Granzyme B-mRNA expression by equine lymphokine activated killer (LAK) cells is associated with the induction of apoptosis in target cells

Lymphokine-activated killer (LAK) cells are a subset of cytotoxic cells capable of lysing freshly isolated tumor cells. While LAK activity is typically measured using the (51)Cr-release assay, here we used a non-radioactive flow cytometric method to demonstrate equine LAK activity. Equine peripheral blood mononuclear cells (PBMC) were stimulated in vitro with recombinant human interleukin 2 (hIL-2) to generate LAK cells. An equine tumor cell line, EqT8888, labeled with carboxyfluorescein succinimidyl ester (CFSE) was used as target cells. Following incubation of the targets with different concentrations of LAK cells, Annexin V was added to identify the early apoptotic cells. With increasing effector to target cell ratios, EqT8888 apoptosis was increased. We also measured interferon-gamma, granzyme B and perforin mRNA expression in the LAK cell cultures as possible surrogate markers for cytotoxic cell activity and found granzyme B mRNA expression correlated best with LAK activity. Also, we found that the reduced LAK activity of young horses was associated with decreased granzyme B mRNA expression. Our results indicate that fluorescence-based detection of LAK cell activity provides a suitable non-radioactive alternative to (51)Cr-release assays and mRNA expression of granzyme B can be used as surrogate marker for these cytotoxic cells.     

Natural killer cells in normal horses and specific-pathogen-free foals infected with equine herpesvirus.

Peripheral blood mononuclear cells (PBMC) from an adult horse and from foals demonstrated natural killer (NK)-type cytotoxicity against a range of xenogeneic and allogeneic cell targets. The human tumour cell line, Chang liver was consistently the most susceptible. Chang liver, rabbit kidney (RK-13), equine sarcoid (ES) and embryonic equine kidney (EEK) cells were more susceptible when presented to horse PBMC than monolayer cultures. Embryonic equine lung (EEL) and murine YAC-1 cells conversely, were more susceptible in a trypsinized state. Horse PBMC demonstrated higher levels of NK-type activity against EEK, EEL and RK-13 cells infected with equine herpesvirus 1 (EHV-1) compared with uninfected cells. Similarly, EEK and EEL cells infected with Semliki forest virus (SFV) were more susceptible. Cytotoxicity against EHV-1-infected EEK cells developed faster, between 4 and 8 h of incubation and reaching a maximum at 24 h. By contrast, cytotoxicity against uninfected fibroblasts was not significant until approximately 16 h of incubation with maximum cytotoxicity observed between 32 h and 48 h. Specific pathogen-free (SPF) foals were inoculated with live EHV-1. PBMC isolated from these foals at different days after inoculation did not display appreciably reduced or elevated NK cytotoxicities against Chang liver cells and EHV-1-infected EEK targets, when compared with that of a PBMC reference from a healthy adult horse.     

Further characterization of cross-reactive anti-human leukocyte mAbs by use of equine leukocyte cell lines EqT8888 and eCAS

The equine leukocyte cell lines, namely eCAS (bone marrow derived) and EqT8888 (lymphoma derived) were used for further analysis, using 40 commercially available mAbs that showed reactivity with equine PBMC. Most mAbs that were detected in these previous studies to react with PBMC, however, did not react with either of these cell lines. Fifteen mAbs were positive on at least one of the cell lines and indicate opportunities to set up further tests using these cells. Notably, two mAbs directed against human CD34 were detected to react with eCAS.     

Functional characterization of equine dendritic cells propagated ex vivo using recombinant human GM-CSF and recombinant equine IL-4

Naive T cells can be activated both in vivo and in vitro by specialized antigen presenting cells, dendritic cells (DC), with potent antigen-specific, immunostimulatory activity. Indeed, DC can provide an extremely powerful and important immunological tool by which to potentiate the immune response for specific recognition of foreign antigens. Until recently, the direct isolation of DC from PBMC required laborious procedures with extremely poor yields (<0.1%). Methods have been developed for the human, lower primate, and murine model systems to propagate large numbers of DC from PBMC or bone marrow ex vivo with various cytokines. However, all other model systems, including equine, still require the laborious isolation procedures to obtain DC. In this study, we have adapted the methods developed for the human system to generate large numbers of equine DC from PBMC precursors using recombinant human GM-CSF and recombinant equine IL-4. Our report is the first documentation of ex vivo generated DC from PBMC in a domesticated animal model system. Equine DC derived from PBMC were rigorously characterized by analyzing morphological, phenotypic, and functional properties and were determined to have similar attributes as DC generated from human PBMC. Equine DC appeared stellate with large projectiles and veils and had cell surface antigens at similar levels as those defined on human and murine DC. Furthermore, functional attributes of the DC included rapidly capturing antigens by pinocytosis, receptor-mediated endocytosis, and phagocytosis, activating naive T cells in a mixed leukocyte reaction to a much greater extent than macrophage or lymphoblasts, presenting soluble and particulate antigen 10-100 fold more effectively to T cells on a per cell basis than macrophage or lymphoblasts, and presenting soluble and particulate antigen to both CD4+ and CD8+ T cells. Taken together, our study provides a framework by which equine DC can now be readily produced from PBMC precursors and presents an impetus for and model by which DC can be simply generated in other animal model systems.     

Tumoricidal effect of interleukin-2-activated killer cells in canines

We investigated the effect of both partially purified (TCGF) and recombinant interleukin-2 (rIL-2) on the tumor-directed cytotoxic activity of canine peripheral blood mononuclear cells (PBMC) using three normal canines. No cytotoxic activity was displayed by unstimulated effector cells at 3 h of incubation; however, the cytotoxic effect was observed in a 16-h assay. PBMC of all canines displayed significant levels of lytic activity after stimulation for 4 to 7 days with both types of IL-2 against a variety of allogeneic and xenogeneic neoplastic cells in 3-h 51Cr release assay. The cytotoxic activity of cultured cells increased proportionally in the 16-h assays. Morphological examination of the May-Grünwald and Giemsa stained cytocentrifuged slides of cultured cells on each day of assay showed an increase in large granular lymphocytes (LGLs) beginning on day 4 and reaching a peak on day 7 of culture.     

Cloning and large-scale expansion of epitope-specific equine cytotoxic T lymphocytes using an anti-equine CD3 monoclonal antibody and human recombinant IL-2

Cytotoxic T lymphocytes are involved in controlling intracellular pathogens in many species, including horses. Particularly, CTL are critical for the control of equine infectious anemia virus (EIAV), a lentivirus that infects horses world-wide. In humans and animal models, CTL clones are valuable for evaluating the fine specificity of epitope recognition, and for adoptive immunotherapy against infectious and neoplastic diseases. Cloned CTL would be equally useful for similar studies in the horse. Here we present the first analysis of a method to generate equine CTL clones. Peripheral blood mononuclear cells were obtained from an EIAV-infected horse and stimulated with the EIAV Rev-QW11 peptide. Sorted CD8+ T cells were cloned by limiting dilution, and expanded without further antigen addition using irradiated PBMC, anti-equine CD3, and human recombinant IL-2. Clones could be frozen and thawed without detrimental effects, and could be subsequently expanded to numbers exceeding 2 x 10(9)cells. Flow cytometry of expanded clones confirmed the CD3+/CD8+ phenotype, and chromium release assays confirmed CTL activity. Finally, sequencing TCR beta chain genes confirmed clonality. Our results provide a reliable means to generate large numbers of epitope-specific equine CTL clones that are suitable for use in downstream applications, including functional assays and adoptive transfer studies.     

Cytotoxicity against autologous, allogeneic, and xenogeneic tumor targets by human recombinant interleukin-2-activated lymphocytes from healthy dogs and dogs with lung tumors

Before dogs with lung tumors were treated by adoptive immunotherapy, the ability of canine blood lymphocytes (PBL) from the peripheral circulation to differentiate in vitro in the presence of human recombinant interleukin-2 (rIL-2) and become tumoricidal was investigated. The PBL from healthy dogs (n = 6) and dogs with lung tumors (n = 5) were grown in culture medium alone, in the presence of rIL-2 to generate lymphokine-activated killer (LAK) cells, or with phytohemagglutinin (PHA) and rIL-2 to generate autologous-stimulated lymphocytes (ASL). After 4 days, cytotoxicity by the ASL, LAK, and PBL was determined in a 4-hour 51chromium-release assay. Target cells in the assay were short-term cultured enzyme digests of autologous (self), allogeneic (genetically different) primary tumors, and Raji, the xenogeneic human lymphoma cell line. The PBL cultured without rIL-2 were not cytotoxic against any tumor. However, when a dog's PBL were activated in vitro, they killed the dog's own tumor, ASL more effectively than LAK cells. Pulmonary adenocarcinomas and an osteosarcoma metastasis to lung were among the autologous tumors assayed. Against an allogeneic canine osteosarcoma, ASL generated from healthy dogs were significantly more cytolytic than LAK from healthy dogs, or than ASL generated from tumor-bearing dogs. Cytotoxicity was greater against allogeneic tumor than against Raji. Lectin-dependent cellular cytotoxicity, tested by including PHA in the assay medium with lymphocytes and Raji cells, by ASL and LAK was greater than cytotoxicity of Raji without PHA. Because ASL were more cytolytic than LAK against all targets in vitro, they may be more beneficial than LAK for immunotherapy of canine tumors.     

Cell mediated immune responses in ponies following infection with equine influenza virus (H3N8): the influence of induction culture conditions on the properties of cytotoxic effector cells

Cytotoxic cell precursors and/or cytotoxic memory cells were demonstrated in the peripheral blood of ponies after aerosol infection with influenza A/equine/Newmarket/79 (H3N8). In order to reveal their cytotoxic potential, peripheral blood mononuclear cells required a secondary antigenic stimulation. In vitro induced cytotoxic cells showed activity against influenza infected target cells in a 3-4 h 51Cr-release assay. The reactivity of cytotoxic cells was markedly influenced by the conditions of the secondary induction culture. If high concentrations of exogenous crude equine IL-2 were used, virus infected target cells were susceptible to lysis by autologous or allogeneic effector cells. However, if IL-2 concentration was reduced, cytotoxic cells were generated which showed features consistent with cytotoxic T cells in that target-cell killing was genetically restricted.     

Natural killer (NK) activity of porcine blood lymphocytes against allogeneic melanoma target cells

Allogeneic PM/86 melanoma cells of Munich Troll miniature swine have been used for the demonstration of porcine peripheral blood NK cell activity. Compared with the specific lysis of xenogeneic K562-, U937- and Vero-target cells, NK cell-mediated cytotoxicity (NK-CMC) against PM/86 melanoma tumor cells was significantly lower in a 16 h chromium release assay. The target cell susceptibility to peripheral blood NK-CMC of both adult Troll miniature swine and German Landrace sows was very similar. Cold target inhibition assays revealed the allogeneic PM/86 melanoma cells to be the most powerful inhibitors of NK-CMC. Nylon wool non-adherent lymphocytes produced interferon (IFN)-alpha in different quantities upon contact with NK susceptible target cells. The NK effector cells could be stimulated to a higher lytic activity against all susceptible targets by a moderate dose of natural human interleukin-2 (nhuIL-2). The role of NK-CMC in melanoma tumor rejection and/or prevention of metastases is yet unknown in swine although porcine melanoma serves as a good model for the disease in man.     

T lymphocyte development in horses. I. Characterization of monoclonal antibodies identifying three stages of T lymphocyte differentiation

Six monoclonal antibodies reacting with equine T lymphocytes at different stages of maturation were selected from antibodies produced against lymphoid cell preparations. EqT12 and EqT13 antibodies identified subsets of cortical thymocytes with high terminal deoxynucleotidyltransferase (TdT) activity and no phytolectin responsiveness. EqT12+ thymocytes were scattered throughout the cortex while EqT13+ thymocytes were located in the subcapsular cortex. EqT12 bound to small numbers of bone marrow cells, splenocytes, and circulating lymphoid cells, but not to mature T lymphocytes. EqT13 bound to very small numbers of bone marrow cells but not to more mature lymphocytes. EqT6 and EqT7 reacted with a large population of cortical thymocytes with high TdT activity and no phytolectin responsiveness. EqT2 and EqT3 bound primarily to medullary thymocytes with low TdT activity. Eq2+ thymocytes responded to phytolectin stimulation while EqT3+ thymocytes did not. EqT2 and EqT3 bound to 33% and 91% of circulating T lymphocytes, respectively. The T lymphocytes bound by both antibodies included cells capable of suppressing a mixed lymphocyte reaction. Thus, EqT12 and EqT13 identify cells with the functional characteristics of prothymocytes. EqT6 and EqT7 identify resident cortical thymocytes, and EqT2 and EqT3 identify a subpopulation of mature T lymphocytes and all mature T lymphocytes, respectively.     

Characterization of monoclonal antibodies to equine interleukin-10 and detection of T regulatory 1 cells in horses

Interleukin-10 (IL-10) terminates inflammatory immune responses and inhibits activation and effector functions of T-cells, monocytes, macrophages and dendritic cells. IL-10 has also been found to be a key cytokine expressed by subpopulations of regulatory T-cells. In this report, we describe the generation and characterization of three monoclonal antibodies (mAbs) to equine IL-10. The antibodies were found to be specific for equine IL-10 using different recombinant equine cytokine/IgG fusion proteins. Two of the anti-equine IL-10 mAbs were selected for ELISA to detect secreted IL-10 in supernatants of mitogen stimulated equine peripheral blood mononuclear cells (PBMC). The sensitivity of the ELISA for detecting secreted IL-10 was found to be around 200pg/ml. The production of intracellular IL-10 was measured in equine PBMC by flow cytometry. PBMC were stimulated with phorbol 12-myristate 13-acetate (PMA) and ionomycin in the presence of the secretion blocker Brefeldin A. All three anti-IL-10 mAbs detected a positive population in PMA stimulated lymphocytes which was absent in the medium controls. Around 80% of the IL-10(+) cells were CD4(+). Another 15% were CD8(+) cells. Double staining with IL-4 or interferon-gamma (IFN-gamma) indicated that PMA and ionomycin stimulation induced 80% IL-10(+)/IFN-gamma(+) lymphocytes, while only 5% IL-10(+)/IL-4(+) cells were observed. By calculation, at least 60% of the IL-10(+)/IFN-gamma(+) cells were CD4(+) lymphocytes. This expression profile corresponds to the recently described T regulatory 1 (T(R)1) cell phenotype. In summary, the new mAbs to equine IL-10 detected native equine IL-10 by ELISA and flow cytometry and can be used for further characterization of this important regulatory cytokine in horses.     

Characterization of porcine xenoreactive cytotoxic T lymphocytes.

Cytotoxic T lymphocytes (CTL) against mouse P815 cells were detected after stimulation of porcine peripheral blood mononuclear cells (PBMC) with irradiated Balb/c splenocytes. In vivo priming prior to in vitro stimulation slightly enhanced CTL activity, but lysis of targets was undetectable from lymphocytes from non-immune or immune animals that were not cultured with mouse splenocytes. After primary culture with Balb/c (H-2d) splenocytes, specific killing of P815 (H-2d) targets and not L929 (H-2k) targets indicated that recognition was specific for the H-2 locus. Similarly, CTL primed by mouse cells from either of two congenic strains recognized targets with alleles homologous to the stimulating cells. The anti-murine CTL was confirmed to be a CD8+ T cell based on studies using specific monoclonal antibodies to the porcine CD4 or CD8 cells. The cells responsible for the cytotoxicity of P815 targets lacked the characteristics of non-specific NK cells because (1) naive PBMC were unable to lyse NK targets (K562 cells) during the 4 h cytotoxic assay and (2) CTL killing of P815 targets increased with time after primary stimulation, whereas killing of K562 cells remained low at all times. These results suggest that porcine CTL can be readily generated against the xenogeneic mouse major histocompatibility complex.     

Interleukin-1 stimulation of equine articular cells

Prostaglandin E2 (PGE2) and stromelysin are produced by equine chondrocytes and synovial cells in vitro in response to recombinant human (rh) interleukin-1 (IL-1) alpha and beta, and equine mononuclear cell supernatants (MCS) containing IL-1. However, culture conditions are important. PGE2 concentrations increase in proportion to the concentration of fetal calf serum (FCS) in the culture medium, whereas stromelysin concentrations are inversely proportional to the concentration of FCS. Equine MCS, containing a lower concentration of IL-1 than the concentration of rhIL-1 used in these experiments, stimulated production of much higher levels of PGE2 than rhIL-1. In addition, equine MCS induced the production of broadly similar levels of PGE2 by both chondrocytes and synovial cells, whereas rhIL-1 was more active on equine synovial cells than equine chondrocytes. Although equine MCS induced both stromelysin and PGE2 production by equine articular cells, on the whole rhIL-1 failed to induce stromelysin production. This supports previous observations of species restrictions in the activity of human IL-1 on equine cells. Therefore, experiments using mammalian cells and heterologous IL-1 should be interpreted with caution.     

Cloning and functional characterization of recombinant equine P-selectin

The recent molecular characterization and sequencing of equine P-selectin (ePsel), and its glycoprotein ligand, P-selectin glycoprotein ligand-1 (PSGL-1), have provided the tools for further investigation into their role in leukocyte trafficking. Here, we report the generation of a genetically engineered chimeric protein (ePsel-IgG) in which the equine P-selectin lectin and epithelial growth factor (EGF) domains were covalently linked to the equine IgG1 heavy chain constant region. The soluble ePsel-IgG was observed to bind to equine monocytes by confocal microscopy and flow cytometry. Furthermore, equine monocytes bound to immobilized ePsel-IgG in a time course and dose dependent manner. Not only did ePsel-IgG act as an adhesion molecule, it was also found to activate ERK1/2 kinase and induce IL-8 mRNA expression in equine monocytes. That all of the aforementioned ePsel-IgG-induced cell binding and cell signaling were abolished by the addition of EDTA, suggested that ePsel-IgG chimera mediated events occurred via the P-selectin ligand, PSGL-1. We were able to demonstrate that 78% of equine monocytes cross-reacted with anti-human HECA-452 antibody, which recognizes the sialy-Lewis X (sLex) epitope, a well-known carbohydrate binding site on human PSGL-1. Pre-incubation of equine PBMC with neuraminidase or O-sialoglycoprotein endopeptidase (OSGP) reduced ePsel-IgG monocyte binding to 36% or 60%, respectively. Taken together, these data suggest that there might be two ligand recognition sites on P-selectin, one of which recognizes sLex and another which recognizes P-selectin ligand core protein. The ePsel-IgG chimera can be a useful as a reagent for further studies on the role of equine P-selectin and signal transduction in inflammatory events in horse.     

Immunoprecipitation of equine CD molecules using anti-human MABs previously analyzed by flow cytometry and immunohistochemistry

Earlier studies investigating the cross-reactivity of antibodies submitted to the HLDA8 had used flow cytometry as a method of choice to screen mAbs for reactivity with equine leukocytes, including two-color flow-cytometry to characterize the lymphocyte population they detect. In addition, immuno-histochemistry (IHC) was used to detect distribution of positive cells in lymphoid tissue sections. In this study we performed immunoprecipitation (IP) to complement the previous results and add valuable information regarding the molecules detected by the cross-reacting antibodies. Surface molecules from primary equine PBMC or the equine cell line T8888 were biotinylated prior to precipitation to determine the molecular weight of the corresponding molecules in a western blot using streptavidin-AP. 21 out of 24 mAbs precipitated the molecules with a MW corresponding to its human orthologue. Positive mAbs were directed against CD2, CD5, CD11a, CD11b, CD14, CD18, CD21, CD44, CD83, CD91, CD172a, MHCI and MHCII. Three mAbs directed against CD49d, CD163, and CD206 which were unambiguously identified earlier by flow cytometry failed to immunoprecipitate the corresponding CD molecule. MAbs detecting CD molecules which are expressed internally like CD68 and mAbs of IgM class could not be included into this approach.     

Modulation of the cytokine responses in equine macrophages following TACE-inhibition

The detrimental effects of the pro-inflammatory cytokine TNF-alpha during equine acute abdominal disease are well known. Its pivotal role in many human diseases has led to various in-depth studies regarding its release mechanism, in particular by TNF-alpha converting enzyme (TACE). In this study we investigated the inhibitory effect of a TACE-inhibitor on cytokine release (TNF-alpha, IL-1alpha and IL-6) in three different cell models, including U937 cells, a recently established equine macrophage cell line, known as eCAS, and primary equine PBMC. The aim was to show the similarity of TNF-alpha release through the TACE mechanism in human and equine cells after stimulation with LPS. Results indicate that, by using a TACE-inhibitor, TNF-alpha, IL-1alpha and IL-6 release can be reduced in both equine cell models and achieved comparable results in the human U937 cells. These results suggest that equine TNF-alpha, like its human homologue, is released from its membrane-bound position by TACE.     

Induction and characterization of endotoxin tolerance in equine peripheral blood mononuclear cells in vitro

Endotoxemia is responsible for severe illness in horses. Individuals can become clinically unresponsive to the endotoxin molecule after an initial exposure; a phenomenon referred to as 'endotoxin tolerance' (ET). ET has been observed clinically in horses in vivo; however, cytokine expression associated with ET has not been investigated. The purpose of this study was to develop and validate a method for inducing ET in equine peripheral blood mononuclear cells (PBMC) in vitro, and to describe selected cytokine responses which are associated with ET. ET was induced by culturing cells with three concentrations of endotoxin, and evaluated after a second dose of endotoxin given to challenge the cells. The relative mRNA expression of IL-10 and IL-12 was measured by use of quantitative PCR. ET was induced in all cells exposed to the 2-step endotoxin challenge. The relative mRNA expression of IL-10 in tolerized cells was not different from positive control cells. In contrast, the relative mRNA expression of IL-12 in tolerized cells was decreased by 15-fold after the second endotoxin challenge. This experiment demonstrated a reliable method for the ex vivo induction of ET in equine PBMC. A marked suppression of IL-12 production is associated with ET.     

Cross-species reactivity of seven monoclonal antibodies with equine lymphocytes by flow cytometry

The recognition of equine lymphocyte antigens by monoclonal antibodies (mAbs) directed against human CD11a, CD18, CD21, CD23, CD29 and DR, as well as mouse CD23 was studied by flow cytometry. Unlike anti-CD11a, -CD21, -CD23 and DR mAbs, anti-CD18 and CD29 mAbs labelled the same percentage of horse peripheral blood lymphocytes (PBL) as human PBL. Double-staining with anti-horse immunoglobulin antibodies showed that anti-CD21 and -CD23 mAbs are mainly bound to peripheral blood B lymphocytes. The seven mAbs were also tested on the lymph node and thymus cells. The molecular targets of anti-CD11a, CD18 and CD29 mAbs were confirmed by immunoprecipitation of the membrane proteins. Our results suggest that anti-CD18, -CD29 and -DR mAbs recognise similarly expressed molecular homologues on equine cells, but that anti-CD11a, -CD21 and -CD23 mAbs recognise either different molecules or homologues that are expressed at different levels on horse cells.     

Generation and characterization of monoclonal antibodies to equine CD16

The low-affinity Fc receptor CD16 plays a central role in the inflammatory and innate immune responses of many species, but has not yet been investigated in the horse. Using the predicted extracellular region of equine CD16 expressed as a recombinant fusion protein with equine IL-4 (rIL-4/CD16), we generated a panel of mouse monoclonal antibodies (mAbs) that recognize equine CD16. Nine mAbs were chosen for characterization based upon recognition of CD16, but not IL-4, in ELISA. All nine mAbs recognized full-length, cell-surface CD16 expressed as a GFP fusion protein by CHO cells, but not the closely related Fc receptor CD32 expressed in the same system. In flow cytometric analysis with equine peripheral leukocytes, the mAbs labeled cells in the granulocyte, monocyte, and lymphocyte populations in a pattern consistent with other species. Monocytes that were strongly labeled with CD16 mAb 9G5 were also positive for the LPS receptor CD14. Cytospins made with peripheral leukocytes were immunohistochemically labeled and showed mAb recognition of primarily mononuclear cells. ELISA revealed that the nine mAbs can be grouped into three patterns of epitope recognition. These new antibodies will serve as useful tools in the investigation of equine immune responses and inflammatory processes.     

Concepts for the clinical use of stem cells in equine medicine

Stem cells from various tissues hold great promise for their therapeutic use in horses, but so far efficacy or proof-of-principle has not been established. The basic characteristics and properties of various equine stem cells remain largely unknown, despite their increasingly widespread experimental and empirical commercial use. A better understanding of equine stem cell biology and concepts is needed in order to develop and evaluate rational clinical applications in the horse. Controlled, well-designed studies of the basic biologic characteristics and properties of these cells are needed to move this new equine research field forward. Stem cell research in the horse has exciting equine specific and comparative perspectives that will most likely benefit the health of horses and, potentially, humans.