Cytomegalovirus-specific cytolytic T-cell lines and clones generated against adenovirus-pp65-infected dendritic cells.

Cytomegalovirus (CMV) infection is a serious complication of allogeneic bone marrow transplantation (BMT). CMV disease can usually be prevented by passive immunization with donor-derived CMV-pp65-specific T-cell clones if provided early post-BMT. The classic method of generating CMV-specific T-cell clones requires donor-derived fibroblast lines infected with CMV as stimulators, thus limiting the availability of CMV immunotherapy to those patients for whom a donor skin biopsy can be obtained 6 to 8 weeks pretransplantation. To overcome this limitation we have used monocyte-derived dendritic cells (DCs) to induce donor anti-CMV cytotoxic T lymphocytes (CTLs). Matured, adeno-pp65-infected DCs were added at day 0 and at day 7 of a 2-week culture of donor peripheral blood mononuclear cells. DC-primed cultures were compared with cultures stimulated in an identical fashion with CMV-infected fibroblasts or with adeno-pp65-infected freshly isolated blood monocytes. Specific killing of CMV-infected fibroblasts was detected in all except the culture stimulated with pp65-infected monocytes. DCs infected after maturation elicited greater CTL activity than did DCs matured after infection. A series of 5 CD8+ clones from a fibroblast-stimulated culture and 7 CD8+ clones from a mature-DC-stimulated culture derived from a single HLA-A*0201+ individual were characterized. All 12 clones lysed autologous CMV-infected fibroblasts. All except 1 clone from the CMV-infected fibroblast arm (fibroblast arm) lysed vaccinia-pp65-infected B-lymphoblastoid cell lines (BLCLs); none lysed vaccinia-pp150-infected or noninfected BLCLs. Ten of 10 CD8+ clones tested were restricted by HLA-A*0201. Seven of the 12 clones were Vbeta6+ (2 from the fibroblast arm and 5 from the DC arm) with an identical Vbeta6.1-J1.4 sequence. Three clones from the fibroblast arm and 5 clones from the DC arm recognized the pp65 peptide NLVPMVATV (amino acids [aa], 495-503). These data show that CMV-specific T-cell clones with similar restriction patterns, T cell-receptor usage, and specificity can be generated using monocyte-derived pp65-infected-DC or CMV-infected-fibroblast stimulators. This approach should broaden the applicability of CMV-specific T-cell immunotherapy to a wider spectrum of patients by reducing the time required to generate CMV-specific T-cell clones.

Rat cytomegalovirus major immediate-early enhancer switching results in altered growth characteristics.

It has been hypothesized that the major immediate-early (MIE) enhancer of cytomegalovirus (CMV) is important in determining virus tropism and latency because of its essential role in initiating the cascade of early gene expression necessary for virus replication. Although rat CMV (RCMV) and murine CMV (MCMV) exhibit extreme species specificity in vivo, they differ in their ability to replicate in tissue culture. MCMV can replicate in a rat embryo fibroblast (REF) cell line while RCMV does not grow in murine fibroblasts. The tropism is not due to a block in virus entry into the cell. We have constructed a recombinant RCMV in which the RCMV MIE enhancer has been replaced with that of MCMV. Growth of the recombinant virus in tissue culture remains restricted to rat cells, suggesting that other viral and/or host factors are more important in determining in vitro tropism. Unlike findings using recombinant MCMV in which the human CMV (HCMV) MIE enhancer substitutes for the native one (A. Angulo, M. Messerle, U. H. Koszinowski, and P. Ghazal, J. Virol. 72:8502-8509, 1998), infection with our recombinant virus at a low multiplicity of infection resulted in a substantial decrease in virus replication. This occurred despite comparable or increased MIE transcription from the recombinant virus. In vivo experiments showed that the recombinant virus replicates normally in the spleen during acute infection. Notably, the recombinant virus appears to be deficient in spreading to the salivary gland, suggesting a role for the MIE enhancer in tropism for certain tissues involved in virus dissemination. Four months after infection, recombinant virus with the foreign MIE enhancer was reactivated from spleen explants.

Identification and characterization of a spliced C-type lectin-like gene encoded by rat cytomegalovirus.

The English isolate of rat cytomegalovirus (RCMV) encodes a 20-kDa protein with a C-type lectin-like domain that is expressed in the delayed-early and late phases of the viral replication cycle. Genomic sequence analysis of the restriction fragment KpnR of RCMV revealed significant homology to several C-type lectin-containing molecules implicated in natural killer (NK) and T-cell interactions, as well as genes from four poxviruses and African swine fever virus. The gene is spliced into five exons and shows a splicing pattern with exon boundaries similar to those observed in the human differentiation antigen CD69. The cap site of the gene was mapped by RNase protection, 5' rapid amplification of cDNA ends, and primer extension experiments. This analysis demonstrated that the core promoter of the RCMV lectin-like gene contains a GATA rather than a TATA box. Splicing patterns were confirmed with isolates from an infected-cell cDNA library. A unique aspect of the protein is that its translation is not initiated by the canonical methionine but rather by alanine. To study its role in virus replication and pathogenesis, a recombinant virus was constructed in which the gene is interrupted. Replication in tissue culture was similar to that of wild-type virus.

Effects of human cytomegalovirus immediate-early proteins on p53-mediated apoptosis in coronary artery smooth muscle cells.

BACKGROUND: Restenotic and atherosclerotic lesions often contain smooth muscle cells (SMCs), which display high rates of proliferation and apoptosis. Human cytomegalovirus (HCMV) may increase the incidence of restenosis and predispose to atherosclerosis. Although the mechanisms contributing to these processes are unclear, studies demonstrate that one of the immediate-early (IE) gene products of HCMV, IE2-84, binds to and inhibits p53 transcriptional activity. Given the role of p53 in mediating apoptosis, we studied the ability of IE2-84 to inhibit p53-dependent apoptosis in human coronary artery SMCs. METHODS AND RESULTS: Apoptosis of SMCs was induced either by use of an adenovirus vector encoding human wild-type p53 protein or by treatment with doxorubicin. HCMV IE1-72 and IE2-84, the major IE proteins of HCMV, were overexpressed separately with adenovirus vectors encoding each protein, and the effects on p53-induced apoptosis were examined by both nick end-labeling (TUNEL) assay and flow cytometry. Expression of IE2-84, but not IE1-72, protected SMCs from p53-mediated apoptosis. CONCLUSIONS: These data indicate that an HCMV IE protein antagonizes p53-mediated apoptosis, suggesting a pathway by which HCMV infection predisposes to SMC accumulation and thereby contributes to restenosis and atherosclerosis.

Rat cytomegalovirus has a unique immediate early gene enhancer.

The immediate early (IE) enhancer/promoter regions of human cytomegalovirus (HCMV), simian cytomegalovirus (SCMV), and murine cytomegalovirus (MCMV) share the following characteristics: (1) they demonstrate high-level transcriptional properties, (2) multiple repetitive elements are found throughout the enhancer region, and (3) consensus binding sites for cellular transcription factors are frequently located within the repetitive elements. Here we characterize the enhancer/promoter region of the major IE gene locus for rat cytomegalovirus (RCMV). The RCMV IE enhancer/promoter region is able to activate transcription at the same level as HCMV or MCMV IE enhancer/promoter regions. In contrast to HCMV, MCMV, and SCMV, the RCMV IE enhancer/promoter region is almost completely lacking in repetitive elements and the only recognized consensus binding sites for cellular transcription factors recognized by sequence are three CTF, two AP1, and one NF-kappa B binding sites. However, enhancing activity does not appear to be dependent on these individual sites in transient transfection experiments, but is additive across the enhancer region.

Characterization of the major locus of immediate-early genes of rat cytomegalovirus.

A major locus of rat cytomegalovirus (RCMV) immediate-early (IE) RNA transcription was identified. A cDNA library from rat embryo fibroblasts infected with RCMV under IE conditions was constructed and screened by using appropriate RCMV DNA probes, revealing at least two IE genes (IE1 and IE2) transcribed from this locus by differential splicing. The first three exons (the first is noncoding) are spliced to exon 4 to form IE1 and to exon 5 to form IE2. The structural organization of the RCMV major IE region is therefore similar to that of human cytomegalovirus (HCMV) and murine cytomegalovirus (MCMV). When we compared the predicted amino acid sequences of the IE1 proteins of RCMV, HCMV, and MCMV, no areas of homology were found across all three proteins, while a few small areas of homology were found between RCMV IE1 and MCMV IE1. In contrast, large areas of homology were found across the carboxyl half of RCMV IE2, HCMV IE2, and MCMV ie3 proteins. In addition, similarities were found at the beginning of exon 5 of RCMV and MCMV. The possible significance of these conserved regions is discussed. Dinucleotide frequency analysis demonstrated a decrease in CpG frequency over the IE region. The IE gene products were able to transactivate heterologous promoters.

Susceptibility of human herpesvirus 6 to antivirals in vitro.

Human herpesvirus 6 (HHV-6) is a recently recognized human herpesvirus isolated from lymphoid cells and thought to be the causative agent for exanthem subitum. Using dot blot hybridization, the HHV-6 sensitivity pattern to several antivirals was compared with those of herpes simplex virus type 1 and human cytomegalovirus. HHV-6 most closely resembled cytomegalovirus in that it was relatively resistant to the antiviral effects of acyclovir and bromovinyl-deoxyuridine but sensitive to ganciclovir and phosphonoacetic acid. From these results, it appears more likely that HHV-6 infections would respond to ganciclovir and foscarnet than to acyclovir should treatment be deemed advisable, although the low toxicity of acyclovir may allow for its use at doses that might affect replication of HHV-6.

Use of temperature-sensitive mutants of mouse cytomegalovirus as vaccines.

Three temperature-sensitive (ts) mutants of mouse cytomegalovirus (MCMV) were compared with parent virus for their ability to produce acute infection, to stimulate protection against a lethal challenge with salivary gland MCMV, and to become latent and then be reactivated. During the acute phase of infection, ts mutant virus demonstrated very limited replication. During the later phase of infection (seven to 14 days after challenge), titers of virus in the pancreas and salivary glands in mice infected with parent virus continued to rise, whereas no virus could be detected in mice infected with the ts mutants. Prior infection with parent virus or the ts mutants protected susceptible mice from a lethal dose of salivary gland MCMV. One year after infection, treatment of mice with an immunosuppressive regimen resulted in reactivation of parent and ts mutant virus. Reactivated virus recovered from mice infected with ts mutant virus remained temperature sensitive.

Molecular cloning and mapping of rat cytomegalovirus DNA.

The DNA of a rat cytomegalovirus (RCMV) isolate from England is cleaved by HindIII into 25 fragments, 20 of which have been cloned into recombinant plasmids. Twenty-two of thirty KpnI fragments were similarly cloned which, together with the HindIII clones, resulted in cloning of 93% of the viral genome. Terminal fragments were identified and restriction site maps were constructed of the viral genome for HindIII, KpnI, and HpaI. The lack of cross-hybridization among fragments derived with each restriction enzyme and the finding of only two termini support the view that RCMV DNA consists of one long unique sequence. The size of the genome of the English strain of RCMV is approximately 206 kbp, much smaller than that of the previously reported "Dutch" strain.

Genetic analysis of the susceptibility of mouse cytomegalovirus to acyclovir.

Eight independently derived mouse cytomegalovirus (MCMV) mutants resistant to acyclovir (ACV) were obtained by the sequential plating of wild-type virus in increasing concentrations of ACV. Results of complementation studies among these eight mutants suggest that all had mutations within the same or closely associated genes. A ninth MCMV mutant resistant to phosphonoacetate (PAA) derived by plating wild-type virus in the presence of 100 micrograms of PAA per ml displayed coresistance to ACV and was unable to complement any of the ACV-derived mutants. Recombination experiments among all combinations of the nine MCMV mutants were performed and supported the complementation data in that no recombination could be detected. Seven of the eight ACV-resistant mutants demonstrated cross-resistance to PAA and hypersensitivity to aphidicolin. The one mutant not coresistant to PAA was more susceptible to PAA than was the parent virus. Only a few mutants demonstrated coresistance when the mutants were tested against 9-beta-D-arabinofuranosyladenine (ara-A). The ACV mutant that demonstrated increased susceptibility to PAA was 30-fold more susceptible to ara-A but remained unchanged in susceptibility to aphidicolin. Two of the parent-mutant combinations were selected for DNA synthesis analysis in the presence of ACV (5 microM). A significant decrease in DNA synthesis was demonstrated for both parent viruses, and there was little effect on mutant virus DNA synthesis at the same drug concentration. These results suggest that susceptibility of MCMV to ACV is confined to a product of a single gene and that a mutation of this gene can lead to an altered phenotype when compared with parent virus in susceptibility of DNA synthesis to PAA, ara-A, and aphidicolin, drugs that are known to inhibit DNA polymerase activity.

Differences in virulence of clinical isolates of Candida tropicalis and Candida albicans in mice.

Prior reports from this institution indicated that Candida tropicalis was more pathogenic than C. albicans in oncology patients. Pairs of clinical isolates of C. tropicalis and C. albicans recovered from similar patients at other institutions were examined to determine their relative virulence. After intravenous inoculation in normal mice, three pairs of isolates had no significant differences in the 50% lethal dose, and one C. tropicalis isolate was less virulent than its companion C. albicans isolate. In contrast, in mice treated with antibiotics and cytarabine, an antineoplastic drug which damages the gastrointestinal mucosa and produces granulocytopenia, oral inoculation of yeast cells produced striking differences in the 50% infective dose: each C. tropicalis isolate was more virulent than the companion C. albicans isolate from the same institution. The increased virulence of the C. tropicalis isolates compared with the C. albicans isolates when given orally to compromised mice parallels clinical observations in compromised patients.

Acyclovir in mouse cytomegalovirus infections.

A virus-specified thymidine kinase appears to be a general requirement for herpes virus susceptibility to the antiviral effect of acyclovir. Surprisingly, mouse cytomegalovirus (MCMV), which does not encode for a thymidine kinase, is exquisitely sensitive to the drug both in vitro and in vivo. The drug is active against the virus in the absence of a cellular thymidine kinase and the antiviral activity is not diminished in the presence of excess thymidine or a variety of nucleosides and deoxynucleosides. Thus, a thymidine phosphorylation pathway is not required for the drug's activation of this infection. The enzyme system responsible for phosphorylation of the drug has not been identified. Mouse cytomegalovirus mutants resistant to the drug have been isolated, indicating that the antiMCMV effect results from selective inhibition of viral replication rather than indirectly through toxicity to the host cell. Eight resistant mutants appear to be in the same complementation group and seven of the mutants demonstrate coresistance to phosphonoacetic acid, a marker for the DNA polymerase locus of herpes viruses. The evidence to date indicates that the MCMV DNA polymerases is the final site of action of the drug. Investigations of the antiMCMV activity of acyclovir should provide insights into the antiviral effects of this drug and other nucleoside analogs in other herpes virus infections in which the virus does not code for a thymidine kinase (for example, human cytomegalovirus and Epstein-Barr virus).

The value of fungal surveillance cultures as predictors of systemic fungal infections.

Fungal surveillance cultures consisting of urine, stool, and respiratory specimens were analyzed from 37 recipients of bone-marrow transplants and 52 patients undergoing chemotherapy for acute leukemia and other hematologic malignancies. All patients had prolonged aplasia. Sixty-seven percent of the patients were colonized by Candida albicans, and 28% were colonized by Candida tropicalis. No patient was colonized with any species of Aspergillus. There were 21 proven systemic fungal infections: three due to C. albicans, 16 due to C. tropicalis, and two due to Aspergillus. Positive surveillance data for C. tropicalis correlate with disease. Multiple positive-culture data yielded high predictive values (67%-83%), and single positive-culture data yielded slightly lower values as a function of body site. Positive surveillance data for C. albicans did not correlate with disease; negative culture data correlate with the absence of systemic disease due to C. tropicalis and C. albicans. Thus, surveillance data for specific fungal species can aid in diagnosis and appropriate therapy.

Pathogenicity of Candida tropicalis and Candida albicans after gastrointestinal inoculation in mice.

The ability of clinical isolates of Candida albicans and candida tropicalis to invade through normal and damaged gastrointestinal mucosa was determined. Adult mice were treated with either gentamicin or gentamicin and cytarabine. Suspensions of yeast cells (10(7)) were administered through a catheter intraesophageally. Invasion was determined by culturing liver, kidney, and lung tissue from mice sacrificed after 48 h. C. albicans and C. tropicalis were incapable of invading through normal gastrointestinal mucosa in mice treated only with gentamicin. Two isolates of C. tropicalis penetrated the damaged gastrointestinal mucosa in 69% (49 of 71) of mice treated with gentamicin and cytarabine. In contrast, three isolates of C. albicans penetrated he damaged gastrointestinal mucosa in only 23% (14 of 62) of mice. These results suggest that C. tropicalis is more capable of invading through damaged gastrointestinal mucosa than C. albicans. The observations in this mouse model parallel those seen in patients on cytotoxic drugs. Therefore, this model offers a tool for investigation of the pathogenicity of these organisms in a model analogous to the compromised host.

Isolation and characterization of a polyene-resistant variant of Candida tropicalis.

An atypical variant of Candida tropicalis was recovered from multiple specimens from a patient who had been a recipient of a bone marrow transplant. This yeast variant showed atypical morphology on corn meal agar distinguishable from typical isolates of C. tropicalis by the production of clusters of blastospores. Isolates of the variant produced acid, but no gas, from maltose and sucrose in fermentation tests. Isolates from blood, pleural fluid, respiratory secretions, and stool specimens were susceptible to amphotericin B and nystatin in an agar dilution system. However, eight isolates of the variant C. tropicalis recovered over a period of 4 weeks from the patient's urine after amphotericin B therapy were found to be resistant to amphotericin B and nystatin. The isolate recovered after 7 days of therapy had minimal inhibitory concentrations of 100 micrograms of amphotericin B and 20 micrograms of nystatin per ml, whereas the seven isolates recovered subsequently had minimal inhibitory concentrations of greater than 500 micrograms of amphotericin B and 50 micrograms of nystatin per ml. The resistant isolates concomitantly lost the capacity to utilize amino acids that susceptible isolates could utilize. Ultraviolet absorption spectra of nonsaponifiable fractions of whole cells showed that resistant isolates lacked ergosterol, which susceptible isolates contained.

Three serologic tests for candidiasis. Diagnostic value in distinguishing deep or disseminated infection from superficial infection or colonization.

One hundred fifty-one sera from 100 hospitalized patients with positive cultures for yeasts were assayed using whole-cell agglutination (Aggl.), agar gel diffusion (AGD), and counterimmunoelectrophoresis (CEP) to determine the relative diagnostic values of three serologic tests for anti-Candida antibodies. Serial samples were obtained from 29 patients. Tests were read blindly; correlations of the three test results with culture results and clinical findings were determined only after all data had been accumulated. Thirty-five of 100 patients had Aggl. titers of 1:160 or greater, although 13/35 had no evidence of deep or disseminated disease. Twenty-four of 100 patients had clinical or autopsy evidence of deep or disseminated candidiasis; 22/24 had Aggl. titers of 1:160 or greater. Twenty of the 24 patients were CEP-positive, whereas 18/24 were AGD-positive. In five patients CEP became positive earlier (10--21 days) than AGD. Three patients had false-positive precipitin tests, two by both CEP and AGD and the third by CEP only. In this population, a positive CEP and a positive AGD test showed good correlation with deep or disseminated candidiasis, whereas a negatvie Aggl. test showed the best correlation for excluding deep or systemic candidiasis.