Role of interferon in the pathogenesis of virus diseases in mice as demonstrated by the use of anti-interferon serum. II. Studies with herpes simplex, Moloney sarcoma, vesicular stomatitis, Newcastle disease, and influenza viruses.

The effect of potent sheep anti-mouse interferon globulin was investigated in several different experimental virus diseases of mice. In anti-interferon globulin-treated mice infected intraperitoneally with herpes simplex virus (HSV) type I, the latent period was shortened, and the overall LD50 was increased several hundredfold compared to virus-infected control mice. When HSV was inoculated subcutaneously all anti-interferon globulin-treated mice died, whereas only 5% of virus-infected control mice died. Subsequent treatment with anti-interferon globulin of previously HSV-infected mice did not result in reactivation of HSV. Treatment of adult mice with anti-interferon globulin resulted in an earlier appearance of MSV-induced tumors, a greater number of mice bearing tumors, an increase in tumor size, and an increase in the duration of tumors. All tumors eventually regressed despite reinjection of anti-interferon globulin. Anti-interferon globulin treatment resulted in a rapid onset of disease and death in adult mice inoculated (intranasal) with VSV and in newborn mice infected with NDV. Anti-interferon globulin exerted no effect on the course of influenza virus infection of mice. We conclude that the early production of interferon is an importane element in the response of the mouse to several viruses exhibiting different pathogeneses.

Antibody to mouse interferon alpha/beta abrogates resistance to the multiplication of Friend erythroleukemia cells in the livers of allogeneic mice.

Friend erythroleukemia cells (FLC) (H-2d) injected intravenously into adult syngeneic DBA/2 or allogeneic C57B1/6 (H-2b) or C3H (H-2k) mice lodge in the liver but only multiply in the liver of syngeneic mice. Our results indicated that endogenous IFN-alpha/beta was a crucial factor in preventing the multiplication of FLC in the liver of adult allogeneic mice. (a) Treatment of allogeneic adult C57B1/6 or C3H mice with polyclonal antibody to mouse IFN-alpha/beta (but not antibody to IFN-gamma) completely abrogated the resistance to the multiplication of FLC in the liver and 87% of tumor-injected, antibody-treated C57B1/6 mice died with extensive tumor involvement of the liver. In contrast, after intravenous inoculation FLC do not multiply at all (or very rarely) in the liver of adult C57B1/6 mice left untreated or treated with a variety of control globulins, and no deaths occurred. (b) 8 h after intravenous inoculation of FLC, poly(A)+ RNA hybridizable with specific DNA probes for mouse IFN-alpha or -beta (but not -gamma) was present in the liver of injected C57B1/6 mice. Using the expression of the Mx protein as an indicator of the presence of IFN-alpha/beta, we showed that Mx+ congenic C57B1/6 mice injected with FLC exhibited a marked increase in the expression of the Mx protein in the liver, spleen, kidney and lung, and this increase was blocked by treatment of mice with antibody to IFN-alpha/beta. The possibility that different host mechanisms are elicited depending on the site of tumor growth in allogeneic mice is discussed. IFN-alpha/beta appears to be of particular importance in determining the resistance of the liver to FLC in allogeneic mice.

Injection of mice with antibody to interferon enhances the growth of transplantable murine tumors.

Injection of DBA/2, C57Bl/6, or BALB/c mice with antibody to mouse interferon alpha/beta enhanced the i.p. transplantability of six different murine tumors, as manifested by an increase in the percentage of tumor-bearing mice and a decrease in survival time. The effect was observed in mice injected with antibody to interferon raised in three sheep, a goat, and a rabbit, but not with sheep antibody to "impurities" present in the mouse interferon preparations or with normal sheep or goat globulins. The enhancement in transplantability was most marked when tumor cells had been previously passaged in vitro and were of low tumorigenicity. Analysis of some of the experimental conditions using interferon-sensitive and interferon-resistant lines of Friend erythroleukemia cells (FLC) showed that the enhancing effect was observed over a wide range of tumor cell inocula, was directly related to the amount of antibody to interferon injected and was most pronounced when antibody was administered at the time of tumor cell injection. Enhancement was also observed when FLC were injected subcutaneously (s.c.). Antibody did not act directly on the tumor cells in vitro. Although we were unable to demonstrate any biologically active interferon in mice before or after tumor cell inoculation, the results suggest that endogenous interferon is present and plays a role in inhibiting the transplantability of some murine tumors in immunocompetent mice.

Role of interferon in the pathogenesis of virus diseases in mice as demonstrated by the use of anti-interferon serum. I. Rapid evolution of encephalomyocarditis virus infection.

The role of interferon in the pathogenesis of encephalomyocarditis (EMC) virus infection was determined by treating mice with potent, partially purified sheep anti-mouse interferon globulin. In control mice, EMC virus was present in low titers in various visceral organs but attained high titers in the brain towards the 4th to 5th day, at which time mice died with signs of central nervous system disease. In mice treated with anti-mouse interferon globulin, virus was present in high titer in visceral organs 24--36 h after viral inoculation and virtually all mice were dead by 45 h. This rapid evolution of EMC virus infection was not observed in mice treated with the globulin fraction prepared from a normal sheep, from a sheep exhibiting a low anti-mouse interferon-neutralizing titer, nor from a sheep having a high titer of antibody to human leukocyte interferon. The experimental results indicated that anti-interferon globulin neutralized the interferon liberated by virus-infected cells, thus permitting extensive virus multiplication in several visceral organs. We conclude that interferon is an important early component of host resistance to this virus infection.

Single-stranded RNA viruses inactivate the transcriptional activity of p53 but induce NOXA-dependent apoptosis via post-translational modifications of IRF-1, IRF-3 and CREB.

To characterize the mechanisms underlying apoptosis induced by viral infection, transcriptional activation of genes encoding members of the 'BH3-only' family of proteins was analysed during the course of virus infection. Among these genes, only NOXA is transcriptionally activated by vesicular stomatitis virus (VSV), sendai virus (SV), measles virus, herpes simplex virus, or dsRNA and required for efficient apoptosis of cells. Transcriptional activation of NOXA by VSV or SV is independent of p53, but requires the presence of interferon regulatory factor 1 (IRF-1), IRF-3 and cAMP-responsive element binding protein (CREB). Binding to and transactivation of the NOXA promoter by each of these transcription factors is governed by post-translational modification involving different pathways for each factor. Thus, SV infection activates IRF-3 and CREB by phosphorylation triggered by Toll like receptor 3 signalling, and a pathway involving calcium-independent phopholipase A2, respectively. In addition transactivation induced by IRF-1 during viral infection correlates with a 10 kDa increase in its molecular weight, suggesting a covalent linkage with a previously unknown regulatory polypeptide.

A sub-population of high proliferative potential-quiescent human mesenchymal stem cells is under the reversible control of interferon alpha/beta.

Type I interferon (IFN) is shown to control the reversible quiescence of a primitive human bone marrow mesenchymal stem cell (MSC) subpopulation. A 24 h pre-treatment of Stro1+/GlycoA- or CD45-/GlycoA- subpopulations with a monoclonal antibody (mAb) against the IFNAR1 chain of the human type I IFN receptor (64G12), or with a polyclonal anti-IFNalpha antibody, resulted in a marked increase in the number of very large colonies (CFU-F >3000 cells) obtained in the presence of low, but necessary, concentrations of bFGF. Over a 2-month culture period, this short activation promoted a faster and greater amplification of mesenchymal progenitors for adipocytes and osteoblasts. Activation correlated with inhibition of STAT1 and STAT2 phosphorylation and of STAT1 nuclear translocation. A non-neutralizing anti-IFNAR1 mAb was ineffective. We demonstrate that control and activated MSCs express ST3GAL3, a sialyltransferase necessary to produce the embryonic antigens SSEA-3 and -4. Interestingly, activated MSC progeny expressed SSEA-3 and -4 at a higher level than control cultures, but this was not correlated with a significant expression of other embryonic markers. As MSCs represent an essential tool in tissue regeneration, the use of 64G12, which rapidly recruits a higher number of primitive cells, might increase amplification safety for cell therapy.

Methicillin-resistant Staphylococcus aureus detection using chromogenic media: the Sheffield experience.

Methicillin-resistant Staphylococcus aureus (MRSA) continues to cause major problems, both in hospitals and the community. Microbiology departments need to review their methodology regularly to ensure that they are contributing in the most appropriate manner to the battle against MRSA. Media employing chromogenic enzymes to aid the isolation and identification of MRSA is a relatively new approach. In this study, 192 swabs from 112 different patients were inoculated on two chromogen-containing media and four other commonly used solid MRSA media to determine which gave the appropriate combination of sensititivity, specificity and speed of result. Methicillin-resistant S. aureus was isolated on at least one of the six media from 102 of the 192 swabs. Both chromogenic media proved to be statistically significantly more sensitive than the other media after overnight incubation and had a sensitivity of 96% after 48 hours' incubation. The recent introduction of chromogen-containing MRSA media offers microbiology laboratories the opportunity to isolate and confirm the majority of MRSA infections/colonisations in 24 hours, which should result in better patient care. The possible slight increase in costs should not provide a valid excuse for using inferior methodologies.

Genomic footprinting: detection of putative regulatory proteins in the promoter region of the interferon alpha-1 gene in normal human tissues.

Dimethyl sulfate (DMS) genomic footprinting revealed the presence of putative regulatory proteins attached to specific sequences of the promoter region of the interferon (IFN) alpha-1 gene in normal human tissue. The pattern of protein-DNA interactions observed for the human alpha-1 promoter in freshly isolated human spleen cells was identical to that seen in DNA from the B-cell line Namalwa. The protein interactions involving the human IFN alpha-1 promoter spanned a region from positions -38 to -174 relative to the cap site which encompasses that part of the IFN alpha-1 promoter previously shown by deletion analysis to confer virus inducibility on the IFN alpha-1 gene. DNase I footprinting performed on isolated nuclei revealed a pattern of protein-DNA interactions for the promoter region of the IFN alpha-1 gene similar to that obtained with DMS footprinting performed on whole cells, with the appearance or disappearance of only a few additional protected nucleotides outside the region identified by the use of DMS. These results provide the first direct evidence for the presence of proteins bound in vivo to those parts of the IFN alpha-1 promoter between positions -64 and -109 previously shown by deletion analysis to confer virus inducibility on the IFN alpha-1 gene. The pattern of protein-DNA interactions observed for the IFN alpha-1 promoter after virus induction was identical to that seen before induction, in keeping with the finding that many transcriptional activators are present in both induced and uninduced cells.

Priming affects the activity of a specific region of the promoter of the human beta interferon gene.

Treatment of Daudi or HeLa cells with human interferon (IFN) alpha 8 before induction with either poly(I)-poly(C) or Sendai virus resulted in an 8- to 100-fold increase in IFN production. The extent of priming in Daudi cells paralleled the increase in the intracellular content of IFN-beta mRNA. IFN-alpha mRNA remained undetectable in poly(I)-poly(C)-treated Daudi cells either before or after priming. An IFN-resistant clone of Daudi cells was found to produce 4- to 20-fold more IFN after priming, indicating that priming was unrelated to the phenotype of IFN sensitivity. IFN treatment of either Daudi or HeLa cells transfected with the human IFN-beta promoter (-282 to -37) linked to the chloramphenicol acetyltransferase (CAT) gene resulted in an increase in CAT activity after induction with poly(I)-poly(C) or Sendai virus. A synthetic double-stranded oligonucleotide corresponding to an authentic 30-base-pair (bp) region of the human IFN-beta promoter between positions -91 and -62 was found to confer virus inducibility upon the reporter CAT gene in HeLa cells. IFN treatment of HeLa cells transfected with this 30-bp region of the IFN-beta promoter in either the correct or reversed orientation also increased CAT activity upon subsequent induction. IFN treatment alone had no detectable effect on the activity of either the 30-bp region or the complete human IFN promoter.

Interferon modulation of c-myc expression in cloned Daudi cells: relationship to the phenotype of interferon resistance.

Treatment of interferon-sensitive Daudi cell with electrophoretically pure human interferon alpha markedly reduced the level of c-myc mRNA, increased the level of class I histocompatibility antigen (HLA) mRNA, and did not affect the level of actin mRNA within the same cells. In contrast, the level of c-myc mRNA or HLA mRNA did not change significantly following interferon treatment in different clones of Daudi cells selected for resistance to the antiproliferative action of interferon. These cells possessed interferon receptors, however, and responded to interferon modulation of other genes, including 2',5' oligoisoadenylate synthetase (M. G. Tovey, M. Dron, K. E. Mogensen, B. Lebleu, N. Metchi, and J. Begon-Lours, Guymarho, J. Gen. Virol., 64:2649-2653, 1983; M. Dron, M. G. Tovey, and P. Eid, J. Gen. Virol., 66:787-795, 1985). A clone of interferon-resistant Daudi cells which had reverted to almost complete sensitivity to both the antiproliferative action of interferon and the interferon-enhanced expression of HLA mRNA remained refractory, however, to interferon modulation of c-myc expression, suggesting that a reduced level of c-myc mRNA may not be a prerequisite for inhibition of cell proliferation in interferon-treated cells. Our results do not exclude the possibility, however, that posttranscriptional modification(s) of c-myc expression may precede an inhibition of cell proliferation in interferon-treated cells.

Structure of a murine alpha interferon pseudogene with a repetitive R-type sequence in the 3' flanking region.

A murine alpha interferon pseudogene was identified in a mouse genomic library. The nucleotide sequence revealed several in-phase termination codons within the gene and repetitive oligonucleotides in the flanking regions. The nucleotide sequences and the amino acids of the peptide signal sequences were compared with known human alpha interferon genes and the pseudogene.

Inhibition of tumor necrosis factor alpha gene transcription by pentoxifylline reduces normothermic liver ischemia-reperfusion injury in rats.

Pentoxifylline (PTX) has been shown to protect the liver against normothermic ischemia-reperfusion (I-R) injury. The aims of this study were to investigate the action of PTX on tumor necrosis factor alpha (TNFalpha) gene transcription following normothermic liver I-R as well as to evaluate the resulting effects on liver function and survival. A segmental normothermic liver ischemia was induced for 90 minutes. Rats were divided into three groups: group 1, control, Ringer lactate administration; group 2, PTX treatment; group 3, sham-operated control rats. PTX (50 mg/kg) was injected intravenously 30 minutes before induction of ischemia and 30 minutes before reperfusion. The nonischemic liver lobes were resected at the end of ischemia. Survival rates were compared and serum activities of TNFalpha, aspartate aminotransferase, alanine aminotransferase, and lactate dehydrogenase were measured. Liver histology was assessed 6 hours after reperfusion. Liver TNFalpha mRNA was assessed by polymerase chain reaction amplification at different times after reperfusion. PTX treatment significantly decreased serum activities of TNFalpha and inhibited liver expression of TNFalpha mRNA. The extent of liver necrosis and serum levels of liver enzymes were significantly decreased by PTX treatment, resulting in a significant increase in 7-day survival compared with nontreated control rats. In conclusion, PTX inhibits liver TNFalpha gene transcription, decreases serum TNFalpha levels, and reduces liver injury following normothermic I-R.

Interleukin-6 repression is associated with a distinctive chromatin structure of the gene.

Expression of the interleukin-6 (IL-6) gene is usually tightly controlled and may be induced in specific tissues only after treatment with appropriate stimuli. The molecular mechanisms responsible for IL-6 gene repression in specific tissues or cell lines remain poorly defined. In order to address this question we have studied two human breast carcinoma cell lines, MDA-MB-231, in which the IL-6 gene is expressed, and MCF-7, in which it is not. The promoter region of the IL-6 gene was analysed in both cell lines with reference to two different parameters: (i) DNase I hypersensitivity; (ii) the in vivo pattern of DNA-protein interactions. We show herein that the mechanism responsible for silencing IL-6 gene expression in MCF-7 cells most probably involves a modification of chromatin structure, as suggested by a decreased sensitivity of the IL-6 promoter to DNase I relative to the IL-6-expressing cell line MDA-MB-231. Moreover, we show that a 'closed' nucleosomal structure in MCF-7 cells does not inhibit the binding of nuclear proteins to IL-6 gene regulatory sequences in vivo. We suggest, therefore, that, in non-expressing cells, local chromatin remodelling at the proximal promoter is inhibited by negative regulators, as suggested by two specific hallmarks of nuclear factor binding that are not observed in expressing cells: an additional in vivo footprint spanning positions -135/-119 and an additional DNase I hypersensitive site far upstream, around position -1400. Furthermore, a specific factor binding in vitro to the -140/-116 region of the IL-6 promoter is found in MCF-7 cells.

Prevention of interferon-induced augmentation of cellular antitumor effector mechanisms by phorbol esters.

Both natural killer cell- and macrophage-mediated spontaneous in vitro cytotoxicity for tumor targets is rapidly and strongly augmented by interferon. Macrophage-activating lymphokines considerably enhance macrophage-tumoricidal activity but did not affect natural killer cell-type cytotoxicity. Augmentation of cytolytic capacity by interferon and by macrophage-activating lymphokines is prevented by the tumor-promoting phorbol ester, 12-O-tetradecanoylphorbol-13-acetate. However, the classical antiviral activity and the specific binding of interferon to cell surface receptors remains unaffected by 12-O-tetradecanoylphorbol-13-acetate.

Direct signal transduction via functional interferon-alphabeta receptors in CD34+ hematopoietic stem cells.

Affinity purified, freshly isolated CD34+ progenitors were shown to express low levels of type I interferon (IFN) receptors (740 +/- 60 binding sites/cell, K(d) 0.7 +/- 0.04 nM) determined by Scatchard's analysis using a radiolabelled, neutralizing, monoclonal antibody directed against the IFNAR1 chain of the human type I IFN receptor. Treatment of freshly isolated (day 0), highly purified (>95% pure) CD34+ cells with recombinant IFN-alpha resulted in rapid tyrosine phosphorylation and activation of STAT1, Tyk2 and JAK1 as shown by Western immunoblotting. Similarly, IFN treatment was shown by confocal microscopy to result in rapid nuclear localization of the transcription factors IRF1 and STAT2, demonstrating the presence of functional IFN receptors on freshly isolated (day 0) CD34+ cells. The number of specific type I IFN receptor binding sites expressed on hematopoietic progenitor cells increased to some 1440 +/- 40 per cell after 11 days of cultivation of CD34+ cells in vitrosuggesting that receptor expression increases with cell differentiation. IFN-mediated signal transduction and the inhibitory effect of IFN-alpha on 7 or 14 days CFU-GM and BFU-E colony formation was abrogated in the presence of the anti-IFNAR1 mAb, indicating that IFN-alpha acts directly on the proliferation of human hematopoietic progenitor cells via receptor activated signal transduction without excluding the induction of other cytokines or growth factors by residual accessory cells.

Constitutive expression of specific interferon isotypes in peripheral blood leukocytes from normal individuals and in promonocytic U937 cells.

Constitutive expression of IFN-alpha5 and IFN-beta was detected in different lymphoid cells including peripheral blood mononuclear cells from normal individuals following amplification of IFN mRNA by reverse transcriptase-polymerase chain reaction and direct sequencing of the amplified product. The activated form of the interferon-induced transcription factor complex ISGF3 was also detected in nuclear extracts from uninduced cells. Culture supernatants from uninduced U937 cells were also found to activate an ISRE cloned upstream of the luciferase reporter gene, indicating the presence of endogenous IFN activity equivalent to approximately 0.3 to 0.5 IU/mL. This endogenous IFN was also shown to play a role in maintaining the basal level of expression of the major histocompatibility class I genes in lymphoid cells. These results suggest that IFN-alpha5 and IFN-beta are produced at low levels in normal tissues and play an important role in the regulation of cell function and in the maintenance of homeostasis.

Role of the type I interferons in allograft rejection.

Type I interferons are potent immuno-modulatory cytokines that enhance expression of the major histocompatibility complex (MHC) class I antigens, T-cell cytotoxicity, and natural killer (NK) cell activity, all of which are implicated in graft rejection. A monoclonal antibody (mAb) directed against the extracellular domain of the human interferon gamma (IFN-gamma) receptor (IFN-alpha R), which inhibits both the binding and biological activity of all the type I IFNs tested, exerted a dose-dependent inhibition of the mixed lymphocyte reaction and induced permanent survival of skin allografts in MHC-divergent Cynomologus monkeys treated with a subeffective dose of cyclosporin A. Marked differences were observed in the composition of T lymphocyte subpopulations in anti-IFN-alpha R mAb-treated animals relative to the various control groups. Skin biopsies from animals treated with anti-IFN-R Mab + cyclosporin A revealed very low levels of MHC class I and class II antigen expression and the absence of histological signs of rejection, whereas skin biopsies from control animals exhibited high levels of MHC antigen expression and the histological signs of acute rejection, including a pronounced lymphocytic infiltrate, edema, and necrosis. No monkey antibodies (IgG) to the mouse anti-human IFN-alpha R mAb were detected in the serum of any of the animals treated with the anti-IFN-alpha R mAb either alone or together with cyclosporin A. Treatment of lethally irradiated Cynomologus monkeys with the anti-IFN-alpha R mAb together with a subeffective dose of cyclosporin A was also found to markedly enhance the survival of animals grafted with allogeneic bone marrow cells from donors differing in both MHC class I and class II antigens. These results show that selective and lasting immunosuppression can be obtained by the short-term administration of an IFN-alpha antagonist together with a subeffective dose of cyclosporin A, and may have important implications for the therapy of human allograft rejection.

Characterization of a domain of a human type I interferon receptor protein involved in ligand binding.

Two monoclonal antibodies that recognize different epitopes of the extracellular domain of one of the proteins that constitute the type I interferon receptor were used to delineate the interferon binding site. Antibody 64G12 both inhibits the binding of radiolabeled interferon-alpha 2 and IFN-alpha 8 to their cell surface receptors and neutralizes the antiviral and antiproliferative actions of all the type I interferons tested, including IFN-beta, IFN-omega, and human leukocyte IFN, a mixture of different interferon-alpha isotypes. Antibody 34F10 recognizes the type I interferon receptor with an affinity similar to that of the MAb 64G12 but does not inhibit either the binding or the biologic activity of any of the type I interferons tested. Both antibodies recognize a protein of 105 +/- 5 kD from either Daudi or Ly28 cells. Immunoprecipitation following surface iodination demonstrated that the neutralizing MAb recognizes a protein of 105 kD and the nonneutralizing MAb a protein of 110 kD in extracts of Daudi cells. A second less intense band was also detected by both antibodies. Cross-linking of IFN-alpha 2 to its receptor before immunoprecipitation prevented the neutralizing antibody from immunoprecipitating the receptor protein, but the nonneutralizing MAb was still able to recognize a 140 kD protein corresponding to the cross-linked interferon-receptor protein complex. Thus, an interferon binding domain appears to be localized in a region between amino acids 23 and 229 of the extracellular domain of a transmembrane protein that forms part of the type I interferon receptor complex containing the epitopes recognized by each antibody.

Oromucosal interferon therapy: marked antiviral and antitumor activity.

Oromucosal administration of murine interferon-alpha/beta (IFN-alpha/beta) or individual recombinant species of murine IFN-alpha, IFN-beta, or IFN-gamma or recombinant human IFN-alpha1-8, which is active in the mouse, exerted a marked antiviral activity in mice challenged systemically with a lethal dose of encephalomyocarditis virus (EMCV), vesicular stomatitis virus (VSV), or varicella zoster virus (VZV). The effects observed were dose dependent and similar in magnitude to those observed following parenteral administration of the same dose of IFN. No antiviral activity was observed after oromucosal administration of murine IFN-alpha/beta in animals in which the IFN receptor had been inactivated by homologous recombination. In contrast to parenteral treatment, oromucosal IFN therapy was found to be ineffective when IFNs were administered before virus infection. Oromucosal administration of IFN-alpha also exerted a marked antitumor activity in mice injected i.v. with highly malignant Friend erythroleukemia cells or other transplantable tumors, such as L1210 leukemia, which has no known viral etiology, the EL4 tumor, or the highly metastatic B16 melanoma. These results show that high doses of IFN can be administered by the oromucosal route apparently without ill effect, raising the possibility that the oromucosal route will prove to be an effective means of administering high doses of IFN that are clinically effective but poorly tolerated.

Induction of tolerance to recombinant therapeutic proteins.

The specific IgM and IgG antibody responses to subcutaneous (s.c.) treatment of mice with recombinant human IFN-alpha2a (rHuIFN-alpha2a) or IFN-beta were inhibited in a dose-dependent manner by prior oromucosal (o.m.) administration of rHuIFN-alpha2a or IFN-beta, respectively. Pretreatment of animals once a day for 7 days by the o.m. route with the highest dose of IFN-alpha2a tested (10(7) IU) resulted in complete inhibition of the peak IFN-alpha2a-specific IgG antibody response detected 28 days after subsequent s.c. injection of IFN-alpha2a (p < 0.001). Similarly, prior o.m. administration of 1-10 microg rHuGM-CSF per day for 7 days resulted in a statistically significant (p < 0.001) inhibition of the peak GM-CSF-specific IgG antibody response detected 28 days after s.c. administration of GM-CSF. In contrast, prior o.m. treatment with a quantity of bovine serum albumin (BSA) (100 microg) or human serum albumin (HSA) (10 microg) equivalent, respectively, to the protein content of the highest dose of IFN-alpha2a or GM-CSF administered by the o.m. route, did not affect significantly the IFN-alpha2a-specific or GM-CSF-specific IgG antibody responses detected on subsequent s.c. administration of IFN-alpha2a or GM-CSF. Oromucosal administration of IFN-alpha2a, IFN-beta, or GM-CSF alone did not induce detectable IFN-alpha2a-specific, IFN-beta-specific, or GM-CSF-specific IgM or IgG antibody responses at any of the time points tested. These results suggest that short-term o.m. administration of a recombinant protein is an effective means of inducing peripheral tolerance to subsequent parenteral administration of a therapeutic protein.