Human cytomegalovirus modulation of signal transduction

An upregulation of cellular signaling pathways is observed in multiple cell types upon human cytomegalovirus (HCMV) infection, suggesting that a global feature of HCMV infection is the activation of the host cell. HCMV initiates and maintains cellular signaling through a multitiered process that is dependent on a series of events: (1) the viral glycoprotein ligand interacts with its cognate receptor, (2) cellular enzymes and viral tegument proteins present in the incoming virion are released and (3) a variety of viral gene products are expressed. Viral-mediated cellular modification has differential outcomes depending on the cell type infected. In permissive cell types, such as diploid fibroblasts, the upregulation of cellular signaling pathways following infection can initiate the viral gene cascade and promote the efficient transcription of multiple viral gene classes. In other cell types, such as endothelial cells and monocytes/macrophages, the upregulation of cellular pathways initiates functional host changes that allow viral spread to multiple organ systems. Together, the modification of signaling processes appears to be part of a thematic strategy deployed by the virus to direct the required functional changes in target cells that ultimately promote viral survival and persistence in the host.

Role of human cytomegalovirus immediate-early proteins in cell growth control

Human cytomegalovirus (HCMV) is a ubiquitous herpesvirus that has been implicated in several disorders, including an association between HCMV reactivation and the overproliferation of arterial smooth muscle cells observed in restenosis. Although HCMV can mediate a growth-arrest phenotype in infected cells, the virus can also promote an environment conducive to proliferation. Here, we present evidence that the HCMV immediate-early (IE) proteins, IE1-72 and IE2-86, may be responsible for inducing this proliferative environment by altering cell cycle control. We find that expression of either of these IE proteins can alter the cell cycle distribution of randomly cycling cells towards S and G(2)/M phases. Additionally, we find that expression of IE2-86, but not IE1-72, induces quiescent cells into S phase and delays cell cycle exit. In the absence of p53, IE1-72 expression can induce S phase and delay cell cycle exit. We also demonstrate that p53 protein levels increase in fibroblasts following the expression of IE1-72. The observed accumulation of p53 protein in IE1-72-expressing cells may account for the inability of IE1-72 to induce S phase and delay cell cycle exit. Our data suggest that expression of HCMV IE1-72 and IE2-86 is sufficient to alter the cell cycle to generate an environment conducive to proliferation.

Domain mapping of the human cytomegalovirus IE1-72 and cellular p107 protein-protein interaction and the possible functional consequences

Our previous work demonstrated that following human cytomegalovirus (HCMV) infection of fibroblasts, there was a protein-protein interaction between the HCMV IE1-72 immediate-early (IE) protein and the cellular p107 protein which resulted in the alleviation of p107-mediated transcriptional repression of E2F-responsive promoters. In a further characterization of this interaction, we now show that IE1-72 binds to the N-terminal portion of p107, not the C-terminal 'pocket' region that binds E2F-4, and where a number of other viral gene products bind. Additionally, we show that exons 2 and 3 of IE1-72 are required for binding to p107. After mapping the binding domains, we next wanted to address the additional functional consequences of this interaction. It is well known that p107 can negatively regulate cell growth. To examine whether IE1-72 can also overcome this growth suppression, we transfected and infected or cotransfected various constructs into SAOS-2 cells. We showed that infection of SAOS-2 cells was capable of alleviating p107-mediated growth suppression. Additionally, we showed that IE1-72 alone is capable of overcoming p107-mediated growth arrest. Alleviation of this repression by IE1-72 is dependent on the protein-protein interaction between p107 and IE1-72 as deletion mutants of either protein which lack the identified binding domains fail to achieve this effect. These data indicate that the IE1-72 protein is capable of overcoming p107-mediated blocks in cellular proliferation, events that occur in both productive and non-productive HCMV infections.

Human cytomegalovirus binding to human monocytes induces immunoregulatory gene expression

To continue our investigation of the cellular events that occur following human CMV (HCMV) infection, we focused on the regulation of cellular activation following viral binding to human monocytes. First, we showed that viral binding induced a number of immunoregulatory genes (IL-1beta, A20, NF-kappaB-p105/p50, and IkappaBalpha) in unactivated monocytes and that neutralizing Abs to the major HCMV glycoproteins, gB (UL55) and gH (UL75), inhibited the induction of these genes. Next, we demonstrated that these viral ligands directly up-regulated monocyte gene expression upon their binding to their appropriate cellular receptors. We then investigated if HCMV binding also resulted in the translation and secretion of cytokines. Our results showed that HCMV binding to monocytes resulted in the production and release of IL-1beta protein. Because these induced gene products have NF-kappaB sites in their promoter regions, we next examined whether there was an up-regulation of nuclear NF-kappaB levels. These experiments showed that, in fact, NF-kappaB was translocated to the nucleus following viral binding or purified viral ligand binding. Changes in IkappaBalpha levels correlated with the changes in NF-kappaB translocation. Lastly, we demonstrated that p38 kinase activity played a central role in IL-1beta production and that it was rapidly up-regulated following infection. These results support our hypothesis that HCMV initiates a signal transduction pathway that leads to monocyte activation and pinpoints a potential mechanism whereby HCMV infection of monocytes can result in profound pathogenesis, especially in chronic inflammatory-type conditions.

Identification of human cytomegalovirus target sequences in the human immunodeficiency virus long terminal repeat. Potential role of IE2-86 binding to sequences between -120 and -20 in promoter transactivation

OBJECTIVE

Because of the important medical consequences of human cytomegalovirus (HCMV) infection in human immunodeficiency virus (HIV)-infected individuals, we wanted to understand the molecular interactions that occur during co-infection. Specifically, in this study, we wanted to identify the transactivating target sequences on the HIV long terminal repeat (LTR) that responded to HCMV infection.

STUDY DESIGN/METHODS

In this study, we transfected the HIV-LTR into human fibroblasts and then mapped the regulation of this promoter following HCMV infection and co-transfection with the HCMV immediate-early (IE) gene product IE2-86. In addition, we examined IE2-86 binding to specific sequences in the HIV-LTR by electrophoretic mobility shift assay.

RESULTS

Our results documented that HCMV and IE2-86 could transactivate the HIV-LTR. In mapping the regions of the HIV-LTR that IE2-86 transactivates, we identified discrete target sequences between -120 and -20 that are the major transactivating regions for the IE2-86-mediated effects and determined that IE2-86 could specifically bind to several discrete sequences within this region of the HIV-LTR.

CONCLUSIONS

Our discovery of the binding of IE2-86 to the HIV-LTR, coupled with its ability to transactivate the HIV-LTR and induce cellular transcription factors, points to potential molecular mechanisms used by HCMV to upregulate the HIV life cycle and, consequently, exacerbate the conditions observed in individuals co-infected with HCMV and HIV.

The human cytomegalovirus UL55 (gB) and UL75 (gH) glycoprotein ligands initiate the rapid activation of Sp1 and NF-kappaB during infection

The cellular transcription factors Sp1 and NF-kappaB were upregulated shortly after the binding of purified live or UV-inactivated human cytomegalovirus (HCMV) to the cell surface. The rapid time frame of transcription factor induction is similar to that seen in other systems in which cellular factors are induced following receptor-ligand engagement. This similarity suggested that a cellular receptor-viral ligand interaction might be involved in Sp1 and NF-kappaB activation during the earliest stages of HCMV infection. To focus on the possible role viral ligands play in initiating cellular events following infection, we first used purified viral membrane extracts to demonstrate that constituents on the membrane are responsible for cellular activation. Additionally, these studies showed, through the use of neutralizing antibodies, that the viral membrane mediators of this activation are the major envelope glycoproteins gB (UL55) and gH (UL75). To confirm these results, neutralizing anti-gB and -gH antibodies were used to block the interactions of these glycoproteins on whole purified virus with their cell surface receptors. In so doing, we found that Sp1 and NF-kappaB induction was inhibited. Lastly, through the use of purified viral gB protein and an anti-idiotypic antibody that mimics the image of the viral gH protein, it was found that the engagement of individual viral ligands with their appropriate cell surface receptors was sufficient to activate cellular Sp1 and NF-kappaB. These results support our hypothesis that HCMV glycoproteins mediate an initial signal transduction pathway which leads to the upregulation of host cell transcription factors and suggests a model wherein the orderly sequence of virus-mediated changes in cellular activation initiates with viral binding via envelope glycoproteins to the cognate cellular receptor(s).

Induction of the transcription factor Sp1 during human cytomegalovirus infection mediates upregulation of the p65 and p105/p50 NF-kappaB promoters

During human cytomegalovirus (HCMV) infection, the promoters for the classical NF-kappaB subunits (p65 and p105/p50) are transactivated. Previously, we demonstrated that the viral immediate-early (IE) proteins (IE1-72, IE2-55, and IE2-86) were involved in this upregulation. These viral factors alone, however, could not account for the entirety of the increased levels of transcription. Because one of the hallmarks of HCMV infection is the induction of cellular transcription factors, we hypothesized that one or more of these induced factors was also critical to the regulation of NF-kappaB during infection. Sp1 was one such factor that might be involved because p65 promoter activity was upregulated by Sp1 and both of the NF-kappaB subunit promoters are GC rich and contain Sp1 binding sites. Therefore, to detail the role that Sp1 plays in the regulation of NF-kappaB during infection, we initially examined Sp1 levels for changes during infection. HCMV infection resulted in increased Sp1 mRNA expression, protein levels, and DNA binding activity. Because both promoters were transactivated by Sp1, we reasoned that the upregulation of Sp1 played a role in p65 and p105/p50 promoter activity during infection. To address the specific role of Sp1 in p65 and p105/p50 promoter transactivation by HCMV, we mutated both promoters. These results demonstrated that the Sp1-specific DNA binding sites were involved in the virus-mediated transactivation. Last, to further dissect the role of HCMV in the Sp1-mediated induction of NF-kappaB, we examined the role that the viral IE genes played in Sp1 regulation. The IE gene products (IE1-72, IE2-55, and IE2-86) cooperated with Sp1 to increase promoter transactivation and physically interacted with Sp1. In addition, the IE2-86 product increased Sp1 DNA binding by possibly freeing up inactive Sp1. These data supported our hypothesis that Sp1 was involved in the upregulation of NF-kappaB during HCMV infection through the Sp1 binding sites in the p65 and p105/p50 promoters and additionally demonstrated a potential viral mechanism that might be responsible for the upregulation of Sp1 activity.

Human cytomegalovirus upregulates NF-kappa B activity by transactivating the NF-kappa B p105/p50 and p65 promoters

During human cytomegalovirus (HCMV) infection, a series of regulated events take place following virus binding and entry into the cell, including the upregulation of cellular transcription factors, such as NF-kappa B, which play an essential role in the viral life cycle. We show here that NF-kappa B message is induced during HCMV infection and that the induction is biphasic, suggesting an initial induction at immediate-early (IE) times and a second round of induction at early times. This hypothesis is supported by experiments using cyclohexamide, which showed that the first tier of induction was drug insensitive, while the second tier was drug sensitive. We then show that virus binding alone is sufficient to stimulate NF-kappa DNA binding activity, supporting its role in the initial induction of NF-kappa B. To begin to elucidate the mechanism(s) for the second tier of NF-kappa B regulation, we examined promoter constructs from the NF-kappa B subunits (p105/p50 and p65) for responsiveness following HCMV infection. HCMV infection transactivated the p105/p50 and p65 promoters. The viral IE proteins (IE1-72, IE2-55, and IE2-86) are expressed during the time we see NF-kappa B induction, so we examined their role in NF-kappa B induction. The IE1-72, IE2-55, and IE2-86 proteins transactivated the p65 promoter, while only the IE2-55 protein transactivated the p105/p50 promoter. The p105/p50 promoter has NF-kappa B sites; therefore, upregulation could also be caused by an autoregulatory mechanism. The p65 promoter, however, has been demonstrated to contain only Sp1 sites. To investigate the potential role of SP1, we examined nuclear extracts from HCMV-infected cells. Here, we show that there is a biphasic increase in SP1 activity during viral infection and that there is apparently an absolute requirement for SP1 in the transactivation of the p65 promoter. In conclusion, we suggest a model in which the initial induction of NF-kappa B occurs through viral modulation of cellular factors and the sustained levels of NF-kappa B induction are regulated by a combination of cellular and viral factors.

Monocyte-induced cytokine expression in cultured human retinal pigment epithelial cells

Monocytes and retinal pigment epithelial cells are intimately associated in membranes of eyes with proliferative vitreoretinopathy and in certain types of uveitis. The goal of this study was to determine whether monocytes modulate cytokine expression in retinal pigment epithelial cells, and if so, to identify the monocyte products responsible for this effect. Cultured human retinal pigment epithelial cells were exposed to varying concentrations of monocyte-conditioned medium from unstimulated human monocytes for 1-48 hr, or from monocytes prestimulated with lipopolysaccharide. mRNA expression of interleukin-1 beta, interleukin-6, interleukin-8, melanoma growth stimulating activity/gro alpha and gamma, macrophage colony stimulating factor, transforming growth factor-beta 2, basic fibroblast growth factor and activin beta A chain was determined by reverse transcription polymerase chain reaction. Protein secretion of selected cytokines, interleukin-1 beta, interleukin-6, interleukin-8, macrophage colony stimulating factor and transforming growth factor-beta 2 was measured in RPE-conditioned medium by ELISA. Retinal pigment epithelial cells constitutively expressed mRNA for interleukin-6, macrophage colony stimulating factor, transforming growth factor-beta 2, basic fibroblast growth factor and activin beta A chain. Interleukin-1 beta, melanoma growth stimulating activity/gro alpha and gamma and interleukin-8 were not expressed under basal conditions. Stimulated monocyte-conditioned medium markedly induced mRNA of all cytokines except basic fibroblast growth factor and transforming growth factor-beta 2 in a dose- and time-dependent manner. Unstimulated monocyte-conditioned medium was a less potent inducing agent, but still enhanced mRNA expression of interleukin-6, interleukin-8 and melanoma growth stimulating activity/gro alpha. Stimulated monocyte-conditioned medium also induced a time-dependent increase in interleukin-6, Interleukin-8, macrophage colony stimulation factor and transforming growth factor-beta 2, but not interleukin-1 beta protein secretion (p < 0.05 for all time points). Neutralizing antibodies to interleukin-1 beta, or tumour necrosis factor alpha, but not interleukin-1 alpha, significantly reduced cytokine mRNA expression induced by stimulated monocyte-conditioned medium. The combination of all three neutralizing antibodies almost entirely eliminated monocyte-induced mRNA expression and protein production of all cytokines studied. Activated monocytes secrete a heterogeneous mixture of products that together strongly induce expression of multiple cytokines in human retinal pigment epithelial cells. Most if not all of the inducing effect can be accounted for by interleukin-1 beta and tumour necrosis factor alpha. Because cytokines have been implicated in proliferative vitreoretinopathy and uveitis, monocyte-mediated cytokine expression by RPE cells may serve to initiate and perpetuate these diseases.

Productive infection of human endometrial stromal cells by human cytomegalovirus

Cultured endometrial stromal cells were susceptible to productive human cytomegalovirus (HCMV) infection. Infection of endometrial stromal cells resulted in pronounced cytopathic effects including cell rounding and aggregation, fusions, and some lysis, although not in the synchronous fashion observed in infected fibroblasts. The aggregation events were reminiscent of normal endometrial stromal cell responses to cyclical estrogen/progesterone levels. Immunofluorescence analysis demonstrated expression of viral gene products suggesting a productive virus infection. One-step growth analysis showed that infectious virus was produced but the titers were two logs lower than those obtained in fibroblasts even though HCMV DNA accumulated to similar levels in both cell types. In contrast, viral DNA replication was greatly reduced in endometrial stromal cells immortalized with a temperature-sensitive SV40 large T gene at both permissive and nonpermissive temperatures. A more detailed analysis of viral gene expression by Northern blotting revealed earlier appearances and greater initial levels of viral transcripts in endometrial stromal cells. No HCMV gene expression was observed at 120 hpi in these cells even though half of the cells were still intact and cellular gene expression was functional. Since this was the time of peak virus production, it seems plausible that reduced viral gene expression at late times p.i. was a major contributor to the reduced titers observed in endometrial stromal cells. These in vitro results coupled with in vivo observations by others of endometritis associated with HCMV suggest that further investigation into the effects of HCMV on the endometrium is warranted.

Ia- macrophages and cytokine networks contribute to tumor-induced suppression of CD4+ autoreactive T cells

Tumor growth changes the functions and phenotypes of macrophages (M phi) and T cells. Suppression of CD4+ T cell autoresponses during tumor growth was contributed primarily by M phi. Tumor-induced alterations in the abilities of these cells to mediate autorecognition were assessed through syngeneic mixed lymphocyte reaction (SMLR) assays. Tumor-bearing host (TBH) M phi were significantly more suppressive (60-90%) than normal host (NH) M phi, and this suppression was caused partly by reduced Ia expression. TBH Ia- M phi were significantly more suppressive (50-80%) than their NH counterparts. The suppression mechanism was controlled partly by prostaglandin E2 (PGE2), because treating cultures with indomethacin and titrated NH and TBH Ia- M phi led to increased T-cell responsiveness, although responsiveness never reached levels of assays containing unseparated M phi. Blocking studies using anti-interferon-gamma (anti-IFN-gamma) monoclonal antibodies (mAb), anti-interleukin 4 (anti-IL-4) mAb, and indomethacin suggested that IFN-gamma, IL-4, and PGE2 contributed to tumor-induced M phi-mediated suppression. Our results suggested that a quantitative shift in M phi phenotype and a qualitative shift in M phi function in addition to differences in cytokine-directed accessory activities are partly responsible for tumor-induced suppression CD4+ T cell autoresponses.

Cytokines and suppressor macrophages cause tumor-bearing host CD8+ T cells to suppress recognition of allogeneic and syngeneic MHC class II molecules

Quantitative and qualitative tumor-associated changes in T cell phenotype and function were identified in CD8+ T cells. Tumor growth changed splenic CD4+/CD8+ T cell ratios and induced the appearance of more cells with the CD8+ phenotype. In comparison to equal concentrations of normal host (NH) counterparts, tumor-bearing host (TBH) CD8+ T cells were highly suppressive to allorecognition and autorecognition. Suppression was not due to quantitative reductions in CD4+ T cells, although minor qualitative differences were observed. Suppression appeared to be mediated partly by prostaglandin E2 (PGE2). Interferon-gamma (IFN-gamma) and interleukin-4 (IL-4) contributed to TBH CD8+ T cell-mediated suppression. Blocking studies using monoclonal antibodies (mAb) in conjunction with indomethacin suggested that cytokine networks involving IFN-gamma, IL-4, and PGE2 were disrupted during tumor growth and promoted TBH CD8+ T cell suppression. Alloresponses and autoresponses were significantly suppressed when TBH CD8+ T cells mediated these reactions simultaneously with TBH Ia- macrophages. Inhibition of PGE2 production was unable to reverse the additive suppression caused by these two cell types. These results collectively suggest that tumor-induced changes in CD8+ T cells lead to suppressed allo-recognition and autorecognition through both soluble mediator molecules and cellular interactions.

Integrins as a primary signal transduction molecule regulating monocyte immediate-early gene induction

Integrins are cell surface receptors found on monocytes that facilitate adhesion to both cellular and extracellular substrates. These integrins are thought to be involved in the selective gene induction observed after monocyte adhesion to various extracellular matrices. To investigate this hypothesis, we stimulated monocytes with monoclonal antibodies to different integrin receptors to specifically mimic the integrin receptor-ligand interactions. Engagement of the common beta chain of the beta 1 subfamily of integrins resulted in expression of the inflammatory mediator genes, interleukin 1 beta, interleukin 1 receptor antagonist, and monocyte adherence-derived inflammatory gene 6 (MAD-6), whereas engagement of the common beta chain of the beta 2 family did not. Furthermore, to characterize integrin-mediated gene induction, we examined the ability of antibodies to the alpha chain of integrin receptors to regulate gene expression. Engagement of the very late antigen 4 (VLA-4) receptor resulted in induction of all the mediator genes. Receptor crosslinking was required because individual Fab fragments were unable to stimulate gene induction whereas the divalent F(ab')2 fragment and the whole IgG molecule could. Interleukin 1 beta secretion was dependent on the anti-integrin antibody used. Some antibodies required a second signal and, for others, direct engagement was sufficient for protein production. In conclusion, engagement of integrin receptors regulated the production of both inflammatory mediator mRNA and protein. These results suggest that integrin-dependent recognition and adherence may provide the key signals for initiation of the inflammatory response during monocyte diapedesis.

Tumor growth changes the contribution of granulocyte-macrophage colony-stimulating factor during macrophage-mediated suppression of allorecognition

Tumor-bearing host (TBH) macrophages (M phi) suppress T cell alloresponses, and this study suggests granulocyte-macrophage colony-stimulating factor (GM-CSF), a molecule associated with suppressive M phi activity during tumor growth, signals more immunosuppression. In the absence of M phi, GM-CSF increased T cell proliferation in response to alloantigen. However, TBH M phi-mediated suppression of allorecogntion was further induced by GM-CSF. Allogeneic mixed lymphocyte reaction (MLR) cultures, containing normal host (NH) M phi, were either unaffected or enhanced. Prostaglandin E2 (PGE2), a highly suppressive monokine that decreases alloreactivity, did not seem to be involved in the suppression caused by the TBH M phi/GM-CSF interaction. M phi-CSF (M-CSF) addition to cultures did not reverse the suppression caused by TBH M phi and GM-CSF, and inhibition of PGE2 synthesis did not change the response to M-CSF. TBH Ia- M phi, a suppressor population that predominates among splenic M phi during tumor growth, demonstrated significantly lower reactivity in the presence of GM-CSF. In contrast, alloresponses suppressed by NH Ia- M phi demonstrated higher reactivity in the presence of GM-CSF. The data collectively suggest that TBH M phi respond differently to GM-CSF, and that tumor-induced changes in GM-CSF responsiveness affect M phi accessory ability.

Macrophages stimulated by receptor-ligand interactions exhibit differences in cell-cycle kinetics during tumor growth: stimulation at Mac-1 and Mac-3 receptors alters DNA synthesis

Phenotypic and functional changes associated with tumor-bearing host (TBH) macrophages (M phi) are partly responsible for immunosuppression during tumor growth. Flow cytofluorometric analyses revealed differences in cell-cycle kinetics between normal host (NH) and TBH M phi that were stimulated at specific receptors. Receptor-ligand interactions were induced by antibodies against Mac-1, -2, -3, and Ia receptors and changes in DNA synthesis were measured over a 12-h time course by incorporation of propidium iodide. TBH M phi showed higher DNA synthesis than NH M phi over this time course irrespective of the receptor induced. NH M phi stimulated at the Mac-1 receptor demonstrated higher DNA synthesis than control NH M phi although TBH M phi stimulated at this receptor and control TBH M phi failed to show any differences. Both NH and TBH M phi exhibited small, short-term decreases in DNA synthesis when stimulated at the Mac-2 receptor. TBH M phi that were stimulated at the Mac-3 receptor demonstrated higher DNA synthesis than their control counterparts while NH M phi stimulated at this receptor and control NH M phi showed identical levels of DNA synthesis. No differences in DNA synthesis were found among normal or TBH M phi that were stimulated through Ia. Differences in DNA synthesis did not appear to be attributable to differences in receptor expression. Further analysis of Mac-1 and Mac-3 stimulated cells revealed that DNA synthesis in NH M phi stimulated at the Mac-1 receptor returned to control levels at 48 h.(ABSTRACT TRUNCATED AT 250 WORDS)

Characterization of an immediate-early gene induced in adherent monocytes that encodes I kappa B-like activity

We have cloned a group of cDNAs representing mRNAs that are rapidly induced following adherence of human monocytes. One of the induced transcripts (MAD-3) encodes a protein of 317 amino acids with one domain containing five tandem repeats of the cdc10/ankyrin motif, which is 60% similar (46% identical) to the ankyrin repeat region of the precursor of NF-kappa B/KBF1 p50. The C-terminus has a putative protein kinase C phosphorylation site. In vitro translated MAD-3 protein was found to specifically inhibit the DNA-binding activity of the p50/p65 NF-kappa B complex but not that of the p50/p50 KBF1 factor or of other DNA-binding proteins. The MAD-3 cDNA encodes an I kappa B-like protein that is likely to be involved in regulation of transcriptional responses to NF-kappa B, including adhesion-dependent pathways of monocyte activation.

Tumor modulation of autoreactivity: decreased macrophage and autoreactive T cell interactions

The autologous mixed lymphocyte reaction (AMLR) is an in vitro measure of autoreactivity, a key mechanism in immune homeostasis. In this system, macrophages (M phi) act as accessory cells to autoreactive L3T4+ T cells by presenting self-Ia and releasing soluble modulators. During tumor growth, changes occur in M phi and T cells. Tumor-bearing host (TBH) M phi have a reduced ability to act as accessory cells. In fact, TBH M phi suppressed autoreactivity by 60-70%. The decrease in TBH M phi or T-cell abilities was not due to differences in cell numbers or incubation time. Because tumor growth causes increased prostaglandin E2 (PGE2) production by M phi, indomethacin was used to assess the contribution of prostaglandins. Normal and TBH T-cell reactivity increased nearly 50% when stimulated by normal host M phi, while normal and TBH T-cell reactivity increased nearly 100% when stimulated by TBH M phi. Thus increased prostaglandin production is partly responsible for the increased TBH suppressor M phi activity and in the normal host, suppressor M phi may be responsible for maintaining immune regulation. To assess the direct role of prostaglandins in T-cell hyporesponsiveness, PGE2 was titrated into the cultures. PGE2 suppressed normal and TBH T-cell responsiveness in a dose-dependent manner. Normal host T cells were suppressed to a greater extent than TBH T cells by PGE2 (66% versus 42% suppression, respectively). Reduced Ia expression and active suppressor mechanisms are not the only mechanisms mediating hypoautoreactivity during tumor growth. TBH autoreactive L3T4+ T cells were less responsive to self-Ia; they were only 60-80% as reactive as their normal counterparts. To address whether the helper T (TH)-cell defect involved cytokines, T cells were treated with interleukin (IL)-1, IL-2, and IL-4. In all cases, the TBH T-cell response to the factors was decreased (only 60-75% as reactive as normal T cells). Because TBH M phi-mediated suppression can override the addition of IL-1, IL-2, and IL-4, indomethacin was also added with the exogenous interleukins. This coaddition significantly enhanced normal host autoreactivity above control levels while TBH autoreactivity (the combination of TBH T cells and TBH M phi) only returned to normal host unstimulated levels. Tumor growth modulates the immune response at least by (i) decreasing the accessory cell abilities of TBH M phi through decreased Ia expression and increased production of suppressive molecules such as prostaglandins; and (ii) decreasing the responsiveness to immune enhancing factors by TH cells.

Normal and tumor-bearing host macrophage responses: variability in accessory function, surface markers, and cell-cycle kinetics

Normal and tumor-bearing host (TBH) peritoneal macrophage (M phi) responses to in vitro lipopolysaccharide (LPS) treatment were measured by assessing functional and phenotypic changes. Both normal and TBH untreated M phi suppressed mixed lymphocyte reaction (MLR) reactivity at all concentrations. Normal host M phi treated with LPS for 3 h were suppressive at all concentrations. TBH M phi treated with LPS for 3 h were not suppressive in the MLR until more than 5% were added. Surprisingly, 24 h treatment of normal and TBH M phi with LPS induced cells that significantly enhanced MLR reactivity when added at 2% or 5%. These cells were not suppressive until a 20% M phi concentration was reached. LPS treatment of normal and TBH M phi changed the percentage of cells expressing the surface markers Mac-1, -2, -3, and Ia as determined by flow cytometry. Normal host peritoneal M phi treated with LPS for 3 h had decreased Mac-1 and -3 expression, but there was no change in Mac-2 or Ia. Plating for 24 h did not change the percentage of M phi expressing Mac-1, -3, or Ia but did cause an increase in Mac-2+ M phi. Treatment of normal host M phi with LPS for 24 h led to a decrease in Mac-1+ and Ia+ M phi, no change in Mac-3+ M phi, but an increase in Mac-2+ M phi. LPS treatment of TBH M phi for 3 h decreased the number of Mac-1+ M phi, but Mac-2+, -3+, or Ia+ M phi numbers did not change. Plating TBH M phi for 24 h caused a decrease in the number of Mac-1+ M phi, no change in Mac-3+ or Ia+ M phi, but an increase in Mac-2+ M phi. Treatment with LPS for 24 h led to no change in the number of Mac-1+, -3+, or Ia+ TBH M phi, but Mac-2+ M phi increased. The phenotypic and functional changes after LPS treatment led us to ask if these changes were detectable at the level of DNA and RNA. Flow cytometric analysis of acridine orange-stained M phi was used to measure DNA and RNA levels. This analysis determines M phi cell-cycle kinetics and estimates their RNA synthesis. In normal host M phi, a 3-h LPS treatment caused a decrease of cells in G0/G1 but an insignificant change in RNA levels.(ABSTRACT TRUNCATED AT 250 WORDS)

Normal and tumor-bearing host splenic macrophage responses to lipopolysaccharide

Lipopolysaccharide (LPS) was used to assess the responses of normal and tumor-bearing host (TBH) macrophages (M phi) to activation signals. M phi were incubated with LPS for either 3 hr or 24 hr and then assayed for phenotypic, functional, or cell-cycle changes. A 3-hr LPS treatment had no significant effect on M phi phenotype. In contrast, the 24-hr LPS treatment caused a significant decrease in the percentage of normal host Mac+ and Ia+ M phi. In the TBH, a 24-hr LPS treatment caused an increase in the percentage of Mac-1+ and -2+ M phi and a decrease in the percentage of Mac-3+ and Ia+ M phi. When normal host M phi were plated for 24 hr (control), there was an increase in Mac-1+ and -2+ M phi and a decrease in Ia+ M phi. The 24-hr TBH control showed a decrease in Mac-1+ and Ia+ M phi. To assess functional changes, LPS-treated normal and TBH M phi were added to the allogeneic mixed lymphocyte reaction (MLR) or the autologous MLR (AMLR). There were no significant differences in M phi accessory activity after a 3-hr LPS treatment. A 24-hr LPS treatment of normal host M phi also had no effect. A 24-hr LPS treatment of TBH M phi led to a significant decrease in allogeneic T-cell reactivity, and even the 24-hr TBH control showed significant suppression of T-cell responsiveness. In the AMLR, a measure of autoreactive T-cell responsiveness, a 3-hr LPS treatment had no affect on normal host M phi but led to increased accessory ability in TBH M phi. TBH M phi, however, were still less than 50% as responsive as normal host M phi even after LPS treatment. The 24-hr LPS treatment caused a significant decrease in normal or TBH M phi accessory activity. A 24-hr plating of normal host M phi decreased their accessory ability. In addition to the phenotypic and functional changes after LPS treatment, M phi cell-cycle kinetics were also investigated. The percentages of normal host M phi in G0/G1 were not changed significantly after a 3-hr LPS treatment. In contrast, M phi plated for 24 hr with or without LPS had a decreased percentage of cells in G0/G1. Normal host M phi showed little change in total RNA levels after a 3-hr treatment but had increased RNA levels after a 24-hr treatment.(ABSTRACT TRUNCATED AT 400 WORDS)

Two-color flow cytometric analysis of the expression of MAC and MHC class II antigens on macrophages during tumor growth

Tumor-bearing host (TBH) macrophages (M phi) exhibit immune dysfunction that is concomitant with phenotypic changes. We examined M phi subpopulations by changes in the expression of surface antigens Mac-1, -2, -3, and Ia on normal and TBH peritoneal and splenic M phi. M phi were double-labeled and analyzed by flow cytometry to observe multiple expression of surface antigens. Tumor growth alters the multiple expression of these M phi markers. Peritoneal and splenic M phi had different Mac+ and Mac+Ia+ population percentages. In TBH, peritoneal M phi had decreased percentages of Mac-1+2+, Mac-1+3+, Mac-2+3+, and Mac+Ia+ M phi. This decrease correlated with functional changes in TBH M phi. In contrast, there was an increase in Mac-2-Ia- TBH peritoneal M phi. Previously undiscovered Mac-1+2-3- and Mac-1-2-3+ populations were found. In contrast to peritoneal M phi, there was an increase in the percentage of Mac-1+2+, Mac-1+3+, and Mac-2+3+ splenic TBH M phi but, like peritoneal M phi, there was a decrease in the percentage of Mac+Ia+ M phi. Also, TBH splenic M phi showed a smaller but more uniform antigen density than normal host splenic M phi. Tumor growth modulated phenotypic alterations in peritoneal and splenic M phi subpopulations. Combined with earlier functional studies of M phi subpopulations, these data suggested a relationship between changes in M phi phenotype and tumor-induced dysfunction of M phi-modulated immune activity.

Changes in macrophage populations: phenotypic differences between normal and tumor-bearing host macrophages

As a tumor grows, changes occur in the function of macrophages (M phi). This is concomitant with changes in their phenotype. Flow cytometric analysis of monoclonal antibody (mAb)-labeled thioglycollate-elicited peritoneal, and resident splenic, M phi showed a tumor-induced shift of Mac-1, -2, -3, and Ia antigen expression. During tumor growth, the percentage of peritoneal Mac-2+, -3+, and Ia+ M phi decreased significantly (22%, 14%, and 58%, respectively), while Mac-1+ M phi remained unchanged. By analyzing the data on two-dimensional histograms and comparing the sizes of M phi to cell-surface antigen expression, we identified distinct subpopulations of peritoneal M phi. Three distinct size versus antigen expression M phi subpopulations were detected by flow cytometry and consisted of 10-16, 17-22, and 23-27 microns for the small-, medium-, and large-sized populations, respectively. Large-sized Mac-1+ and -2+ M phi decreased (37% and 38%), while large-sized Mac-3+ M phi did not decrease during tumor growth. Medium-sized Mac-3+ M phi decreased 33% during tumor growth, while no differences could be seen in medium-sized Mac-1+ or -2+ M phi. Concomitant with the decrease in large-sized Mac-1+ M phi was an increase in small-sized Mac-1+ M phi. Peritoneal Ia+ M phi were mostly small-sized (4-7-fold increase over the medium-sized and none in the large-sized population). M phi Ia antigen expression was nearly absent in the 21-day tumor-bearing host, with less than 4% of the cells labeling positive (a 73% drop from normal host M phi). In splenic M phi, the percentage of Mac-1+ M phi significantly increased (90%) during tumor growth, while Mac-2+ and -3+ M phi showed a smaller, but still significant, increase (48% and 40%, respectively). Additionally, splenic Ia+ M phi significantly decreased (29%) during tumor growth. More important than the decreased cell numbers was the significant decrease in Ia antigen expression per cell. Unlike the peritoneal M phi, the splenic M phi did not show distinct size versus antigen expression subpopulations, although there was an overall difference in M phi size between normal and TBH. These data suggested that M phi from different anatomical sites are phenotypically different and tumor growth mediates phenotypic alterations in peritoneal and splenic M phi populations. This may be the source of tumor-induced dysfunction of M phi-mediated immune activity.

Tumor-induced variations in a high molecular weight inhibitory monokine

A soluble inhibitor of T cell proliferation was demonstrated in splenic and peritoneal macrophage (M phi) culture supernatants and was variably glycosylated during tumor development. This factor(s) inhibited mixed lymphocyte reaction- (MLR) and concanavalin A-induced T cell proliferation and was not prostaglandin E2 (PGE2), as demonstrated by size (nondialyzable, found in a 30 kilodalton [kd] ultrafiltration retentate and in fractions corresponding to greater than 67 kd on S-200 Sephacryl chromatography) and time course of activity in the MLR. Concentrated supernatants were PGE2-free and yet inhibited proliferation in the MLR. Isoelectric focusing (IEF) revealed normal and tumor-bearing host (TBH) concentrated M phi supernatants' major peaks of inhibitory activity differed in charge, with a pI of 6.5-7.6 for normal hosts and 4.0-6.0 for TBH. Activity in TBH M phi supernatants was found primarily in fractions eluting from hydroxylapatite at 0.3 M sodium phosphate buffer, pH 7.3, was resistant to proteolytic enzymes, but was sensitive to neuraminidase. In contrast, inhibitory activity in normal host M phi supernatants eluted from hydroxylapatite at 0.6 M sodium phosphate buffer and was not sensitive to neuraminidase. Thus, variable glycosylation (presence and absence of sialic residues) could account for the charge difference seen in IEF and may have a bearing on tumor-induced hyporesponsiveness. Kinetic addition of supernatants to the MLR revealed PGE2 may be required for inhibitory activity to be manifested early (0 and 24 h) but not if the high molecular weight (mw) inhibitor was added late (48 and 72 h post initiation). Both normal and TBH M phi supernatants suppressed the generation of interleukin 2 (IL 2) with a dose- and time-dependent difference. Cell-cycle analysis of mitogen-stimulated cells treated with normal and TBH M phi supernatants revealed that TBH M phi supernatants enhanced cell-cycle progression when measured early, but that both normal and TBH M phi supernatants suppressed the number of cells in S phase when measured late in the assay. The production of a high mw inhibitor by both normal and TBH M phi could suggest a homeostatic mechanism, which was upset by increased PGE2 production and production of an inhibitor containing sialic acid, tipping the balance in favor of immunosuppression.

Tumor-induced alteration in macrophage accessory cell activity on autoreactive T cells

Using a tumor-model system, differences in the accessory cell capabilities on autoreactive T cells of splenic macrophages from normal and tumor-bearing hosts (TBH) were assessed in the syngeneic mixed lymphocyte reaction. Tumor development caused a drop in autoreactivity. At 0 and 7 days of tumor growth, no drop in reactivity occurred when TBH macrophages were used as accessory cells and L3T4+ autoreactive T cells from normal mice were used as responder cells. However, by day 14, there was a 32% drop in reactivity, and by day 21 only 22% of the T cell reactivity remained when TBH macrophages were used as accessory cells. Alterations in macrophage Ia antigen during tumor growth were first investigated as the potential cause of reduced autoreactivity. Before tumor growth (day 0) 59% of the splenic macrophages were found to be Ia+. Day-7 TBH macrophages showed no difference in Ia antigen expression when compared to day 0 macrophages. However, by day 14, TBH macrophages showed a 9% decrease, and by day 21 they showed a 36% decrease in the number which were Ia+. Concomitant with the decrease in the number of Ia+ cells was a decrease in the density of Ia antigen expression on day-14 and -21 TBH macrophages. In day-14 and -21 TBH macrophages, two populations were seen that were Ia+. The first had a 10%-20% decrease in Ia antigen expression per cell while the second population had a greater than 50% drop in Ia antigen expression per cell. By titrating and mixing TBH macrophages with normal host macrophages, we assessed whether they could actively mediate suppression of autoreactive T cells. A titratable suppressive phenomenon was demonstrated using day-21 TBH macrophages. In contrast, day-7 and -14 TBH macrophages titrated with normal host macrophages had no effect on the syngeneic mixed lymphocyte reactivity. Lastly, we investigated whether the macrophage-mediated suppression was caused by increased prostaglandin secretion. Addition of indomethacin to cultures increased autoreactive T cell reactivity stimulated by normal or TBH macrophages (59% and 99% increase, respectively). Although indomethacin reduced suppression mediated by TBH macrophages, autoreactivity did not return to levels induced by untreated or indomethacin-treated cells from a normal host. Taken together, the data suggested that tumor growth modulates the function of macrophage accessory cells with autoreactive T cells in at least two ways: by decreasing Ia antigen expression and by increasing suppressor activity.

Shifts in macrophage (M phi) surface phenotypes during tumor growth: association of Mac-2+ and Mac-3+ M phi with immunosuppressive activity

Rat anti-mouse monoclonal antibodies (mAb), anti-Mac-1, -2, and -3, directed against macrophage (M phi) glycoprotein surface antigens, were used to demonstrate a tumor-induced shift in peritoneal M phi subpopulations. This study of the tumor-induced shift was approached in two steps. First, to show that separate phenotypic M phi subpopulations existed and second, to show that a shift in these populations was involved in immunosuppression of the host during tumor growth. Endogenous peroxidase activity was examined among normal and tumor-bearing host (TBH) M phi. A significant increase in the number of peroxidase-positive M phi occurred during tumor growth. Indirect immunofluorescence showed a decrease in Mac-2+ cells and an increase in Mac-3+ cells in TBH M phi populations. When the mAb, anti-Mac-1,-2, and -3 were used in the presence of complement (C), they were cytotoxic for M phi and showed differential depletion of normal and TBH M phi. Peroxidase-positive TBH M phi were susceptible to C-mediated lysis by anti-Mac-1 and -3 but not by anti-Mac-2, whereas no direct relationship was observed among normal host M phi. To demonstrate differences between normal and TBH M phi subpopulations, soluble inhibitory factors were examined from mAb plus C-modified M phi populations. Anti-Mac plus C-treated normal and TBH M phi produced supernatants with different regulatory capabilities as assessed in the mixed-lymphocyte reaction (MLR). Anti-Mac-2 plus C treatment significantly reduced the ability of TBH M phi to produce a soluble suppressor(s) but did not alter normal host M phi-derived suppressor production. In contrast, anti-Mac-1 and -3 plus C treatment of normal host M phi significantly reduced suppressor production. In the TBH, however, anti-Mac-1 plus C had no effect, while anti-Mac-3 plus C had only a limited reduction as compared to the normal host. Determination of levels of prostaglandin E2 (PGE2) in M phi supernatants showed that normal host Mac-1+ M phi were involved in down regulation of PGE2 production. This control was missing in the TBH M phi. Mac-2+ M phi were the apparent producers of PGE2 which accounts for the factor-mediated MLR suppression attributed to TBH Mac-2+ M phi. Collectively, these data suggest that tumor-induced aberrations in immunoregulation can in part be attributed to differences in anti-Mac mAb-defined M phi subpopulations.