Search

The composition of the T lymphocyte population in swine is special in that in addition to the CD4-CD8+ subpopulations, CD4-CD8- and CD4+CD8- and CD4+CD8+ subpopulations are prominent in the peripheral circulating as well as in the resident T lymphocyte pools. Since the same phenotypes are characteristic of thymic populations, it was asked whether the unusual distribution in swine may result from an emigration of thymic precursor phenotypes to the periphery. This explanation was refuted, as all thymic subpopulations were found to express CD1, albeit with differences in antigen density, whereas all extrathymic subpopulations lack CD1. The cellular distribution of CD2 in swine is without precedent among all species studied. Whereas in sheep and cattle the extrathymic CD4-CD8- subpopulation is known to entirely lack CD2 and to have a low propensity for homing to lymphoid tissues, the CD4-CD8- subpopulation in swine splits into CD2+ and CD2- subsets, both of which do reside in lymphoid tissues. While CD2+CD4-CD8- T lymphocytes are rare in the circulating pool, this subset accumulates in spleen and lymph nodes. This may indicate a role for CD2 in homing. Thus the species swine is immunologically unique, not only because of having CD1-CD2+CD4+CD8+ T lymphocytes in the periphery, but also with regard to subdivision and homing behavior of its CD4-CD8- T lymphocyte subpopulation.

The cloned murine cytolytic T-lymphocyte line IE1-IL and several sublines detect a murine cytomegalovirus immediate-early (IE) membrane determinant in conjunction with Ld class I major histocompatibility glycoprotein. The lines retained cytolytic activity, strict antigen specificity, and self-restriction even when adapted to long-term, antigen-independent growth in the presence of interleukin-2 only (M. J. Reddehase, H.-J. Bühring, and U. H. Koszinowski, J. Virol. 57:408-412). These attributes allowed us to use IE1-IL as a stable, monospecific probe for tracing the expression of the IE membrane antigen throughout the viral replication cycle. Presentation of the antigen at the cell membrane proved to be most effective when expression of IE genes in infected mouse embryo fibroblasts was selectively enhanced by consecutive cycloheximide-actinomycin D treatment, whereas without enhancement high numbers of IE1-IL cytolytic T lymphocytes were required to demonstrate the antigen in the IE phase. In the early phase of infection when IE genes were no longer transcribed, cytolysis was not observed, although IE proteins were detectable in the nuclei of the infected cells. Without application of inhibitors IE membrane antigen expression was most prominent during the late phase of infection. Reinitiation of transcription from the genomic region encoding the major IE protein (pp89) and de novo synthesis of pp89 correlated with this reexpression of the IE membrane antigen.

The ability of memory T lymphocytes derived from latently infected mice to control murine cytomegalovirus disease in the immunocompromised host was studied by adoptive transfer experiments. At a stage of pathogenesis when virus had already colonized target tissues, a therapeutic antiviral function could be ascribed to the CD8+ subset. This in vivo function was not restricted to sites in which intravenously infused lymphocytes usually are trapped or home in, such as the lungs or the spleen, respectively, but was also evident in the adrenal glands, a site to which antiviral effector cells have to specifically migrate. Specific infiltration of adrenal gland cortical tissue by donor-derived CD8+ memory T lymphocytes was demonstrated. CD4+ memory T lymphocytes had no antiviral effect by themselves and also were not required for the function of the CD8+ effector cells in this short-term immunotherapy model. These findings should help settle the debate about which subset of T lymphocytes comprises the effector cells that can directly control cytomegalovirus infection in the murine model system.

The nonstructural immediate-early protein pp89 of murine cytomegalovirus (MCMV) is the first viral protein synthesized after infection and has a regulatory function in viral gene expression. Despite its localization in the nucleus of infected cells, pp89 is also the dominant antigen recognized by MCMV-specific cytolytic T lymphocytes. The recombinant vaccinia virus MCMV-ieI-VAC, which expresses pp89, was used to study the capacity of this protein to induce protective immunity in BALB/c mice. Vaccination with MCMV-ieI-VAC induced a long-lasting immunity that protected mice against challenge with a lethal dose of MCMV but did not prevent infection and morbidity. In vivo depletion of CD8+ T lymphocytes before challenge completely abrogated the protective immunity. CD8+ T lymphocytes derived from MCMV-ieI-VAC-primed donors and adoptively transferred into sublethally irradiated and MCMV-infected recipients were found to limit viral replication in host tissues, whereas CD4+ T lymphocytes and pp89-specific antiserum had no protective effect. The data demonstrate for the first time that a single nonstructural viral protein can confer protection against a lethal cytolytic infection and that this immunity is entirely mediated by the CD8+ subpopulation of T lymphocytes.

During the acute cytolytic T lymphocyte (CTL) response of mice to infection with the murine cytomegalovirus two independent populations of activated interleukin-receptive CTL precursors can be demonstrated. One population is specific for cell membrane-incorporated viral structural antigens, whereas the second population detects an antigen, whose appearance is correlated with the synthesis of viral immediate early proteins. Since this new type of antigen is only defined by lymphocyte recognition, it is referred to as the lymphocyte-detected immediate early antigen (LYDIEA). Expression of immediate early antigen precedes the production of viral progeny and, therefore, it is possible that LYDIEA-specific CTL could serve as indicator cells for the very first activities of the viral genome, even during nonproductive infection.

The murine cytomegalovirus protein pp89, which is encoded by gene ieI, is a nonstructural regulatory protein expressed in the immediate-early phase of the viral replication cycle and located mainly in the nucleus of infected cells. Protection of BALB/c (H-2d) mice against a lethal murine cytomegalovirus challenge infection is achieved by vaccination with a recombinant vaccinia virus, MCMV-ieI-VAC, expressing pp89 as the only murine cytomegalovirus gene product. The protection is entirely mediated by T lymphocytes of the CD8+ subset. In the present report, we analyzed the molecular basis of the recognition of pp89 by BALB/c CD8+ cytolytic T lymphocytes. A series of internal and terminal deletion mutants of gene ieI was constructed and cloned in vaccinia virus, and the antigenicity and immunogenicity of the fragments of pp89 expressed by the recombinants were studied. A region of only one-sixth of the protein, from amino acids 154 to 249 and encoded by the fourth exon of gene ieI, was sufficient for both the recognition in vitro of the protein by pp89-specific cytotoxic T lymphocytes and the induction in vivo of pp89-specific cytotoxic T lymphocytes. By using synthetic peptides, the sequence between residues 161 and 179, which is located within the defined domain, was identified as an epitope presented to BALB/C cytotoxic T lymphocytes by the class I major histocompatibility antigen Ld.

Cytolytic T lymphocyte precursors (CTL-P) were sensitized in vivo by intraplantar infection of C57BL/6 mice with a lethal dose of rabies virus, strain ERA (ERA). As a result of sensitization CTL-P matured to interleukin-receptive CTL-P (IL-CTL-P) that could be expanded in vitro to Thy-1+, Lyt-2+ CTL clones in the presence of IL without subjection to antigen-driven selection. After infection with ERA, IL-CTL-P-derived CTL lysed fibroblasts infected with rabies virus but not those infected with another rhabdovirus, the vesicular stomatitis virus. These CTL, however, did not discriminate between fibroblasts infected with the serologically closely related laboratory strains of classic rabies virus, ERA and HEP-Flury, and the serologically distinct rabies-related African isolate Mokola. This finding implies that in vivo sensitized IL-CTL-P recognize common genus-specific determinants expressed on cells infected with members of the lyssavirus genus.

We have shown in a murine model system for cytomegalovirus (CMV) disease in the immunocompromised host that in vivo application of recombinant human IL-2 (rhIL-2) can enhance the antiviral effect of a limited number of CD8+T lymphocytes, not only in prophylaxis, but also in therapy, when virus has already colonized host tissues. The observed net effect of IL-2 was consistent with the assumption of daily effector population doublings. The prospects for IL-2-supported immunotherapy of established CMV infection depend upon the tissues involved in disease. It appears that the prospects for controlling established CMV adrenalitis are less promising than for a therapy of interstitial CMV pneumonia.

We have constructed target cells by cotransfection of the MHC gene Ld and fragments of murine cytomegalovirus (MCMV) DNA coding for nonstructural immediate-early (IE) proteins. Transfectants were tested by using CTL clone IE1 with specificity for an IE epitope presented in association with Ld. Data show that clone IE1 recognizes a product of the ie1 transcription unit of MCMV, and that its specificity is shared by approximately 25% of polyclonal IE-specific CTL. The results provide the first definite evidence that expression of a herpes virus IE gene encoding a regulatory protein gives rise to antigen expression detectable by specific CTL.

We have shown in a murine model system for acute, lethal cytomegalovirus (CMV) disease in the immunocompromised natural host that control of virus multiplication in tissues, protection from virus-caused tissue destruction, and survival are mediated by virus-specific CD8+ CD4-T lymphocytes. Protection from a lethal course of disease did not result in a rapid establishment of virus latency, but led to a long-lasting, persistent state of infection. The CD8- CD4+ subset of T lymphocytes was not effective by itself in controlling murine CMV (MCMV) multiplication in tissue or essential for the protective function of the CD8+ CD4- effector cells. The antiviral efficacy of the purified CD8+ CD4- subset was not impaired by preincubation with fibroblasts that presented viral structural antigens, but was significantly reduced after depletion of effector cells specific for the nonstructural immediate-early antigens of MCMV, which are specified by the first among a multitude of viral genes expressed during MCMV replication in permissive cells. Thus, MCMV disease provides the first example of a role for nonstructural herpesvirus immediate-early antigens in protective immunity.

Interstitial pneumonia linked with reactivation of latent human cytomegalovirus due to iatrogenic immunosuppression can be a serious complication of bone marrow transplantation therapy of aplastic anaemia and acute leukaemia. Cellular immunity plays a critical role in the immune surveillance of inapparent cytomegalovirus infections in man and the mouse. The molecular basis of latency, however, and the interaction between latently or recurrently infected cells and the immune system of the host are poorly understood. We have detected a so far unknown antigen in the mouse model. This antigen is found in infected cells in association with the expression of the herpesvirus 'immediate early' genes and is recognized by cytolytic T lymphocytes (CTL). We now demonstrate that an unexpectedly high proportion of the CTL precursors generated in vivo during acute murine cytomegalovirus infection are specific for cells that selectively synthesize immediate early proteins, indicating an immunodominant role of viral non-structural proteins.

The regulation of antigen processing and presentation to MHC class I-restricted cytolytic T lymphocytes was studied in cells infected with murine cytomegalovirus. Recognition by cytolytic T lymphocytes of the phosphoprotein pp89, the immunodominant viral antigen expressed in the immediate-early phase of infection, was selectively prevented during the subsequent expression of viral early genes. The surface expression of MHC class I glycoproteins and their capacity to present externally added pp89-derived antigenic peptides were not affected. Because recognition of several other antigens occurred during the early phase, a general failure in processing and presentation was excluded. Since neither rate of synthesis, amount, stability, nor nuclear transport of pp89 was modified, the failure in recognition indicates a selective interference with pp89 antigen processing and presentation.

The aim of this study was to determine the number and state of activity of cytolytic T lymphocytes (CTL) and their precursors (CTL-P) present in vivo during the early stages of viral infection. The local response to lethal infection with rabies virus was used as a model system that is not accessible to analysis by secondary activation in vitro. The local response to alloantigen served as a control. Experimental protocols were established that allow frequency estimates of in vivo antigen-triggered CTL-P. Data allow a distinction between CTL-P activated in vivo by alloantigen and viral antigen with respect to their different capacity to utilize T cell growth factors (interleukins). In vivo alloantigen-primed CTL-P generate, in vitro, an active effector progeny in the presence of interleukins of xenogeneic origin, whereas the majority of virus-specific CTL-P, in spite of considerable expansion in vivo, fail to generate CTL in vitro unless antigen is added.

The expression of an antigen on porcine T lymphocytes detected by murine monoclonal antibody (mAb) 8/1 was investigated by functional studies and dual-parameter immunofluorescence. mAb 8/1 reacts with greater than 95% of thymocytes and in peripheral blood with all T lymphocytes and with cells of the monocyte/macrophage lineage, but not with B cells, erythrocytes, and platelets. Pretreatment of peripheral blood lymphocytes with mAb 8/1 plus complement abrogated the proliferative response in vitro to mitogen, soluble antigen, and MHC determinants. Dual-parameter immunofluorescence revealed that resting porcine T8+ as well as T4+ lymphocytes express the 8/1 antigen, whereas after in vitro activation, cell surface expression of the antigen was low or absent in both T cell subsets. Thus, the 8/1 antigen represents a marker that discriminates between resting and activated T lymphocytes. Distribution and functional criteria indicate that 8/1 represents a novel marker not described before for any other mammalian species.

The existence of four subpopulations in resting porcine T lymphocytes is documented. In addition to the two known subpopulations which are typified by a mutually exclusive expression of either the CD8 or the CD4 differentiation antigen, CD4-CD8+ and CD4+CD8- T lymphocytes, respectively, two unusual subpopulations were prominent not only in peripheral blood, but also in lymphoid tissues: CD4-CD8- T lymphocytes expressing neither of these antigens and CD4+CD8+ T lymphocytes coexpressing both antigens. While CD4+CD8+ lymphoblasts have been found also in other species, resting T lymphocytes with that phenotype are without precedent among all species analyzed to date. This unique composition of the porcine T lymphocyte population has to be taken into consideration when the swine is used as a large animal model in experimental medicine.

The infection of the mouse with murine cytomegalovirus (MCMV) served as a model system to understand the biology of human CMV infection. The contribution of cytolytic T lymphocytes (CTL) to the recovery from infection was studied. Protection against lethal MCMV disease could be conferred on immunodepleted hosts by adoptive transfer of lymphocytes. The antiviral effect was mediated by specifically sensitized T lymphocytes of the CD8+ subset. These cells limited viral spread, prevented tissue destruction by viral cytopathic effects, and protected from lethal disease. Transferred cells have protective therapeutic function even when the virus has already colonized host tissues. CD8+ cells do not require the contribution of CD4+ cells for in vivo function. Selective expression of immediate-early (IE) phase genes in target cells allowed the detection of the immunodominant IE antigen recognized by CTL. The major IE gene ieI encodes a non-structural viral phosphoprotein, pp89, which resides in the nucleus of infected cells where it acts as transcriptional regulator. Expression of gene ieI is under temporal control, and membrane presentation of the protein domain detected by CTL is down-regulated by MCMV early-phase products. A recombinant vaccinia virus expressing gene ieI induced immunity that protected mice against a subsequent challenge with a lethal dose of MCMV. The protective effect was entirely mediated by CD8+ T lymphocytes. Thus, an experimental vaccine expressing a single nonstructural herpesvirus protein can induce a protective cellular immune response.

Expression of the differentiation antigen 8/1, a porcine leucocyte marker for which no counterpart is known as yet in other species, was tested in relation to the expression of surface Ig and the monocyte marker 74-22-15. B lymphocytes and T lymphocytes can be typified by the exclusive expression of surface Ig and 8/1, respectively, while all monocytes defined by the expression of the 74-22-15 antigen co-express 8/1 with the highest staining density observed for this marker. 8/1+ T lymphocyte subsets defined by the expression of CD4 and CD8 cell surface antigens differed in the expression of 8/1 with the staining density being high in CD4+ T lymphocytes, intermediate in CD8+ as well as in CD4+CD8+ T lymphocytes, and low in the CD4-CD8- subset. Practically all mononuclear leucocytes present in peripheral blood are included in the subpopulations defined by this kit of reagents, and thus it can be predicted that if mononuclear leucocytes which do not belong to any of these subpopulations exist at all, these must be rare and will not be accessible to an analysis without enrichment.

We have established a murine model system for exploring the ability of a CD4 subset-deficient host to cope with cytomegalovirus infection, and reported three findings. First, an antiviral response of the CD8 subset of T lymphocytes could be not only initiated but also maintained for a long period of time despite a continued absence of the CD4 subset, whereas the production of antiviral antibody proved strictly dependent upon help provided by the CD4 subset. Second, no function in the defense against infection could be ascribed as yet to CD4-CD8- T lymphocytes, which were seen to accumulate to a new subset as a result of depletion of the CD4 subset. This newly arising subset did not substitute for CD4+ T lymphocytes in providing help to B lymphocytes, and was also not effective in controlling the spread of virus in host tissues. As long as a function of these cells in the generation and maintenance of a CD8 subset-mediated response is not disproved, caution is indicated with concern to an autonomy of the CD8 subset. Third, even though with delay, the CD8+ effector cells raised in the CD4 subset-deficient host were able of clear vital tissues from productive infection and to restrict asymptomatic, persistent infection to acinar glandular epithelial cells in salivary gland tissue.

Long-term cytolytic T-lymphocyte (CTL) lines that are specific for distinct antigens associated with different phases of the replicative cycle of the murine cytomegalovirus (MCMV) were established by cloning of CTL lines derived from lymph nodes of latently infected BALB/c mice. Two CTL clones were characterized in detail. Both displayed the Lyt-2+, L3T4- surface phenotype, and the recognition of their respective target antigens was class I (DLd) major histocompatibility complex antigen restricted. Clone S1 was specific for a structural antigen of MCMV, and clone IE1 detected an MCMV-specified immediate-early (IE) membrane antigen. Clone IE1 retained lytic activity, antigen specificity, and self-restriction after prolonged propagation in the presence of recombinant human interleukin-2 without restimulation by antigen. This interleukin-2-dependent line of the clone IE1, line IE1-IL, can serve as a reference line for the definition of the antigenic determinant IE1 of an IE membrane antigen.

Mutagen treatment of mouse P815 tumor cells produces immunogenic mutants that express new transplantation antigens (tum- antigens) recognized by cytolytic T cells. We found that the gene conferring expression of tum- antigen P91A contains 12 exons, encoding a 60 kd protein lacking a typical N-terminal signal sequence. The sequence shows no significant similarity with sequences in current data bases. A mutation that causes expression of the antigen is located in exon 4; it is the only apparent difference between the normal and the antigenic alleles. A short synthetic peptide corresponding to a region of exon 4 located around this mutation makes P815 cells sensitive to lysis by anti-P91A cytolytic T cells. The mutation creates a strong aggretope enabling the peptide to bind the H-2 Ld molecule. Several secondary tumor cell variants that no longer express tum- antigen P91A were found to carry deletions in the gene.

Limiting dilution (LD) analysis with two modifications, the expansion and the restimulation LD assay, led to the detection and quantification of two distinct in vivo maturation stages within the lineage of virus-specific self-restricted CTL after infection of mice with the murine cytomegalovirus (MCMV). A low frequency set, representing an average of 15% of the specifically activated CTL-P in a draining lymph node, generated virus-specific lytic activity in the absence of antigen, solely under expansion conditions provided by growth and differentiation interleukins. These cells were considered to be active and were denoted antigen-independent or interleukin-receptive CTL-P (IL-CTL-P). A high frequency set required additional antigen in vitro to generate functionally active clones, and therefore the cells were termed antigen-dependent. Both sets are present in vivo simultaneously at the peak of the acute immune response and represent antigen-activated cells because their existence strictly depends on a preceding priming event. IL-CTL-P disappear quickly after acute infection and are absent during the memory state. It is proposed that the isolation of IL-CTL-P could serve to detect viral antigen expression during persistent and/or recurrent herpes virus infections.

We have shown in a murine model system for cytomegalovirus (CMV) disease in the immunocompromised host that CMV infection interferes with the earliest detectable step in hemopoiesis, the generation of the stem cell CFU-S-I, and thereby prevents the autoreconstitution of bone marrow after sublethal irradiation. The antihemopoietic effect could not be ascribed to a direct infection of stem cells. The failure in hemopoiesis was prevented by adoptive transfer of antiviral CD8+ T lymphocytes and could be overcome by syngeneic bone marrow transplantation. CD8+ T lymphocytes and bone marrow cells both mediated survival, although only CD8+ T lymphocytes were able to limit virus multiplication in host tissues. We concluded that not the cytopathic effect of virus replication in host tissues, but the failure in hemopoiesis, is the primary cause of death in murine CMV disease.

Peptides that are antigenic for T lymphocytes are ligands for two receptors, the class I or II glycoproteins that are encoded by genes in the major histocompatibility complex, and the idiotypic alpha/beta chain T-cell antigen receptor. That a peptide must bind to an MHC molecule to interact with a T-cell antigen receptor is the molecular basis of the MHC restriction of antigen-recognition by T lymphocytes. In such a trimolecular interaction the amino-acid sequence of the peptide must specify the contact with both receptors: agretope residues bind to the MHC receptor and epitope residues bind to the T-cell antigen receptor. From a compilation of known antigenic peptides, two algorithms have been proposed to predict antigenic sites in proteins. One algorithm uses linear motifs in the sequence, whereas the other considers peptide conformation and predicts antigenicity for amphipathic alpha-helices. We report here that a systematic delimitation of an antigenic site precisely identifies a predicted pentapeptide motif as the minimal antigenic determinant presented by a class I MHC molecule and recognized by a cytolytic T lymphocyte clone.

Interstitial pneumonia associated with viral replication in lung tissue was observed after cytomegalovirus infection of total-body gamma-irradiated mice, whereas in noncompromised hosts the lungs were not affected and virus multiplication was restricted to the salivary glands. The radiation damage could either predispose normally nonpermissive cell types for productive infection or abrogate an immune control of the tissue manifestation of infection by elimination of lymphocytes. Adoptive transfer of lymphoid cells into irradiated, infected recipients supported the second alternative. Even when infection was established in the lungs, as manifested by the presence of infected lung tissue cells in the alveolar septa, an antiviral effect could be assigned to the Lyt-2+, L3T4- subset of T lymphocytes specifically sensitized in the immunocompetent donor. These cells did not require in vitro propagation to perform effector cell functions in vivo and were operative under physiological conditions in comparatively low numbers. Hence, there is reason to assume that T lymphocytes are responsible for the tissue distribution of cytomegalovirus replication during infection.

To confirm that immediate-early (IE) genes of murine cytomegalovirus (MCMV) give rise to antigens recognized by specific cytolytic T lymphocytes (CTL), a 10.8-kilobase fragment of MCMV DNA which is abundantly transcribed at IE times was transfected into L cells expressing the Ld class I major histocompatibility glycoprotein. The viral genome fragment contains sequences of the three IE transcription units of MCMV: ie1, ie2, and ie3. In the transfected cell lines, only the predominant 2.75-kilobase transcript of ie1 and its translation product pp89 could be detected. The transfectants were analyzed for membrane expression of an IE antigen by employing clone IE1, an IE-specific CTL clone, as the probe. Only cells that expressed both the MCMV IE gene(s) and the Ld gene were recognized by the CTL clone.