Cytomegalovirus vaccine. Specific humoral and cellular immune responses in human volunteers.

Four seronegative adult male volunteers were immunized with Towne strain cytomegalovirus (CMV) vaccine. The only reaction was transient pain and swelling at the inoculation site. Viral cultures were performed during the first 12 weeks after immunization, and CMV was not recovered from throat, urine, or peripheral blood mononuclear cells. Both CMV-specific humoral and cellular immunity developed within three weeks of vaccination. Whereas humoral antibody titers declined steadily with time, the cellular immune responses seemed biphasic. An early peak in lymphocyte proliferation to CMV antigen occurred three to six weeks after immunization. Responses then diminished but increased again six to ten months after immunization. This study in a small group of normal male volunteers indicated that CMV vaccine was safe and immunogenic. That CMV vaccine elicited CMV-specific humoral and cell-mediated immunity is important, because there is evidence that both are necessary for protection from CMV infections.

Identification and molecular characterization of the murine cytomegalovirus homolog of the human cytomegalovirus UL100 gene.

The UL100 gene of human cytomegalovirus (HCMV) has been shown to encode an envelope glycoprotein that might play an important role in HCMV infection. Using the HCMV UL100 gene as a probe in low stringency hybridization studies, we were able to identify the putative UL100 homologous gene of murine cytomegalovirus (MCMV), strain Smith. The MCMV gene was sequenced and localized to the left end of the EcoRI fragment H on the MCMV physical map. This genomic location is similar to that found for the HCMV UL100 gene. The MCMV UL100 mRNA is 1.6 kb in size and is expressed exclusively in the late stages of infection. The 5' and 3' boundaries of the transcript were determined. The open reading frame (ORF) of the UL100 gene could encode a protein of 371 amino acid residues with a calculated molecular mass of 42 kDa. Computer analysis of the deduced amino acid sequence of this gene predicted the presence of eight transmembrane domains and four N-linked glycosylation sites in the protein. Sequence comparison revealed that this putative protein shares similarity with the predicted UL100 homologs of several other herpesviruses, and is most similar to the HCMV UL100 protein (47% identity).

Dw subtypes of serologically defined DR-DQ specificities restrict recognition of cytomegalovirus.

We have investigated the relationship between serologically defined (Ia) and T lymphocyte-defined (LD/Dw) determinants in restricted recognition of cytomegalovirus (CMV) by human T lymphocytes. T lymphocytes isolated from CMV seropositive individuals expressing DQw3/DR4/Dw4 antigens were "sensitized" to CMV in vitro; CMV-specific blasts were isolated and tested for their ability to recognize CMV presented by cells expressing different DR4-associated Dw antigens (i.e., Dw4, Dw10, Dw13, Dw14, and Dw15). Similar studies were also performed using T lymphocytes from individuals expressing DQw1/DR2/Dw2 specificities and antigen presenting cells (APC) expressing the DR2-associated Dw/LD subtypes, Dw2, Dw12, and LD-MN2. CMV-specific T cell blasts were used as responding cells in order to reduce nonspecific background alloresponses which occur with allogeneic APC. In all cases it was found that the determinants involved in restricted recognition of CMV were subtypic to the DR-associated Ia specificities. To distinguish whether Dw specificities associated with DQ or with DR molecules, or both, were involved in these responses, we used anti-DR (L243) and an anti-DQwl (S3/4) monoclonal antibodies (MoAb) to block CMV-specific responses. Both MoAb significantly blocked responses, suggesting that determinants associated with both DR and DQ molecules are involved in restricted recognition of CMV by T cells.

A combination of human cytomegalovirus (HCMV)-specific murine monoclonal antibodies exhibits synergistic antiviral activity in vitro.

A combination of HCMV-specific monoclonal antibodies (MAbs) reactive with glycoproteins in gcI complexes which exhibit synergistic antiviral activity in vitro is described. MAbs directed against different structural and biological properties of HCMV have been selected to increase the antiviral activity against all possible strains, and to reduce the likelihood that resistant strains will emerge with prolonged exposure. Furthermore, in vitro analysis demonstrates that certain of the MAbs in the combination augment the virus-neutralizing activity of other component antibodies, thereby decreasing the amount of total antibody protein required to inhibit HCMV infection. Certain MAbs have been selected to inactivate extracellular virus during the early phase of HCMV infection, whereas others have been selected to prevent its spread once cells have been infected. These data suggest that a MAb cocktail may be useful for prophylaxis and treatment of patients at risk of life-threatening HCMV infections.

Varicella-zoster-specific immune responses in acute herpes zoster during a placebo-controlled trial of oral acyclovir therapy.

During a placebo-controlled trial of oral acyclovir therapy for acute zoster in immunocompetent patients, we examined the blastogenic response of peripheral blood mononuclear cells and antibody titers in both placebo and acyclovir recipients to determine whether the drug affected the cell-mediated or humoral immune responses. Proliferative responses to mitogens and two dilutions of varicella-zoster virus antigen were not inhibited when fresh peripheral blood mononuclear cells were simultaneously tested in autologous sera collected before and on day 7 of a 10-day course of 2 g/day of oral acyclovir (plasma drug levels averaged 4.6 microM). Using cryopreserved cells from study days 0, 3, 7, 14, and 30, thawed and tested simultaneously, there was no significant difference at the p less than or equal to 0.05 level between the net proliferative responses at each time point for the two groups. On day 14, however, the proliferative response of the acyclovir group was approximately 50% lower than that of the placebo group. Geometric mean antibody titer rises to varicella-zoster virus were also lower among drug recipients but not significantly so. Although this dose of acyclovir did not have a statistically significant effect on lymphocyte proliferative responses to varicella-zoster virus antigen or antibody titers, the lower values in drug recipients may be a reflection of the ability of acyclovir to terminate viral replication, thus reducing the patient's antigenic burden.

Characterization of the role of mononuclear cell subpopulations in the in vitro lymphocyte proliferation assay.

Purified human mononuclear cell subpopulations have been evaluated in the in vitro lymphocyte proliferation assay. Monocyte-depleted mononuclear cells had reduced or absent responses to mitogens and antigens which could be restored to the original mononuclear cell response by addition of purified plastic adherent cells. Purified T lymphocytes obtained by density gradient centrifugation of E-rosette-forming cells demonstrated low but significant proliferative responses to mitogens, but no significant response to antigens. The addition of monocytes potentiated the response of purified T cells to mitogens and antigens, but did not fully reconstitute the original mononuclear cell response unless non-T lymphocytes were also present. It is concluded that mononuclear cell proliferation represents a complex mechanism of cellular interaction involving multiple subpopulations of cells.

Cytomegalovirus (CMV): specific lymphokine production in congenital CMV infection.

The studies reported here were designed to examine cytomegalovirus (CMV)-specific lymphokine production in infants with congenital CMV infection to elucidate the role of the efferent limb of the immune response to this virus. Mononuclear cells (MNC) from adult control donors and congenitally infected infants were stimulated with mitogen (PHA) or antigens (CMV, mumps), and the supernatants were assayed for production of migration inhibitory factor (MIF) and interferon (IFN). Concordance was observed between lymphocyte proliferative responses to CMV antigen and production of MIF in both control donors and congenitally infected infants. PHA-stimulated cultures from both controls and patients resulted in immune-specific IFN-gamma production in all cases. CMV-stimulated MNC from controls produced IFN when lymphocyte proliferation responses to the antigen were present, whereas those with absent proliferation did not produce IFN. CMV-induced IFN was detected in several patients who had reduced lymphocyte proliferative responses to CMV. The predominant species of IFN stimulated by CMV was neutralized by anti-IFN-gamma, suggesting that CMV induced primarily immune-specific IFN-gamma. In some cases, IFN activity was also reduced to a lesser degree by anti-IFN-alpha, suggesting either the presence of IFN-alpha or of cross-reacting antigenic determinants shared by the two species of IFN. Mumps viral antigen induced primarily IFN-alpha regardless of the immunological status of the donor. We conclude that lymphokine production in congenital CMV reflects the state of activation and/or proliferation of CMV-specific T helper cells rather than an intrinsic defect in the efferent limb of the immune response.

Cytomegalovirus (CMV)-specific lysis of CMV-infected target cells can be mediated by both NK-like and virus-specific cytotoxic T lymphocytes.

The host response to cytomegalovirus (CMV) involves both humoral and cell-mediated immunity. Because virus-specific cytotoxicity appears to be associated with recovery from CMV infection, we have investigated spontaneous (NK) and cytotoxic T cell (Tc) activity against CMV-infected target cells in normal donors and infants with active CMV infection. Fresh mononuclear cells from seropositive donors expressed low-level, non-MHC-restricted cytotoxic activity that preferentially lysed CMV-infected but not uninfected autologous and allogeneic fibroblasts. No cytotoxic activity was observed when mononuclear cells from seronegative donors were studied. Mononuclear cells from infants with active CMV infection and mononuclear cells from seropositive adults that were pre-incubated with cell-free CMV antigen in bulk culture for 6 days had enhanced cytotoxic activity against CMV-infected target cells sharing one or more class I MHC determinants with the effector cell populations. Selective enrichment or depletion experiments were performed to characterize the effector cell populations involved in the augmented cytotoxic response following in vitro induction with CMV antigen. Purified E-rosette-forming cells expressed increased cytotoxic activity against both virus-infected fibroblasts and the NK target cell, K562. Depletion experiments with monoclonal antibodies and complement indicated that the CMV-specific cytotoxic response involves both an E-rosette-positive, T8+, M1- T-cell subset that most likely represents CMV-specific Tc, and an NK population that can be induced to differentiate by in vitro stimulation with CMV antigen following quantitative NK-cell depletion with monoclonal antibody, B73.1.

Characterization of a human cytomegalovirus glycoprotein complex (gcI).

Three distinct families of glycoprotein complexes present in the envelopes of human cytomegalovirus and designated gcI, gcII and gcIII have been described recently. The synthesis of the gcI family was analysed using either inhibitors of glycoprotein processing and transport or endoglycosidase treatments of purified glycoproteins. The initial step in gcI synthesis involved the glycosylation of a 95K protein (p95) to form a high-mannose, simple N-linked glycoprotein of Mr 158K (gp158), which was detected only in the presence of the glycoprotein processing inhibitor castanospermine. This intermediate was rapidly trimmed in the virus-infected cell to form a more stable simple N-linked precursor glycoprotein of Mr 138K (gp138). Treatment of either gp158 or gp138 with endoglycosidase H produced p95. Both molecules, gp158 and gp138, were found in disulphide-linked complexes which are presumably infected cell precursors to gcI since they were not found in virions. The processing of these complexes involved complete cleavage of gp138 and conversion of some but not all of its oligosaccharide to complex N-linked chains. Both processing events were inhibited by the ionophore monensin. Mature gcI contained the gp138 cleavage product, gp93-130. The latter glycoprotein could be separated into two electrophoretic forms, gp93 and gp130. The deglycosylated form of gp55 had a discrete banding pattern with an apparent Mr of 46K (p46). In contrast, the deglycosylated forms of gp93 and gp130 had diffuse banding patterns with apparent Mr values of 46K to 56K (p46-56) and 60K to 70K (p60-70) respectively. Peptide profiles comparing gp93 with gp130 indicated that they have highly similar polypeptide backbones. Since the deglycosylated forms of gp55 and gp130, 46K and 60K to 70K, respectively, together exceed the 95K precursor/deglycosylated intermediate in Mr, we propose that the above glycoproteins are derived by an alternative proteolytic cleavage of the precursor. The heterogeneous electrophoretic properties of the deglycosylated forms of gp93 and gp130 may be due to additional post-translational modifications other than glycosylation.

Human immune responses to major human cytomegalovirus glycoprotein complexes.

Sera from both human cytomegalovirus (HCMV)-seropositive adults and infants with congenital HCMV infection recognized two major HCMV glycoprotein complexes. However, proliferative responses of peripheral blood mononuclear cells to these complexes varied among seropositive adults and were not detected in any of the infants. Thus, these glycoproteins alone may not be sufficient to develop a subviral HCMV vaccine.

Immune mechanisms in congenital cytomegalovirus infection: activation of CMV-specific T helper cells (CMV-Th) by exogenous IL-2.

Infants with congenital CMV infection have a specific defect in CMV-induced lymphocyte proliferation, providing a model for investigation of mechanisms of viral immune recognition and immune response. In the present study the possible role of a defect in lymphokine activation of CMV-specific T helper cells (Th) was examined. IL-1 activity was detected in supernatants of patient mononuclear cell (MNC) cultures stimulated with CMV. In contrast, no IL-2 activity could be detected in supernatants of CMV-stimulated MNC cultures, whereas PHA induced normal IL-2 production. Addition of low concentrations of either crude TCGF or recombinant IL-2 (rIL-2) resulted in 2-4 fold augmentation of CMV-specific lymphocyte proliferation; exogenous IL-2 had no effect on MNC responses to HSV. CMV-specific Th lines/clones were established from three congenital CMV patients by initial stimulation of MNCs with CMV antigen and 0.1 U/ml rIL-2, followed by repeated stimulation with CMV, HLA-matched allogeneic feeder cells and 10% TCGF. The resulting CMV Th lines/clones proliferated specifically in response to stimulation with CMV antigen and produced endogenous IL-2. Thus, the immune deficiency associated with congenital CMV may either be due to an intrinsic defect in CMV-Th activation or CMV-specific suppressor cell activity.

Epstein-Barr virus-transformed lymphoblastoid cell lines as antigen-presenting cells and "augmenting" cells for human CMV-specific Th clones.

The ability of Epstein-Barr virus-transformed lymphoblastoid cell lines (LCL) to present human cytomegalovirus (HCMV) antigen to a panel of HCMV-specific T helper (Th) clones was evaluated. Among the seven Th clones studied, only one clone (SP-CN/T3-16) proliferated well to HCMV presented by both autologous mononuclear cells (MNC) and LCL, and one clone (SP-CN/T3-9) proliferated significantly better to HCMV presented by autologous LCL than by autologous MNC. The majority of the HCMV-specific Th clones tested (five out of seven) responded much better to HCMV presented by MNC than to HCMV presented by LCL. The mechanism(s) responsible for the inefficiency of LCL to present HCMV to certain clones was studied. Our results suggested that the defect of LCL is not due to insufficient interleukin 1 production, insufficient MHC class II molecule expression, nor an inhibitory mechanism or factor. In this report, we also demonstrate that by adding a minimum amount of LCL along with MNC as antigen-presenting cells (APC), one can restimulate and expand Th clones much more efficiently than by using MNC alone as APC.