Risk of Bias Assessments and Evidence Syntheses for Observational Epidemiologic Studies of Environmental and Occupational Exposures: Strengths and Limitations.

BACKGROUND: Increasingly, risk of bias tools are used to evaluate epidemiologic studies as part of evidence synthesis (evidence integration), often involving meta-analyses. Some of these tools consider hypothetical randomized controlled trials (RCTs) as gold standards. METHODS: We review the strengths and limitations of risk of bias assessments, in particular, for reviews of observational studies of environmental exposures, and we also comment more generally on methods of evidence synthesis. RESULTS: Although RCTs may provide a useful starting point to think about bias, they do not provide a gold standard for environmental studies. Observational studies should not be considered inherently biased vs. a hypothetical RCT. Rather than a checklist approach when evaluating individual studies using risk of bias tools, we call for identifying and quantifying possible biases, their direction, and their impacts on parameter estimates. As is recognized in many guidelines, evidence synthesis requires a broader approach than simply evaluating risk of bias in individual studies followed by synthesis of studies judged unbiased, or with studies given more weight if judged less biased. It should include the use of classical considerations for judging causality in human studies, as well as triangulation and integration of animal and mechanistic data. CONCLUSIONS: Bias assessments are important in evidence synthesis, but we argue they can and should be improved to address the concerns we raise here. Simplistic, mechanical approaches to risk of bias assessments, which may particularly occur when these tools are used by nonexperts, can result in erroneous conclusions and sometimes may be used to dismiss important evidence. Evidence synthesis requires a broad approach that goes beyond assessing bias in individual human studies and then including a narrow range of human studies judged to be unbiased in evidence synthesis. https://doi.org/10.1289/EHP6980.

Polymorphisms of the DNA repair genes XPD (Lys751Gln) and XRCC1 (Arg399Gln and Arg194Trp): relationship to breast cancer risk and familial predisposition to breast cancer.

Family history is a risk factor for breast cancer and could be due to shared environmental factors or polymorphisms of cancer susceptibility genes. Deficient function of DNA repair enzymes may partially explain familial risk as polymorphisms of DNA repair genes have been associated, although inconsistently, with breast cancer. This population based case-control study examined the association between polymorphisms in XPD (Lys751Gln) and XRCC1 (Arg399Gln and Arg194Trp) genes, and breast cancer. Breast cancer cases (n=321) and controls (n=321) were matched on age and menopausal status. Conditional logistic regression was used to estimate odds ratios (OR) and 95% confidence intervals (CI). The analysis was conducted omitting observations with missing data, and by using imputation methods to handle missing data. No significant association was observed between the XPD 751Gln/Lys (OR 1.37, 95% CI 0.96-1.96) and Gln/Gln genotypes (OR 1.08, 95% CI 0.62-1.86) (referent Lys/Lys), XRCC1 399Arg/Gln (OR 1.48, 95% CI 0.92-2.38) and Gln/Gln genotypes (1.11, 95% CI 0.67-1.83) (referent Arg/Arg) or the XRCC1 Arg/Trp and Trp/Trp genotypes (OR 1.12, 95% CI 0.69-1.83) (referent Arg/Arg) and breast cancer. In multivariate analysis, the adjusted odds ratios for the XPD and XRCC1 399 polymorphisms increased and became statistically significant, however, were attenuated when imputation methods were used to handle missing data. There was no interaction with family history. These results indicate that these polymorphisms in XPD and XRCC1 genes are only weakly associated with breast cancer. Without imputation methods for handling missing data, a statistically significant association was observed between the genotypes and breast cancer, illustrating the potential for bias in studies that inadequately handle missing data.

Polymorphisms in the DNA repair gene XRCC1 and breast cancer.

X-ray repair cross complementing group 1 (XRCC1) encodes a protein involved in base excision repair. We examined the association of polymorphisms in XRCC1 (codon 194 Arg-->Trp and codon 399 Arg-->Gln) and breast cancer in the Carolina Breast Cancer Study, a population-based case-control study in North Carolina. No association was observed between XRCC1 codon 194 genotype and breast cancer, and odds ratios (ORs) were not modified by smoking or radiation exposure. A positive association for XRCC1 codon 399 Arg/Gln or Gln/Gln genotypes compared with Arg/Arg was found among African Americans (253 cases, 266 controls; OR = 1.7, 95% confidence interval, 1.1-2.4) but not whites (386 cases, 381 controls; OR =1.0, 95% confidence interval, 0.8-1.4). Among African-American women, ORs for the duration of smoking were elevated among women with XRCC1 codon 399 Arg/Arg genotype (trend test; P < 0.001) but not Arg/Gln or Gln/Gln (P = 0.23). There was no difference in OR for smoking according to XRCC1 codon 399 genotype in white women. ORs for occupational exposure to ionizing radiation were stronger for African-American and white women with codon 399 Arg/Arg genotype. High-dose radiation to the chest was more strongly associated with breast cancer among white women with XRCC1 codon 399 Arg/Arg genotype. Our results suggest that XRRC1 codon 399 genotype may influence breast cancer risk, perhaps by modifying the effects of environmental exposures. However, interpretation of our results is limited by incomplete knowledge regarding the biological function of XRCC1 alleles.

DNA repair gene XRCC1 polymorphisms, smoking, and bladder cancer risk.

Bladder cancer is the sixth most common cancer in the United States. The main identified risk factor is cigarette smoking, which is estimated to contribute to up to 50% of new cases in men and 20% in women. Besides containing other carcinogens, cigarette smoke is a rich source of reactive oxygen species (ROS) that can induce a variety of DNA damage, some of which is repaired by the base excision repair (BER) pathway. The XRCC1 gene protein plays an important role in BER by serving as a scaffold for other repair enzymes and by recognizing single-strand DNA breaks. Three polymorphisms that induce amino acid changes have been found in codon 194 (exon 6), codon 280 (exon 9), and codon 399 (exon 10) of this gene. We tested whether polymorphisms in XRCC1 were associated with bladder cancer risk and whether this association was modified by cigarette smoking. Therefore, we genotyped for the three polymorphisms in 235 bladder cancer cases and 213 controls who had been frequency matched to cases on age, sex, and ethnicity. We found no evidence of an association between the codon 280 variant and bladder cancer risk [odds ratio (OR), 1.2; 95% confidence interval (CI), 0.6-2.6]. We found some evidence of a protective effect for subjects that carried at least one copy of the codon 194 variant allele relative to those homozygous for the common allele (OR, 0.59; 95% CI, 0.3-1.0). The combined analysis with smoking history suggested a possible gene-exposure interaction; however, the results were not statistically significant. Similarly, for the codon 399 polymorphism, our data suggested a protective effect of the homozygous variant genotype relative to carriers of either one or two copies of the common allele (OR, 0.70; 95% CI, 0.4-1.3), and provided limited evidence, albeit not statistically significant, for a gene-smoking interaction.

XPD polymorphisms: effects on DNA repair proficiency.

XPD codes for a DNA helicase involved in transcription and nucleotide excision repair. Rare XPD mutations diminish nucleotide excision repair resulting in hypersensitivity to UV light and increased risk of skin cancer. Several polymorphisms in this gene have been identified but their impact on DNA repair is not known. We compared XPD genotypes at codons 312 and 751 with DNA repair proficiency in 31 women. XPD genotypes were measured by PCR-RFLP. DNA repair proficiency was assessed using a cytogenetic assay that detects X-ray induced chromatid aberrations (breaks and gaps). Chromatid aberrations were scored per 100 metaphase cells following incubation at 37 degrees C (1.5 h after irradiation) to allow for repair of DNA damage. Individuals with the Lys/Lys codon 751 XPD genotype had a higher number of chromatid aberrations (132/100 metaphase cells) than those having a 751Gln allele (34/100 metaphase cells). Individuals having greater than 60 chromatid breaks plus gaps were categorized as having sub-optimal repair. Possessing a Lys/Lys751 genotype increased the risk of sub-optimal DNA repair (odds ratio = 7.2, 95% confidence interval = 1.01-87.7). The Asp312Asn XPD polymorphism did not appear to affect DNA repair proficiency. These results suggest that the Lys751 (common) allele may alter the XPD protein product resulting in sub-optimal repair of X-ray-induced DNA damage.

Expression of cytochrome P450 1A1/2 and 3A4 in liver tissues of hepatocellular carcinoma cases and controls from Taiwan and their relationship to hepatitis B virus and aflatoxin B1-and 4-aminobiphenyl-DNA adducts.

Cytochrome P450 enzymes play a major role in the metabolism of several of the chemical carcinogens involved in the development of hepatocellular carcinoma (HCC). To investigate by immunohistochemistry interindividual differences in these enzymes, polyclonal antisera and immunoperoxidase staining were used to detect the expression of CYP1A1/2 and 3A4 in 37 surgical control tissues and 105 tumour and adjacent nontumour tissues of HCC cases from Taiwan. There was variability in the expression and staining pattern for both CYP1A1/2 and 3A4 in all tissue types. In tissues from controls, there was no correlation between P450 expression and smoking history or hepatitis B virus antigen status. Since these samples had been previously analysed for the DNA adducts of aflatoxin B1 (AFB1), a dietary mould contaminant, and 4-aminobiphenyl (4-ABP), a component of cigarette smoke, we also investigated the relationship between P450 levels and DNA adducts. 4-ABP-DNA adducts were higher in tissues with elevated levels of CYP1A1/2 (p = 0.02). Overall there was no relationship between CYP1A1/2 or CYP3A4 and AFB1-DNA adducts in control tissues. Staining intensity for CYP1A1/2 and 3A4 followed the order: tumour tissues < control tissues < adjacent non-tumour tissues. CYP1A1/2 levels tended to be lower in tumour and adjacent non-tumour tissues than for CYP3A4. In HCC cases, 4-ABP-DNA adducts were higher in subjects with higher levels of CYP1A1/2, stratified by tissue type, but these differences were not significant. For CYP3A4, in contrast to control tissues, there was a significant association with AFB1-DNA adducts in tumour and adjacent non-tumour tissue of HCC cases. These results suggest that one factor influencing carcinogen-DNA adducts is levels of specific P450 enzymes. However, adduct formation in vivo is a complex processes dependent upon numerous genetic and environmental factors.

Prostate cancer risk and polymorphism in 17 hydroxylase (CYP17) and steroid reductase (SRD5A2).

Prostate cancer is the most common malignancy in males and is the second most common cause of cancer mortality in American men. Polymorphisms have been identified in two genes, the 17-hydroxylase cytochrome P450 gene (CYP17) and the steroid 5-reductase type II gene (SRD5A2) that are involved with androgen biosynthesis and metabolism. The CYP17 A2 allele contains a T-->C transition in the 5' promoter region that creates an additional Sp1-type (CCACC box) promoter site. The SRD5A2 valine to leucine (V89L) polymorphism has been correlated with lower dihydroxytestosterone levels. We tested genotypes in 108 prostate cases and 167 controls along with samples (n = 340) from several different ethnic groups. The CYP17 A2 allele (combined A1/A2 and A2/A2 genotypes) occurred at a higher frequency in Caucasian patients with prostate cancer (70%) than in Caucasian clinical control urology patients (57%), suggesting that the A2 allele may convey increased risk for prostate cancer [odds ratio (OR) = 1.7, 95% confidence interval (CI) = 1.0-3.0]. Blacks and Caucasians had a similar frequency of the A2 genotype (16 and 17%, respectively) while Taiwanese had the highest frequency (27%). The SRD5A2 leucine genotype was most frequent in Taiwanese (28%), intermediate in Caucasians (8.5%) and lowest in Blacks (2.5%). Genotypes having a SRD5A2 leucine allele were somewhat more common in prostate cancer cases (56%) than in controls (49%) (OR = 1.4, 95% CI = 0.8-2.2) but this difference was not significant. These results support the hypothesis that some allelic variants of genes involved in androgen biosynthesis and metabolism may be associated with prostate cancer risk.

XRCC1 polymorphisms: effects on aflatoxin B1-DNA adducts and glycophorin A variant frequency.

Hereditary genetic defects in DNA repair lead to increased risk of cancer. Polymorphisms in several DNA repair genes have been identified; however, the impact on repair phenotype has not been elucidated. We explored the relationship between polymorphisms in the DNA repair enzyme, XRCC1 (codons 194, 280, and 399), and genotoxic end points measured in two populations: (a) placental aflatoxin B1 DNA (AFB1-DNA) adducts in a group of Taiwanese maternity subjects (n = 120); and (b) somatic glycophorin A (GPA) variants in erythrocytes from a group of North Carolina smokers and nonsmokers (n = 59). AFB1-DNA adducts were measured by ELISA, and erythrocyte GPA variant frequency (NN and NO) was assessed in MN heterozygotes with a flow cytometric assay. XRCC1 genotypes were identified by PCR-RFLPs. The XRCC1 399Gln allele was significantly associated with higher levels of both AFB1-DNA adducts and GPA NN mutations. Individuals with the 399Gln allele were at risk for detectable adducts (odds ratio, 2.4; 95% confidence interval, 1.1-5.4; P = 0.03). GPA NN variant frequency was significantly higher in 399Gln homozygotes (19.6 x 10(-6)) than in Gln/Arg heterozygotes (11.4 x 10(-6); P < 0.05) or Arg/Arg homozygotes (10.1 x 10(-6); P = 0.01). No significant effects were observed for other XRCC1 polymorphisms. These results suggest that the Arg399Gln amino acid change may alter the phenotype of the XRCC1 protein, resulting in deficient DNA repair.

Growth inhibition induced by Ro 31-8220 and calphostin C in human glioblastoma cell lines is associated with apoptosis and inhibition of CDC2 kinase.

Protein kinase C (PKC) is a central component in signal transduction and growth control and might be an appropriate target for the chemotherapy of human brain tumors. This study demonstrates that the staurosporine derivative Ro 31-8220, a potent PKC inhibitor, inhibited the growth of 7 human brain tumor cell lines with an IC50 of about 2 microM. Calphostin C, a structurally unrelated PKC inhibitor, inhibited the growth of two of these cell lines with an IC50 of about 100 to 300 nM. Drug withdrawal and clonogenicity assays indicated that the growth inhibition by both of these compounds was irreversible. Morphologic studies, DNA fragmentation studies and flow cytometric assays showed that the treated glioblastoma cells underwent apoptosis. Treatment of glioblastoma cells with Ro 31-8220 lead to a rapid decline in the level of the anti-apoptosis protein bcl-2. At least three of the glioblastoma cell lines carried mutant p53 alleles with missense mutations in the DNA binding domain of p53. Therefore, the induction of apoptosis in these cell lines occurred through a p53-independent mechanism. Furthermore treatment of these glioblastoma cell lines with Ro 31-8220 or calphostin C led to an increase of cells in the G2-M phase of the cell cycle. This correlated with a decrease in CDC2-associated histone H1 kinase activity, as well as a decrease in the level of the CDC2 protein as shown by immunoblotting. When added to subcellular assays Ro 31-8220 markedly inhibited CDC2 histone H1 kinase activity with an IC50 of 100 nM, but calphostin C directly inhibited this kinase activity only at very high concentrations (above 100 microM). Thus these compounds inhibit the growth of glioblastoma cells through novel mechanisms. Ro 31-8220, in particular, might be a useful agent for the treatment of human brain tumors.

p53 mutations, chronic hepatitis B virus infection, and aflatoxin exposure in hepatocellular carcinoma in Taiwan.

Recent studies have implicated aflatoxin B1 (AFB1) exposure as an etiological agent in hepatocellular carcinoma (HCC) and suggested an interaction with chronic hepatitis B virus (HBV) infection. Worldwide AFB1 exposure correlates with a specific mutation at codon 249 in the p53 tumor suppressor gene in liver tumors. This study investigated the roles of HBV and AFB1 in the HCC carcinogenic pathway involving p53 mutations. In cases and controls, chronic HBV infection was assessed by serum hepatitis B surface antigen (HBsAg) and AFB1 exposure by immunohistochemical detection of AFB1-DNA adduct in liver tissue. p53 protein mutations in tumor tissues of HCC cases were identified by immunohistochemistry and DNA mutations by single-stranded conformational polymorphism and sequencing analysis. Both chronic HBsAg carrier status and liver AFB1-DNA adducts were significantly higher in cases than in controls with odds ratios (OR) of 8.4 and 3.9, respectively (P < 0.01). Moreover, HCC risk was greatest in individuals with both AFB1-DNA adducts and HBsAg, suggesting a viral-chemical interaction. Mutant p53 protein, mutations in the p53 gene, and specific codon 249 mutations were detected in 37, 29, and 13%, respectively, of the HCC cases. Most of the DNA mutations were transversions, and the only major clustering site for mutations was codon 249. AFB1-DNA adducts were associated with p53 protein (OR = 2.9, P = 0.054) and DNA mutations (OR = 2.9, P = 0.082) but with borderline significance. All of the codon 249 mutations (n = 12) occurred in HBsAg-seropositive carriers, resulting in an OR of 10.0 (P < 0.05), suggesting that HBV may be involved in the selection of these mutations. The ORs between HBsAg and p53 DNA and protein mutations were 2.6 (P = 0.077) and 1.8 (P > 0.05), respectively. Both p53 DNA and protein mutations were related to tumor stage, suggesting that they are late events. These studies provided further support for the role of aflatoxin exposure in HCC in Taiwan and insight into viral-chemical interactions and molecular pathogenesis.

Isolation of cDNA clones for the interferon-induced 67,000-dalton protein: direct induction of a family of mRNAs by human interferon-alpha and interferon-gamma.

A partial cDNA clone for the interferon (IFN)-induced 67,000-dalton (67K) protein was isolated by immunological screening and used as a probe to study the expression of mRNAs encoding this protein. Northern blot analyses of RNA from IFN-treated GM2767 cells revealed the presence of two major 67K-specific RNA species, 2.7 and 4.3 kb in length, and two minor RNA species, 5.7 and 7.2 kb long. All of these 67K-specific RNAs were polyadenylated. Multiple 67K-specific mRNAs were observed to be induced in several cell lines. IFN-gamma was more effective at inducing these mRNAs than was IFN-alpha. In IFN-alpha-treated GM2767 cells, the 67K-specific mRNAs were detectable 6 h following IFN treatment, but not 12, 18, or 24 h following treatment. In IFN-gamma-treated cells, these mRNAs were detectable 6 h after treatment and continued to be present 24 h after treatment. The induction of the 67K-specific mRNAs in GM2767 cells did not require protein synthesis as the RNAs were induced by IFN-alpha or IFN-gamma in the presence of cycloheximide (CHX, 50 micrograms/ml). Treatment of cells with the combination of CHX and IFN-alpha mediated an enhanced accumulation of the 67K-specific mRNAs, suggesting that ongoing protein synthesis may downregulate the induction or accumulation of the IFN-alpha-induced 67K-specific mRNAs. Western blot analysis employing a monoclonal antibody to the 67K protein revealed that several distinctly sized but immunologically related proteins were induced in IFN-treated cells.

Fragmentation of cellular DNA is a nonspecific indicator of responsiveness to tumor necrosis factor.

Tumor necrosis factor (TNF) or lymphotoxin (LT) treatment of cells sensitive to the anticellular action of TNF results in the degradation of their cellular DNA into fragments that are multiples of about 200 base pairs. The specificity of this DNA fragmenting effect was examined. The DNA of cells dying either as a result of exposure to interferon-gamma (IFN-gamma) or as a result of having exhausted their culture media was observed to be fragmented into multiples of 200 base pairs. Antibody to TNF or LT failed to block the IFN-gamma-mediated DNA fragmentation and antibodies to IFN-gamma, TNF, and LT failed to block the DNA fragmentation observed in the cells dying as a result of having exhausted their culture media. Thus the fragmentation of cellular DNA appears to be nonspecific effect of cell death that can be induced by a variety of treatments.

Induction of proteins in interferon-alpha- and interferon-gamma-treated polymorphonuclear leukocytes.

Interferon-gamma (IFN-gamma) treatment of polymorphonuclear leukocytes (PMNs) results in an activation of their functions. Studying the IFN responsiveness of PMNs and using antibodies to the IFN-induced proteins, we have observed the ability of IFN-alpha to stimulate the production of the IFN-induced 67,000 and 56,000 dalton proteins and the ability of IFN-gamma to induce the synthesis of the 67,000, 56,000, and 42,000 dalton proteins. The induction of these proteins is dependent on de novo RNA synthesis, as its induction is inhibited if the IFNs and actinomycin D are added to the cells simultaneously. The results of this study confirm the ability of PMNs to carry out gene activation and demonstrate the ability of PMNs to respond to both IFN-alpha and IFN-gamma.

Correlation between the anticellular and DNA fragmenting activities of tumor necrosis factor.

The treatment of cells sensitive to the anticellular effect of tumor necrosis factor (TNF) with TNF results in a degradation of their cellular DNA into DNA fragments that are multiples of 200 base pairs. TNF treatment of cells resistant to the anticellular effect of TNF, but bearing receptors for TNF, fails to result in any DNA fragmentation. Incubation conditions, such as temperature, the presence of metabolic inhibitors or amino acid deprivation, that modulate the effectiveness of TNF or affect the rate at which TNF exerts its anticellular effect have a similar effect on the ability of the TNF to generate DNA fragments. Thus the TNF-mediated DNA fragmentation and the rate at which it occurs correlates with the rate at which cells respond to the anticellular effect of TNF and, as such, might serve as a marker for the responsiveness of cells to TNF.

The development of antibody to the interferon-induced indoleamine 2,3-dioxygenase and the study of the regulation of its synthesis.

Interferon (IFN) treatment of cells induces the synthesis of several new proteins. Antibody to the IFN-induced 42,000 dalton protein has been prepared and used in the study of this protein. The synthesis of the 42,000 dalton protein is dependent on de novo RNA synthesis, as its induction can be blocked if actinomycin D and the IFN are added to the cells simultaneously. Several lines of evidence suggest that the IFN-induced 42,000 dalton protein is the IFN-induced indoleamine 2,3-dioxygenase (IDO). These are as follows: 1) antibody to the 42,000 dalton protein neutralizes the activity of the IDO; 2) examination of a variety of cell lines reveals a correlation between the presence of this protein and the presence of the IDO; and 3) the induction of both the IDO and the 42,000 dalton protein is blocked under conditions in which the IFN treatment is performed in the presence of cycloheximide, and actinomycin D is added to the cells prior to the removal of the cycloheximide. A study of a variety of cell lines has revealed that the induction of the IDO occurs primarily in response to IFN-gamma. Peripheral blood mononuclear cells (PBMC) were the only cell population in which IFN-alpha and IFN-gamma were observed to produce the IDO. The IDO activity induced in IFN-alpha and IFN-gamma treated PBMC is neutralized by antibody to the 42,000 dalton protein, thus demonstrating that the IDO activity induced in these cells by IFN-alpha and IFN-gamma is mediated by the same molecule or antigenically related molecules. Fractionation of the PBMC populations reveals that it is the monocyte that produces the IFN-induced IDO.

Tumor necrosis factor and IFN induce a common set of proteins.

The treatment of cells with TNF or IFN results in the development of an antiviral state and in the induction of a common set of proteins with m.w. of 80,000, 67,000, and 56,000. The induction of the 80,000- and 56,000-Da proteins after TNF treatment is dependent on the synthesis of an intermediary protein, whereas the induction of the 67,000-Da protein appears to occur as a direct result of the TNF treatment. The effects of antibodies to IFN on the TNF-mediated effects have been evaluated and reveal that the incubation of TNF-treated cells with antibody to rIFN-beta 1 greatly reduces the antiviral effectiveness of the TNF treatment and blocks the ability of TNF to induce the 80,000-Da protein. Incubation with antibodies to either IFN-alpha or IFN-gamma failed to affect the TNF-mediated responses. Thus, the induction by TNF of each of the proteins is regulated differently and is mediated through both IFN-dependent and IFN-independent mechanisms.

Production of a monoclonal antibody directed against an interferon-induced 56,000-dalton protein and its use in the study of this protein.

Interferon (IFN) treatment of cells induces the synthesis of several new proteins. A hybridoma cell line producing monoclonal antibody to the IFN-induced 56,000-dalton protein has been developed. The IFN-induced 56,000-dalton protein is synthesized by a variety of different cells and in response to IFN-alpha, IFN-beta, and IFN-gamma. The induction of this protein is dependent on de novo RNA synthesis, since its induction is inhibited if actinomycin D and the IFNs are added to the cells simultaneously. Labeling of IFN-treated cells at 4-h intervals at various times after the addition of the IFNs reveals that the synthesis of the 56,000-dalton protein in IFN-alpha-treated cells peaks within 12 h after the addition of the IFN and is no longer enhanced 20 h after exposure to the IFN. In contrast, IFN-gamma-treated cells continue to show an enhanced synthesis of this IFN-induced protein even after 20 h of exposure to the IFN. Thus, the synthesis of the IFN-induced 56,000-dalton protein is regulated differently by the different IFNs. When cells are treated with IFN-alpha or IFN-gamma in the presence of cycloheximide, and actinomycin D is added prior to the removal of the cycloheximide, the cells produce the IFN-induced 56,000-dalton protein and develop an antiviral state in response to both IFN-alpha and IFN-gamma. These results demonstrate that the synthesis of the 56,000-dalton protein is not dependent on the synthesis of an intermediary protein and that the establishment of an antiviral state occurs in the absence of multiple transcriptional events.