Tonsillar B cells do not express PSGL-1, but a significant fraction displays the cutaneous lymphocyte antigen and exhibits effective E- and P-selectin ligand activity.

Skin-homing T cells are defined by the expression of the cutaneous lymphocyte-associated antigen (CLA) which enables the cells to selectively bind to vascular endothelial E-selectin close to sites of cutaneous inflammation, an initial step in the effective extravasation from blood into the inflamed tissue. Essentially all CLA on T cells decorates the backbone of the P-selectin glycoprotein ligand-1 (PSGL-1). In this study we show that human peripheral blood B cells (PBBC) and tonsillar B cells (TBC) do not display PSGL-1 in fluorescence-activated cell sorter analysis using different murine monoclonal antibodies and polyclonal rabbit anti-PSGL-1 antiserum. A significant population of TBC, however, expresses a HECA-452-reactive epitope. These cells represent nonactivated IgM(+)/IgG(-) mature B lymphocytes. Up to 50% of the TBC in a given preparation strongly bind to E- and up to 79% to P-selectin. The shear stress resistance in a parallel-plate flow chamber system was high. Neuraminidase treatment of TBC totally and O-sialoglycoprotein endopeptidase partially diminished HECA-452 reactivity and reduced E- but not P-selectin ligand activities. Mocarhagin had no effect in the assays. The data suggest a different ligand for P-selectin and a distinct glycoprotein carrier for the E-selectin ligand as compared to T cells or other leukocytes. Adhesion to P-selectin, however, still required sulfation of the ligand for function. Western blots of TBC cell lysates detected a >240-kD HECA-452-reactive material that was resistant to reducing conditions. Anti-PSGL-1 did not reveal immunoreactive material in these cell lysates. B cell activation did neither significantly change HECA positivity nor induce PSGL-1 expression. Cultured, activated TBC, however, maintained expression of the integrin alpha4beta7. Human peripheral blood B cells had similar cell surface characteristics to TBC. Our observations suggest that several adhesion molecules may be involved in B cell homing which include CLA, the P-selectin ligand, and structures such as alpha4beta7.

Neutrophils, monocytes, and dendritic cells express the same specialized form of PSGL-1 as do skin-homing memory T cells: cutaneous lymphocyte antigen.

Memory T cells in inflamed skin express the cutaneous lymphocyte-associated antigen (CLA), a glycosylated epitope defined by the mAb HECA-452. We previously reported that on T cells, CLA occurs almost exclusively on the protein backbone of P-selectin glycoprotein ligand-1 (PSGL-1). T cells exhibiting the CLA isoform of PSGL-1 can tether and roll on both E- and P-selectin, while T cells expressing PSGL-1 without the CLA epitope do not bind E-selectin, though they may bind P-selectin. We show here that circulating neutrophils and monocytes, and cultured blood dendritic cells, also express CLA almost entirely as an isoform of PSGL-1. These cells all tether and roll on both E- and P-selectin. A chimeric fusion protein incorporating the 19 N-terminal amino acids of mature PSGL-1 exhibited HECA-452 immunoreactivity and supported rolling of CHO cells expressing either E- or P-selectin. These findings indicate a site for the CLA modification within the distal tip of PSGL-1, previously shown to be critical for P-selectin binding and to mediate some, but not all, of the E-selectin binding of PSGL-1. We hypothesize that the types of circulating leukocytes discussed above all use CLA/PSGL-1 to tether and roll on E- and P-selectin along the vascular endothelium.

Functional cutaneous lymphocyte antigen can be induced in essentially all peripheral blood T lymphocytes.

The cutaneous lymphocyte-associated antigen (CLA) is a skin-homing receptor expressed on a minority of memory-type peripheral blood T (PBT) lymphocytes. Induction of high-level CLA expression in PBT has previously been difficult to accomplish in vitro. Here we report that constitutive CLA expression could be readily induced in virtually all PBT by various polyclonal activators. There was no requirement for accessory cells or addition of other mediators except for IL-2 for maintaining cell survival. Absence of serum in the culture medium was important for optimal induction of CLA. The number of T cells adhering to E-selectin as well as tethering and shear stress resistance under hydrodynamic flow increased in correlation with the level of cell surface CLA expressed. Clonal analysis of CLA induction revealed that in serum-containing medium, which permits the majority of T cells to expand, only a minority of clones did not express CLA. Such T cells could be induced to highly express CLA within 8 days by switching from serum-containing to serum-free medium. This cell-surface phenotype change was closely associated with acquisition of E-selectin ligand activity. Fucosyltransferase VII, which is believed to be important for the generation of the CLA epitope on the P-selectin glycoprotein ligand-1 (PSGL-1) backbone, was shown to be significantly increased in CLA-positive versus CLA-negative T cell populations by PCR analysis. Our findings are consistent with the idea that restriction of CLA expression after activation, rather than positive selection of predetermined T cell subpopulations exposed to restrictive stimulatory conditions in unique microenvironments, may be important in vivo.

Cutaneous lymphocyte antigen is a specialized form of PSGL-1 expressed on skin-homing T cells.

T cells play a pathogenic role in many inflammatory and certain malignant skin diseases, including psoriasis, atopic and allergic contact dermatitis, and cutaneous T-cell lymphoma. Memory T cells that infiltrate the skin express a unique skin-homing receptor called cutaneous lymphocyte-associated antigen (CLA), a carbohydrate epitope that facilitates the targeting of T cells to inflamed skin. CLA is defined by both its reactivity with a unique monoclonal antibody, HECA-452, and its activity as a ligand for E-selectin, but the structure of the protein component of CLA has not previously been defined. Here we report that CLA is an inducible carbohydrate modification of P-selectin glycoprotein ligand-1 (PSGL-1), a known surface glycoprotein that is expressed constitutively on all human peripheral-blood T cells. Cultured peripheral-blood T cells can be differentiated into CLA-bearing cells, which bind both E-selectin and P-selectin, or CLA-negative cells, which bind P-selectin but do not bind E-selectin, suggesting that there is independent regulation of selectin-binding phenotypes. We propose that differential post-translational modification of a single cell-surface receptor, PSGL-1, mediated by fucosyltransferase VII, serves as a mechanism for regulating tissue-specific homing of memory T cells.

Induction of cognate and non-cognate T-cell help for B-cell IgE production in relation to CD40 ligand expression.

Nonactivated, fixed peripheral blood T cells (PBT) from healthy donors or patients with X-linked-hyper-IgM (HIGM) syndrome, or cloned T cells provided effective help for tonsillar B lymphocytes for induction of IgE or other immunoglobulin (Ig) isotypes. Helper activity was mediated by staphylococcal superantigens adsorbed to the T cells prior to fixation and required presence of IL-4 in the cultures. We demonstrated that the T cells neither expressed detectable CD40 ligand at the beginning of the superantigen treatment nor 24 h later. Phorbol ester (PMA) plus Ca-ionophore treatment efficiently induced CD40L. Such T cells did not, however, provide any help for B-cell activation in some experiments or stimulated only low responses in others. Antibodies against CD2, CD3 and ICAM-1 adsorbed to fixed T cells prior to coculturing inhibited helper activity. A soluble CTLA4 construct was also inhibitory. Our results suggest a pathway of B-cell activation independent of CD40L expressed on T cells.

Induction of IgE and IgG1 in human B cell cultures with staphylococcal superantigens: role of helper T cell interaction, resistance to interferon-gamma.

Non-antigen-specific activation of human B lymphocytes for IgE production in vitro requires the presence of interleukin 4 and non-cognate physical interaction with T cells. The latter can be replaced by antibodies directed against the B cells' CD40 structure. Antigen-specific induction of immunoglobulin responses, including IgE, is difficult in human lymphocyte cultures. Thus, we developed a model system which might resemble physiological B lymphocyte stimulation by antigen. Co-cultures of purified tonsillar B cells from normal donors with non-HLA matched T helper clones obtained from the skin of atopic dermatitis patients produced significant levels of IgE and IgG1 after stimulation with appropriate types of staphylococcal exotoxins, provided that IL-4 was also induced in the T cells. Such responses were further enhanced by addition of low doses of anti-CD40 antibodies. Concentrations of anti-CD40, optimal for stimulation of B cells in the absence of T helper lymphocytes, were less effective in this regard and even inhibitory in some experiments. Most powerful immunoglobulin induction was observed when the cultures were spiked with low amounts of IL-4 and anti-CD40 which did not elicit substantial immunoglobulin production in the absence of the staphylococcal exotoxins. Induction of IL-2 in T/B cell cultures by superantigens without production of appreciable quantities of IL-4 provoked considerable IgG1 titer but no IgE. High amounts of interferon-gamma generated by the T cells in vitro in the presence of superantigens did not appear to interfere with immunoglobulin induction. Addition of recombinant interferon at the beginning of the culture period at doses which effectively suppressed IL-4 plus anti-CD40 induced immunoglobulin responses did not inhibit T helper and superantigen dependent B cell activation. Superantigen mediated B cell stimulation for immunoglobulin production was dependent on cell-cell contact. The experimental results presented suggest that this cellular interaction did not necessarily involve T-B cell bridging by superantigens.

Phorbol ester, prostaglandin E2, forskolin and okadaic acid differentially modulate interleukin-4-versus interleukin-2-dependent immunoglobulin induction in human cellular models, in contrast to other selected modifiers of cellular activation.

Interleukin 2 (IL2) and 4 (IL4) are the most important mediators for immunoglobulin (Ig) synthesis of human B lymphocytes. There is no obvious difference with regard to Ig isotypes induced by either lymphokine except for IgE: only IL4 induces this allergic antibody type. Monoclonal anti-CD40 antibodies enhance both IL2- and IL4-dependent Ig induction. Searching for drugs which may inhibit induction of IgE but not of rather non-pathogenic immunoglobulins, we selected commercial compounds which are commonly used as probes for transmembrane signalling pathways in other cellular systems. They included modulators of protein kinase C and intracellular calcium, inducers of cAMP, and inhibitors of protein tyrosine kinase, protein serine/threonine phosphatases and phosphodiesterases. The data presented suggest that IL2- and IL4-mediated B cell activation can be differentially modulated. Phorbol ester at non-cell-toxic doses inhibited IL4- but not IL2-dependent Ig induction. Prostaglandin E2 potently enhanced IgE production stimulated with IL4 alone but was inhibitory in the presence of anti-CD40 as a co-stimulatory signal. IgG1 responses elicited with IL2 plus anti-CD40, in contrast, were not affected. All other compounds did not discriminate between IL2- versus IL4-mediated Ig induction.

Recombinant interleukin 2 rapidly augments human natural killer cell activity.

Recombinant human interleukin 2 (r-IL-2) rapidly stimulated human natural killer cell activity in vitro. Augmentation of NK activity occurred within 1 hr of preincubation with r-IL-2. Responsive killer cells were typical NK cells as shown by cell fractionation procedures. These included Percoll density gradient separation and depletion of OKT3+ T cells by an indirect rosetting method. Analysis with a panel of polyclonal and monoclonal antibodies against alpha and gamma interferon revealed that this early enhancement of NK activity by r-IL-2 was independent of the production of both types of interferon.

In vivo activation of murine natural killer cell functions by human recombinant DNA interleukin 2.

Intraperitoneal (i.p.) injections of purified human recombinant DNA (rDNA)-interleukin 2 (IL 2) resulted in in vivo activation of local natural killer (NK) cell activities in wild-type and congenitally athymic mice. NK cells were identified by short-term cytotoxicity assays against YAC tumor targets and by cell-surface phenotyping. The magnitude of the cytolytic responses was dependent on the IL 2 dose (greater than or equal to 0.1 microgram per injection) and the time period of treatment (the maximum response was on days 3 to 4 after daily treatment). In vivo application of antisera against the murine NK marker asialo GM1 (asGM1) and against interferon-alpha/beta and -gamma (IFN) significantly inhibited NK cell activation. Limiting dilution analysis revealed high frequencies (up to 1 in 1.8 X 10(3)) of in vitro IL 2 reactive mononuclear cells among the peritoneal exudate cells (PEC) of normal mice. rDNA-IL 2 activated non-adherent PEC to proliferate. The majority of these cultures also displayed cytotoxicity against YAC targets. No exogenous IFN was required for either response. Endogenous IFN production did not appear to play an important role for induction of cytotoxicity in this system either. Only a minority of cultures produced measurable levels of IFN without showing excessive cytotoxic activity. In vivo IL 2 treatment resulted in a rapid increase of the total numbers and frequencies of the IL 2 reactive PEC. Hence, IL 2 alone was apparently sufficient for in vitro activation of NK-like activities, whereas IFN-induction by IL 2 was required for in vivo elicitation of similar responses in perhaps the same cell populations.

Modulation by cyclosporin A of murine natural resistance against herpes simplex virus infection. II. Influence on the HSV-induced natural killer cell responses, macrophage activities and interferon levels.

Cyclosporin A (CyA) interfered locally at the site of injection with several resistance functions which are of potential importance in experimental herpes simplex virus (HSV) infections of mice. HSV-induced stimulation of macrophage phagocytosis was reduced by CyA when the mice were infected 5 days before the assay. The in vitro replication of the virus in macrophages, however, was enhanced. Natural killer (NK) cell response were severely impaired. To some extent this could be attribute to the induction of suppressive macrophages by the drug treatment. Interferon levels induced by HSV were not diminished but rather enhanced in some experiments. Inhibitory effects ceased after termination of CyA treatment and could be prevented by presensitization of the mice with attenuated HSV type 2.

Modulation by cyclosporin A of murine natural resistance against herpes simplex virus infection. I. Interference with the susceptibility to herpes simplex virus infection.

Adult BALB/c mice which are medium-high resistant against intraperitoneal (i.p.) infection with herpes simplex virus type 2 (HSV-2) manifested a drastic increase in susceptibility to the virus when treated locally with cyclosporin A (CyA) during infection. Oral application of the drug had no effect on the natural resistance status. Mice appeared normal 2 weeks after CyA treatment with regard to their ability to resist i.p. infections. CyA did not interfere with established specific immune protection nor with the induction in immune responses to HSV-2.

Studies on the interrelation of resistance and immunity in a mouse model system of herpes-simplex type 2 infection.

Intraperitoneal (i.p.) vaccination of mice with attenuated herpes simplex virus type 2 (HSV 2) induced solid protection to i.p. infection with pathogenic virus within two days. Protection was non-virus-specific until day four after sensitization but increased in specificity thereafter. Normal mice could be protected by adoptively transferred spleen cells, serum, and peritoneal fluid from donors vaccinated seven days before. Virus-specific effector cells induced in the spleen by in vivo i.p. sensitization with either live, pathogenic, or attenuated virus and tested in a cytotoxicity assay were exclusively B lymphocytes. No functional B cells, but natural killer (NK) cells, could be detected in the unseparated peritoneal exudate cell (PEC) population. Ability to generate HSV 2 specific antibody responses did not correlate with natural resistance.

Evaluation of live and inactivated influenza A virus vaccines in a mouse model.

Induction of cross-protective immunity against serologically distinct subtypes of influenza A virus in mice was examined in an attempt to correlate cross-protection with heterotypic lymphocyte responses. Live and inactivated virus vaccines protected against the homologous subtype, but only whole virus protected against heterologous subtypes. Live virus vaccines provided better cross-protection than inactivated virus vaccines. A weak defense against heterotypic challenge generated by live H0N1 virus could be boosted by cross-stimulation with whole H3N2 virus and by restimulation with pathogenic H0N1 virus. Heterotypic protection persisted for at least five months. Live viruses induced cross-reactive cytotoxic T cells in normal mice. However, cross-stimulation with heterologous virus was required to generate secondary cytotoxicity. Cross-reactive B lymphocytes were evident after inoculation with whole virus.

Killer T cell responses to influenza A during a drift period: studies in mice.

After intravenous immunization of mice with any influenza A H3N2 drift strain attempts to restimulation of cytotoxic T cell (CTL) activities with the same virus or other drift period variants were unsuccessful for up to 6 weeks. Cross-stimulation 4-5 months after primary sensitization yielded, in most situations, positive but lower--as compared to primary--secondary cytotoxic T cell responses. Homotypic challenge was also effective after priming with some influenza A subtypes (A/E/72, A/PC/73, A/T/77) at this time.

Experimental infection of inbred mice with herpes simplex virus. IV. Comparison of interferon production and natural killer cell activity in susceptible and resistant adult mice.

Inbred mouse strains differ in susceptibility to infection with herpes simplex virus type 1 or type 2 (HSV-1, HSV-2). In this study interferon production was tested in the peritoneal exudate of mice after intraperitoneal (i.p.) injection of HSV-1 or HSV-2. In HSV-resistant mice (C57 BL/6, C3 H/HeJ) high titers of interferon were already present 2 to 4 hours after injection. In comparison, less resistant mice (DBA/2, AKR) lacked this early response. There was no correlation between interferon titers and resistance at post-infection times later than twelve hours. At twelve hours, however, high titers of HSV were detected in the peritoneum of DBA/2 mice and significantly lower titers in C57 BL/6 mice. In a comparative analysis of eight different inbred mouse strains, again early (2 to 4 hours) interferon production was correlated to resistance. In assays of HSV-stimulated early (24 hours) NK cell responses not only the good interferon producer strains but also one of the less resistant low interferon producers (BALB/c) showed significant cytotoxic activities. Conversely, SJL mice that are very low in HSV-induced NK cell activity are resistant and show high early interferon responses at the local site.

Selective induction of immunological tolerance in antiviral T killer cells of inbred mice after treatment with cyclosporin A.

Primary anti-influenza A cytotoxic thymus-derived (T) and bone marrow (B) lymphocyte-dependent responses in inbred mice were used as an in vivo model system to study the effects of the immunosuppressive fungus metabolite cyclosporin A (CyA). Five consecutive daily oral applications of CyA, with the first being given 1 or 2 h before virus inoculation of the animals, caused a complete blockage of induction of anti-influenza T killer cells and a partial reduction of cytotoxic B lymphocyte activities. Adoptive cell transfer experiments revealed that incapability to respond was due neither to humoral factors nor to the generation of suppressor cells. The tolerance state appeared to be specific for influenza A; cytotoxic T lymphocytes against allogeneic cell surface determinants could be stimulated in immunosuppressed mice. CyA treatment abolished virus-specific and cross-reactive anti-influenza killer T cell responses. Suppression was of short duration: less than 1 week for B cell-dependent functions, and between 1 and 2 weeks for T killer cell responses. Animals appeared to be normal with regard to both of these cellular activities for 4 weeks after tolerance induction. Thus, the data indicate that CyA exerted preferential effects on killer T cells. Moreover, evidence was presented that CyA treatment during an ongoing influenza infection did not increase sensitivity to that virus. Mice with no measurable cytolytic anti-influenza T killer cell activities but significant B cell responses, although partially diminished by the drug, were completely protected against the lethal effects of influenza infection.

In-vivo modulation of macrophage functions by herpes simplex virus type 2 in resistant and sensitive inbred mouse strains.

Intra-peritoneal (i.p.) infection of mice with herpes simplex virus type 2 (HSV 2) attracted macrophages into the peritoneum. Macrophages from moderately and highly HSV 2 resistant mouse strains expressed elevated phagocytosis activity 24 hours after injection. Stimulation of phagocytosis in low resistant strains was generally less effective or absent. This was, in some experiments, due to the fact that macrophages were already highly activated before the experimental infection. I.p. infection also caused HSV replication in the adherent peritoneal exudate cell (PEC) population. The capacity of macrophages supporting HSV 2 replication was low in three of four resistant mouse strains and high in all moderately and highly susceptible and in one of the resistant (SJL) strains when determined 24 hours after infection. Four different F1 hybrids between resistant and susceptible strains exhibited significantly lower yields of virus-producing macrophages than the HSV-sensitive parent. One hybrid between two HSV-susceptible lines restricted virus replication in the PEC populations better than both parental strains.

Induction of natural killer cells by herpes-simplex virus type 2 in resistant and sensitive inbred mouse strains.

Infection of mice with herpes simplex virus type 2 (HSV 2) stimulated natural killer (NK) cells in a variety of inbred mouse strains including athymic nude mice. Essentially all mouse strains tested exhibited high NK activity on day four after virus inoculation. Assayed 24 hours after infection, SWR/J, AKR/J, SJL/J and C57B1/10J mice were low or negative for these non-virus-specific cytotoxic responses. Whereas the first two mouse strains were most sensitive to the lethal effects of HSV 2, the latter two were highly resistant. Three lines with intermediate susceptibility and three highly resistant strains were all efficient with regard to early NK-cell.

Function, target-cell preference and cell-surface characteristics of herpes-simplex virus type-2-induced non-antigen-specific killer cells.

Wild-type and congenitally athymic nude mice injected with herpes-simplex virus type 2 (HSV 2) responded with a local outburst of non-antigen-specific killer cells masking any virus-specific response. Cytolytic activity could be assayed on mouse-tumor cell lines and on syngeneic or allogeneic non-transformed cells from various sources. Some of the tumor cell lines and proteose-peptone-induced peritoneal exudate cells were lysed more efficiently after infection with either HSV 2, vaccinia or influenza A virus. Preference for virus-infected target cells was already expressed 24 hours after HSV-2 injection. Killing activity was not H-2-restricted, not complement- or immunoglobulin-dependent and did not involve Fc receptors. The cytotoxic cells were non-adherent and could be shown to express Thy1, Quat4, and Quat4 cell-surface antigens. They lacked immunoglobulin and Lyt1: Lyt2,3 determinants. The functional and serological characteristics identify the HSV-2-induced cytolytic cells as natural killer (NK) cells. The potential importance of this cell population for natural resistance will be discussed.