpha 2b is most similar to the structure of muIFN-beta. Unique to huIFN-alpha 2b is a 3(10) helix in the AB loop which is held to the core of the molecule by a disulfide bond. CONCLUSIONS: The structure of huIFN-alpha 2b provides an accurate model for analysis of the > 15 related type 1 interferon molecules. HuIFN-alpha 2b displays considerable structural similarity with muIFN-beta, interleukin-10 and interferon-gamma, which also bind related class 2 cytokine receptors. From these structural comparisons and numerous studies on the effects of mutations on biological activity, we have identified protein surfaces that appear to be important in receptor activation. This study also reveals the potential biological importance of the huIFN-alpha 2b dimer.

Evaluation of the antitumor activity of recombinant human gamma-interferon employing human melanoma xenografts in athymic nude mice.

Three human melanoma cell lines (G-361, HT-144, and SK-MEL-3) that were highly sensitive to growth inhibition in vitro by recombinant human interferon (rHuIFN-gamma) were adapted to grow as s.c. xenografts in athymic nude mice. Take rates were greater than 90% for all three tumors, with in vivo doubling times of 15, 4, and 15 days for G-361, HT-144, and SK-MEL-3, respectively. Commencing 3 days posttumor implantation mice were treated with daily s.c. injections of rHuIFN-gamma for 30 days over a dose range of 2.4-326 megaunits/mouse/day at a site distinct from the tumor implant. Tumors were measured twice weekly and mice were observed daily for deaths and morbidity until day 60 postimplantation. No apparent antitumor activity was observed in either the G-361 or HT-144 tumors at any dose despite the achievement of high rHuIFN-gamma blood levels. Intralesional treatment of the HT-144 xenograft with rHuIFN-gamma at 240 megaunits/mouse/day did not significantly retard tumor growth or increase lifespan. However, the SK-MEL-3 tumor showed a significant response at the 326-megaunit dose as noted by a tumor growth delay of 11.8 days in treated versus control animals and by an increased number of 60-day survivors. The tumor growth suppression appeared to be greater during the treatment period than during the subsequent observation period. Other experiments employing 326, 530, and 860 megaunits rHuIFN-gamma/mouse/day in mice bearing the SK-MEL-3 tumor demonstrated tumor growth delays of 4.2, 4.8, and 19.8 days, respectively, suggesting a dose response. These data support the conclusions that (a) in vitro antiproliferative activity of rHuIFN-gamma is not necessarily predictive of in vivo efficacy; and (b) relatively high doses of rHuIFN-gamma appear to be required for demonstrating an in vivo antitumor effect in this model.

In vitro assessment of the effects of granulocyte-macrophage colony-stimulating factor on primary human tumors and derived lines.

One hundred eighty-nine human tumor specimens were tested in a human tumor cloning assay to determine their growth response to human recombinant granulocyte-macrophage colony-stimulating factor. Of these samples 48 were evaluable for response. Growth stimulation to greater than 150% of controls was noted in 1 of 12 lung cancers (8%) and 1 of 14 breast cancers (7%) but in no other instances for an overall rate of 2 of 48 (4.2%). A dose-response effect was not seen with each of the two stimulated samples responding only at the two lowest concentrations tested. In addition, 7 cell lines derived from human tumors were tested using a metabolic CO2 production assay without evidence of growth stimulation. Samples of normal bone marrow displayed the usual dose-dependent stimulation whether grown in agar or assayed metabolically. We conclude that human recombinant granulocyte-macrophage colony-stimulating factor has minimal effect on the growth of the solid tumors tested and that clinical trials to reduce chemotherapy-associated myelo-suppression may proceed without undue concern for enhancement of tumor growth.

Specific immunosuppressive effects of constant infusion of 2'-deoxycoformycin.

The effect of continuous infusion into C57BL/6J mice of 2'-deoxycoformycin (DCF), a tight-binding inhibitor of adenosine deaminase, on the biological function of bone marrow stem cells and T- and B-lymphocytes was evaluated. Greater than 85% inhibition of adenosine deaminase in erythrocytes, thymus, and bone marrow was noted after DCF infusion at 0.4 mg per kg body weight per day, while lesser extents of inhibition were characteristic of spleen and lymph nodes. The reconstitution of lethally irradiated C57BL/6J mice with bone marrow cells from DCF- and 0.9% NaCl infused mice of the same strain was compared. The two groups of animals were virtually identical with respect to (a) the number of spleen colony-forming units, (b) the response of splenic lymphocytes to both B- and T-cell mitogens, (c) hematological analysis of peripheral blood elements, and (d) survival time, thus strongly supporting the lack of effect of DCF infusion on the capacity of stem cells to differentiate. In contradistinction, DCF infusion was highly lymphocytotoxic as noted by the severe necrosis in both B- and T-cell regions in lymph nodes and spleen and by the dramatic weight reduction in spleen and thymus. Histopathology of other tissues including bone marrow was normal except for the occurrence of hepatitis. A striking decrease in blastogenesis induced by the mitogens concanavalin A, phytohemagglutinin, and Escherichia coli lipopolysaccharides was also observed after DCF infusion. Consistent with these data, in vitro incubation of bone marrow cells with DCF did not impair the number of spleen colony-forming units produced in lethally irradiated mice. These data suggest a potential use for adenosine deaminase inhibitors in the prevention of graft-versus-host disease in hematopoietic transplantation.

Biological properties of recombinant alpha-interferons: 40th anniversary of the discovery of interferons.

IFNs were first described as potent antiviral agents 40 years ago, and recombinant IFN-alpha2a and IFN-alpha2b were approved for the treatment of hairy cell leukemia just 11 years ago. Today, alpha-IFNs are approved worldwide for the treatment of a variety of malignancies and virologic diseases. Although the exact mechanism of action of IFN-alpha in the treatment of such diseases is not fully understood, many advances have been made in the characterization of the physicochemical and diverse biological properties of this highly pleiotropic cytokine. Here we review recent developments in our understanding of the antiviral and immunoregulatory properties of IFN-alpha, the nature of the multisubunit IFN-alpha receptor, and the molecular mechanisms of signal transduction. Where available, we have included comparative data on recombinant alpha-IFNs derived from both naturally occurring and nonnaturally occurring synthetic genes. We also review clinical data and data on the side effects and antigenicity of different sources of recombinant alpha-IFNs in humans. These latter topics are of clinical interest, because they may potentially affect the efficacy of these various products. Hopefully, what is already known about IFN will prompt further exploration into the mechanism(s) of action of IFN-alpha and thus deliver new applications for this prototypic cytokine, whose full therapeutic potential is yet to be realized.

Purification and properties of a novel recombinant human hybrid interferon, delta-4 alpha 2/alpha 1.

The human interferon (huIFN) delta-4 alpha 2(5-62)/alpha 1(64-166) is a genetically engineered hybrid that consists of residues 5-62 of huIFN alpha 2 and residues 64-166 of huIFN alpha 1. This variant contains four cysteine residues at positions 29, 86, 99 and 139, but does not contain the cysteine at position 1 that is characteristic of naturally occurring huIFN alpha subtypes. This novel recombinant hybrid was purified from Escherichia coli to greater than 95% homogeneity. The purification was based on ethanol extraction of a trichloroacetic acid precipitate and Matrex Gel Blue A chromatography followed by either a selective precipitation or DEAE-Sepharose chromatography. The purified protein that was treated with 2-mercaptoethanol exhibited two closely migrating bands on sodium dodecyl sulfate-polyacrylamide gel electrophoresis with apparent molecular weight values of 17,800 and 17,100, both of which exhibited antiviral activity. Electrophoresis performed without prior reduction with 2-mercaptoethanol indicated only a minor extent of intermolecular disulfide bonding. The purified protein exhibited a high specific antiviral activity of 7 x 10(7) units/mg when assayed on human fibroblast cells and, in distinction to the parental huIFN alpha 2, it also demonstrated antiviral activity on human fibroblast cells and, in distinction to the parental huIFN alpha 2, it also demonstrated antiviral activity on murine L929 cells. The level of antiproliferative activity of huIFN delta-4 alpha 2(5-62)/alpha 1(64-166) on various cell lines of different histological origin appeared to be more comparable to that of huIFN alpha 1 than huIFN alpha 2. The data suggest that huIFN delta-4 alpha 2(5-62)/alpha 1(64-166) hybrid may be a useful tool for understanding huIFN structure-function relations.

Multiple forms of recombinant murine interleukin-4 expressed in COS-7 monkey kidney cells.

Recombinant murine interleukin-4 (muIL-4) expressed in COS-7 monkey kidney cells was purified to homogeneity by sequential CM-Sepharose, Sephadex G-100 chromatography and mono-S FPLC to a specific activity of 6.10(7) units per mg of protein based on an in vitro HT-2 cell proliferation assay. Two electrophoretic variants, designated a and b, which migrated on SDS-PAGE as a closely spaced doublet with Mr 19,000, were present in the final product. Gas phase sequencing of the purified protein revealed the presence of an N-terminus corresponding to the mature protein predicted from the cDNA sequence and sequencing of a cyanogen bromide digest confirmed 75 of the 120 predicted amino acids. Elution behavior on gel filtration corresponded to that of a monomer of Mr 19,000. Since there are three potential sites of N-glycosylation predicted by the cDNA sequence, the contribution of glycosylation to the observed heterogeneity was examined by treatment with endoglycosidases. Variant b was digested by either endo-beta-N-acetylglucosaminidase H (endo H) or endo-beta-N-acetylglucosaminidase F (endo F) to protein of Mr 15,000 on SDS-PAGE but was unaffected by treatment with endo-beta-N-acetylglucosaminidase D (endo D), thus indicating the presence of high mannose type of N-glycan. In contrast, variant a was resistant to endo H, F and D. Complete conversion of a mixture of variants a and b to a single protein of Mr 15,000 on SDS-PAGE was obtained only after treatment with N-glycanase. Both variants were resistant to neuraminidase and O-glycanase treatment. These data show that the microheterogeneity observed in purified muIL-4 preparations is due to differences in the nature of the N-linked oligosaccharides. The availability of purified recombinant muIL-4 and a methodology for both total and selective deglycosylation provides a basis for the initiation of structure-function studies of this novel T-cell lymphokine.

Subunits and their interactions.

There is fairly general agreement that myosin isolated from rabbit skeletal muscle has a molecular weight of about 500,000. The higher values that have been reported apparently reflect protein aggregation related to the method of preparation. On the basis of present evidence, the myosin molecule has an elongate helical core of two f subunits (average weight about 215,000) that extend into a globular head region containing three g subunits (average weight about 20,000). Myosin may be dissociated into subunits by a number of methods. In 5 M guanidine, the myosin molecule is dissociated into f and g subunits, while at pH above 10, the g subunits are dissociated from the intact fibrous core of myosin. The dissociation of g subunits at pH 10 is accompanied by the loss of both ATPase activity and actin-binding capacity; however, the exact biological significance of the g subunits is presently uncertain. In preliminary studies, the f subunits appear to contain the sulfhydryl residues currently implicated in myosin ATPase, and there is some indication of allosteric regulation of enzymic activity.

Three-dimensional structure of recombinant human granulocyte-macrophage colony-stimulating factor.

The crystal structure of recombinant human granulocyte-macrophage colony-stimulating factor (rhGM-CSF) has been determined at 2.8 A resolution using multiple isomorphous replacement techniques. There are two molecules in the crystallographic asymmetric unit, which are related by an approximate non-crystallographic 2-fold axis. The overall structure is highly compact and globular with a predominantly hydrophobic core. The main structural feature of rhGM-CSF is a four alpha-helix bundle, which represents approximately 42% of the structure. The helices are arranged in a left-handed antiparallel bundle with two overhand connections. Within the connections is a two-stranded antiparallel beta-sheet. The tertiary structure of rhGM-CSF has a topology similar to that of porcine growth factor and interferon-beta. Most of the proposed critical regions for receptor binding are located on a continuous surface at one end of the molecule that includes the C terminus.

Crystallization and preliminary X-ray investigation of recombinant human interleukin 4.

Crystals of recombinant human interleukin 4 have been grown from solutions of ammonium sulfate. The crystals are tetragonal, space-group P4(1)2(1)2 or P4(3)2(1)2; the unit cell axes are a = 92.2(1) A and c = 46.4(1) A. The crystals are stable to X-rays for at least three days and diffract beyond 2.8 A resolution. The crystals contain approximately 63% solvent, assuming there is one molecule in the asymmetric unit.

Isolation and characterization of a resistant core peptide of recombinant human granulocyte-macrophage colony-stimulating factor (GM-CSF); confirmation of the GM-CSF amino acid sequence by mass spectrometry.

A trypsin-resistant core peptide of recombinant human granulocyte-macrophage colony-stimulating factor (rhGM-CSF) was isolated and analyzed by high-energy Cs+ liquid secondary-ion (LSI) mass spectrometric analysis. This analysis provided successful detection of the high-mass disulfide-linked core peptide as well as information confirming the existence of disulfide pairing. Similarly, LSI mass spectrometric analysis of the peptide fragments isolated chromatographically from a Staphylococcus aureus V8 protease digest of rhGM-CSF provided rapid confirmation of the cDNA-derived sequence and determination of the existing disulfide bonds between cysteine residues 54-96 and 88-121. Electrospray ionization mass spectrometry was employed to measure the molecular weight of the intact protein and to determine the number of the disulfide bonds in the protein molecule by comparative analysis of the protein before and after reduction with beta-mercaptoethanol.

Preclinical biology of alpha interferons.

The availability of large quantities of purified recombinant human alpha interferons has permitted an expansion of studies on the preclinical biology of interferon. These purified preparations have definitively established that interferon exhibits pleiotypic effects on cellular function, including antiviral, antiproliferative, and immunomodulatory activities. Thus, interferon can exert therapeutic activity by both direct effects on the growth of the tumor and by modulation of the biologic response of the host. The exact mechanism of action will most probably vary from patient to patient. The variety of biologic activities that interferon displays, as well as the fact that it is highly species-specific, makes it difficult to design preclinical studies that can provide guidance for the clinical application of interferon as an antineoplastic agent. Two models that have provided useful preclinical data on human cell lines or fresh biopsies are reviewed: the human tumor clonogenic assay and human tumor xenografts in immunodeficient mice. These models indicate that interferon is likely to be an effective treatment for a broad range of malignancies, and that the response will be highly dependent on the type of interferon as well as its dose, schedule, and route of administration. Both models have provided evidence that alpha interferon may demonstrate a synergistic or additive interaction with standard chemotherapeutic agents like cyclophosphamide or doxorubicin. The utility of any preclinical model for predicting an individual patient's response to interferon is yet to be established.

Molecular and biologic characterization of recombinant interferon-alpha2b.

Recombinant IFN-alpha2b (Intron A, Schering-Plough Corp, Kenilworth, NJ), derived from Escherichia coli, is currently approved worldwide for the therapy of various cancers and chronic hepatitis B and C. We describe here the purity and identity tests for both physicochemical properties and bioactivity that have been developed for the characterization of highly purified IFN-alpha2b with a specific activity of 2.5 x 10(8) IU/mg protein. The data indicate that a product of high purity can be consistently produced without DNA contamination. The antiviral bioassay chosen to measure potency is based on the ability of IFN-alpha2b to protect human foreskin fibroblast FS-71 cells from the cytopathic effects of encephalomyocarditis virus. Various immunoassays are described that have been used to quantitate the presence of binding and neutralizing antibodies in patients treated with IFN-alpha2b. Overall, the available data indicate a very low incidence of neutralizing antibody formation. We also summarize the current state of knowledge with respect to IFN reference standards for the biologic assay as well as recommendations for the calculation of antibody neutralization titers.

Application of electrospray mass spectrometry in probing protein-protein and protein-ligand noncovalent interactions.

A novel mass spectrometry-based methodology using electrospray ionization (ESI) is described for the detection of protein-protein [interferon (IFN)-γ dimer] and protein-ligand [ras-guanosine diphosphate (GDP)] noncovalent interactions. The method utilizes ESI from aqueous solution at appropriate pH. The presence of the noncovalent complex of the IFN-γ dimer was confirmed by the observed average molecular weight of 33,819 Da. The key to the detection of the IFN-γ dimer is the use of an alkaline solution (pH ≈ 9) for sample preparation and for mass spectrornetry analysis. The effect of the declustering energy in the region of the ion sampling orifice and focusing quadrupole on the preservation of the gas-phase noncovalent complex (IFN-γ dimer) was also studied. The effect of the declustering energy on complex dissociation was further extended to probe the noncovalent protein-ligand association of ras-GDP. It was found that little energy is required to dissociate the IFN-γ dimer, whereas a substantial amount of energy is required to dissociate the gas-phase ras-GDP complex.

Regulation of natural killer cytotoxicity by recombinant alpha interferons. Augmentation by IFN-alpha 7, an interferon similar to IFN-alpha J.

The structure-function relationship of several recombinant human alpha interferons (IFN-alpha) (IFN-alpha 1, IFN-alpha 2, IFN-alpha 4, IFN-alpha 7, IFN-alpha 2/alpha 1 and IFN-delta 4 alpha 1) was investigated with respect to their ability to augment natural killer (NK) cytotoxicity of human peripheral blood mononuclear cells (PBMC) against hemopoietic tumor cell lines. Although all these IFNs significantly augmented NK cytotoxicity against the K562, Daudi and U937 targets, significant quantitave differences were observed in their ability to augment NK. INF-alpha 4, IFN-alpha 2 and IFN-alpha 2/alpha 1 were able to augment NK at low concentrations (less than 0.1 ng/ml), whereas IFN-alpha 7, IFN-alpha 1 and IFN-delta 4 alpha 1 required significantly higher concentrations (3 ng/ml or higher). The cumulative rank order of INFs on the basis of NK augmenting ability was found to be: IFN-alpha 4 approximately IFN-alpha 2 approximately IFN-alpha 2/alpha 1 greater than IFN-alpha 7 greater than IFN-alpha 1 approximately IFN-delta 4 alpha 1. To determine synergism or potentiation in the ability of IFNs to augment NK cytotoxicity, we investigated the effect of simultaneous, sequential and reversed order of treatment of human PBMC by these IFNs. Such potentiation or synergism was not observed. In addition, all these IFNs were able to augment NK cytotoxicity against targets from malignant melanoma cell lines. IFN-alpha 7 augmented regularly and reproducibly NK cytotoxicity in 15 of 19 normal donors examined (79%). This augmentation was blocked by an anti-IFN-alpha antibody. Concentrations of IFN-alpha 7 as low as 0.06 ng/ml were able significantly to augment NK cytotoxicity of PBMC after incubation for one hour at 37 degrees C. In contrast to these findings, IFN-alpha J, an interferon similar to IFN-alpha 7, has been report to be incapable of augmenting NK cytotoxicity and also of interfering with augmentation of NK by other IFNs. Sequential treatment of PBMC first with IFN-alpha 7 and then with other interferons did not prevent the augmentation of NK. Similarly, simultaneous treatment with IFN-alpha 7 and other interferons did not prevent augmentation of NK. In both treatments IFN-alpha J has been reported to prevent augmentation of NK. IFN alpha J and IFN-alpha 7 differ only by one amino acid, at position 107, where a lysine in IFN-alpha J has been replaced by a glutamic acid in the IFN-alpha 7.(ABSTRACT TRUNCATED AT 250 WORDS)

Review of recent developments in the molecular characterization of recombinant alfa interferons on the 40th anniversary of the discovery of interferon.

Recombinant alfa interferons (IFN-alpha s) are approved worldwide for the treatment of a variety of cancers and diseases of virologic origin. A series of recent advances in the molecular characterization of recombinant IFN-alpha s have allowed the determination of the three-dimensional IFN-alpha 2b structure by high-resolution x-ray crystallography. We review here recent developments in our understanding of the molecular and physicochemical properties of recombinant IFN-alpha, including our current state of knowledge of the IFN-alpha gene family and the multiple species of human leukocyte IFN. Based on the reported three-dimensional structure of IFN-alpha 2b, we propose a molecular model for the IFN-alpha 2b receptor complex and predict models for the naturally occurring subtypes IFN-alpha 1 and IFN-alpha 8, as well as the synthetic, non-naturally occurring consensus IFN. Such models provide molecular insights into the mechanism of action of IFN-alpha.