The immune system initially controls tumors or invading

Transcription

1 JOURNAL OF INTERFERON & CYTOKINE RESEARCH Volume 9, Number 3, 9 Mary Ann Liebert, Inc. DOI:.89/jir.8.58 Activation of Porcine Natural Killer Cells and Lysis of Foot-and-Mouth Disease Virus Infected Cells Felix N. Toka,, Charles K. Nfon, Harry Dawson, 3 D. Mark Estes,,5 and William T. Golde Natural killer (NK) cells play a vital role in innate response against viral infections and cellular transformation. In vivo modulation of their response may enhance their antiviral function. Here we describe the phenotype of porcine NK cells, test potential proinflammatory cytokines for activation of these cells and assess the capability of porcine NK cells to kill virus-infected or tumor cells in vitro. The CD + /CD8 + /CD3 cell compartment contained porcine NK cells, which at the resting stage were minimally cytotoxic toward foot-and-mouth disease virus (FMDV)-infected porcine cells or tumor cell lines. Direct stimulation of NK cells with proinflammatory cytokines induced efficient lysis of FMDV-infected cells with interleukin (IL)- or IL-5 showing the highest stimulatory capacity. Lower levels of NK cell activation were induced by IL-, IL-8, or interferon (IFN)-α, however, IL- and IL-8 synergistically activated NK cells. Combinations of IL-5 and IL- or IL-5 and IL-8 did not further increase the porcine NK cell lytic capability over IL-5 alone. Natural killer cells expressed IFN-γ regardless of the cytokine used for stimulation while expression of perforin increased modestly. The enhancement of porcine NK cell activity by proinflammatory cytokines offers a promising tool for development of antiviral approaches against virus infection. Introduction The immune system initially controls tumors or invading pathogens by engaging innate immune mechanisms among which, the natural killer (NK) cells play a critical role. Natural killer cells form a small fraction of circulating lymphocytes that are neither B nor T cells. Natural killer cells can destroy transformed or infected cells of the host in a rapid, antigen nonspecific manner and there is ample evidence for their role in immunosurveillance (Klein and others 98; Kobayashi 989) and protection against certain viral, bacterial, and protozoan infections (Hirahashi and others ; Byrne and others ; Feng and others 6). Mechanisms engaged by NK cells to eliminate the target cells, as described for humans and mice, involve exocytosis of cytotoxic granules containing perforin and granzymes and induction of apoptosis through appropriate receptor ligand interactions (Miyaji and others ; Grossman and others ; Ochi and others ). Besides the cytotoxic function, NK cells are a major source of interferon-γ (IFN-γ) (Martin-Fontecha and others ; Thale and Kiderlen 5; Montoya and others 6), a cytokine integral to the development of the adaptive immune response following infection. Reactivity of NK cells is highly regulated by an array of stimulatory and inhibitory receptors on their cell surface and the balance of signals between these receptors determines the ultimate reactivity of NK cells [reviewed in Lanier (5)]. Little is known about the expression and distribution profile of activating and inhibitory receptors on porcine NK cells. Natural killer cells are known to act on demand, at least in humans (Price and others 98). Contrarily in mice (Fehniger and others 7) and pigs (Pinto and Ferguson 988), these cells may require activation to effectively carry out their function. Several cytokines have been reported to activate NK cells and most are part of the inflammatory response to virus infection. Interferon-α, interleukin (IL)-, and IL-8 are well characterized for this function (Lauwerys and others ). Interleukin- functions to enhance NK Plum Island Animal Disease Center, Agricultural Research Service, US Department of Agriculture, Greenport, New York. Department of Preclinical Science, Faculty of Veterinary Medicine, Warsaw University of Life Science, Warsaw, Poland. 3 Beltsville Area Research Center, Agricultural Research Service, US Department of Agriculture, Beltsville, Maryland. Department of Pediatrics and 5 Department of Microbiology and Immunology, Sealy Center for Vaccine Development, University of Texas Medical Branch, Galveston, Texas. 79

2 8 cells, lymphocyte-activated killing and cytotoxic T lymphocytes (CTLs). Also IL- induces production of IFN-γ and may mediate proliferation of T cells in the presence of a mitogen (Gately and others 99). Interleukin 8 enhances NK cell activity and promotes Th immune responses. Originally, it was described as an IFN-γ-inducing factor. Recently, IL-5 has also been described to participate in development and survival of NK cells in mice (Liu and others ; Golden- Mason and others ). Natural killer cells are less characterized in nonrodent species such as porcine and bovine. The unavailability of reagents has slowed progress in dissecting the precise phenotype, function and mechanisms of activation of NK cells in large animal species. However, recently a commercially available antibody against bovine NKp6 (Storset and others ) has rekindled the interest in further characterization of NK cells in the bovine. Recent reports describe a possible role of bovine NK cells in controlling infections such as Mycobacteria bovis (Endsley and others 6) and possible mechanisms of activation of bovine NK cells have been reported (Endsley and others 6). Denyer and others (6) recently reported the classification of porcine T cells based on their expression of perforin. Porcine NK cells appear to be contained in the CD + /CD8 + /CD3 cell population, however, CD3 + cells were also observed to contain perforin and had capacity to lyse targets in major histocompatibility complex unrestricted manner. Therefore, it is likely that porcine NK cells may share characteristics of a wide range of lymphocytes including a small population of CD3 + T cells, possibly NKT cells. Similarly, a small portion of porcine NK cell activity likely resides in the CD +, CD3, CD, CD8a, LGL population as anti-cd8 + complement treatment eliminates most but not all NK cell activity (Pescovitz and others 988). Recently an antibody against porcine NK cells was described that binds to CDR, an integrin that is similar to human CDb but has a different expression and distribution pattern in porcine (Antigenix America, Inc., Huntington Station, NY, USA). To date no in-depth characterization has been done on porcine cells expressing CDR. The role of NK cells in controlling viral infections, especially acute viral infections, is likely very significant. Footand-mouth disease virus (FMDV) is an important pathogen of cloven hoofed animals, particularly bovine and swine. It spreads rapidly causing an economically devastating disease due to highly acute pathology and being extremely contagious. In regions where the virus is endemic, vaccination is practiced and this vaccine has been shown to work in as little as 7 days in cattle (Golde and others 5). However, control of FMD outbreaks require antiviral treatments that work in as little as h. Therefore, it is important to understand the function of NK cells in domestic species and how activation of these cells may be regulated. Protection of livestock against this and other acute viral infections may be enhanced by the use of interventional vaccines designed to activate the innate response before the adaptive immunity develops. We propose that rapid acting vaccines against FMDV should be designed to stimulate innate responses for early protection against infection while allowing normal or possibly accelerated development of effective adaptive immune responses. Here, we have characterized the in vitro cytotoxic capacity of porcine NK cells and analyzed their response TOKA ET AL. to stimulation with proinflammatory cytokines including human (h)il-, porcine (p)il-, pil-5, pil-8, and pifn-α. Freshly isolated porcine NK cells are less cytotoxic, but a number of the cytokines tested activate the NK cell function. These cytokines increased the level of killing, IFN-γ secretion and perforin expression in porcine NK cells. Interleukin5 was the most potent in activating porcine NK cells, whereas IL- and IL-8 exhibited synergism in inducing cytolytic activity. These inflammatory and proinflammatory cytokines activated the porcine NK cells most likely through cytokine receptor interaction and not through an intermediary cell types. Importantly, the cytokines enhanced cytotoxicity against FMDV-infected SK6 cells. Exogenous proinflammatory cytokines may likely augment the in vivo function of porcine NK cells and molecular methods may be manipulated to enhance the innate response against infection with FMDV. Materials and Methods Animals Yorkshire pigs aged between 3 and months were purchased from Animal Biotech, Inc. (Danboro, PA, USA) and acclimated for week before use in the experiments described here. All procedures performed on the pigs were approved by the Institutional Animal Care and Use Committee. Preparation of peripheral blood mononuclear cells Assays involving cells from pigs were performed on peripheral blood mononuclear cells (PBMCs) isolated as follows. Blood was drawn into heparin-containing vacutainer tubes, diluted : with phosphate-buffered saline (PBS), layered onto Lymphoprep (Axis-Shield, PoC AS, Oslo, Norway) and spun at g for min at ºC. PBMCs were collected from the gradient interface and washed in PBS three times to remove platelets followed by wash in Rosewell Park Memorial Institute (RPMI)-6 medium (Invitrogen, Carlsbad, CA) and finally suspended in RPMI-6 supplemented with % fetal bovine serum (FBS, Hyclone, Logan, UT, USA), mm Hepes, mm l-glutamine, and antibioticantimycotic (Invitrogen, Carlsbad, CA). Enrichment of porcine NK cells and cell sorting Natural Killer cells were enriched based on the characterization by Denyer and others (6). Negative selection was performed by first removing adherent macrophages (MΦ), followed by CD7 +, CD3 +, CD +, and CD + cells. Magnetic-activated cell sorting (MACS) (Miltenyi Biotech, Bergisch Gladbach, Germany) or the Dynabead cell-sorting systems (Dynal Biotech ASA, Oslo, Norway) were used to separate the cells according to the manufacturers instructions. Mouse antibodies against porcine CD3 (clone PPT3), CD (clone 7--), CD (clone BB6-C9.6), and CD7 (clone 7--5) were purchased from Southern Biotech (Burmingham, AL). All antibodies used for cell enrichment were low endotoxin and azide free as assured by the manufacturer. Briefly, PBMCs were labeled with antibodies for min at ºC. Subsequently, cells were bound by microbeads coated with goat anti-mouse IgG and further incubated for 5 min at ºC. Cells were washed and passed through a magnetic field for separation. After separation, cells were

3 PORCINE NK CELL ACTIVATION 8 washed twice and resuspended in RPMI-6 supplemented with % FBS. Enrichment usually reached 85 9% of CD + / CD8 + /CD3 cells containing the majority of porcine NK cells. Adherent monocyte/macrophages were isolated by incubating PBMCs in plastic flasks for h at 37ºC followed by removal of nonadherent cells. The flasks were then washed once and cell dissociation solution (Sigma, St. Louis, MO, USA) was added and flasks incubated at 37ºC for 5 min. Detached cells were washed in RPMI-6 and used in the assays. CD3 +, CD +, CD +, and CD7 + cells were positively selected using MACS (see above) and rested for 8 h before performing any assays on them. This sorting procedure did not cause changes in reactivity of cells. NK cell cytotoxicity assay A flow cytometry technique was used to assess the NK cell lytic activity. YAC- (Moloney murine leukemia virus induced lymphoma) and K56 (human erythroleukemia cell line) tumor cell lines were obtained from the American Type Culture Collection (Manassas, VA, USA). K56 tumor cells stably transfected with the green fluorescent protein (GFP), referred to here as K56-GFP, were provided by Dr. Mike Olin (University of Minnesota, School of Veterinary Medicine, St. Paul, MN, USA). Bulk PBMCs or CD + /CD8 + / CD3 enriched cells from pigs were added to target cells harvested from cultures in exponential growth. Cells were distributed in 96-well microtiter plates at effector:target (E:T) ratios of 5:, 5:, :, and 6: followed by addition of 5 μl/ well 7AAD (.5 μg, dead/live discriminating dye, 7 aminoactinomycin D, BD Bioscience, San Diego CA, USA). Cells were mixed, spun for min at 5g and incubated at 37ºC in 5% CO for h. When YAC- cells were used as target cells, carboxy fluorescein succinimidyl ester (CFSE) was used as a tracking dye. Labeling of YAC- cells was performed as described for SK6 cells in the next section. Data were acquired using a FACSCalibur flow cytometer with a high throughput system (HTS) and later analyzed in CellQuest Pro software (BD Biosciences, San Jose, CA, USA) by establishing a gate on K56-GFP to display the double-positive (GFP+7AAD) cells representing the killed population. The lysis level was determined by the formula R/(R+R) = %Lysis, where; R = double-positive cells; R = GFP-positive cells. Natural killer cells were further analyzed by flow cytometry for degranulation by staining with mouse anti-porcine CD7a (Serotec, Raliegh, NC). NK cell cytotoxicity assay for FMDV-infected SK6 cells To determine the capability of porcine NK cells to lyse virus-infected target cells, SK6 cells (porcine kidney fibroblast cell line) were infected with an attenuated strain of FMD, LL-KGE, at a multiplicity of infection of for h. LL-KGE is a FMDV O Campos structural proteins (icosehedral capsid) inserted into the strain A backbone with a positive charge mutation, and another mutation in the receptor binding (RGD) sequence to KGE to allow for binding of the heparan sulfate receptor. This allows infection of all cell types rather than only cells expressing the physiological receptor for FMDV. Additionally, this contruct has no leader protease that determines the in vitro attenuation. Following infection, the cells were washed twice and resuspended in PBS then labeled with 5 μm of CFSE for 5 min at room temperature. Labeling was stopped by adding RPMI-6 supplemented with % FBS. Cells were washed once and allowed to stand for 3 min at 37ºC. Finally, the labeled cells were washed three times, counted and used as target cells in the killing assay at appropriate E:T ratios. The time of incubation with NK cells, data acquisition, and correction were as described for K56-GFP cells. CFSE-labeling procedure was also performed on CD + /CD8 + /CD3 cells that were later stimulated with cytokines for 8 h to determine proliferation of porcine NK cells. Stimulation of porcine NK cells To measure the direct stimulatory effect of cytokines on porcine NK cells, hil- (Hemagen Diagnostics, Inc., Columbia, MD, USA), pil-, pil-8 (R&D Systems, Minneapolis, MN, USA) pil-5 (Invitrogen, Carlsbad, CA, USA), and pifn-α (in-house preparation as described by Moraes and others (7) were used to stimulate the PBMCs or NK cells. Cytokines were used at a concentration of ng/ml hil-, ng/ml for pil-, 5 ng/ml for pil-5, ng/ml for pil-8, and 5 U for pifn-α, predetermined in preliminary experiments. Cells were stimulated for 8 h and later used in NK cell cytotoxicity assay or intracellular cytokine measurement. Intracellular staining for IFN-γ and perforin Cells stimulated as described earlier were placed in 96-well microtiter plates at 6 cells/well and incubated for 5 6 h in the presence of GolgiPlug containing brefeldin A ( μl/ 6 cells; BD Bioscience, San Diego CA). Phorbol myristate acetate and Ionomycin (Sigma, St Louis, MO, USA) were added to cells designated for positive controls at ng/ml and ng/ml, respectively. After 5 6 h cells were stained on the surface for CD (clone MSA, Accurate Chemicals, Westbury, NY, USA) and CD8 (clone 76--, Biotin-labeled, Southern Biotech). Secondary antibodies, anti-mouse IgGa- was used to detect CD and streptavidin-percp to detect CD8-biotin. Intracellular staining for IFN-γ or perforin was performed according to BD Bioscience intracellular staining protocol. Anti-porcine IFN-γ-PE clone PG was purchased from BD Bioscience and used at μg/ ml. Perforin was detected with anti-human perforin-pe antibody clone δg9 (BD Biosciences, San Diego, CA, USA) that cross reacts with porcine perforin at the concentration suggested by the manufacturer. Data were acquired and analyzed in FACSCalibur HTS and CellQuest Pro, respectively (BD Biosciences, San Diego, CA, USA). NK cell proliferation CD + /CD8 + /CD3 cells were sorted by negative selection with magnetic bead labeling as described earlier and rested for 8 h in RPMI-6. Later the cells were washed once and resuspended in PBS, and labeled with CFSE at 5 μm for 5 min at room temperature. Labeling was quenched by addition of RPMI-6 containing % FBS for 3 min then washed three times and finally resuspended in RPMI-6. Before setting up the assays, cells were counted and viability

4 8 assessed by trypan blue exclusion. 5 cells were plated per well and cytokines added to the cells at the same concentration as used previously and incubated for 8 or 7 h. Dilution of CFSE was monitored by flow cytometry (FACSCalibur, CellQuest Pro, BD Biosciences, San Diego, CA, USA). Quantitative real-time RT-PCR The CD3 + cells were selected positively and rested for 8 h before RNA isolation as described earlier. RNA was isolated from non-stimulated or cytokine-stimulated CD + /CD8 + /CD3 - cells, treated with DNAse then transcribed into cdna using the following reaction mixture; 5x reaction buffer, mm dntp, 5 ng/μl Random Primers,. M DTT, U RNAse inhibitor and U M-MLV RT. The cdna templates were later used in the Quantitative realtime reverse transcriptase polymerase chain reaction (qrrt- PCR). The reactions were done in duplicate. TaqMan RT-PCR system was used for amplification. Primers and probes were designed at the Beltsville Human Nutrition Research Center (Beltsville, ARS USDA, docs.htm?docid=665) and sequences are provided in Table. Polymerase chain reactions were run in the 77 ABI Prism Sequence Detector (Applied Biosystems, Foster City, CA, USA) with the following cycling parameters: hold at 5ºC for min, hold at 95ºC for min and then run cycles at 95ºC for 5 s (denature), and 6ºC for min (anneal). Ubiquitin or peptidylprolyl isomerase A were used as reference genes to normalize each gene expression. Relative expression was calculated using REST5 version.9. ( and is expressed as ratios relative to the calibrator (purified CD3 + T cells or nonstimulated CD + / CD8 + /CD3 cells). Statistical evaluation Where appropriate, a test of significance in differences between means of treatment groups was performed by a t-test for paired sample means. A p value of.5 was considered significant. Results Resting porcine NK cells have low cytotoxicity We determined the resting lytic activity of porcine NK cells by a flow cytometry based cytotoxicity assay. Data presented in Figure A and B show that resting porcine NK cells are capable of killing K56-GFP or YAC- cells, respectively, albeit at a very low level when a standard incubation time of h was used. Low or even complete lack of lytic activity of porcine NK cells has been reported previously (Pinto and Ferguson 988) and may be associated with the breed of pigs being analyzed. Because relatively low cytotoxicity could easily be interpreted as lack of killing capability, we treated a set of cells with chloroquine, a lysosomotropic agent that raises the ph of perforin granule compartment blocking the maturation of perforin thus inhibiting cytotoxicity. Treatment with chloroquine diminished the cytotoxicity against both target cell lines (Fig. A and B). Additionally, to show whether this low-level cytotoxicity involved exocytosis of cytotoxicity granules, PBMCs were cultured with K56 cells lacking GFP expression, with addition of -labeled TOKA ET AL. mouse anti-porcine CD7a, a lysosomal-associated membrane protein (LAMP-) expressed mainly in endosome-lysosome membranes of cells. CD7a has been described as a marker for CTL or NK cell degranulation (Alter and others ). PBMCs have a low expression of CD7a (Fig. C), but upon culture with K56 cells for h the expression of CD7a increases considerably (Fig. D). Addition of CHQ inhibited further expression CD7a on the cell surface (Fig. E). These data indicate that the low cytotoxicity observed is due to degranulation of the NK cells, further confirming that freshly isolated porcine NK cells have reduced cytotoxicity against tumor cells lines. Although the bulk assay indicated a low level of lytic activity of freshly isolated PBMCs we sought to localize the effector function to a particular cell type. Thus, PBMCs were sorted into CD + /CD8 +, CD3 +, CD +, CD + (B cells), CD7 + cells, and adherent MΦ and then tested for cytotoxicity against K56-GFP cells. Residual killing was preserved at very low levels in CD3 +, but not in B cells, CD7 + cells and MΦ (Fig. F). However, the highest cytolytic potential was detected in the CD + /CD8 + /CD3 cell fraction only and the difference was significant compared to CD3 + cells (p.5). Thus, we have demonstrated that the NK cell phenotype in the porcine likely resides in the CD + /CD8 + /CD3 - cell population. Proinfl ammatory cytokines activate porcine NK cells The relatively low level of lytic activity of NK cells against sensitive targets suggested that porcine NK cells require activation before engagement in target cell recognition and lysis. During infection with certain viruses, cytokines such as IFN-α, IL-, IL-5, and IL-8 are abundantly secreted, and activation of NK cells by these cytokines has been reported in humans and mice (Naume and others 99; Perussia and others 99; Lauwerys and others ). Therefore, we investigated whether direct cytokine stimulation of porcine NK cells from freshly isolated PBMCs influenced their lytic capability. Human IL-, pil-, pil-5, pil-8, or pifn-α were added to cultures of PBMCs or enriched NK cells as described in Material and Methods for 8 h and then NK capacity was tested. Four of the proinflammatory cytokines tested enhanced porcine NK cells lytic activity of CD + / CD8 + /CD3 cells or bulk PBMCs against K56-GFP cells compared to nonstimulated NK cells (Fig. A and B). The only cytokine that did not increase the NK cell lytic activity was IL-, although addition of excess amounts, (three times the concentration recommended to achieve biological response by the manufacturer) enhanced NK cell lytic activity (data not shown). The difference in lytic activity between stimulated and nonstimulated cells was statistically significant (p.5), apart from cells stimulated with pil-. However, there were no significant differences between hil-, pil-5, pil-8, and pifn-α. IL-, whose activating influence on human and mouse T cells is known, and IL-5 activate porcine NK cells to a 5- and -fold increase, respectively, in percentage of K56-GFP cell lysis compared to nonstimulated cells. The pattern of NK cell reactivity in both PBMCs and CD + /CD8 + /CD3 cells was similar, which indicates that the cytokines activated the same effector cells. Thus, similar to human and mice, the cytokines tested induced porcine NK cells to kill sensitive tumor cell targets.

5 PORCINE NK CELL ACTIVATION 83 Gene Table. Primers and Probes Used in the qrrt-pcr Assay Primers and probes IL- receptor-β IL-5 receptor-α IL-8 receptor-α IFN-α receptor Tumor necrosis factor (ligand) subfamily member TRAIL Ubiquitin house-keeping gene Peptidylprolyl isomerase A housekeeping gene NCR (NKp6) NCR3 (NKp3) KLRF (NKp8) Granzyme A Granzyme B Perforin KLRK (NKGD) KLRB (CD6) KLRC (NKGA) SH domain containing B KLRA (Ly9) p-ilrb-36f: GGCCAGGAAAGGGACAAAG p-ilrb-7r: CCCCAGCACCTTGTACAGATC p-ilrb-3t: AAGTCCACCACCTCCAAGGGCTCTCAC p-il5ra-36f: GCTGGGTTCAAGCGGAAAG p-il5ra-3r: GCCACGTTCATGGTCTCGTT p-il5ra-9 revt: AACACGCACTCCGTCAGGCTGGA p-il8rap-8f: CTCCTCGCCACCGTCATG p-il8rap-596r: GCGTCTCATCCTTGCTCTGGTA p-il8rap-5t: TGTGCTGACCACAGGTGCTCTCCTCT p-ifnar-3f: AAAAATTAAATTGCGTATAAGAGCAGAA p-ifnar-385r: TTGAAATGGTATAAACGGCTCAAC p-ifnar-33t: ACAGGAAACAGCACTTCTCCGTGGTATGA p-tnfsf-77f: TGTGAGAAAGATGATTTTGAGAACCT p-tnfsf-38r: CTCTCTGTGGACCTTTTTCTCTTTCTA p-tnfsf-36t: TCAGAAAAGCAACAAGGCATTCCTCACCT p-ubc-7f: GCGCACCCTGTCTGACTACA p-ubc -6R: AGATCTGCATCCCACCTCTGA p-ubc-79t: AGTCCACCCTGCACCTGGTCCTCC p/h-ppia-63f: GCCATGGAGCGCTTTGG p-ppia-5r: TTATTAGATTTGTCCACAGTCAGCAAT p/h-ppia-5 revt: TGATCTTCTTGCTGGTCTTGCCATTCCT p-ncr-f: CCTTTGACCAGGAGCTTCACA p-ncr-7r: AGATGTTTTGGCTCCTACAACAAG p-ncr-3t: GCTCACTGGGGAAAGACCATGCGT p-ncr3-97f: TGATCAGGGTCCATCCAGGAT p-ncr3-68r: TGCCCTCCTGAGTACAGATCTCA p/h-ncr3-9t: CTGTGCTCTCTGGGTGTCCCAGCC p-klrf-3f: GATTGGACTTAACTTTACCTTCCAGAA p-klrf-65r: TCTTTGATGGTAGCACAGCTATTTTC p-klrf-95 revt: AACCATCCACCCAGGTCCATGTCCT p-gzma-79f: GGAGCTCACTCGATAACCAAGAAA p-gzma-396r: GCTTTAGAAGTTTAAGGTCACCCTCAT p-gzma-3t: TCCTTATCCATGCTTTGACCAGGACACAC p-gzmb-56f: TCTCCTATGGAAGAAAGGATGGAA p-gzmb-65r: ATCCAGGGCAGGAAACTTGA p/h-gzmb-57t: CCTCCACGAGCCTGCACCAAAGTCT p-perf-f: TTCGCGGCCCAGAAGAC p-perf-3r: CTGTAGAAGCGACACTCCACTGA p-perf-59t: CACCAGGACCAGTACCGCTTCAGCC p-klrk-38f: TCTCAAAATTCCAGTCTTCTGAAGATATA p-klrk-9r: AGGATCTGTTTGTTGGAATTTGTACTA p-klrk-63 revt: CCCATCCAATGATATGACTTCACCAATTTGA p-klrb-377f: TTTATAAACACTGGAATGACAGTCTAGCT p-klrb-7r: TGTATGAGTCTCAATTCCTCATTATCTTG p-klrb-8t: CTGTTCCACAAAAGAATCCAGTCTGCTGC p-klrc-39f: GTAATTGTCCAAAGGAATGGTTTACA p-klrc-57r: CGAAGTAGAGTAGAATTCCGTGAAGC p-klrc-8 revt: CACAGGCCATCAAACTCTCATTCCATGTC p-shdb-9f: TGTTAAGGGACAGCGAGTCCAT p-shdb-73r: GAAGATTCGGTATGTGTAGACAAAATTT p/h-shdb-5t: CAGGAGTCCTGTGCCTCTGTGTCTCGTTTA p-klra-85f: TGGCAAGACTGATGAAAAAGAGTT p-klra-57r: GAAACAGAGGATCCCAAGAATCAC p-klra-t: CAGTGCTCTGGCATCGCATTGCA Probes are indicated in bold.

6 8 TOKA ET AL. A 8 YAC- Target Cells Nontreated CHQ Treated B 8 K56-GFP Target Cells Nontreated CHQ Treated % Lysis 6 % Lysis 6 Donor Donor Donor 3 Donor Donor Donor 3 C D E 88C 88A7 88A.73% 5.7%.5% F 8 K56-GFP Target Cells 3 CD PE 3 CD7a- 3 CD PE 3 CD7a- 3 CD PE 3 CD7a- % Lysis 6 CD3 CD NK CD7 Mac CD FIG.. Resting porcine NK cells exhibit low cytotoxicity. Bulk PBMCs or enriched CD + /CD8 + /CD3 cells were cultured with K56-GFP or YAC- cells as targets for h. Lysis levels were determined by a flow cytometry based NK assay. (A) NK cell assay with bulk PBMCs against YAC- and (B) K56-GFP cells; (C) surface expression of CD7a on CD + cells; (D) surface expression of CD7a on CD + cells following a -h incubation with K56 cells; (E) surface expression of CD7a on CD + cells after incubation with CHQ; (F) cytotoxic activity of enriched CD + /CD8 + /CD3 cells and other cell populations in PBMCs against K56-GFP at E:T ratio of 5:. YAC- cells were labeled with 5-μM CFSE to enable quantification of lytic activity by flow cytometry. CHQ = chloroquine, p.5 compared to nonstimulated cells, CD + /CD8 + /CD3. Data are presented as means ± SD of at least three individual experiments. Synergism between innate cytokines in activating porcine NK cells Naturally, cytokines work in concert and we were interested in finding cytokine combinations that optimally stimulated porcine NK cells. We therefore examined the stimulatory capacity of cytokine combinations including pairing pil- with pil-8 or pil-5, pil-8 paired with pil-5 and the combination of all three cytokines. There was an increase in lytic activity of CD + /CD8 + /CD3 cells cultured in the presence of all combinations of cytokines, with a - to 5-fold increase compared to nonstimulated cells (Fig. 3). Synergistic activation was observed between pil- and pil-8 since the level of lysis achieved with the combination was much higher compared to that obtained with individual cytokines. Similarly, stimulation with pil- and pil-5, or pil-5 and pil-8, induced high cytolysis of K56-GFP cells, compared to nonstimulated cells. However, no synergism was observed in these combinations of stimulatory cytokines, because the difference in lytic levels between the combined stimulation and pil-5 alone appeared to be additive. The same was observed when all three cytokines were used together. Moreover, all differences between stimulated and nonstimulated cells were statistically significant (p.5), but differences between cytokine combinations were not. Altogether, significant enhancement of porcine NK cell cytotoxicity may be achieved by the combination of pil-/ pil-8 or pil-5/pil-8, but IL-5 alone is as effective. Cytokines affect induction of IFN-γ and perforin in porcine NK cells In human and mouse studies, NK cells are reported to be the major producers of IFN-γ and also store perforin in their cytotoxic granules (Clark and Griffiths 3). Initially, we investigated the expression of IFN-γ and perforin by intracellular staining in resting PBMCs and CD + / CD8 + /CD3 cells. Only a small percentage of resting NK cells secreted IFN-γ with ~5% expressing perforin (Fig. ). Stimulation with pil-, pil-5, and pil-8 but not IFN-α increased secretion of IFN-γ by CD + /CD8 + /CD3 cells. No synergism was observed when the cytokines were used in combinations. Surprisingly, there was only a -fold increase in perforin expression in response to these cytokines.

7 PORCINE NK CELL ACTIVATION 85 A 8 8 % Lysis 6 % Lysis 6 B % Lysis 8 6 Nonstim IL- Nonstim IL- IL- IL-5 IL-8 IFN-α Overall, stimulation with pil-, pil-5, and pil-8 causes increased production of IFN-γ and perforin in porcine NK cells. Proinfl ammatory cytokines induce proliferation of CD + /CD8 + /CD3 cells To investigate if the increase in the cytotoxicity and cytokine secretion by porcine NK cells could partly result from increase in cell number following culture in the presence of cytokines, we labeled the CD + /CD8 + /CD3 cells with CFSE and cultured them with hil-, pil-, pil-5, pil-8 or pifn-α, or combinations pil-/pil-5, pil-/pil-8, or pil-5/pil-8 for 8 h or 7 h. Cells were cultured for 8 h so that any changes in CFSE dilution could be related to the changes observed in other assays of NK cell function reported here. We found that pil- or pil-8 used at and ng/ml each, respectively, were not sufficient to cause a shift in CFSE dilution, indicating inability to induce proliferation at least in the first 8 h (Fig. 5A). No dilution of IL- IL-5 IL-8 IFN-α FIG.. Proinflammtory cytokines directly activate porcine NK cells. PBMCs or enriched CD + /CD8 + /CD3 cells cultured in the presence of hil-, pil-, pil-5, pil-8, or pifn-α for 8 h followed by NK cell assay against K56-GFP cells. (A) Bulk PBMCs, (B) sorted CD + /CD8 + /CD3 cells. Lysis levels are presented as the mean ±SD of % Lysis at the E:T ratio of 5:. These data are representative of at least three separate determinations. Nonstim IL-/5 IL-/8 IL-5/8 IL-/8 FIG. 3. Cytokine synergism in activating porcine NK cells. Cytokines were added in pairs to enriched CD + /CD8 + / CD3 cells and incubated for 8 h, and the NK assay was performed against K56-GFP. Lysis levels are presented as the mean ±SD of % Lysis at the E:T ratio of 5:. These data are representative of at least three separate determinations. CFSE was observed in nonstimulated cells and as expected pifn-α did not induce proliferation of cells. However, hil- or pil-5 caused CFSE dilution as did the combinations between all cytokines (Fig. 5A). Again pil- and pil-8 showed synergistic activity in inducing proliferation of CD + /CD8 + /CD3 cells (Fig. 5B). Further incubation up to 7 h showed a similar pattern but with pronounced dilution of CFSE indicating that the proinflammatory cytokines caused proliferation of NK cells (Fig 5C and D). Thus, innate cytokine stimulation of porcine NK cells may cause an increase in cell number, which possibly maintains the level of lytic activity and cytokine secretion by porcine NK cells. Cytokine receptor messenger RNA expression We have shown here that porcine NK cells react to stimulation with cytokines. Subsequently, we sought to determine the presence of cytokine receptors, likely responsible for mediating the observed responses. Because there are no reagents available to detect protein expression levels of these cytokine receptors on porcine lymphocytes, we measured the level of messenger RNA (mrna) expression for cytokine receptor genes such as pil-rβ, pil-5rα, pil-8rα, and pifn-αr in purified CD + /CD8 + /CD3 cells. Quantitative real-time RT-PCR assay (Fig. 6A) revealed mrna expression ratios relative to purified CD3 + T cells. pil-8rα was the only receptor mrna highly expressed in resting porcine NK cells. The remaining receptors mrna were also expressed, with lowest expression exhibited by IL-5Rα and IFN-αR mrna. These data indicated that porcine NK cells do express mrna for the cytokine receptors albeit at various levels. Next, NK cells were stimulated with rpil-, rpil-5, or rpil-8 for 8 h using similar cytokine concentrations as those used to assess cytotoxicity enhancement. Stimulation with rpil- strongly induced the expression of pil-8rα mrna and only moderately influenced expression of its respective receptor mrna (Fig. 6B). Similarly, stimulation with rpil-5 upregulated the expression of pil-5rα mrna

8 86 TOKA ET AL. % CD/CD8 Cells Positive for cytokine 8 6 IFN-γ Perforin Nonstim IL- IL-5 IL-8 IL-/5 IL-/8 IL-5/8 IL-/5/8 IFN-α FIG.. Interferon (IFN)-γ production and perforin storage by CD + /CD8 + /CD3 cells. Natural killer (NK) cells were enriched as described earlier and stimulated or not with hil-, pil-, pil-5, pil-8, pifn-α, or their combinations. Intracellular staining was performed for perforin and IFN-γ. Data are presented as means ± SD of CD + positive for the cytokine from four separate experiments. and only minimally affected the mrna for pil-rβ, pil-8rα, and pifn-αr (Fig. 6C). Addition of rpil-8 had little effect on expression of the measured cytokine receptor mrna (Fig. 6D) but only moderately increased expression of its receptor mrna in relation to nonstimulated CD + / CD8 + /CD3 cells. Cytokine combinations (rpil-8/ or rpil-8/5) significantly upregulated expression of IL-5Rα mrna (6-fold, data not shown). We were interested examining the effect of cytokine stimulation on mrna expression of NK cell associated genes. Therefore, a selected set of NK cell activating and inhibitory receptors, and effector molecules was analyzed. As shown in Figure 7 rpil- moderately upregulated most of the genes analyzed, but significantly increased the expression of KLRC (inhibitory receptor) and SHB (negative regulator of NK cell activation) mrna. Recombinant pil-5 only A Non-stim. cells IL- 3 IL-5 IL-8 IFNa IL B IL-/IL-5 Non-stim. cells IL- IL-/IL-8 IL-5/IL-8 3 IL-/5/8 3 3 CFSE C IL IL-8 IFNa IL D IL-/IL-5 IL-/IL-8 IL-5/IL-8 IL-/5/8 3 CFSE FIG. 5. Cytokine stimulation induces proliferation of CD + /CD8 + /CD3 cells. PBMCs were enriched for CD + /CD8 + /CD3 (NK cells) by magnetic separation and labeled with CFSE then cultured with hil-, pil-, pil-5, pil-8, pifn-α and various combinations of these cytokines for 8 h (A and B) and 7 h (C and D). Dilution of CFSE that indicates proliferation was monitored by flow cytometry. Histograms represent a single experiment of three separate determinations.

9 PORCINE NK CELL ACTIVATION 87 highly upregulated KLRF (NKp8) mrna while rpil-8 increased the expression of SHDB mrna. In each case, all other genes were only moderately increased. Cytokine combinations additionally increased expression of granzyme A (GZMA) mrna (data not shown). In summary, it is likely that the observed response of NK cells following treatment with cytokines may have been mediated through their respective receptors. Additionally, stimulation with one cytokine appeared to influence expression of other cytokine receptor mrna which could explain why a synergistic effect in cytotoxicity of porcine NK cells is observed when rpil- is used together with rpil-8. Cytokine-activated porcine NK cells lyse FMDVinfected targets A primary interest was to determine the potential of porcine NK cells, activated with proinflammatory cytokines, to lyse target cells infected with FMDV. We infected SK6 cells with FMDV lacking the leader protease that result in significant attenuation (Mason and others 997). LL-KGE virus is a molecularly derived virus from an infectious clone of FMDV strain A (Rieder and others 996) with the leader protease gene deleted and a mutation in the integrin-binding domain from RGD to KGE. This attenuated virus allowed for infection without killing such that the infected target cells could be used in our NK cytotoxicity assay. Results presented in Figure 8A show that nonstimulated porcine NK cells did not lyse SK6 cells and had very low lytic activity against FMDV-infected SK6 cells. A similar effect was observed when stimulated porcine NK cells were incubated with non-fmdv-infected SK6 cells (Fig. 8B and C). In this instance, lytic levels were comparable to those achieved with nonstimulated porcine NK cells versus K56- GFP cells. Finally, data in Figure 8D and E show at least a - to 5-fold increase in lytic activity against FMDV-infected SK6 cells when porcine NK cells were cultured in the presence of hil- or pil-5. Other cytokines such as pil-8 only induced a -fold increase and pil- or pifn-α did not enhance the lytic activity of porcine NK cells against FMDVinfected SK6 cells. All combinations (Fig. 8E) of cytokines were effective in activating porcine NK cells against FMDVinfected SK6 cells, similar to the effect exerted on K56-GFP cells. Synergism between IL- and IL-8 in activating porcine NK cells to kill infected target cells was not observed. It should be mentioned that the lytic levels of stimulated NK cells against FMDV-infected target cells were slightly lower compared to those obtained in killing assays with K56-GFP cells as targets. Taken together, porcine NK cells stimulated with hil-, pil-5, or pil-8 or combinations between pil-, A Expression ratio IL-8R IL-R IL-5R TNFSF IFN R UBC mrna B Expression ratio IL-8R IL-R IL-5R TNFSF IFN R UBC mrna IL- C Expression ratio 6 8 IL-5 D Expression ratio IL IL-8R IL-R IL-5R TNFSF IFN R UBC mrna IL-8R IL-R IL-5R TNFSF IFN R UBC mrna FIG. 6. Expression of cytokine receptor mrna in CD + /CD8 + /CD3 cells. CD + /CD8 + /CD3 cells were sorted and RNA isolated for qrrt-pcr as described in Materials and Methods. (A) Expression of cytokine receptor mrna in resting CD + / CD8 + /CD3 cells relative to CD3 + T cells; (B) expression of cytokine receptor mrna following stimulation with rpil-; (C) rpil-5; (D) rpil-8 relative to nonstimulated CD + /CD8 + /CD3 cells. The box area in a whisker-box plot encompasses 5% of all observations, the dotted line represents the sample median, and the whiskers represent the outer 5% of observations. Data were collected from three separate experiments encompassing four pigs.

10 88 A 6 Expression ratio B Expression ratio C Expression ratio FIG. 7. Expression of NK cell associated mrna in CD + / CD8 + /CD3 cells following stimulation with cytokines. (A) Messenger RNA expression following 8 h stimulation with rpil-; (B) rpil-5; (C) rpil-8. The box area in a whiskerbox plot encompasses 5% of all observations, the dotted line represents the sample median and the whiskers represent the outer 5% of observations. Data were collected from three separate experiments encompassing four pigs. pil-5, and pil-8 are capable of killing targets infected with FMDV in vitro and thus limiting viral replication. Discussion IL- NCR NCR3 KLRF GZMA GZMB PRF KLRK KLRBKLRCSHDBKLRA UBC mrna IL-5 NCR NCR3 KLRF GZMA GZMB PRF KLRK KLRBKLRCSHDBKLRA UBC mrna IL-8 NCR NCR3 KLRF GZMA GZMB PRF KLRK KLRBKLRCSHDBKLRA UBC mrna In this study, we have characterized porcine NK cells as co-expressing CD and CD8 while lacking CD3 expression. These cells exhibit low, if any, spontaneous cytotoxicity against common tumor cell targets. Activation by TOKA ET AL. proinflammatory cytokines such as IL-5 or IL-8 induce porcine NK cells to become efficient at lysing susceptible tumor cell lines. In addition, combinations of IL-, IL-5, and IL-8 further stimulated porcine NK cells to lysis of tumor cells in vitro. Interestingly, activation with IFN-α had little positive effect on NK cell activity. Heightened lytic activity was accompanied by increased levels of both IFN-γ and perforin. Therefore, in this species, activation of NK cells is required for ex vivo achievement of functional capability and it is likely that in vivo, innate, and adaptive cytokines such as the ones studied here regulate the function of porcine NK cells. The potential role of porcine NK cells in viral infections was assessed knowing that there is little killing of porcine fibroblast cells. Infection of these cells with strains of FMDV that bind through heparin sulfate (Neff and others 998) rather than the physiological integrin receptors (Jackson and others 996; Neff and others 998). for the three receptors, only marginally better targets for killing by naive porcine NK cells. However, as with K56-GFP cell targets, activation of the NK cells with IL-5 or IL-8 leads to significant killing responses against FMDV-infected porcine fibroblasts and these activated cells still have a low-killing capacity when assayed against uninfected targets. These data clearly implicate the activation of NK cells as a potential mechanism of viral clearance of FMDV from infected swine, before the development of antibody responses strong enough to neutralize 5 live viral particles/ml during the peak of viremia. The phenotype of porcine NK cells has not been dissected fully, mostly due to lack of definition of exclusive NK cell surface markers. Initially, Pescovitz and others (988) described the porcine NK cells as being CD + /CD8 +, and recently Denyer and others (6) has confirmed that these cells belong to the CD + /CD8 + /CD3 cell population that exhibit high level of perforin expression, observations confirmed in this study. Some T lymphocytes such as CD3 + T cells also showed the expression of perforin (Denyer and others 6) and a marginal degree of cytotoxicity for tumor cell targets (this study). Both CD and CD8 are expressed on other cell types in the pig such that the precise phenotype of porcine NK cells is still only partially defined. Porcine NK cells appear quiescent at rest compared to murine and human cells and require activation by cytokines. The low cytotoxicity of porcine NK cells observed was characteristic in Yorkshire pigs and was in agreement with data reported by Knoblock and Canning (99). Animals raised in livestock operations may require a higher threshold for innate responses than the laboratory mouse, given the constant exposure to environmental insults, both inert and biological, particularly pathogens. Pinto and others(988) described comparison of porcine NK cells with mouse NK cells and found similar low or lack of lytic activity in pigs after h of incubation. Interestingly, these same swine cells showed considerably higher lytic activity after 8 h of co-culture. This highlights the issue of NK cell reactivity against tumor cell targets being complicated by the lack of consensus in various reports as to the standard incubation time at which to measure lytic activity. As reported by Richerson and Misfeldt (989), environmental factors appear to play a role in the level of cytotoxicity of porcine NK cells, but Sutherland and others (6) showed no influence of breed. Animals bred conventionally

11 PORCINE NK CELL ACTIVATION 89 A 8 Non-stim NK/non-infect.SK6 cells Non-stim NK/infect.SK6 cells B 8 6 IL- IL- IL-5 IL-8 IFNa D 8 6 IL- IL- IL-5 IL-8 IFNa Lysis % 6 5: : E:T Ratio 6: C 8 6 5: : E:T Ratio 6: IL-/5 IL-/8 IL-5/8 IL-/5/8 E 8 6 5: : E:T Ratio 6: IL-/8 IL-/5 IL-5/8 IL-/5/8 5: : E:T Ratio 6: 5: : E:T Ratio 6: FIG. 8. Proinflammatory cytokines activate porcine NK cells to lyse FMDV-infected target cells. SK6 cells were first infected with LL-KGE virus, a live attenuated laboratory strain of FMDV and labeled with CFSE, then incubated with NK cells previously stimulated with the innate cytokines (hil-, pil-, pil-5, pil-8, pifn-α) or not. Cytotoxicity was measured by a flow cytometry assay. (A) Nonstimulated NK cells versus noninfected SK6 cells and nonstimulated NK cells versus FMDV-infected SK6 cells; (B) single cytokine stimulated NK cells versus noninfected SK6 cells; (C) cytokine combination stimulated NK cells versus noninfected SK6 cells; (D) single cytokine stimulated NK cells versus FMDV-infected SK6 cells; (E) cytokine combination stimulated NK cells versus FMDV-infected SK6 cells. Data are presented as means ± SD of percentage lysis at each E:T ratio from three separate determinations. had higher NK cell cytotoxicity at rest compared to pigs raised in specific pathogen-free conditions. In studies described here we show that if the porcine NK cells are stimulated with proinflammatory cytokines for 8 h in vitro, their lytic activity can be demonstrated within h of incubation with tumor target cells. A likely explanation could be that activating receptor expression on resting porcine NK cells is a limiting factor regarding their spontaneous cytotoxicity. Natural killer cell receptor contribution to reactivity in resting NK cells was observed in studies performed on human NK cells. In redirected NK assays, only CD6 mediated lysis in resting NK cells whereas NKp6, NKGD, B, DNAM-I, or CD were all involved in enhanced activity of NK cells following activation with IL- (Bryceson and others 6). Our results are similar to the recent findings of Fehniger and others (7) with murine NK cells, where there was low cytotoxicity of resting murine NK cells which was enhanced by stimulation with cytokines. They observed similar patterns of activation of NK cells by IL-5 or IL-8. These data point to a possible role of proinflammatory cytokines in regulating the activation of porcine NK cells, however, investigations in vivo have yet to confirm such a role. Other studies have reported the possibility that IL- or IL-8 activate porcine NK cells minimally but act synergistically when used in tandem (Pintaric and others 8). We confirm synergism between IL- and IL-8, and add to that body of data that IL- or IL-5 are more potent inducers of porcine NK cell proliferation and cytotoxicity when used alone or in combination with IL- or IL-8. On the other hand IL-5 supported proliferation and was consistent with reports in other species that IL-5 is essential for survival of NK cells and NK dendritic cells (DCs), as these cells were barely detectable in IL-5 / gene knockout animals. Such animals reconstituted by exogenous administration of IL-5 have normal NK cell profiles. This indicates the important role of IL-5 in proliferation and survival of these cells mediating innate responses (Chaudhry and others 7). Indeed, IL-5 deficient mice were not protected against HSV challenge emphasizing the critical role of IL-5 in innate responses (Gill and Ashkar 7). Here, we present the initial report of enhanced reactivity of porcine NK cells following stimulation with IL-5. Because IL- is secreted later in infection and impacts the adaptive immune response, it has not received attention for in vivo enhancement of NK cell activity. In the mouse model not only T cells, but DCs produce IL- which is reported to activate NK cells (Granucci and others ). Previous studies (Knoblock and Canning 99; Duchene and others 8; Pintaric and others 8) analyzing the effect of IL- on porcine NK cells do not explain what may be the in vivo source of this cytokine potentially impacting the function of porcine NK cells. As such, it remains unknown if porcine DCs produce IL- and under what circumstances it may be produced which will lead to activation of NK cells. Interferon-α

12 9 activated porcine NK cells from 3 to -month-old pigs, although not to the same level as IL- or IL-5. Derbyshire and Lesnick (99) reported that this innate cytokine, produced in response to piclc in newborn piglets, activated NK cells only briefly and concluded that prolonged stimulation with IFN-α caused hyporeactivity of porcine NK cells. However, treatment of such pigs with prostaglandin F-α restored the activity of NK cells. Although we could not determine surface expression levels of the cytokine receptors on porcine NK cells as there are no antibodies to these receptors, the quantitative realtime PCR assay showed that there was a detectable level of cytokine receptor mrna expression. Resting stage mrna expression levels were influenced by addition of cytokines to the NK cell culture environment. This could suggest a mechanism through which cytokine synergism may be achieved, e.g., stimulation with IL- achieved only low levels of cytotoxicity while in combination with IL-8 increased cytotoxic activity of porcine NK cells against K56-GFP cells was observed. Further, stimulation of NK cells with cytokines influenced the expression of other NK-associated genes which presumably contributed to the functional activity of porcine NK cells. However, because of the lack of reagents and the narrow spectrum of analyzed genes it is difficult to speculate on the extent of cytokine influence on these genes. It is established that the balance of signals from both activating and inhibitory receptors on NK cells dictate the functional outcome of encounter with a target cell [reviewed in Lanier (5)]. Few examples exist in the porcine system to show how the exogenous cytokines may impact the innate response to infection with viruses. Administration of ril- protected pigs from porcine reproductive and respiratory syndrome virus (Carter and Curiel 5), however, this work did not assess whether NK cell activation occurred or not. Perhaps the best demonstration of the influence of exogenous cytokines on enhancing the porcine innate immunity is that described recently by Moraes and others (7) in which they showed that inoculation of pigs with a combination of had5 vectors encoding pifn-α and pifn-γ completely protected the animals from FMDV infection and progression to clinical disease. Although they showed possible molecular pathways for induction of immunity in this setting, they did not describe the cellular basis of the protection. It is likely that virus clearance could have been mediated by exogenous cytokineactivated NK cells, however, in comparison to proinflammatory cytokines, IFN-α modestly stimulated NK cells. Evaluation of the efficacy of IL-5 or IL- plus IL-8 in protection against FMD virus challenge is presently under development. There are likely many pathways working in concert to activate NK cells in vitro. Short-term over expression of any single cytokine may boost the innate response capable of protecting against challenge with FMDV or potentially interfere with the innate response. The role of viral immune evasion during the acute infection caused by FMDV in swine may also be impacting NK cell activity, especially relative to different DC populations (Nfon and others 8). These cells are a common source of the cytokines that we show activate NK cells and the virus may interfere with the DC/NK interaction even though neither cell is infected by FMDV in vitro or in vivo (Bautista and others 3; Bautista and others 5; Nfon and others 8). Planned in vivo TOKA ET AL. experiments will determine the outcome of exogenous cytokine administration. Conclusion In summary, porcine NK cells can be identified as contained within the CD + /CD8 + /CD3 cell population and naive cells require prior activation to elicit their cytotoxic function. Proinflammatory cytokines such as IL-, IL-, IL-5, IL-8, and IFN-α activate porcine NK cells and enhance their cytotoxicity. Notably, porcine NK cells are noncytotoxic for porcine fibroblasts but activated NK cells kill FMDV-infected fibroblasts in vitro. These studies imply that it may be possible to use proinflammatory cytokines to enhance the porcine innate responses against infection with viruses, a strategy that specifically may be applicable for interventional vaccine design. Studies are currently underway to incorporate proinflammatory cytokine genes in suitable vectors to evaluate counter measures designed for early protection against FMDV. Acknowledgments This work was supported by grants CRIS #9-3-5-D from the Agricultural Research Service, US Department of Agriculture (USDA) (to W.T.G.), CRIS # from the Agricultural Research Service, USDA (to H.D.), and interagency agreement number , between the Department of Homeland Security, Science and Technology Directorate and USDA (to W.T.G. and D.M.E.). C.K.N. was the recipient of a Plum Island Animal Disease Center Research Participation Program fellowship, administered by the Oak Ridge Institute for Science and Education (ORISE) through an interagency agreement between the US Department of Energy (DOE) and the USDA. We thank Dr. Janice Endsley, UTMB Galveston, for her assistance in initiating these studies and the animal care staff at the Plum Island Animal Disease Center for their professional support and assistance. Finally, we thank the members of the Foreign Animal Disease Research Unit for their support, consultation, and discussion of this work. References Alter G, Malenfant JM, Altfeld M.. CD7a as a functional marker for the identification of natural killer cell activity. J Immunol Methods 9( ):5. Bautista EM, Ferman GS, Golde WT. 3. Induction of lymphopenia and inhibition of T cell function during acute infection of swine with foot and mouth disease virus (FMDV). Vet Immunol Immunopathol 9( ):6 73. Bautista EM, Ferman GS, Gregg D, Brum MC, Grubman MJ, Golde WT. 5. Constitutive expression of alpha interferon by skin dendritic cells confers resistance to infection by foot-and-mouth disease virus. J Virol 79(8): Bryceson YT, March ME, Ljunggren HG, Long EO. 6. Synergy among receptors on resting NK cells for the activation of natural cytotoxicity and cytokine secretion. Blood 7(): Byrne P, McGuirk P, Todryk S, Mills KH.. Depletion of NK cells results in disseminating lethal infection with Bordetella pertussis associated with a reduction of antigen-specific Th and enhancement of Th, but not Tr cells. Eur J Immunol 3(9): Carter QL, Curiel RE. 5. Interleukin- (IL-) ameliorates the effects of porcine respiratory and reproductive