Journal of Immunological Methods 294 (2004) 15 – 22 www.elsevier.com/locate/jim

Research paper

CD107a as a functional marker for the identification of natural killer cell activity
Galit Alter, Jessica M. Malenfant, Marcus Altfeld*
Partners AIDS Research Center, Massachusetts General Hospital and Division of AIDS, Harvard Medical School, Massachusetts, USA Received 17 June 2004; received in revised form 5 August 2004; accepted 10 August 2004 Available online 25 September 2004

Abstract Natural killer (NK) cells are a subset of lymphocytes that play a central role in the innate immune response to tumors and infections. An important limitation in the field of NK research is attributable to the deficit of assays available for the detection of the functional activity of NK cells. Recently, lysosomal-associated membrane protein-1 (LAMP-1 or CD107a) has been described as a marker of CD8+ T-cell degranulation following stimulation. Here we describe CD107a as a marker of NK cell functional activity using multi-parameter flow cytometry. CD107a is significantly upregulated on the surface of NK cells following stimulation with MHC devoid targets. Additionally, CD107a expression correlates with both cytokine secretion and NK cell-mediated lysis of target cells. However, as well as being coordinately expressed on nearly all cytokine secreting cells, CD107a was also expressed on a large subset of NK cells that did not secrete cytokine following stimulation. These data suggest that employing CD107a as a marker of NK cell functional activity may allow for the identification of a large fraction of activated NK cells that may degranulate in the absence of cytokine secretion. Cumulatively, the data presented here demonstrate that CD107a is a sensitive marker of NK cell activity. D 2004 Elsevier B.V. All rights reserved.
Keywords: Natural Killer Cells; IFN-g; TNF-a; CD107a; Cytotoxicity; Degranulation

1. Introduction Natural killer (NK) cells are a subset of large granular lymphocytes defined by a lack of T-cell receptor (CD3) and by the surface expression of

* Corresponding author. Tel.: +1 617 724 2461; fax: +1 617 724 8586. E-mail address: maltfeld@partners.org (M. Altfeld). 0022-1759/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jim.2004.08.008

CD56 (Cooper et al., 2001). This subset of lymphocytes plays an integral role in the control of a number of viral infections and in tumor cell clearance (Yokoyama and Scalzo, 2002). NK cells are an important component of the innate immune response as they are able to lyse tumor cells or virally infected cells without prior antigen sensitization. NK cells are also a critical bridge between the innate and adaptive immune response as they secrete large amounts of cytokines and chemokines that can shape and drive

16

G. Alter et al. / Journal of Immunological Methods 294 (2004) 15–22

the ensuing adaptive immune response (Moretta et al., 2002). While NK cells are present predominantly in the peripheral circulation, they home to tissues following chemoattractant release by infected or malignant cells (Moretta et al., 2002). NK cells contain high concentrations of preformed cytolytic granules in their cytoplasm as they circulate in the periphery (Cooper et al., 2001). These lytic vesicles contain a number of cytolytic proteins such as perforin and granzyme uniquely designed to induce death in target cells upon release (Cooper et al., 2001; Burkhardt et al., 1989; Tschopp and Nabholz, 1990). Subsequent to activation, following the integration of complex signals from both activating and inhibitory receptors on the surface of these cells, NK cells rapidly release these granules at the immunological synapse inducing death of the target cell (Cooper et al., 2001; Moretta et al., 2002; Cerwenka and Lanier, 2001). Lining the membrane of these cytolytic granules is the lysosomal-associated membrane protein-1 (LAMP-1 or CD107a) (Winchester, 2001; Peters et al., 1991). This and other LAMP family member proteins are highly glycosylated membrane proteins that represent approximately 50% of the proteins in the lysosomal membrane (Fukuda, 1991; Kannan et al., 1996). Members of the LAMP family have short cytosolic tails, which are thought to interact with trans Golgi mediators that are involved in sorting and targeting proteins to the lysosomal pathway (Winchester, 2001). The highly glycosylated portion of the molecule on the luminal side of the vesicle has also been proposed to be involved in protecting the cellular membrane from attack by the lytic enzymes contained in the granules, and thus may subsequently protect the extracellular membrane of the effector cell following degranulation (Fukuda, 1991). Yet their precise function is still largely unclear. Recently, CD107a expression on the cell surface has been described as a marker of cytotoxic CD8+ T-cell degranulation and was shown to be strongly upregulated on the cell surface following stimulation in concordance with a loss of perforin (Betts et al., 2003). Given the strong cytotoxic capacity of NK cells, we chose to investigate the potential role of CD107a as a marker of NK cell activation and function. We demonstrate that CD107a is significantly upregulated on the surface of NK cells

following stimulation with MHC devoid target cells and following phorbol-12-myristate-13-acetate/ionomycin stimulation. This marker is expressed within 2 h of stimulation, and correlates strongly with both cytokine secretion and NK cell-mediated lysis of target cells.

2. Materials and methods 2.1. Study subjects Peripheral blood mononuclear cells were isolated from whole blood by Ficoll-Hypaque density gradient centrifugation from 12 healthy volunteers. Each subject gave informed consent for participation in the study. 2.2. Cell lines The K562 cell line, a highly undifferentiated human erythroleukemic cell line which does not express MHC class 1 molecules, was provided by the American Type Culture Collection (ATCC; Manassas, VA) and maintained in RPMI 1640 (Sigma, St. Louis, MO) containing 10% FBS (Atlanta Biologicals, Norcross, GA), 2 mM l-glutamine (Cellgro, Herndon, VA), 50 IU/ml penicillin (Cellgro) at 0.25Â106 cells/ml. 2.3. Multi-parameter flow cytometry The frequency of degranulating and cytokine secreting NK cells was quantitated by multi-parameter intracellular cytokine staining. PBMCs were resuspended at 106 cells/ml in RPMI 1640 (Sigma) containing 10% FBS (Atlanta Biologicals), 2 mM lglutamine (Cellgro), 50 IU/ml penicillin (Cellgro). Cells were then stimulated with MHC devoid, K562 cells (ATCC), at an effector to target ratio of 10:1, medium alone served as the negative control. Cells were stimulated with phorbol-12-myristate-13-acetate (PMA) (2.5 Ag/ml) and ionomycin (0.5 Ag/ml) (Sigma) as a positive control. CD107a-PeCy5 antibody (BD Biosciences, San Jose, CA) was added directly to the tubes at 20 Al/ml. Cells were incubated for 1 h at 37 8C in 5% CO2 after which brefeldin A (Sigma) was added at a final concen-

G. Alter et al. / Journal of Immunological Methods 294 (2004) 15–22

17

tration of 10 Ag/mL as well as 6 Al of monensin (Golgi-Stop, BD Biosciences) at a final concentration of 6 Ag/mL and incubated for an additional 5 h at 37 8C in 5% CO2. While brefeldin A prevents the exocytosis of cytokine containing vesicles allowing for the visualization of cytokine production following stimulation, monensin prevents the acidification of endocytic vesicles avoiding the degradation of reinternalized CD107a proteins from the surface and allowing for the visualization of this marker following stimulation. PBMCs were stained for surface NK cell markers CD56-PECy7 (BD Biosciences) and CD3-PECy5.5 (Caltag, Burlingame, CA) for 30 min. Samples were then fixed and permeabilized according to manufacturer’s directions (Caltag), and stained for intracellular interferon-g (IFN-g)-APC and tumor necrosis factor-a (TNF-a)PE (BD Biosciences) for an additional 30 min. After washing, cells were resuspended in 1% paraformaldehyde (Sigma) until five-color flow cytometric analysis was performed on a LSRII instrument (BD Biosciences). A total of 50,000 to 250,000 events were acquired and analyzed using FlowJo software. The analysis was performed on gated cells that fell within the lymphocyte population. We then gated on the CD3À/CD56+ population. Within this population of cells, we noted the independent expression of CD107a, IFN-g, or TNFa for each sample following stimulation. A response was considered positive if the frequency of CD107a expressing or cytokine secreting CD3À/CD56+ cells following stimulation with MHC devoid targets was at least three-fold greater than in unstimulated controls. 2.4. Chromium release cytotoxicity assay The ability of NK cells to lyse MHC devoid target cells was examined using a standard chromium release cytotoxicity assay (Kiessling et al., 1975). Two million K562 cells were labeled with 50 ACi Na2(51Cr04) (1 Ci=37 GBq; New England Nuclear) for 1 h at 37 8C, 5% CO2. PBMCs were added as effectors at E/T ratios of 10:1 and 50:1. Supernatant was harvested after a 4-h incubation at 37 8C and 5% CO2. For E/T titration assays, effectors were plated at 1:1, 5:1, 10:1, 25:1, 50:1, and 100:1 for the 4-h incubation. For the time course experiment,

PBMCs and 51Cr labeled K562 cells were plated at 10:1 and 50:1 and were incubated for variable lengths of time. The percent lysis was calculated as [((sample countÀspontaneous release)/(maximal releaseÀspontaneous release))Â100]. 2.5. Statistical analysis Unpaired T-tests were employed to assess differences between the level of NK activity at different time points for each functional marker. Spearman correlation was used to examine the relation between CD107a expression and a number of NK cell functional responses following stimulation with MHC devoid targets. P-values of less then 0.05 were considered significant.

3. Results 3.1. CD107a is expressed at high levels on NK cells following stimulation Recently, CD107a has been shown to be a marker for cytotoxic CD8+ T-cell activity as its expression was associated with a loss of perforin following antigen stimulation (Betts et al., 2003). Although NK cells are also cytolytic effector cells, the role of this marker has not been addressed for this subset of lymphocytes. Here, we assessed whether CD107a was expressed on the surface of NK cells following stimulation. Freshly isolated PBMCs were incubated with MHC devoid target cells or PMA/ionomycin. Following a 6-h incubation in the presence of monensin and CD107a antibody, NK cells were stained for CD3 and CD56. Representative data from one subject is shown in Fig. 1A–C. Surface expression of CD107a was low in unstimulated NK cells (0.33%) (Fig. 1A). Following stimulation with MHC devoid target cells, CD107a expression on the surface of NK cells increased 52-fold, resulting in 17.4% of CD3ÀCD56+ NK cells expressing CD107a (Fig. 1B). Following maximal stimulation of NK cells with PMA/ionomycin, CD107a expression on the surface of NK cells was up-regulated 88-fold, reaching frequencies of 29.17% of NK cells (Fig. 1C). Overall results for CD107a expression in the 12 subjects tested are shown in Fig. 1D. There was an

18

G. Alter et al. / Journal of Immunological Methods 294 (2004) 15–22

Fig. 1. CD107a is expressed at high levels on the surface of NK cells following stimulation. Flow cytometry figures represent the percent positive CD3À/CD56+ cells that express CD107a following no stimulation (A), stimulation with K562 cells (B) or with PMA/ionomycin (C) for a single representative subject. The dot plot represents the percent CD3À/CD56+ NK cells that express CD107a following no stimulation (n), stimulation with K562 cells (.) or PMA/ionomycin (z) for all subjects tested in this study (D).

average 25-fold increase in CD107a expression compared to baseline activity following co-incubation of NK cells with MHC devoid K562 cells for all subjects tested ( pb0.001, Student’s T-test as compared to baseline CD107a expression). Following PMA/ionomycin stimulation, we observed an average 32-fold increase in CD107a expression, reaching an average frequency of 21.5% ( pb0.001, Student’s Ttest as compared to baseline CD107a expression) (Fig. 1D). Taken together, these data demonstrate that CD107a is significantly upregulated on the surface of NK cells following activation. Traditionally, one means by which NK cell functional activity can be quantitated is via intracellular cytokine secretion profiles following stimulation. Thus we next characterized the distribution of NK cells that expressed CD107a with respect to those that secreted IFN-g and TNF-a following stimulation with K562 cells. As few as 0.66% of IFN-g positive NK cells were not also expressing CD107a following stimulation, and the number of TNF-a positive NK cells not expressing CD107a was

negligible (0.08%). In contrast, significantly more NK cells expressed CD107a following stimulation than either IFN-g or TNF-a (Fig. 2A). Of the CD107a+CD3-CD56+ NK cells, an average of 66% expressed CD107a alone, 11% expressed both TNFa and CD107a, 13% expressed CD107a and IFN-g, and 10% of NK cells expressed all three functional markers (Fig. 2B). Taken together, these data demonstrate that CD107a represents a more comprehensive marker of NK activity as it is expressed on a larger proportion of functionally active cells that are not detected using simple assessment of cytokine secretion. 3.2. Kinetics of CD107a expression on NK cells following stimulation Following the demonstration that CD107a staining allows for the identification of a significantly larger fraction of activated NK cells than intracellular cytokine staining, we studied the kinetics by which this marker is expressed on activated NK

G. Alter et al. / Journal of Immunological Methods 294 (2004) 15–22

19

Fig. 2. Coordinate expression of C107a on cytokine secreting cells following stimulation with K562 cells. Flow cytometry data demonstrates the distribution of cytokine positive cells among the CD107a expressing CD3À/CD56+ NK cells for a single subject. The four flow cytometry figures represent the proportion of NK cells expressing CD107a along the x-axis and IFN-g (top two panels) or TNF-a (bottom two panels) following no stimulation (two panels on the left) or stimulation with K562 cells (two panels on the right) (A). The doughnut chart is representative of the average proportion of CD3À/CD56+/CD107a expressing cells that expressed neither cytokine n, both IFN-g and TNF-a 5, or either IFN-g, or TNF-a n for the whole study population (B).

cells. Thus, the kinetics of NK cell cytokine secretion, NK cell-mediated lysis of target cells and CD107a expression was examined in parallel for three subjects following co-incubation of PBMCs with K562 cells at an effector to target ratio of 10:1. Degranulation and cytokine secretion were performed on PBMCs that were incubated with MHC devoid targets for 15 min to 6 h in the presence of brefeldin A, monensin and CD107a antibody.

Following incubation, cells were stained for NK cell surface markers and intracellular cytokines. Small but significant amounts of intracellular TNFa were detectable on 0.49% of NK cells as early as 30 min following incubation, and increased during the subsequent hours following stimulation ( pb0.05; at time 3 and 6 h; Student’s T-test, Fig. 3A). Intracellular IFN-g production was not detectable at 30 min, but was significantly higher than baseline

Fig. 3. Kinetics of NK cell responses following stimulation with MHC devoid target cells. Bar graphs represent, in three subjects, the level and standard deviation of CD107a, TNF-a, or IFN-g expression for all CD3À/CD56+ NK cells following a 15-min to 6-h incubation with K562 cells at an effector/target ratio of 10:1 (A). In parallel for the same three subjects, we examined the kinetics of K562 target lysis over a 15-min to 4-h period as depicted in the bar graphs with standard deviations (B).

20

G. Alter et al. / Journal of Immunological Methods 294 (2004) 15–22

Fig. 4. Determining the optimal effector to target ratio for CD107a expression, cytokine secretion and target cell lysis. Bar graphs represent the average and standard deviation for the expression of CD107a, TNF-a, or IFN-g following a 6-h incubation at multiple effector/target ratios tested in three subjects(A). In parallel, the average and standard deviation for target cell lysis were measured at identical effector to target ratios following a 4-h co-incubation of PBMCs with K562 target cells (B).

following 1 h of incubation of NK cells and target cells ( p=0.04, Student’s T-test), and increased steadily up to 6 h following stimulation ( pb0.05; at times 1, 3, and 6 h; Student’s T-test, Fig. 3A). CD107a expression was detectable as early as 2 h following incubation (Fig. 3A). Expression of this marker increased over the following 4 h reaching an average of 17.8% after 6 h (Fig. 3A). In line with the kinetics of CD107a expression, NK cell lysis of radioactive-labeled K562 cells was detectable as early as 2 h following co-incubation and subsequently increased, reaching an average of 22% target cell lysis, following co-incubation of targets and effectors at both a 1:10 and 1:50 E/T ratio (Fig. 3B). Overall, while cytokine secretion preceded degranulation, CD107a expression represents a sensitive marker of NK cell activity detectable as early as 2 h following stimulation. In addition, the kinetics of CD107a expression on NK cells corresponded closely to the kinetics of target cell lysis.

3.3. CD107a staining can be performed simultaneously with intracellular cytokine staining by multiparameter flow cytometry at an effector to target ratio of 10:1 To determine the optimal effector to target ratio for inducing maximal NK cell activation, PBMCs were stimulated with different concentrations of K562 target cells. IFN-g, TNF-a, and CD107a were detectable at all E/T ratios tested (Fig. 4A). Maximal cytokine secretion was detected at an E/T ratio of 10:1 (Fig. 4A). Although, CD107a induction was detected at a 5:1 ratio (Fig. 4A), induction was not significantly different at a 10:1 ratio ( p=0.83, Student’s T-test). Thus, it is possible to quantify cytokine secretion and CD107a expression simultaneously on the same cell using multi-parameter flow cytometry at an optimal effector-to-target ratio of 10:1. In contrast, lysis of target cells in an NK cytotoxicity assay peaked at an E/T ratio of 50:1

Fig. 5. CD107a is a marker of various NK cell functional readouts. The scatter diagrams represent the relationship between the expression of CD107a and IFN-g secretion (A), TNF-a secretion (B), or target cell lysis (C) for the each individual tested.

G. Alter et al. / Journal of Immunological Methods 294 (2004) 15–22

21

(Fig. 4B). Taken together, these data demonstrate that both CD107a and cytokine secretion can be done in parallel to assess the level of NK activity at the single cell level. 3.4. CD107a is a general marker of NK cell activity To determine if CD107a expression could be employed as a general marker of NK activity, we assessed the relationship between the induction of this marker and both cytokine secretion and NK cellmediated lysis for each individual. CD107a expression was significantly associated with IFN-g (r=0.59, p=0.02; Spearman correlation) (Fig. 5A) and TNF-a cytokine production (r=0.71, p=0.04; Spearman correlation) (Fig. 5B). Additionally, CD107a staining was also related to the level of NK cell-mediated lysis of chromium labeled K562 cells (r=0.83, p=0.005; Spearman correlation) (Fig. 5C). Overall, the significant correlation between CD107a and a number of different readouts for NK functional activity demonstrate that this marker is a useful tool to assess the level of NK cell activity following stimulation with MHC devoid target cells.

4. Discussion NK cells are an important component of the innate immune response as well as a bridge to the adaptive arm of the immune response (Cooper et al., 2001; Farag et al., 2003; French and Yokoyama, 2003). An important limitation in the field of NK research is attributable to the deficit of assays available for the detection the functional activity of NK cells. It is to this end that we sought to develop a new method to quantitate NK cell functional activity. Given the recent data demonstrating the role of CD107a as a marker of CD8+ T-cell degranulation (Betts et al., 2003), we set out to illustrate a similar role for this marker in NK cells. In this study, we demonstrate that CD107a is upregulated on NK cells following stimulation. CD107a induction is expressed in concert with cytokine secretion and target cell lysis. Moreover, multi-parameter flow cytometry can be performed to detect simultaneous degranulation and cytokine secreting NK cells on a single cell level. CD107a

expression, following stimulation with MHC devoid targets, correlated significantly with cytokine secretion. Furthermore, similar to the correlation that was observed between the expression of this marker on CD8+ T cells and T cell-mediated lysis of target cells (Betts et al., 2003; Rubio et al., 2003), the induction of CD107a on the surface of NK cells was closely related to the level of target cell lysis by NK cells. While assays such as the chromium release assay provide information about the end stage lysis of target cells, CD107a provides data on the level of activation of the effector population. Yet, as a strong relationship exists between this marker and target cell lysis by NK cells, it is still possible to make inferences about the level of potential target cell lysis given the expression of CD107a following stimulation. Furthermore, the use of CD107a as readout of NK activity permits the discrimination of multiple populations of NK cells based on their ability to respond to different stimuli. Given the expression of this marker on the surface of both cytokine secreting and non-secreting cells, it will be possible to sort out the role of these two subsets of NK effector populations and explore their role in diverse models. Thus, this marker will provide us with a means to investigate the diversity of effector NK cell populations that may be differentially affected in infections or malignant states. In conclusion, although the biological function of CD107a remains unclear, it is evident that this marker is a more sensitive marker of NK activity than the traditionally used intracellular cytokine assay or the chromium release assay. Given that CD107a correlates with both cytokine secretion and target cell lysis, this study supports the use of CD107a as a general marker of NK functional activity.

References
Betts, M.R., Brenchley, J.M., Price, D.A., De Rosa, S.C., Douek, D.C., Roederer, M., Koup, R.A., 2003. Sensitive and viable identification of antigen-specific CD8+ T cells by a flow cytometric assay for degranulation. J. Immunol. Methods, 65. Burkhardt, J.K., Hester, S., Argon, Y., 1989. Two proteins targeted to the same lytic granule compartment undergo very different posttranslational processing. Proc. Natl. Acad. Sci. U. S. A., 7128.

22

G. Alter et al. / Journal of Immunological Methods 294 (2004) 15–22 leukemia cells. Specificity and distribution according to genotype. Eur. J. Immunol., 112. Moretta, A., Bottino, C., Mingari, M.C., Biassoni, R., Moretta, L., 2002. What is a natural killer cell? Nat. Immunol., 6. Peters, P.J., Borst, J., Oorschot, V., Fukuda, M., Krahenbuhl, O., Tschopp, J., Slot, J.W., Geuze, H.J., 1991. Cytotoxic T lymphocyte granules are secretory lysosomes, containing both perforin and granzymes. J. Exp. Med., 1099. Rubio, V., Stuge, T.B., Singh, N., Betts, M.R., Weber, J.S., Roederer, M., Lee, P.P., 2003. Ex vivo identification, isolation and analysis of tumor-cytolytic T cells. Nat. Med., 1377. Tschopp, J., Nabholz, M., 1990. Perforin-mediated target cell lysis by cytolytic T lymphocytes. Annu. Rev. Immunol., 279. Winchester, B.G., 2001. Lysosomal membrane proteins. Eur. J. Paediatr. Neurol., 11. Yokoyama, W.M., Scalzo, A.A., 2002. Natural killer cell activation receptors in innate immunity to infection. Microbes Infect., 1513.

Cerwenka, A., Lanier, L.L., 2001. Ligands for natural killer cell receptors: redundancy or specificity. Immunol. Rev., 158. Cooper, M.A., Fehniger, T.A., Caligiuri, M.A., 2001. The biology of human natural killer-cell subsets. Trends Immunol., 633. Farag, S.S., VanDeusen, J.B., Fehniger, T.A., Caligiuri, M.A., 2003. Biology and clinical impact of human natural killer cells. Int. J. Hematol., 7. French, A.R., Yokoyama, W.M., 2003. Natural killer cells and viral infections. Curr. Opin. Immunol., 45. Fukuda, M., 1991. Lysosomal membrane glycoproteins. Structure, biosynthesis, and intracellular trafficking. J. Biol. Chem., 21327. Kannan, K., Stewart, R.M., Bounds, W., Carlsson, S.R., Fukuda, M., Betzing, K.W., Holcombe, R.F., 1996. Lysosome-associated membrane proteins h-LAMP1 (CD107a) and h-LAMP2 (CD107b) are activation-dependent cell surface glycoproteins in human peripheral blood mononuclear cells which mediate cell adhesion to vascular endothelium. Cell. Immunol., 10. Kiessling, R., Klein, E., Wigzell, H., 1975. Natural killer cells in the mouse. I: cytotoxic cells with specificity for mouse Moloney