International Immunology Advance Access originally published online on July 18, 2006
International Immunology 2006 18(9):1347-1354; doi:10.1093/intimm/dxl071
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Constitutively polarized granules prime KHYG-1 NK cells
1 Department of Medical Oncology and Hematology, Princess Margaret Hospital/Ontario Cancer Institute, 610 University Avenue, Suite 5-211, Toronto, ON M5G2M9, Canada
2 Division of Experimental Therapeutics, Toronto General Research Institute, University of Toronto, Toronto, ON M5G2M1, Canada
3 Institute of Medical Science, University of Toronto, Toronto, ON M5G2M1, Canada
4 Research and Development, Canadian Blood Services, Toronto Centre, Toronto, ON M5G2M1, Canada
5 Present address: Division of Biomedical Sciences, Johns Hopkins in Singapore, 31 Biopolis Way, #02-01, The Nanos, Singapore 138669
Correspondence to: G. Suck; E-mail: garnet.suck{at}uhn.on.ca
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResults and discussionReferences |
|---|
The major mechanism for NK cell lysis of tumor cells is granule-mediated cytotoxicity. Polarization of granules is a prelude to the release of their cytotoxic contents in response to target-cell binding. We describe the novel observation of constitutive granule polarization in the cytotoxic NK cell line, KHYG-1. Continuous degranulation of KHYG-1 cells, however, does not occur and still requires target-cell contact. Disruption of microtubules with colcemid is sufficient to disperse the granules in KHYG-1 and significantly decreases cytotoxicity. A similar effect is not obtained by inhibiting extracellular signal-related kinase 2 (ERK2), the most distal kinase investigated in the cytolytic pathway. Disruption of microtubules significantly down-regulates activation receptors, NKp44 and NKG2D, implicating them as potential microtubule-trafficking receptors. Such changes in upstream receptor expression may have caused deactivation of ERK2, since NKG2D cross-linking also leads to receptor down-regulation and diminished ERK phosphorylation. Thus, a functional role for NKG2D in KHYG-1 cytotoxicity is demonstrated. Moreover, the novel primed state may contribute to the high cytotoxicity exhibited by KHYG-1.
Keywords: colcemid, cytotoxic granules, ERK, KHYG-1, microtubules, NK cell, NKG2D, NKp44, perforin
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResults and discussionReferences |
|---|
Granule-mediated lysis of virally infected and tumor cells is the major mechanism of NK cell cytotoxicity (1). NK target cell conjugation triggers the formation of an immunological synapse and initiates a series of events in a specific temporal manner, including recruitment of receptors, re-orientation of the microtubule-organizing center (MTOC) and polarization of cytotoxic granules (2). ERK2 is the most downstream regulatory kinase investigated as demonstrated in an elegant granule polarization assay (3). Although significant progress has been made in elucidating signaling pathways involved in granule exocytosis, the exact sequence of events resulting in target-cell lysis is not yet fully defined. We employed the highly cytotoxic NK cell line KHYG-1 (4, 5) as a model to further elucidate mechanisms of enhanced killing by NK cells.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResults and discussionReferences |
|---|
Cell lines
K562 (American Type Culture Collection, Manassass, VA, USA) was maintained in RPMI 1640 medium/glutamine (RPMI, Invitrogen, Grand Island, NY, USA)/10% (v/v) heat-inactivated fetal bovine serum (HyClone, South Logan, UT, USA). KHYG-1 (4), JCRB0156 (HSRRB, Tokyo, Japan), was cultured in RPMI/2% (v/v) human low-toxicity AB serum (from a single donor after informed consent)/450 U ml–1 recombinant human IL-2 (rIL-2, Chiron, Quebec, Canada). NK-92 was kindly provided by Hans Klingemann (then at Rush University Medical Center, Chicago, USA) and grown in X-Vivo 10 medium (Bio Whittaker, Walkersville, MD, USA)/1.8 mM L-serine, 3 mM L-glutamine, 0.6 mM L-asparagine, 2.5% human plasma (AB, SeraCare, Life Sciences, Oceanville, CA, USA) and 450 U ml–1 rIL-2.
Reagents
The mAb against phospho-extracellular signal-related kinase (phospho-ERK, clone E-4) was from Santa Cruz Biotechnology (Santa Cruz, CA, USA), against ß-actin (clone AC-15) from Sigma–Aldrich (St Louis, MO, USA), against perforin (PFP) (clone 
G9) for confocal microscopy from BD Biosciences PharMingen (BD, San Jose, CA, USA) and against PFP (clone 2D4) for immunoblotting (6) was kindly provided by Richard Miller (Ontario Cancer Institute, Toronto, Ontario, Canada). mAbs against 
-tubulin (TUB) (clones B-5-1-2 and YOL1/34) for microscopy were from Sigma–Aldrich and Abcam (Cambridge, MA, USA), respectively; mAb against NKG2D (clone 1D11, BD Biosciences PharMingen) was used for blocking studies and mouse IgG1 (Dako, Mississauga, Ontario, Canada) as control. Polyclonal antibody against ERK1/2 was from Upstate Biotech (Lake Placid, NY, USA). mAbs for flow cytometry, PE-labeled anti-CD2, allophycocyanin-labeled leukocyte function-associated antigen 1 (LFA-1; CD11a) and FITC-conjugated CD44 were from BD. Anti-NKG2D–PE was from R&D Systems (Minneapolis, MN, USA) and anti-NKp44–PE from Beckman Coulter (Fullerton, CA, USA).
Immunostaining and confocal microscopy
KHYG-1 cells were immunostained according to Wei et al. (7), but fixed in 4% PFA, 20 min at 4°C. Dilutions were 1:100 for anti-TUB (Sigma–Aldrich), 1:200 for anti-TUB (Abcam), 1:200 for anti-PFP (BD) and 1:200 for goat anti-mouse (GAM)–FITC (Sigma–Aldrich) antibodies. Cells were counterstained with 4',6-diamidino-2-phenylindole (DAPI) (EMD Biosciences, San Diego, CA, USA). For double staining, primary antibodies anti-PFP (BD Biosciences PharMingen) and anti--TUB (Abcam) were co-incubated, followed by sequential incubation with GAM-Alexa Fluor 568-labeled (Invitrogen) and donkey anti-rat Alexa Fluor 488-labeled (Invitrogen). Control slides (data not shown) showed no cross-reactivity of antibodies or staining of secondary antibodies alone. Slides were analyzed with a Zeiss confocal laser-scanning microscope LSM510, using the x63 water-immersion objective lens. Quantitative analysis of polarized cells was performed by randomly scanning the slide and imaging sufficient areas, for final analyses of 
100–300 cells.
Pharmacological inhibitor treatments
Cells (2.5 x 105–5 x 105 ml–1) were treated with the mitogen-activated protein kinase (MAPK) pathway inhibitors PD98059 (Sigma) and U0126 (Sigma) in PBS for 30 min at 37°C, washed 3–4x in PBS for subsequent applications. For colcemid treatment, either cells were treated with KaryoMax (Invitrogen), 106 cells per 200 µl solution (10 µg ml–1 in PBS), and incubated for 7 h at 37°C or 2.5 x 105–5 x 105 ml–1 cells were treated for 12–15 h with Demecolcemid (Sigma), in culture medium (or with gentamicin 10 µg ml–1), concentrations as indicated in the text. Treatments were followed by 3–4x washes in PBS prior to use in subsequent experiments.
Flow cytometry/cytotoxicity assay
Flow cytometry and the flow cytometric cytotoxicity assay were performed as previously published (5). Briefly, for the cytotoxicity assay, effector NK cells and K562 target cells were plated at 10:1, 20:1 or 40:1 ratios in triplicate into 96-well U-bottom plates, in RPMI 1640, 0.1% BSA and 225 U ml–1 rIL-2 for 3.5–4.5 h at 37°C. As controls for a 0-h time point, the effector and target cells were plated in parallel at identical concentrations, also in triplicate, but separated from each other and pooled together at the time of harvest. Cell mixtures were stained (see above) for 100% detection of KHYG-1 with anti-CD2–PE, followed by Annexin V–FITC and 7-amino actinomycin D (7AAD) staining, according to the manufacturer's instructions, and subjected to flow cytometry (5). A total of 10 000 events/reaction were analyzed for each sample. Percent lysis was reverse calculated from the percentage of viable target cells at each time point as compared with time 0 h, excluding 7AAD and Annexin V-positive cells and debris, according to the formula [time 0 h (mean) – time point h (e.g. 4 h) (mean)/ time 0 (mean) x 100] and the standard error of the mean (SEM) was calculated. Significance for the results is expressed as P values determined by analysis of variance (ANOVA) or Student t-test as indicated in the figure legends.
Immunoblotting
Immunoblotting and cell lysis were performed as previously published (5, 8). For phosphorylation studies, cells were serum deprived for 2–5.5 h (PD98059 or U0126), or 12–15 h (colcemid), in RPMI/ 0.1% BSA and 450 U ml–1 rIL-2, before lysis. Protein concentrations were determined using the BCATM Protein Assay Kit (Pierce, Rockford, IL, USA). A total of 60–100 µg of protein were separated on 10% SDS-polyacrylamide gels.
Target-cell contact
K562 cells were fixed in 1% PFA (106 cells ml–1) for 30 min on ice, followed by four washes in PBS. KHYG-1 or NK-92 cells were mixed with K562 cells in a 5:1 ratio, centrifuged for 1 min and co-incubated (106 cells ml–1) for 5, 10 or 30 min as indicated in the figure legend, at 37°C in assay medium (see above, RPMI 1640, 0.1% BSA and 225 U ml–1 rIL-2), washed and lysed.
Antibody blocking and receptor down-regulation
(i) For the effect of antibody blocking on KHYG-1 cytotoxicity and ERK phosphorylation, KHYG-1 cells were washed with PBS and 106 cells ml–1 were incubated with 0.3, 1 or 3 µg ml–1 anti-NKG2D (clone 1D11, BD Biosciences PharMingen), or 3 µg ml–1 mouse IgG1 (Dako) as control, for 45 min at 37°C and washed 2x in PBS for subsequent experiments; (ii) For NKG2DR down-regulation, two sets of KHYG-1 cells were incubated with 3 µg ml–1 mouse mAb anti-NKG2D (clone 1D11, BD Biosciences PharMingen) or 3 µg ml–1 mouse IgG1 (Dako) as control. One set of KHYG-1 cells was left at 4°C, while the second set was moved to 37°C to monitor whether or not antibody binding induces down-regulation of the receptor. Cells were washed 2x and then stained with secondary antibody GAM–FITC (Sigma) and analyzed by flow cytometry (see above).
Densitometry
Densitometry was performed on scanned autoradiographs using Quantity One software (Bio-Rad, Mississauga, Ontario, Canada) and further analyzed with Excel software.
Statistical analysis
Statistical analysis was performed using Excel software. Significance for the results is expressed as P values determined by ANOVA or Student t-test as indicated in the figure legends.
|
Results and discussion |
|---|
|
TopAbstractIntroductionMethodsResults and discussionReferences |
|---|
Granules are constitutively polarized in KHYG-1 and reversibly dispersed by colcemid
Granule polarization is known as a consequence of target-cell contact, at least in NK cells and cytotoxic T cells (3, 7, 9, 10). Confocal imaging of KHYG-1, however, revealed that granules are constitutively polarized in the absence of target cells. Figure 1(A) shows KHYG-1 compared with the NK-92 NK cell line with dispersed granules as previously published (3). Granules are secretory vesicles trafficking on microtubules (2). We were able to disperse granules in KHYG-1 reversibly by disrupting the microtubule system with colcemid (Fig. 1B), from 94 to 10% polarized cells in the population, with re-polarization of 90% compared with 91% control cells (Fig. 1B). Thin sectioning of KHYG-1 cells double stained for PFP and
-TUB revealed partial co-localization of the two proteins in the putative MTOC area (Fig. 1C, upper panel, yellow areas), indicating the presence of PFP not only contained in vesicles but also attached to the microtubule fiber. Attachment to microtubules is known for microtubule-associated proteins, which can play a role in fiber stabilization (11). Colcemid treatment disrupted the observed prominent co-localization of PFP and -TUB in KHYG-1 (Fig. 1C, lower panel). Our data infer a novel potential function for PFP in stabilizing microtubules and thereby granule polarization in KHYG-1.
|
Microtubule disruption by colcemid reduces KHYG-1 cytotoxicity
We next tested the effect of colcemid treatment on KHYG-1 cytotoxicity. To focus on granule-mediated cytolysis, we used K562 as target cells and kept assay duration short (
4 h). KHYG-1 cells were pre-treated with colcemid to exclude any drug-mediated effects on K562. Figure 2(A) shows a significant concentration-dependent decrease in KHYG-1 cytotoxicity as measured in our flow cytometric cytotoxicity assay. A plateau effect is notable at doses 
1 µM. Target-cell killing, however, was not completely abrogated, although polarization was significantly disrupted (Fig. 2B). Cell viability was partially affected by colcemid treatment (Fig. 2C); however, the E:T ratios for the cytotoxicity assay were based on viable cell numbers in the flow cytometric assay, supporting the notion of a direct colcemid effect. This is further demonstrated at different E:T ratios in Fig. 2(D). Moreover, these data are in accordance with previous studies with the related drug colchicine, revealing time-dependent (2 h) reversibility of suppressed NK cell cytotoxicity (12). We found that a proportion of KHYG-1 cells, incubated in parallel with the assay, re-polarized their granules during assay duration (Fig. 2B). Considering that in KHYG-1 70% of the maximum killing was already achieved after 2 h (Fig. 2E); this could in part be attributed to the observed lysis after colcemid treatment (Fig. 2A). Furthermore, target-cell binding and granule recycling, two events unrelated to microtubule function (12), could be involved. Our data demonstrate a dynamic correlation between constitutive granule polarization, microtubule disruption, re-assembly and target-cell lysis in KHYG-1 cells.
|
Inhibition of ERK2 is insufficient to significantly interfere with granule polarization and cytotoxicity in KHYG-1
The most distal kinase currently known to mediate NK cell cytotoxicity and granule polarization is MAPK/ERK2. ERK2 is transiently phosphorylated, resulting in its activation within minutes following susceptible target-cell contact (13). Pre-treatment of NK cells with the MAPK pathway inhibitor PD98059 inhibited granule polarization and markedly reduced cytotoxicity of NK cells (7, 13). In contrast to other NK cells, KHYG-1 exhibits constitutively activated ERK2 (5), which could induce constitutive granule polarization. Although two different MAPK pathway inhibitors, PD98059 and U0126, abolished constitutive ERK2 activation in KHYG-1 (Fig. 3A), only a partial effect on granule polarization was detectable with the stronger inhibitor, U0126 (Fig. 3B). Furthermore, ERK2 inhibition decreased KHYG-1 cytotoxicity only insignificantly at the highest doses (Fig. 3C), which at least for PD98059, was likely due to impaired viability (Fig. 3D). We conclude from these data that inhibiting ERK2 activation in KHYG-1 is insufficient to significantly interfere with granule polarization and cytotoxicity. This is in marked contrast to findings with other NK cell lines (NK-92, YT) and primary NK cells (3, 13). Our findings may be directly related to the constitutive activation of ERK2 in KHYG-1, implying the existence of at least one unknown target downstream of ERK2 responsible for granule polarization and cytotoxicity. KHYG-1 may therefore serve as a suitable model to identify such targets.
|
PFP is released after target-cell contact, but not after treatment with higher doses of colcemid
KHYG-1 is a cell line likely arrested at a later stage of the granule release pathway, normally initiated by target-cell contact. KHYG-1 seems exceptionally primed for cytolytic activity. Interestingly, despite the presence of polarized granules in KHYG-1, continuous degranulation does not occur. When KHYG-1 cells and NK-92 control cells were exposed to fixed K562 target cells, significant levels of the granule constituent PFP were released as indicated by the reduced protein levels in the immunoblots shown in Fig. 4(A). In KHYG-1, target-cell contact is, therefore, required to trigger the final events resulting in granule release. Our findings with KHYG-1 are consistent with the notion of separate regulatory mechanisms for polarization and degranulation, as recently described in NK cells (14).
|
We further tested the effect of colcemid on PFP release after target-cell contact. As shown in Fig. 4(B), PFP release is inhibited by colcemid treatment at least at the highest doses. These data are in accordance with the above-described inhibitory effect of colcemid on KHYG-1 cytotoxicity (Fig. 2A and D). Our findings with KHYG-1 indicate that granule polarization could be a prelude for degranulation in granule-mediated cytotoxicity.
Constitutive NKp44 and NKG2DR expression and ERK2 activation are down-modulated by microtubule disruption
It has been shown with fibroblasts that colcemid treatment induces the formation of focal adhesions (15). We measured the expression levels of two adhesion receptors, LFA-1 (CD11a) and CD44, known for their cooperative function in NK cells (16). A minor induction in expression levels, similar for both receptors, was evident, but was not statistically significantly different for either (Fig. 5A). We previously suggested a potential link between the expression of the activation receptors, NKp44 and NKG2D, and enhanced cytotoxicity in KHYG-1. NKp44, an IL-2 inducible NK cell marker, was constitutively expressed in KHYG-1 and unaffected by IL-2 deprivation (5). Here, we investigate the effect of microtubule disruption on NKp44 and NKG2D by flow cytometry. Figure 5(B) shows a significant decrease in NKp44 and NKG2D expression. Furthermore, colcemid significantly deactivated ERK2 in a dose-dependent fashion, likely a consequence of the observed down-regulation of the upstream NKp44 and NKG2D receptors (Fig. 5C). A plateau effect at doses 1 µM was notable as described for our cytotoxicity studies above. These findings indicate a correlation between constitutive ERK2 activation and constitutive NKp44/NKG2D expression in KHYG-1 that is linked to the microtubule cytoskeleton. Constitutive receptor expression has been described for other receptors, for example as a result of gain of function mutations (17). Blocking NKG2D diminishes KHYG-1 cytotoxicity (Fig. 6A) as similarly recently shown by Hanaoka et al. (18). Furthermore, we observe a decrease in ERK phosphorylation after NKG2D antibody treatment (Fig. 6B), which is likely a direct consequence of receptor down-regulation (Fig. 6C). These data are in accordance with previous findings in endogenous NK cells showing that NKG2D is down-regulated after engagement with its ligands or after antibody cross-linking (19). These experiments also identify NKp44 and NKG2D as two potential microtubule-trafficking receptors. It is, therefore, feasible to postulate from our findings that NKp44 and NKG2D are up-regulated upon susceptible target-cell contact in a microtubule-dependent fashion and might then trigger degranulation. A recent report of a xenogeneic model also showed a role for both NKp44 and NKG2D receptors in triggering NK cytotoxicity (20).
|
|
In conclusion, we show that the highly cytotoxic NK cell line KHYG-1 appears to be ‘primed for action’. Only the final events of the granule release-signaling pathway are to be triggered by target-cell contact. Studies with KHYG-1 may not only provide insights into further enhancing the cytotoxicity of NK cells but also help to elucidate the key regulatory mechanisms of granule secretion relevant for other granular cells, including cytotoxic T cells,
-T cells, macrophages, neutrophils, eosinophils and mast cells.
|
Acknowledgements |
|---|
A.K. holds the Gloria and Seymour Epstein Chair in Cell Therapy and Transplantation at the University Health Network and University of Toronto.
|
Abbreviations |
|---|
| 7AAD, 7-amino actinomycin D |
| ANOVA, analysis of variance |
| DAPI, 4',6-diamidino-2-phenylindole |
| DMSO, dimethylsulfoxide |
| ERK, extracellular signal-related kinase |
| GAM, goat anti-mouse |
| LFA-1, leukocyte function-associated antigen 1 |
| MAPK, mitogen-activated protein kinase |
| MTOC, microtubule-organizing center |
| PFP, perforin |
| TUB, tubulin |
|
Notes |
|---|
* These authors contributed equally to this work.
Received 14 January 2006, accepted 22 June 2006.
|
References |
|---|
|
TopAbstractIntroductionMethodsResults and discussionReferences |
|---|
- Orange JS and Ballas ZK. (2006) Natural killer cells in human health and disease. Clin. Immunol. 118:1.[CrossRef][Web of Science][Medline]
- Vyas YM, Maniar H, Dupont B. (2002) Visualization of signaling pathways and cortical cytoskeleton in cytolytic and noncytolytic natural killer cell immune synapses. Immunol. Rev. 189:161.[CrossRef][Web of Science][Medline]
- Jiang K, Zhong B, Gilvary DL, et al. (2000) Pivotal role of phosphoinositide-3 kinase in regulation of cytotoxicity in natural killer cells. Nat. Immunol. 1:419.[CrossRef][Web of Science][Medline]
- Yagita M, Huang CL, Umehara H, et al. (2000) A novel natural killer cell line (KHYG-1) from a patient with aggressive natural killer cell leukemia carrying a p53 point mutation. Leukemia 14:922.[CrossRef][Web of Science][Medline]
- Suck G, Branch DR, Smyth MJ, et al. (2005) KHYG-1, a model for the study of enhanced natural killer cell cytotoxicity. Exp. Hematol. 33:1160.[CrossRef][Web of Science][Medline]
- Baetz K, Isaaz S, Griffiths GM. (1995) Loss of cytotoxic T lymphocyte function in Chediak-Higashi syndrome arises from a secretory defect that prevents lytic granule exocytosis. J. Immunol. 154:6122.[Abstract]
- Wei S, Gamero AM, Liu JH, et al. (1998) Control of lytic function by mitogen-activated protein kinase/extracellular regulatory kinase 2 (ERK2) in a human natural killer cell line: identification of perforin and granzyme B mobilization by functional ERK2. J. Exp. Med. 187:1753.
[Abstract/Free Full Text] - Branch DR and Mills GB. (1995) pp60c-src expression is induced by activation of normal human T lymphocytes. J. Immunol. 154:3678.[Abstract]
- Kuhn JR and Poenie M. (2002) Dynamic polarization of the microtubule cytoskeleton during CTL-mediated killing. Immunity 16:111.[CrossRef][Web of Science][Medline]
- Bossi G and Griffiths GM. (2005) CTL secretory lysosomes: biogenesis and secretion of a harmful organelle. Semin. Immunol. 17:87.[CrossRef][Web of Science][Medline]
- Dehmelt L and Halpain S. (2005) The MAP2/Tau family of microtubule-associated proteins. Genome Biol. 6:204.[CrossRef][Medline]
- Katz P, Zaytoun AM, Lee JH Jr. (1982) Mechanisms of human cell-mediated cytotoxicity. III. Dependence of natural killing on microtubule and microfilament integrity. J. Immunol. 129:2816.[Abstract]
- Trotta R, Puorro KA, Paroli M, et al. (1998) Dependence of both spontaneous and antibody-dependent, granule exocytosis-mediated NK cell cytotoxicity on extracellular signal-regulated kinases. J. Immunol. 161:6648.
[Abstract/Free Full Text] - Bryceson YT, March ME, Barber DF, Ljunggren HG, Long EO. (2005) Cytolytic granule polarization and degranulation controlled by different receptors in resting NK cells. J. Exp. Med. 202:1001.
[Abstract/Free Full Text] - Sablina AA, Chumakov PM, Levine AJ, Kopnin BP. (2001) p53 activation in response to microtubule disruption is mediated by integrin-Erk signaling. Oncogene 20:899.[CrossRef][Web of Science][Medline]
- Matsumoto G, Nghiem MP, Nozaki N, Schmits R, Penninger JM. (1998) Cooperation between CD44 and LFA-1/CD11a adhesion receptors in lymphokine-activated killer cell cytotoxicity. J. Immunol 160:5781.
[Abstract/Free Full Text] - Thompson MD, Burnham WM, Cole DE. (2005) The G protein-coupled receptors: pharmacogenetics and disease. Crit. Rev. Clin. Lab. Sci. 42:311.[CrossRef][Web of Science][Medline]
- Hanaoka N, Kawaguchi T, Horikawa K, Nagakura S, Mitsuya H, Nakakuma H. (2006) Immunoselection by natural killer cells of PIGA mutant cells missing stress-inducible ULBP. Blood 107:1184.
[Abstract/Free Full Text] - Hanna J, Gonen-Gross T, Fitchett J, et al. (2004) Novel APC-like properties of human NK cells directly regulate T cell activation. J. Clin. Invest. 114:1612.[CrossRef][Web of Science][Medline]
- Forte P, Lilienfeld BG, Baumann BC, Seebach JD. (2005) Human NK cytotoxicity against porcine cells is triggered by NKp44 and NKG2D. J. Immunol. 175:5463.
[Abstract/Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. K. Purdy and K. S. Campbell SHP-2 Expression Negatively Regulates NK Cell Function J. Immunol., December 1, 2009; 183(11): 7234 - 7243. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||