Search

Some orthopoxviruses including cowpox virus embed virus particles in dense bodies, comprised of the A-type inclusion (ATI) protein, which may provide long-term environmental protection. This strategy could be beneficial if the host population is sparse or spread is inefficient or indirect. However, the formation of ATI may be neutral or disadvantageous for orthopoxviruses that rely on direct respiratory spread. Disrupted ATI open reading frames in orthopoxviruses such as variola virus, the agent of smallpox, and monkeypox virus suggests that loss of this feature provided positive selection. To test this hypothesis, we constructed cowpox virus mutants with deletion of the ATI gene or another gene required for embedding virions. The ATI deletion mutant caused greater weight loss and higher replication in the respiratory tract than control viruses, supporting our hypothesis. Deletion of the gene for embedding virions had a lesser effect, possibly due to known additional functions of the encoded protein., (Published by Elsevier Inc.)

Classical inbred mice are extensively used for virus research. However, we recently found that some wild-derived inbred mouse strains are more susceptible than classical strains to monkeypox virus. Experiments described here indicated that the 50% lethal dose of vaccinia virus (VACV) and cowpox virus (CPXV) were two logs lower in wild-derived inbred CAST/Ei mice than classical inbred BALB/c mice, whereas there was little difference in the susceptibility of the mouse strains to herpes simplex virus. Live bioluminescence imaging was used to follow spread of pathogenic and attenuated VACV strains and CPXV virus from nasal passages to organs in the chest and abdomen of CAST/Ei mice. Luminescence increased first in the head and then simultaneously in the chest and abdomen in a dose-dependent manner. The spreading kinetics was more rapid with VACV than CPXV although the peak photon flux was similar. These data suggest advantages of CAST/Ei mice for orthopoxvirus studies., (© 2013 Published by Elsevier Inc.)

The licensed smallpox vaccine, comprised of infectious vaccinia, is no longer popular as it is associated with a variety of adverse events. Safer vaccines have been explored such as further attenuated viruses and component designs. However, these alternatives typically provide compromised breadth and strength of protection. We conducted a genome-level screening of cowpox, the ancestral poxvirus, in the broadly immune-presenting C57BL/6 mouse as an approach to discovering novel components with protective capacities. Cowpox coding sequences were synthetically built and directly assayed by genetic immunization for open-reading frames that protect against lethal pulmonary infection. Membrane and non-membrane antigens were identified that partially protect C57BL/6 mice against cowpox and vaccinia challenges without adjuvant or regimen optimization, whereas the 4-pox vaccine did not. New vaccines might be developed from productive combinations of these new and existing antigens to confer potent, broadly efficacious protection and be contraindicated for none.

Intratracheal (i.t.) infection of mice with cowpox virus (CPXV), is lethal at a lower dose than intranasal (i.n.) inoculation. CPXV deleted for cytokine response modifier A (CPXVDeltacrmA) was attenuated compared to CPXV after i.t. inoculation. This attenuation could not be attributed to differences in virus replication, immunomodulators, or cells infiltrating the lungs. Deletion of crmA also caused attenuation during intradermal (i.d.) infection. In contrast to i.t.-inoculated virus, deletion of crmA reduced virus replication at the site of infection. This difference correlated to increased numbers of CD3(+) cells in CPXVDeltacrmA-associated dermal lesions. Thus, crmA is a virulence factor in mice during either pulmonary or dermal cowpox infection; however the influence of crmA is more evident during i.d. inoculation. This suggests that the host immune response differs in the two routes of infection and emphasizes the need to consider the effect of route of infection when examining functions of virulence factors in vivo.

The serpin SPI-3 and the hemagglutinin (HA) encoded by cowpox virus (CPV) block cell-cell fusion, and colocalize at the cell surface. wtCPV does not fuse cells, but inactivation of either gene leads to fusion. SPI-3 mAb added to wtCPV-infected cells caused fusion, confirming that SPI-3 protein at the cell surface prevents fusion. The SPI-3 mAb epitope mapped to an 85-amino acid region at the C-terminus. Removal of either 44 residues from the SPI-3 C-terminus or 48 residues following the N-terminal signal sequence resulted in fusion. Interaction between SPI-3 and HA proteins in infected cells was shown by coimmunoprecipitation. SPI-3/HA was not associated with the A27L "fusion" protein. SPI-3 and HA were able to associate in uninfected cells in the absence of other viral proteins. The HA-binding domain in SPI-3 resided in the C-terminal 229 residues, and did not include helix D, which mediates cofactor interaction in many other serpins.

Host restriction of vaccinia virus has been previously described in CHO and RK13 cells in which a cowpox virus CP77 gene rescues vaccinia virus growth at the viral protein translation level. Here we investigate the restrictive stage of vaccinia virus in HeLa cells using a vaccinia mutant virus (VV-hr) that contains a deletion of 18-kb genome sequences resulting in no growth in HeLa cells. Insertion of CP77 gene into VV-hr generated a recombinant virus (VV-36hr) that multiplied well in HeLa cells. Both viruses could enter cells, initiate viral DNA replication and intermediate gene transcription. However, translation of viral intermediate gene was only detected in cells infected with VV-36hr, indicating that CP77 relieves host restriction at the intermediate gene translation stage in HeLa cells. Caspase-2 and -3 activation was observed in HeLa cells infected with VV-hr coupled with dramatic morphological alterations and cleavage of the translation initiation factor eIF4G. Caspase activation was reduced in HeLa cells infected with VV-36hr, indicating that CP77 acts upstream of caspase activation. Enhanced phosphorylation of PKR and eIF2alpha was also observed in cells infected with VV-hr and was suppressed by CP77. Suppression of eIF4G cleavage with the caspase inhibitor ZVAD did not rescue virus translation, whereas expression of a mutant eIF2alpha protein with an alanine substitution of serine at amino acid position 51 (eIF2alphaS51A) partially restored viral translation and moderately increased virus growth in HeLa cells.

The cowpox virus (CPV) glycoprotein serpin SPI-3, a functional protease inhibitor, and the viral hemagglutinin (HA) are required to prevent fusion of wt CPV infected cells. SPI-3 and HA from CPV infected cells co-localize to the plasma membrane and are found in extracellular enveloped virus (EEV). We also show that an N-terminal SPI-3 signal sequence, but not glycosylation, is required for membrane localization and fusion inhibition. In the absence of HA (CPVDeltaHA), no SPI-3 is found on the membrane and infected cells fuse. Conversely, HA from both wt CPV and CPVDeltaSPI-3 infections is on the membrane, indicating a requirement of HA for SPI-3 plasma membrane localization. In the absence of HA, secretion of SPI-3 or SPI-3 N-glyc(-) was markedly enhanced, suggesting HA serves to retain SPI-3 on the plasma membrane,thereby preventing cell fusion.

The affinities of purified recombinant human IL-18 binding protein (BP) and ectromelia and cowpox virus homologs for human and murine IL-18 were compared by plasmon resonance. The dissociation constants of human IL-18BP were similar for murine and human IL-18. By contrast, the dissociation constants of the viral proteins for murine IL-18 were 12- to 50-fold lower than that for human IL-18. The ectromelia and cowpox virus proteins were biologically active, as judged by their ability to inhibit induction of interferon-gamma by murine and human IL-18. The relative affinities of the orthopoxvirus IL-18BPs are consistent with the rodent host range of the viruses., (Copyright 2001 Academic Press.)

Alastrim variola minor virus, which causes mild smallpox, was first recognized in Florida and South America in the late 19th century. Genome linear double-stranded DNA sequences (186,986 bp) of the alastrim virus Garcia-1966, a laboratory reference strain from an outbreak associated with 0.8% case fatalities in Brazil in 1966, were determined except for a 530-bp fragment of hairpin-loop sequences at each terminus. The DNA sequences (EMBL Accession No. Y16780) showed 206 potential open reading frames for proteins containing >/=60 amino acids. The amino acid sequences of the putative proteins were compared with those reported for vaccinia virus strain Copenhagen and the Asian variola major strains India-1967 and Bangladesh-1975. About one-third of the alastrim viral proteins were 100% identical to correlates in the variola major strains and the remainder were >/=95% identical. Compared with variola major virus DNA, alastrim virus DNA has additional segments of 898 and 627 bp, respectively, within the left and right terminal regions. The former segment aligns well with sequences in other orthopoxviruses, particularly cowpox and vaccinia viruses, and the latter is apparently alastrim-specific., (Copyright 2000 Academic Press.)

The 3' ends of late mRNAs of the ati gene, encoding the major component of the A-type inclusions, are generated by endoribonucleolytic cleavage at a specific site in the primary transcript [Antczak et al., (1992), Proc. Natl. Acad. Sci. USA 89, 12033-12037]. In this study, sequence analysis of cDNAs of the 3' ends of ati mRNAs showed these mRNAs are 3' polyadenylated at the RNA cleavage site. This suggests that ati mRNA 3' end formation involves cleavage of a late transcript, with subsequent 3' polyadenylation of the 5' cleavage product. The RNA cis-acting element, the AX element, directing orientation-dependent formation of these mRNA 3' ends, was mapped to a 345-bp AluI-XbaI fragment. Deletion analyses of this fragment showed that the boundaries of the AX element are within -5 and +38 of the RNA cleavage site. Scanning mutagenesis showed that the AX element contains at least two subelements: subelement I, 5'-UUUAU downward arrowCCGAUAAUUC-3', containing the cleavage site ( downward arrow), separated from the downstream subelement II, 5'-AAUUUCGGAUUUGAAUGC-3', by a 10-nucleotide region, whose composition may be altered without effect on RNA 3' end formation. These features, which differ from those of other elements controlling RNA processing, suggest that the AX element is a component of a novel mechanism of RNA 3' end formation., (Copyright 1999 Academic Press.)

Sequencing and computer analysis of the left (52,283 bp) and right (49,649 bp) variable DNA regions of the cowpox virus strain GRI-90 (CPV-GRI) has revealed 51 and 37 potential open reading frames (ORFs), respectively. Comparison of the structure-function organization of these DNA regions of CPV-GRI with those previously published for corresponding regions of genomes of vaccinia virus, strains Copenhagen (VAC-COP) and Western Reserve (VAC-WR); and variola major virus, strains India-1967 (VAR-IND), Bangladesh-1975 (VAR-BSH); and alastrim variola minor virus, strain Garcia-1966 (VAR-GAR), was performed. Within the left terminal region under study, an extended DNA sequence (14,171 bp), unique to CPV, has been found. Within the right region of the CPV-GRI genome two segments, which are unique to CPV DNA (1579 and 3585 bp) have been found. Numerous differences have been revealed in the genetic structure of CPV-GRI DNA regions, homologous to fragments of the genomes of the above-mentioned orthopoxvirus strains. A cluster of ORFs with structural similarity ot immunomodulatory and host range function of other poxviruses have also been detected. A comparison of the sequences of ORF B, crmA, crmB, crmC, IMP, and CHO hr genes of CPV Brighton strain (CPV-BRI) with the corresponding genes in strain GRI-90 have revealed an identity at the amino acid level ranging from 82 to 96% between the two strains. The findings are significant in light of the recent demonstration of CPV as an important poxvirus model system to probe the precise in vivo role(s) of the unique virally encoded immunomodulatory proteins. Also, the presence of a complete and intact repertoire of immunomodulatory proteins, ring canal proteins family, and host range genes indicates that CPV may have been the most ancient of all studied orthopoxviruses.

Poxvirus genomes encode several proteins which inhibit specific elements of the host immune response. We show the "35K" virulence gene in variola and cowpox viruses, whose vaccinia and Shope fibroma virus equivalents are strongly conserved in sequence, actually encodes a secreted soluble protein with high-affinity binding to virtually all known beta chemokines, but only weak or no affinity to the alpha and gamma classes. The viral protein completely inhibits the biological activity of monocyte chemotactic protein-1 (MCP-1) by competitive inhibition of chemokine binding to cellular receptors. As all beta chemokines are also shown to cross-compete with MCP1 binding to the viral protein, we conclude that this viral chemokine inhibitor (vCCI) not only interacts through a common binding site, but is likely a potent general inhibitor of beta chemokine activity. Unlike many poxvirus virulence genes to date, which are clearly altered forms of acquired cellular genes of the vertebrate immune system, this viral chemokine inhibitor (vCCI) shares no sequence homology with known proteins, including known cellular chemokine receptors, all of which are multiple membrane-spanning proteins. Thus, vCCI presumably has no cellular analogue and instead may be the product of unrelenting sequence variations which gave rise to a completely new protein with similar binding properties to native chemokine receptors. The proposed function of vCCI is inhibition of the proinflammatory (antiviral) activities of beta chemokines.

Vaccinia virus complement control protein (VCP) is encoded by vaccinia virus, with its homolog encoded by other pathogenic poxviruses including variola virus. Since rodents are the primary reservoir hosts of cowpox virus (CPV) and since CPV encodes a highly conserved functional homolog of VCP, termed here the inflammation modulatory protein (IMP), the effects of injection of CPV into the footpads of mice was determined in order to study the precise in vivo effects of IMP. Macroscopic examination of the site of injection with a recombinant virus lacking IMP (CPV-IMP) showed greater tissue damage, with more hemorrhage and induration, than sites injected with the wild-type cowpox virus. In addition, the measurement of the specific swelling response carried out for several weeks revealed significantly greater swelling in mice injected with CPV-IMP. Thus, IMP modulates the complement-activated inflammatory response in vivo. Furthermore, the diminished destruction of host tissue observed in the presence of IMP indicates symbiosis in which the virus ensures the preservation of surrounding host tissue, possibly to support the growth of its progeny.

The Amsacta moorei entomopoxvirus (AmEPv) spheroidin is the most highly expressed late viral gene product in infected insect cells. However, when a cassette containing the spheroidin gene and putative promoter was inserted into cowpox (CPV) or vaccinia viruses, only very low levels of spheroldin gene expression were observed. Primer extension analysis suggests much lower spheroidin gene transcript levels than seen either for the highly expressed CPV A-type inclusion gene or for the spheroidin gene within infected insect cells, indicating that in vertebrate cells, the spheroidin promoter functions poorly if at all. Examination of the spheroidin mRNA synthesized in recombinant CPV shows that the 5' start site of the spheroidin transcript was also unexpectedly imprecise and upstream (approximately 31 bp) of the well-defined start site normally observed in AmEPV-infected insect cells. Sequencing of the 5' terminus of the CPV recombinant spheroidin mRNA suggested that 5' poly(A), a characteristic feature of late poxvirus mRNAs and spheroidin mRNA derived from insect cells, was absent, despite the presence of the typical vertebrate poxvirus late promoter consensus sequence, TAAATG. Our results indicate that insect and vertebrate poxvirus promoters may not be universally interchangeable and imply that there are regulatory features of gene expression unique to the infected insect cell environment.

We show the cowpox genome (Brighton Red strain) contains a single copy gene, crmC, expressed at late times during viral infection, encoding a soluble, secreted protein whose sequence marks it as a new member of the TNF receptor family. The cysteine-rich protein contains 186 amino acids, the N-terminal 21 of which constitute a signal peptide, and two potential N-linked glycosylation sites. The approximately 25-kDa recombinant protein binds TNF specifically and completely inhibits TNF-mediated cytolysis. The strongest sequence homologues are the ligand-binding regions of the type II cellular TNF receptor (TNFRII) and CrmB, a distinct pox virus gene also encoding a soluble TNF binding protein. Unlike TNFRII and CrmB, CrmC does not bind lymphotoxin (LT alpha, TNF beta) and lacks the conserved (but nonhomologous) approximately 150-residue C-terminal domain of CrmB proteins. The presumed function of CrmC is viral inhibition of host-elicited TNF.

We investigated the host-range restriction of a vaccinia virus (VV) K1L deletion mutant in rabbit kidney RK13 cells and the ability of the nonhomologous cowpox virus CP77 gene to overcome this block. Viral early mRNAs were made by K1L- VV but early protein synthesis was arrested consistent with a translational block. Replication of viral DNA did not occur and neither intermediate nor late mRNAs or proteins were detected. These results indicated that host-range restriction occurs earlier in RK13 cells than in Chinese hamster ovary cells (CHO) cells infected with CP77- VV, where the block occurs at translation of intermediate stage mRNA. We confirmed a report (Perkus et al., Virology 179, 276-286, 1990) that the CP77 gene, which allows VV replication in CHO cells, could replace the K1L gene for plaque formation in RK13 cells. However, the size of the plaques formed by K1L-CP77+ VV was smaller than those formed by K1L+CP77- VV. Single-step growth curves also showed that the CP77 gene could functionally replace the K1L gene, although formation of infectious virus was delayed and did not reach the same level as that of K1L+ VV. Most surprisingly, the dramatic shutoff of viral and host gene expression was similar in RK13 cells infected with K1L-CP77- VV and K1L-CP77+ VV and little difference was noted for the first 6 hr. Subsequently, in cells infected with the K1L-CP77+ VV, viral early protein synthesis was spontaneously resurrected and the replication cycle proceeded. Despite the absence of homology, K1L and CP77 gene products appear to be acting in a common virus/cell interaction pathway.

Pig kidney cells (LLC-PK1) were infected with one of three viruses: wild-type cowpox virus (Brighton red strain) expressing the crmA gene; recombinant cowpox virus A602, lacking the crmA gene; or cowpox virus A604, a revertant of virus A602, expressing the crmA gene. The wild-type virus and virus A604 produced identical cytopathic effects consistent with death by necrosis. In these cells, the structural features of the plasma membrane, the nuclear membrane, and the chromatin were maintained until lysis of the cells. In contrast, cowpox virus A602 produced cytopathic effects consistent with death by apoptosis. These effects included loss of microvilli on the cell surface, margination and condensation of the chromatin, progressive convolution of the nuclear membrane, release of dense chromatin masses on disintegration of the nucleus, fragmentation of the DNA, and the generation of apoptotic bodies. These results suggest that the crmA gene is necessary to inhibit processes of apoptosis induced in LLC-PK1 cells by infection with cowpox virus. Thus in cells of certain types, the crmA gene can act with other viral genes to control the mode of death of the virus-infected cell. This capability may be advantageous to virus replication in vivo, potentially facilitating both virus trafficking and interference with antiviral immune defenses.

The inverted terminal repeats of the DNA of cowpox virus (Brighton Red strain) contain the crmB gene, an additional member of a family of viral genes that modify cytokine responses to infection. The crmB gene is transcribed from an early promoter. The primary product is a 355-amino-acid protein containing a signal peptide sequence and three potential N-linked glycosylation sites. The mature gene product is a secreted soluble protein that has an apparent molecular mass of 48 kDa. TNF alpha and TNF beta bind to this protein in a competitive manner, consistent with the sequence of its N-terminal 176 amino acids, which closely resembles the ligand-binding domains of the type II (75-kDa) human TNF receptor. The sequence of the C-terminal 161 amino acids of the CrmB protein is unlike that of human TNF receptors, but overall, the CrmB protein is similar to the T2 proteins of the leporipoxviruses (48% identity) and the predicted product of the G4R/G2R open reading frame of variola virus (85% identity), suggesting that not only the TNF-binding domains but also the C-terminal regions contribute to the functions of these viral proteins. These results show that orthopoxiviruses such as cowpox virus encode secreted forms of TNF receptors that can contribute to the modification of TNF-mediated antiviral processes.

Wild-type rabbitpox virus (RPV) and cowpox virus (CPV) produce red hemorrhagic lesions or pocks upon infection of the chicken chorioallantoic membrane (CAM) of 11-day-old embryonated chicken eggs. However, white, nonhemorrhagic pock variants arise spontaneously within wild-type (wt) populations of either virus at a frequency of about 1%, reflective of complex deletions/rearrangements in the termini of the viral DNA. A subpopulation of the RPV white-pock mutants fail to plaque on pig kidney (PK-15) cells and are referred to as host-range (hr) mutants. In the case of CPV, white-pock formation has been linked to mutations in the SPI-2 (crmA) gene. We show that five spontaneous RPV white-pock host-range mutants (RPV mu hr8sm, RPV mu hr23, RPV mu hr28, RPV mu hr30, and RPV mu hr31) each contain a SPI-2 (crmA) gene and express the crmA protein but lack instead a functional SPI-1 gene. Two other spontaneous RPV white-pock mutants, RPV mu 9 and RPV mu 12, which plaque on PK-15 cells (nonhost-range mutants) contain and express a SPI-1 gene but lack instead a functional SPI-2 gene. Targeted disruption of either the SPI-1 or SPI-2 genes of wtRPV, but only the SPI-2 gene of wtCPV, generates mutants which produce white pocks. The RPV delta SPI-1 mutant fails to plaque on PK-15 or human A549 cells, whereas the RPV delta SPI-2 mutant has a normal host range. No changes in host range compared to wtCPV for either the CPV delta SPI-1 or CPV delta SPI-2 mutants were noted. These differences in phenotypes observed between the two viruses may be reflective of either small sequence variations between the highly conserved SPI-1 or SPI-2 genes or the aggregate phenotypes provided by the other remaining genes.

Intranasal infection of Balb/c mice with 106 plaque forming units (PFU) of wild-type cowpox virus (CPV) and rabbitpox virus (RPV) induced strikingly different pulmonary pathology despite nearly identical clinical signs of illness and LD50. Intranasal infection with CPV induced severe peribronchial, peribronchiolar, and perivascular hemorrhage with a mixed inflammatory cell infiltrate, bronchial and bronchiolar epithelial cell hyperplasia with intracytoplasmic acidophilic inclusion bodies, and alveolar hemorrhage and edema. In contrast, infection with RPV induced a mixed peribronchial and peribronchiolar inflammatory cell infiltrate, multifocal areas of bronchiolar epithelial cell coagulation necrosis, alveolar edema, and a conspicuous absence of pulmonary hemorrhage. Viremia was not detected following CPV infection and only 1 of 11 mice had brain-associated virus at death. Mice infected with RPV exhibited a viremia 2-3 days after infection and all mice had virus associated with the brain at death. Mice infected with viruses containing certain serine protease inhibitor (SPI) gene mutations (CPV delta SPI-1, CPV delta SPI-3, and RPV SPI-1-) exhibited no difference in clinical disease manifestation when compared with those infected with wild-type viruses. However, inactivation of the SPI-2 genes in either CPV or RPV resulted in disease attenuation and alteration of pulmonary pathology. Mice infected with the CPV delta SPI-2 mutant showed decreased pulmonary hemorrhage, reduced inflammation, and an absence of alveolar edema, while mice infected with the RPV delta SPI-2 mutant had a marked increase in intrapulmonary inflammatory cells and only a transient viremia.

The chick embryo chorioallantoic membrane was used to study the acute inflammatory response in the absence of contributions from the immune system. In preliminary experiments, lesions of wild-type cowpox virus strain Brighton (CPV-BR) and a 38K gene deletion mutant of CPV-BR (CPV-BR.D1) were compared with vaccinia virus (strains WR and Copenhagen), fowlpox virus, laryngotracheitis virus, and infectious tenosynovitis virus, and were ranked for degree of induced inflammation. The maximal and minimal inflammatory responses were observed with CPV-BR.D1 and CPV-BR viruses, respectively. CPV-BR.D1 lacks a 38K gene which encodes an anti-inflammatory 38-kDa protein that has homology to SERPINs. The kinetics and character of the inflammatory response were examined further in the wild-type CPV-BR and mutant CPV-BR.D1 infections using cell counts, electron microscopy, and assays for inflammatory cell activation. CPV-BR virus infection rapidly spread through the ectoderm, uniformly infecting all cells with the production of large amounts of virions and viral-induced cytopathic effect, but evoking little or no inflammatory response until 144 hr p.i. The CPV-BR.D1 infection, on the other hand, was rapidly contained by a dexamethasone-sensitive inflammatory response mainly of activated heterophils which was advanced by 36 hr p.i. Both infections resulted in disseminated disease with similar numbers of liver lesions and only a slight difference in the LD50, with the CPV-BR.D1 values being higher than that for CPV-BR virus. In this model, the acute inflammatory response alone is unable to prevent disseminated disease and associated mortality.

We have characterized one of the most highly expressed genes of vaccinia virus, WR strain, in the wild type and in several spontaneous mutants isolated from persistently infected cells. This gene encodes the 94-kDa inclusion protein, which is the vaccinia virus counterpart of the 160-kDa A-type inclusion (ATI) protein of cowpox virus. The homology index between both genes is greater than 95%. A deletion of two consecutive adenylate residues is responsible for a frameshift mutation and premature translational termination in the vaccinia virus gene. In addition, several point mutations and small deletions occur in the 94K gene. The deduced protein contains 725 amino acids, and 4 of the 10 repeated motifs present in the carboxyl terminus of the cowpox virus 160-kDa protein are conserved. In several mutants independently isolated from untreated and interferon-treated persistently infected cells, the gene encodes a 40-kDa protein. In mutant 87-4, this truncated protein is due to the insertion of a cytidilate residue that produces a frameshift mutation and premature translational termination. The deduced protein contains 366 amino acids and has lost all the repetitions. Transcriptional analysis has shown that the steady-state levels of mRNAs in cells infected with the mutants or wild-type vaccinia virus are similar. However, the accumulation of this protein in cells infected with the mutants is reduced indicating some instability. In addition the mutated protein is not recognized by polyclonal antisera. Existence of tandemly repeated sequences at the carboxyl terminus of this family of inclusion proteins correlates with their antigenicity. These results indicate a high degree of mutability of the ATI gene and products, which apparently has no consequence on replication in vitro, but could have relevance to control of the infection by immune responses in animal hosts.

We have used plasmid clones spanning the region encoding the 132-kDa subunit of the cowpox virus RNA polymerase (CPV rpo 132) to marker rescue each of five vaccinia virus (VV) temperature sensitive (ts) mutants, ts 27, ts 29, ts 32, ts 47, and ts 62, which together constitute a single complementation group. The experiments fine-map the vaccinia mutations to a 1.3-kb region containing the 3' end of the CPV rpo 132 gene. Phenotypic characterization shows that all five mutants are affected to varying extents in their ability to synthesize late viral proteins at the nonpermissive temperature, similar to other ts mutants with lesions in the 22- and the 147-kDa subunits of the VV RNA polymerase. Two mutants, ts 27 and ts 32, exhibit a delay in the synthesis of late viral proteins at both the permissive and the nonpermissive temperatures. We conclude that the five VV mutants affect the 132-kDa subunit of the VV RNA polymerase. Additional genetic experiments demonstrate intragenic complementation between ts 62 and three other members of this complementation group, ts 27, ts 29, and ts 32.