Suppression of CCR5- but not CXCR4-tropic HIV-1 in lymphoid tissue by human herpesvirus 6.

HIV-1 infects target cells via a receptor complex formed by CD4 and a chemokine receptor, primarily CCR5 or CXCR4 (ref. 1). Commonly, HIV-1 transmission is mediated by CCR5-tropic variants, also designated slow/low, non-syncytia-inducer or macrophage-tropic, which dominate the early stages of HIV-1 infection and frequently persist during the entire course of the disease. In contrast, HIV-1 variants that use CXCR4 are typically detected at the later stages, and are associated with a rapid decline in CD4+ T cells and progression to AIDS (refs. 2,7-11). Disease progression is also associated with the emergence of concurrent infections that may affect the course of HIV disease by unknown mechanisms. A lymphotropic agent frequently reactivated in HIV-infected patients is human herpesvirus 6 (HHV-6), which has been proposed as a cofactor in AIDS progression. Here we show that in human lymphoid tissue ex vivo, HHV-6 affects HIV-1 infection in a coreceptor-dependent manner, suppressing CCR5-tropic but not CXCR4-tropic HIV-1 replication, as shown with both uncloned viral isolates and isogenic molecular chimeras. Furthermore, we demonstrate that HHV-6 increases the production of the CCR5 ligand RANTES ('regulated upon activation, normal T-cell expressed and secreted'), the most potent HIV-inhibitory CC chemokine, and that exogenous RANTES mimics the effects of HHV-6 on HIV-1, providing a mechanism for the selective blockade of CCR5-tropic HIV-1. Our data suggest that HHV-6 may profoundly influence the course of HIV-1 infection.

In vivo evolution of HIV-1 co-receptor usage and sensitivity to chemokine-mediated suppression.

Following the identification of the C-C chemokines RANTES, MIP-1alpha and MIP-1beta as major human immunodeficiency virus (HIV)-suppressive factors produced by CD8+ T cells, several chemokine receptors were found to serve as membrane co-receptors for primate immunodeficiency lentiretroviruses. The two most widely used co-receptors thus far recognized, CCR5 and CXCR4, are expressed by both activated T lymphocytes and mononuclear phagocytes. CCR5, a specific RANTES, MIP-1alpha and MIP-1 receptor, is used preferentially by non-MT2-tropic HIV-1 and HIV-2 strains and by simian immunodeficiency virus (SIV), whereas CXCR4, a receptor for the C-X-C chemokine SDF-1, is used by MT2-tropic HIV-1 and HIV-2, but not by SIV. Other receptors with a more restricted cellular distribution, such as CCR2b, CCR3 and STRL33, can also function as co-receptors for selected viral isolates. The third variable region (V3) of the gp120 envelope glycoprotein of HIV-1 has been fingered as a critical determinant of the co-receptor choice. Here, we document a consistent pattern of evolution of viral co-receptor usage and sensitivity to chemokine-mediated suppression in a longitudinal follow-up of children with progressive HIV-1 infection. Viral isolates obtained during the asymptomatic stages generally used only CCR5 as a co-receptor and were inhibited by RANTES, MIP-1alpha and MIP-1beta, but not by SDF-1. By contrast, the majority of the isolates derived after the progression of the disease were resistant to C-C chemokines, having acquired the ability to use CXCR4 and, in some cases, CCR3, while gradually losing CCR5 usage. Surprisingly, most of these isolates were also insensitive to SDF-1, even when used in combination with RANTES. An early acquisition of CXCR4 usage predicted a poor prognosis. In children who progressed to AIDS without a shift to CXCR4 usage, all the sequential isolates were CCR5-dependent but showed a reduced sensitivity to C-C chemokines. Discrete changes in the V3 domain of gp120 were associated with the loss of sensitivity to C-C chemokines and the shift in co-receptor usage. These results suggest an adaptive evolution of HIV-1 in vivo, leading to escape from the control of the antiviral C-C chemokines.

Infection of gamma/delta T lymphocytes by human herpesvirus 6: transcriptional induction of CD4 and susceptibility to HIV infection.

Human herpesvirus 6 (HHV-6), a T-lymphotropic human herpesvirus, is a potentially immunosuppressive agent that has been suggested to play a role as a cofactor in the natural history of human immunodeficiency virus (HIV) infection. We studied the interactions between HHV-6 and gamma/delta T lymphocytes, a subset of T cells involved in the protective immune response against specific microorganisms. Polyclonal gamma/delta T cell populations, purified from the peripheral blood of healthy adults and activated in vitro with phytohemagglutinin, were exposed to HHV-6, strain GS (subgroup A), at the approximate multiplicity of infection (MOI) of 1. Signs of virus replication were detected as early as 72 h after infection, as documented by immunofluorescence, electron microscopy, and transmission of extracellular virus. Progression of the infection was associated with the appearance of typical cytomorphological changes and, eventually, massive cell death. In contrast, no signs of infection or cytopathic effects were detected after exposure of gamma/delta T lymphocytes to HHV-7, a CD4+ T-lymphotropic virus closely related to HHV-6. Polyclonal gamma/delta T cells displayed cytolytic activity against both autologous and heterologous target cells infected with HHV-6 and maintained this activity for at least 72 h after infection with HHV-6, despite the high MOI used. As previously documented in mature CD8+ alpha/beta T cells and natural killer cells, HHV-6 infection induced gamma/delta T lymphocytes to express de novo CD4 messenger RNA and protein, as detected by reverse transcriptase-polymerase chain reaction and fluorocytometry, respectively. Whereas purified CD4- gamma/delta T cell populations were per se refractory to HIV infection, they became susceptible to productive infection by HIV-1, strain IIIB, after induction of CD4 expression by HHV-6. These results demonstrate that gamma/delta T cells can be directly targeted and killed by a herpesvirus and may have implications for the potential role of HHV-6 in AIDS.

Recognition of virus-infected cells by natural killer cell clones is controlled by polymorphic target cell elements.

Natural killer (NK) cells provide a first line of defense against viral infections. The mechanisms by which NK cells recognize and eliminate infected cells are still largely unknown. To test whether target cell elements contribute to NK cell recognition of virus-infected cells, human NK cells were cloned from two unrelated donors and assayed for their ability to kill normal autologous or allogeneic cells before and after infection by human herpesvirus 6 (HHV-6), a T-lymphotropic herpesvirus. Of 132 NK clones isolated from donor 1, all displayed strong cytolytic activity against the NK-sensitive cell line K562, none killed uninfected autologous T cells, and 65 (49%) killed autologous T cells infected with HHV-6. A panel of representative NK clones from donors 1 and 2 was tested on targets obtained from four donors. A wide heterogeneity was observed in the specificity of lysis of infected target cells among the NK clones. Some clones killed none, some killed only one, and others killed more than one of the different HHV-6-infected target cells. Killing of infected targets was not due to complete absence of class I molecules because class I surface levels were only partially affected by HHV-6 infection. Thus, target cell recognition is not controlled by the effector NK cell alone, but also by polymorphic elements on the target cell that restrict NK cell recognition. Furthermore, NK clones from different donors display a variable range of specificities in their recognition of infected target cells.

Human herpesvirus 6 (HHV-6) causes severe thymocyte depletion in SCID-hu Thy/Liv mice.

Human herpesvirus 6 (HHV-6) is a potentially immunosuppressive agent that may act as a cofactor in the progression of AIDS. Here, we describe the first small animal model of HHV-6 infection. HHV-6 subgroup A, strain GS, efficiently infected the human thymic tissue implanted in SCID-hu Thy/Liv mice, leading to the destruction of the graft. Viral DNA was detected in Thy/Liv implants by quantitative polymerase chain reaction (PCR) as early as 4 d after inoculation and peaked at day 14. The productive nature of the infection was confirmed by electron microscopy and immunohistochemical staining. Atypical thymocytes with prominent nuclear inclusions were detected by histopathology. HHV-6 replication was associated with severe, progressive thymocyte depletion involving all major cellular subsets. However, intrathymic T progenitor cells (ITTPs) appeared to be more severely depleted than the other subpopulations, and a preferred tropism of HHV-6 for ITTPs was demonstrated by quantitative PCR on purified thymocyte subsets. These findings suggest that thymocyte depletion by HHV-6 may be due to infection and destruction of these immature T cell precursors. Similar results were obtained with strain PL-1, a primary isolate belonging to subgroup B. The severity of the lesions observed in this animal model underscores the possibility that HHV-6 may indeed be immunosuppressive in humans.

Peptide specificity in the recognition of MHC class I by natural killer cell clones.

Recognition by natural killer (NK) cells of major histocompatibility complex (MHC) class I molecules on target cells inhibits NK-mediated lysis. Here, inhibition of NK clones by HLA-B*2705 molecules mutated at single amino acids in the peptide binding site varied among HLA-B*2705-specific NK clones. In addition, a subset of such NK clones was inhibited by only one of several self peptides loaded onto HLA-B*2705 molecules expressed in peptide transporter-deficient cells, showing that recognition was peptide-specific. These data demonstrate that specific self peptides, complexed with MHC class I, provide protection from NK-mediated lysis.

Productive infection of HUVEC by HHV-8 is associated with changes compatible with angiogenic transformations.

Kaposi's Sarcoma (KS) is an angioproliferative disease associated with human herpesvirus 8 (HHV-8) infection. We have characterized the morphologic and phenotypic modifications of HUVEC in a model of productive HHV-8 infection. HHV-8 replication was associated with ultra-structural changes, flattened soma and a loss of marginal folds and intercellular contacts, and morphologic features, spindle cell conversion and cordon-like structures formation. Phenotypic changes observed on cordon-like structures included partial loss and redistribution of CD31/PECAM-1 and VE-cadherin, uPAR up-regulation and de novo expression of CD13/APN. Such changes demonstrate the induction, in HUVEC, of an angiogenic profile. Most of these findings are directly linked to HHV-8-encoded proteins expression, suggesting that HHV-8 itself may participate to the initial steps of the angiogenic transformation in KS.

Longitudinal analysis of serum chemokine levels in the course of HIV-1 infection.

OBJECTIVES: To investigate the correlation between the serum levels of the CC-chemokines RANTES, macrophage inflammatory protein (MIP)-1alpha and MIP-1beta, and the progression of HIV-1 disease. DESIGN: Retrospective analysis of serial serum samples from HIV-1 seroconverters selected according to clinical outcome. METHODS: Twenty-one patients, derived from a cohort recruited between 1985 and 1996 for a prospective study of the natural history of HIV infection, were analysed. All patients had at least one HIV-1-seronegative sample within 1 year prior to the first seropositive test and were followed longitudinally throughout the course of HIV-1 infection (mean follow-up, 73.5 months). Nine were rapid progressors (RP; patients who developed AIDS within 60 months of antibody seroconversion), seven were slow progressors (SP; patients who developed AIDS after 60 months), and five were long-term asymptomatic (LTA; patients with circulating CD4+ cells higher than 400 x 10(6)/l, no signs of HIV disease, no antiretroviral therapy for more than 96 months). A total of 339 serum samples was studied (mean, 16.1 per patient). The levels of RANTES, MIP-1alpha and MIP-1beta were measured by enzyme-linked immunosorbent assay and correlated with different immunological and clinical parameters. RESULTS: Over the entire follow-up period, the geometric mean of serum RANTES was significantly higher in RP [68.6 ng/ml; 95% confidence interval (CI), 56.9-82.7] than in SP (23.7 ng/ml; 95% CI, 20.0-28.2; P < 0.001) and LTA (19.5 ng/ml; 95% CI, 15.5-24.5; P < 0.001). This difference was already significant during the early clinical stages, when patients had peripheral blood CD4+ cell counts still greater than 400 x 10(6)/l (P < 0.001). By contrast, the mean serum levels of MIP-1alpha and MIP-1beta did not differ significantly between the three study groups. Multivariate analysis using the Cox proportional hazard model demonstrated that the mean serum concentration of RANTES before the development of AIDS was independently associated with the time to AIDS (relative risk, 4.5; 95% CI, 1.1-18.2; P = 0.035). In patients with low versus high mean serum RANTES before the fall of CD4+ cells below 400 x 10(6)/l, the median AIDS-free time was 117.5 and 42.7 months, respectively (P = 0.037). CONCLUSION: These data suggest that an elevation of serum RANTES predicts a rapid progression of the disease since the early stages of HIV-1 infection.

Analysis of serum and plasma beta chemokines in primary HIV infection (PHI).

The levels of certain beta-chemokines in biological fluids do not necessarily reflect their circulating concentrations as they may be dramatically influenced by ex vivo release during sample manipulation. In the present study beta-chemochine levels were evaluated in sequential paired plasma and serum samples collected from a cohort of 18 patients with primary HIV infection (PHI), as well as from 17 HIV-seronegative individuals. In plasma of PHI patients, a significant increase of RANTES (mean 119.1 vs 15.85 ng/ml; p=0.0001) and MIP-1beta (mean 53.4 pg/ml vs 33.6 pg/ml; p=0.0001) was documented. Intra-patient covariance analysis demonstrated no significant association between the variations of RANTES in plasma and serum or between RANTES levels and platelet counts. Reproducibility tests of RANTES measurements in plasma from PHI patients showed a mean coefficient of variation of 1.8%. These data demonstrate that the plasma levels of RANTES and, to a lesser extent, MIP-1beta are persistently perturbed during the early phase of HIV infection. Furthermore they indicate that plasma and serum levels are not directly correlated, being influenced by different physiological phenomena that occur during the ex vivo preparation procedures of the two biological fluids.

Increased plasma levels of the C-C chemokine RANTES in patients with primary HIV-1 infection.

To investigate the role played by chemokines in the natural history of human immunodeficiency virus (HIV) infection, we measured the plasma levels of RANTES. MIP-1 alpha and MIP-1 beta in a cohort of patients with primary HIV-1 infection (PHI) followed longitudinally. The cohort included 17 patients with well-documented history of acute HIV syndrome within two months of the first observation. The mean plasma concentration of RANTES, but not that of MIP-1 alpha or MIP-1 beta, was significantly higher in patients with PHI (192.3 ng/ml) than in five HIV-seronegative controls (8.0 ng/ml) studied during the same time period. Treatment of blood with a cocktail of drugs preventing platelet activation, followed by high-speed centrifugation, reduced the levels of RANTES by approximately 2 logs both in patients and in controls, indicating that the bulk of RANTES was released by platelets, which are known to store this chemokine in their alpha-granules, in the immediate aftermath of blood drawing. No correlation was seen between the levels of RANTES and the number of HIV genome equivalents in plasma. These data suggest that large amounts of pre-formed RANTES are stored in platelets and, possibly, in other blood cells during the early phases of HIV infection. The possible role of this HIV-suppressive chemokine in the control of viral replication during PHI remains to be established.

Two processing pathways for the MHC class II-restricted presentation of exogenous influenza virus antigen.

The natural Ag influenza virus A was used to test the requirements for the HLA-DR1-restricted presentation of the epitopes 18-29 in the matrix protein and 307-318 in the hemagglutinin protein. CD4+ cytotoxic T cell clones of similar efficiency were used to detect presentation of these two epitopes. Presentation of the matrix epitope by APC pulsed with either inactivated virus particles or purified soluble protein followed the classical pathway in that 1) it required invariant chain expression, 2) it was blocked by inhibition of protein synthesis, and 3) it was dependent on a function(s) encoded in the MHC class II region. These characteristics suggest that peptides corresponding to the matrix epitope can only load onto newly synthesized class II molecules that were targeted to a processing compartment by the invariant chain. In contrast, presentation of the hemagglutinin epitope processed from virus particles followed a different pathway. First, presentation of hemagglutinin was independent of invariant chain expression. Second, a human B lymphoblastoid cell line in which protein synthesis was inhibited for 9 h was still able to present hemagglutinin even at very low doses of Ag. Third, a DR1-transfected mutant B cell line missing the MHC class II region was able to present hemagglutinin. Thus, mature class II alpha beta molecules can acquire immunogenic peptides derived from intact natural Ags for presentation to CD4+ T cells. This pathway may be useful for the binding of peptides derived from Ags that are rapidly degraded upon uptake into APC.

Presentation of cytosolic antigen by HLA-DR requires a function encoded in the class II region of the MHC.

The processing pathway for the MHC class II-restricted presentation of endogenous cytosolic Ag is distinct from the class I pathway since a cytosolic form of the influenza virus A hemagglutinin, expressed by a recombinant vaccinia virus, was presented by HLA-DR in a B cell mutant lacking the TAP1 subunit of the transporter for Ag presentation (TAP). In this report, two additional B cell mutants have been used to define the requirements of this TAP1-independent processing pathway. The first mutant, .61, lacks expression of both TAP1 and TAP2 genes, and of both LMP2 and LMP7 genes encoding proteasome subunits. As expected, class I-restricted presentation of the influenza virus matrix protein was totally deficient in mutant .61. In contrast, class II-restricted presentation of both the natural cytosolic matrix and the engineered cytosolic hemagglutinin proteins was functional in mutant .61. Thus, presentation of cytosolic Ag by class II molecules is independent of both TAP subunits and of the two MHC-encoded proteasome subunits. However, this endogenous processing pathway is dependent on at least one other function encoded in the class II region of the MHC as demonstrated with the second mutant, .174, in which a large deletion eliminates all expressed class II genes. Mutant .174 transfected with HLA-DR1 genes was previously shown to be defective in the presentation of exogenous Ag but normal in the presentation of short exogenous peptides. We show here that .174(DR1) is also defective in the presentation of cytosolic matrix and hemagglutinin proteins. This similar requirement for the class II-restricted presentation of either cytosolic Ag or internalized exogenous Ag suggests that both forms of Ag are ultimately targeted to the same cellular compartment for association with class II molecules.

Peptide sequence requirements for the recognition of HLA-B*2705 by specific natural killer cells.

NK clones that were inhibited by target cell expression of HLA-B*2705 displayed peptide-specific recognition of HLA-B*2705. To evaluate the specificity of this recognition, synthetic versions of 14 endogenous ligands of HLA-B*2705 were tested for their ability to provide protection from NK-mediated lysis by binding to surface HLA-B*2705 molecules on RMA-S cells deficient in the transporter for Ag presentation. Several unrelated peptides inhibited lysis by the same NK clones. Despite similar capacities to stabilize HLA-B*2705 molecules on RMA-S cells, the 14 peptides differed widely in their abilities to provide protection. Single amino acid substitutions in both a protective and a nonprotective peptide revealed the importance of residues 7 and 8 in the peptide for recognition by NK clones, thus localizing the peptide influence to a polymorphic region of the alpha-helix of HLA class I molecules known to control discrimination among allelic variants of HLA-B and HLA-C by NK cells.

Real-time quantitative PCR for human herpesvirus 6 DNA.

The diagnosis of human herpesvirus 6 (HHV-6) infection represents a complex issue because the most widely used diagnostic tools, such as immunoglobulin G antibody titer determination and qualitative DNA PCR with blood cells, are unable to distinguish between latent (clinically silent) and active (often clinically relevant) infection. We have developed a new, highly sensitive, quantitative PCR assay for the accurate measurement of HHV-6 DNA in tissue-derived cell suspensions and body fluids. The test uses a 5' nuclease, fluorogenic assay combined with real-time detection of PCR amplification products with the ABI PRISM 7700 sequence detector system. The sensitivity of this method is equal to the sensitivity of a nested PCR protocol (lower detection limit, 1 viral genome equivalent/test) for both the A and the B HHV-6 subgroups and shows a wider dynamic range of detection (from 1 to 10(6) viral genome equivalents/test) and a higher degree of accuracy, repeatability, and reproducibility compared to those of a standard quantitative-competitive PCR assay developed with the same reference DNA molecule. The novel technique is versatile, showing the same sensitivity and dynamic range with viral DNA extracted from different fluids (i.e., culture medium or plasma) or from tissue-derived cell suspensions. Furthermore, by virtue of its high-throughput format, this method is well suited for large epidemiological surveys.

Infection of natural killer cells by human herpesvirus 6.

Natural killer (NK) cells are a functionally defined subset of non-T, non-B lymphocytes of bone marrow origin, which induce lysis of selected target cells, including neoplastic and virus-infected cells. The NK cell function provides an important mechanism of primary defence against viruses in vivo, as demonstrated by the occurrence of multiple herpesvirus infections in patients congenitally lacking NK cells. Here we show that functionally competent CD3- NK clones can be productively infected by human herpesvirus 6 (HHV-6), a T-lymphotropic DNA virus that may play a role in the acquired immunodeficiency syndrome (AIDS) and in the chronic fatigue syndrome, two disorders associated with a defective NK cell activity. The infection is cytopathic and induces de novo expression of CD4, an antigen not expressed within the NK lineage, thereby predisposing NK cells to infection by human immunodeficiency virus type 1 (HIV-1). These results provide evidence that a herpesvirus can directly target and kill NK cells, a potential strategy to suppress the natural anti-viral immunity of the host.

Processing pathways for presentation of cytosolic antigen to MHC class II-restricted T cells.

Antigens presented to CD4+ T cells derive primarily from exogenous proteins that are processed into peptides capable of binding to class II major histocompatibility complex (MHC) molecules in an endocytic compartment. In contrast, antigens presented to CD8+ T cells derive mostly from proteins processed in the cytosol, and peptide loading onto class I MHC molecules in an early exocytic compartment is dependent on a transporter for antigen presentation encoded in the class II MHC region. Endogenous cytosolic antigen can also be presented by class II molecules. Here we show that, unlike class I-restricted recognition of antigen, HLA-DR1-restricted recognition of cytosolic antigen occurs in mutant cells without a transporter for antigen presentation. In contrast, DR1-restricted recognition of a short cytosolic peptide is dependent on such a transporter. Thus helper T-cell epitopes can be generated from cytosolic antigens by several mechanisms, one of which is distinct from the classical class I pathway.

CD46 is a cellular receptor for human herpesvirus 6.

Human herpesvirus 6 (HHV-6) is the etiologic agent of exanthema subitum, causes opportunistic infections in immunocompromised patients, and has been implicated in multiple sclerosis and in the progression of AIDS. Here, we show that the two major HHV-6 subgroups (A and B) use human CD46 as a cellular receptor. Downregulation of surface CD46 was documented during the course of HHV-6 infection. Both acute infection and cell fusion mediated by HHV-6 were specifically inhibited by a monoclonal antibody to CD46; fusion was also blocked by soluble CD46. Nonhuman cells that were resistant to HHV-6 fusion and entry became susceptible upon expression of recombinant human CD46. The use of a ubiquitous immunoregulatory receptor opens novel perspectives for understanding the tropism and pathogenicity of HHV-6.

Clonal analysis of joint fluid T lymphocytes in patients with juvenile rheumatoid arthritis.

Synovial fluid (SF) lymphocytes from 4 patients with pauciarticular juvenile rheumatoid arthritis (JRA) and 4 patients with polyarticular JRA were examined for their phenotypic and functional characteristics. In all 8 patients there was a high proportion of activated SF T cells, together with an increased proportion of CD2+CD3- and the presence of CD3+CD4-CD8-WT31- lymphocytes. The functional analysis at the clonal level in 5 patients (427 clones) showed a relevant proportion of cytotoxic T cell clones, which were not confined to typically cytolytic phenotypes, but were also present among CD3+CD4+CD8- cultures. Compared to those with pauciarticular JRA, patients with polyarticular disease had a significantly higher proportion of T cell clones with cytotoxic activity. Although derived from a limited number of patients, our data suggest a direct involvement of T cells in the pathogenetic mechanisms that originate and maintain the articular damage, and the possibility of different or more pronounced T cell reactivities in the clinically more diffuse JRA types.

Molecular clones of the p58 NK cell receptor reveal immunoglobulin-related molecules with diversity in both the extra- and intracellular domains.

Recognition of major histocompatibility class I molecules on target cells by natural killer (NK) cells confers selective protection from NK-mediated lysis. Cross-linking of the p58 NK receptor, involved in the recognition of HLA-C alleles, delivers a negative signal that prevents target cell lysis. Molecular cloning of the p58 NK receptor reported here revealed a new member of the immunoglobulin superfamily. Five distinct p58 receptors, with sequence diversity in the immunoglobulin-related domains, were identified in a single individual. All NK clones tested expressed at least one p58 member. Three different types of transmembrane and cytoplasmic domains exist, even among receptors with closely related extracellular domains. These data revealed a repertoire of NK cells with clonally distributed p58 receptors exhibiting diversity in both extracellular and intracellular domains.

Coinfection of SCID-hu Thy/Liv mice with human herpesvirus 6 and human immunodeficiency virus type 1.

Human herpesvirus 6 (HHV-6) has been proposed as a potential cofactor in the progression of human immunodeficiency virus type 1 (HIV-1) disease. We used the SCID-hu Thy/Liv mouse model to evaluate the in vivo interactions between HHV-6 and HIV-1. Our results demonstrate that HHV-6 and HIV-1 can simultaneously replicate in the human thymus in vivo. In this model, however, the presence of one virus appears not to modify the replication or cytopathicity of the other.