Serine Protease Inhibitor 6 Protects iNKT Cells from Self-Inflicted Damage A. Wahid Ansari, Jeff N. Temblay, Syarifah H. Alyahya and Philip G. Ashton-Rickardt This information is current as of September 28, 2021. J Immunol 2010; 185:877-883; Prepublished online 11 June 2010; doi: 10.4049/jimmunol.1000651 http://www.jimmunol.org/content/185/2/877 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2010/06/09/jimmunol.100065 Material 1.DC1

References This article cites 36 articles, 15 of which you can access for free at: http://www.jimmunol.org/ http://www.jimmunol.org/content/185/2/877.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists by guest on September 28, 2021

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2010 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Serine Protease Inhibitor 6 Protects iNKT Cells from Self-Inflicted Damage

A. Wahid Ansari, Jeff N. Temblay, Syarifah H. Alyahya, and Philip G. Ashton-Rickardt

The role played by apoptosis in the homeostasis of effector cells of the innate is unclear. Serine protease inhibitor 6 (Spi6) is an inhibitor of granzyme B (GrB) that protects cytotoxic T lymphocytes of the from apoptosis. To determine whether Spi6 also protects cells of the from self-inflicted damage we have examined invariant NKT (iNKT) cells. Spi6-deficient iNKT cells harbored increased levels of GrB after TCR stimulation with the PBS-57 glycolipid Ag and were susceptible to apoptosis. The increased apoptosis of Spi6 knock-out (KO) iNKT cells lead to a complete loss in the production of IL-4 and IFN-g by Spi6 KO iNKT cells after PBS-57 challenge. The increased activation-induced apoptosis resulted in impaired survival and a decreased clonal burst size of Spi6 KO iNKT cells, which could be corrected by GrB deficiency. However, the clonal

burst of Spi6 KO iNKT cells after TCR-independent activation with lymphocytic choriomeningitis virus was not affected. Our Downloaded from findings demonstrate that Spi6 protects cytotoxic cells of the innate immune system from GrB-mediated self-inflicted triggered by the recognition of Ag. The Journal of Immunology, 2010, 185: 877–883.

nvariant NKT (iNKT) cells are a subset of T lymphocytes that In addition to an immunomodulatory role, iNKT cells also kill express an invariant TCR (Va 14) and the NK cell surface cancer cells through contact-mediated cytotoxicity (2, 7, 16). In I marker NK1.1 (1). Recognition of glycolipid Ags presented this process, the exocytosis of perforin (17, 18) facilitates the entry http://www.jimmunol.org/ in context of the CD1d MHC molecule leads to the activation of of serine proteases called granzymes from cytotoxic granules of iNKT cells and the rapid production of both IFN-g (Th1) and lymphocytes, such as iNKT cells, into the cytoplasm of target cells IL-4 (Th2) , which govern a wide variety of immune where they trigger apoptosis (16). Granzyme B (GrB) is required for reactions (2–7). Activation of iNKT cells occurs early in the im- the induction of target cell death (19) through the activation of mune response and so provides a critical link between the innate caspase 3 and the induction of nuclear DNA fragmentation (16). and adaptive immune response (2–7). The most commonly used Given the effectiveness of the exocytosis CTL pathway, mecha- ligand to study iNKT cells is a-galactoceramide, a glycolipid (5, nisms for the protection from self-inflicted damage have been pro- 8). PBS-57 is a functional analog of a-galactoceramide that shares posed for some time (20). the same specificity and stimulatory capacity for iNKT cells (9). Serine protease inhibitor 6 (Spi6) (SERPINB9) is an intracellular by guest on September 28, 2021 The invivo stimulation of iNKT cells displays the peculiar feature serine protease inhibitor of the OVA family (OVA-serpin) (21) that of disappearance within 3–12 h and reappearance 24–48 h later (10). is specific for GrB and expressed in the cytoplasm of mouse CTL Expansion of up to 10-fold typically results in a maximal iNKT cell and NK cells (22, 23). Using Spi6 knock-out (KO) mice, we have number on day 3 (clonal burst), which is followed by a contraction in shown that Spi6 protects CTLs from apoptosis by inactivating GrB cell numbers for 6–9 d to the same level as before stimulation. that leaks from cytotoxic granules into cytoplasm (24). Spi6 acts Studies from different laboratories have shown that transient down- a classical OVA-serpin substrate for GrB resulting in the denatur- regulation of TCR and NK1.1 is the major cause of iNKT disap- ation and destruction of the apoptotic serine protease (24). pearance (11–13). However, there are also reports suggesting that Although the cytolytic lymphocytes of the innate (iNKT cells) the contraction in iNKT cell number after stimulation is caused by and adaptive (CTL) immune systems both use the exocytosis path- apoptosis (10) mediated by Bim (14) and classical Fas–FasL- way for cell killing, there are significant differences in their im- mediated activated-induced cell death (15). munobiology. The most obvious being the rapidity of effector re- sponses after first Ag encounter and the relatively short life span of iNKT cells compared with CTL (2, 3). Therefore, it is not clear whether protection from GrB by intracellular OVA-serpins is like- Section of Immunobiology, Division of Immunology and Inflammation, Department ly to be a general homeostatic mechanism shared between CTL of Medicine, Imperial College London, London, United Kingdom of the adaptive immune system and iNKT cells of the innate im- Received for publication February 24, 2010. Accepted for publication May 5, 2010. mune system. This work was supported by National Institutes of Health Grant AI45108 (to P.G.A-R). Address correspondence and reprint requests to Prof. P.G. Ashton-Rickardt, Section Materials and Methods of Immunobiology, Department of Medicine, Imperial College London, Hammer- Mice smith Campus, Du Cane Road, London, W12 0NN, U.K. E-mail address: p.ashton- [email protected] C57 BL/6 Spi6 KO (24), C57BL/6 GrB KO (24), and control wild-type The online version of this article contains supplemental material. (Wt) (purchased from Charles River Laboratories, Margate, U.K.) were bred and maintained under the appropriate conditions in the Imperial Abbreviations used in this paper: BM, bone marrow; CIM, confocal immunofluores- College animal facility. Mice were injected i.p. with 2 mgofPBS-57 cence microscopy; GrB, granzyme B; ICS, intracellular staining; iNKT, invariant NKT; KO, knock-out; LCMV, lymphocytic choriomeningitis virus; liv, liver; MFI, mean and control mice were treated with PBS. Mice were infected with lym- 3 5 fluorescence intensity; PI9, proteinase inhibitor 9; Spi6, serine protease inhibitor 6; phocytic choriomeningitis virus (LCMV) Armstrong (2 10 PFU/- spl, spleen; thy, thymus; Wt, wild-type. mouse i.p.)(24). All studies were approved by the Home Office (London, U.K.). PBS-57 (9) was kindly provided by Paul Savage (Brigham Copyright Ó 2010 by The American Association of Immunologists, Inc. 0022-1767/10/$16.00 Young University, Provo, UT). www.jimmunol.org/cgi/doi/10.4049/jimmunol.1000651 878 Spi6 AND iNKT CELLS

Cell preparation ELISA Splenocytes were prepared as described previously (24). Perfused liver Serum was collected by tail vein bleeding, 24 h prior to (prebleed), and 2 h was minced, pressed through stainless steal mesh, and suspended in RPMI postinjection with PBS-57 and stored at 280˚C for further analysis. Col- 10 (Invitrogen, Paisley, U.K.) supplemented with 5% FBS (Invitrogen). lected samples were then heated to 37˚C for 10 min and centrifuged at After washing twice with medium, the pellet was resuspended in 5 ml 5000 3 g for 5 min, and the resulting supernatant extracted for use in a mul- medium and passed through a 70-m cell strainer. Cells were washed and tiplex sandwich ELISA detection for mouse IL-4 and IFN-g (Luminex, resuspended in 40% Percoll (GE Healthcare, Chalfont St. Giles, U.K.) and Austin, TX/Invitrogen). This was performed under the manufacturer’s overlaid on 70% Percoll solution, then centrifuged at 2000 rpm for 20 min instructions. Serum concentrations were calculated from standard curves at room temperature. Interface layer containing mononuclear cells were with both cytokines. The level of IL-4 in serum from prebleeds was removed carefully and washed. Cell pellets were treated with RBC lysis 200 pg/ml (Wt) and ,100 pg/ml (Spi6 KO), and for IFN-g 200 pg/ml buffer (eBiosciences, San Diego, CA), and washed twice with and re- (Wt) and ,50 pg/ml (Spi6 KO). suspended in medium. Confocal immunofluorescence microscopy Flow cytometry Spleens were harvested on day 3 after Ag challenge, stained with CD1d tet ε . PBS-57 loaded/CD1d tetramer allophycocyanin was provided by the Na- and CD3 and purified on a Dako Moflo FACS ( 90% pure). iNKT cells tional Institutes of Health Tetramer Facility (Atlanta, GA). The following were subjected to ICS for GrB as cytospins and examined with a Leica Abs were used: anti-mouse FITC or PE TCRb and PE-NK1.1 from (BD TCS SP5 confocal microscope. To localize nucleus cells were mounted Pharmingen, San Diego, CA), and PE-Cy7-CD3ε, Pacific Blue-CD24, allo- with DAPI-containing medium (Vectashield, Vector Laboratories, Burlin- phycocyanin-CD44, allophycocyanin-NK1.1, PE Cy7, and Pacific Blue- game, CA). CD69, FITC-CD1d, FITC-CD122 (eBiosciences), anti- PE-GrB (Cal- Statistical analysis tag Laboratories, Burlingame, CA), brefeldin A, PE-active caspase 3, Annexin Downloaded from V , and BrdU detection kit (BD Pharmingen). Cells were stained with Data are expressed as SEM. Statistical significance was determined by CD1d tet and TCR-b, followed by intracellular staining (ICS) with rabbit Student t test. All analyses were performed using GraphPad Prism 5 soft- anti-Spi6 antiserum (1/1000 dilution) as described previously (24). For GrB ware (GraphPad, San Diego, CA). staining, cells were cultured for 2 h in complete RPMI 1640 medium in presence of brefeldin A (BD Pharmingen). For ICS of Spi6, GrB, and active caspase 3 we used BD FACS/Perm reagents. Cells were analyzed on a Cyan Results ADP (DakoCytomation, Carpinteria, CA) flow cytometer and analyzed us- Normal development of iNKT cells in Spi6 KO mice ing FlowJo software (TreeStar, Ashland, OR). For annexin V staining, cells Spi6 is highly expressed by cytotoxic lymphocytes, such as CTL and http://www.jimmunol.org/ were surface stained with CD1d tet and TCR-b, washed, and labeled with either FITC-conjugated (BD Pharmingen) or Pacific Blue-conjugated (Invi- NK cells, that kill using GrB (22, 25). To determine whether GrB- trogen) annexin V, according to manufacturer’s protocol (BD Pharmingen containing iNKT cells also express Spi6, we performed ICS on Wt Annexin V kit). For activated iNKT cells, mice were injected i.p. with BrdU and Spi6 KO cells with our anti-Spi6 antiserum (25). We found in PBS (1 mg, BD Biosciences), then fed BrdU (0.8 mg/ml) drinking water, iNKT cells constitutively express Spi6 (Fig. 1A). The mean fluo- supplemented with 5% (weight/volume ) continuously for 3 d. Mice ∼ were fed with BrdU in drinking water for 4 d to study homeostatic pro- rescence intensity (MFI) of anti-Spi6 staining was 14-fold higher liferation. Cells were surface stained and BrdU staining was performed in Wt compared with Spi6 KO iNKT cells (Fig. 1B). To study the according to manufacturer’s protocol (BD Pharmingen BrdU Flow Kit). role of Spi6 in the development, maturation, and organ distribution by guest on September 28, 2021

FIGURE 1. Spi6 KO iNKT cell development and homeostatic proliferation. Naive Wt and Spi6 KO thy, spl, and liv mononuclear cells were surface stained with PBS-57 loaded/CD1d tetramer and TCRb. A, Spleen cells were subjected to ICS with rabbit anti-Spi6 antiserum and preimmune serum. The MFI (6 SEM) of anti-Spi6 staining is indicated. B, The MFI of anti-Spi6 staining was subtracted for preimmune serum staining and was ∼14 times higher for Wt than for Spi6 KO iNKT cells. n = 5 mice. C, Dot plot showing percentage of CD1d tet+ TCRb+ iNKT cells (indicated top left corner) in the thy, spl, and liv. D, Mean percentage (6 SEM) of iNKT cells. E, iNKT cell homeostatic proliferation in BM, thy, spl, and liv. n = 5 mice. Dot plots showing BrdU incorporation during S-phase after 7-AAD staining. F, Mean percentage (6 SEM) of BrdU+ iNKT cells. n = 5 mice. Data are representative of at least two to three independent experiments. pppp , 0.0001. liv, liver; spl, spleen; thy, thymus. The Journal of Immunology 879 of iNKT cells, a comprehensive phenotypic analysis was performed 57 in both the liver (p = 0.03) and the spleen (p = 0.04) of Spi6 in Spi6 KO mice. We found that the thymic development and mat- KO mice, we observed a significant increase in the percentage of uration of iNKT cells, as shown by the expression of heat stable Ag annexin V+ iNKT cells compared with Wt (Fig. 2C,2D). Staining (CD24), CD44, and NK1.1, was unaffected in Spi6 KO mice (Sup- for the CD69 activation marker revealed the activation status of Spi6 plemental Fig. 1). Moreover, the levels of iNKT cells (CD1d+ KO iNKT cells was no different to Wt (Fig. 2E). We conclude that TCRb+) in the in thymus, spleen, and liver were also not affected Spi6 KO iNKT cells undergo increased apoptosis after Ag stimula- in Spi6 KO mice (Fig. 1C,1D). iNKTs display an activated pheno- tion because of increased levels of GrB. type as indicated by the expression of activation makers, such as To determine the functional consequences of the increased apo- CD122, and undergo cell division in the absence of overt antigenic ptosis of Spi6 KO iNKT cells, we measured the production of the stimulation (2, 5). We found the activation status of iNKT cells, as cytokines IFN-g and IL-4 after challenging Spi6 KO mice with indicated by CD122 expression, was unaffected in Spi6 KO mice PBS-57 (2–7). The levels of both IFN-g and IL-4 were dramatically (Supplemental Fig.1D). Furthermore, the proliferation of Spi6 KO reduced to barely detectable levels (.100-fold decrease) in the se- iNKT cells, as measured by BrdU incorporation into the DNA of S- rum of Spi6 KO mice compared with Wt controls 2 h after PBS-57 phase cells, was also unaffected (Fig. 1E,1F). We conclude that Spi6 challenge (Fig. 2F). In fact, the levels of IL-4 and IFN-g in Spi6 deficiency has no major effect on the development, distribution, and KO mice were no higher than the levels observed in prebleeds homeostasis of iNKT cells in the absence of overt Ag stimulation. before PBS-57 challenge. Therefore, there was essentially a com- plete loss in the production of IL-4 and IFN-g by iNKT cells in Spi6 GrB mediates activation-induced apoptosis of Spi6 KO KO mice after Ag challenge. We conclude that protection from iNKT cells GrB-mediated death by Spi6 is required for immunoregulatory func- Downloaded from We have shown previously that Spi6 protects CTL from apoptosis tion of iNKT cells. by binding to and destroying GrB in the cytoplasm after it leaks from cytotoxic granules (24). To determine whether Spi6 similarly Clonal burst of Spi6 KO iNKT cells is impaired after glycolipid protects iNKT cells, we measured the level of GrB in Ag- Ag activation stimulated Spi6 KO iNKT cells. Wt and Spi6 KO mice were chal- We next examined the effect of the increase in self-inflicted damage

lenged with PBS-57 and the level of intracellular GrB measured in on the response of iNKT cells to Ag stimulation in vivo. Challenge http://www.jimmunol.org/ iNKT cells after 2 h by ICS with anti-GrB Ab. Flow cytometry of Wt mice with PBS-57 resulted in the rapid expansion of iNKT revealed that after activation, all Wt iNKT cells stained positive cells with a peak of response (clonal burst) on day 3, followed by for GrB with ∼20% expressing high levels of GrB (GrBhi) (Fig. a steep decline (contraction) in the frequency of iNKT cells before 2A). There were ∼2-fold more GrBhi iNKT cells from Spi6 KO returning tothe original level on day6 inboththe spleen (Fig.3A) and compared with Wt (p = 0.02) (Fig. 2B). Therefore, Ag-stimulated the liver (Fig. 3B) (11–13). In both the spleen and the liver, we Spi6 KO iNKT cells harbor increased levels of GrB presumably be- observed a .2-fold decrease in the frequency and absolute number cause Spi6 acts to inhibit and destroy misdirected GrB in the cyto- of iNKT cells in Spi6 KO mice at the clonal burst on day 3 (Fig. 3A, plasm (24). 3B). Therefore, Spi6 KO mice have an impaired clonal burst of

Stimulation in vivo of the TCR with glycolipd Ag or anti-CD3ε iNKT cells. by guest on September 28, 2021 Ab induces the apoptosis of iNKT cells (10, 11, 26). We detected To determine the cause of the impaired clonal burst, we exam- apoptosis of iNKT cells by measuring the expression of phos- ined GrB-mediated apoptosis of Spi6 KO iNKT cells on day 3 after photidyl serine by annexin V binding. After challenge with PBS- PBS-57 challenge. As we observed for 2 h after activation

FIGURE 2. Glycolipid-Ag–induced cell death of Spi6 KO iNKT cells. Wt and Spi6 KO mice were injected i.p. with 2 mg PBS-57 or vehicle (PBS, naive) then analyzed after 2 h. Splenocytes were stained with CD1d tet and TCR-b, followed by ICS with anti-GrB Ab. A, Percentage of GrBhi population in Wt and Spi6 KO iNKTeither unstimulated (naive) or stimulated (ac- tivated) with PBS-57 cells is indicated in top right cor- ner. B, Mean percentage (6 SEM) of GrBhi iNKT cells. n = 3–5 mice. p = 0.02. C, Percentage of annexin V+ iNKT cells (gated on CD1d tet+ TCRb+) in the spl and liv from naive or PBS-57 challenged (activated) Wt (right hand number) and Spi6 KO (left hand number) mice. D, Mean percentage (6 SEM) of annexin V + iNKT cells in spl and liv of naive and activated mice. n = 3–4 mice. E, Flow cytometric analysis of CD69 expression on naive and activated iNKT cells (CD1d tet+ gated) from the spl and liv. Percentage (6 SEM) of Wt (right hand number) and Spi6 KO (left hand number) iNKT cells that were CD69+ is indicated. F, Mean concentration (6 SEM) of cytokines in the serum of mice 2 h after challenge with PBS-57. n = 5 mice. Data are representative of at least two to three indepen- dent experiments. pp , 0.05; pppp , 0.0005. liv, liver; spl, spleen. 880 Spi6 AND iNKT CELLS

FIGURE 3. Clonal burst of Spi6 KO iNKT after glycolipid Ag challenge in vivo. Wt and Spi6 KO mice were injected i.p. with PBS-57 (2 mg) and iNKT cells harvested on day 0, day 3, and day 6. n = 3–5 mice, n = 4–5 mice, and n = 2–4 mice, respectively. Per- centage of CD1d tet+ iNKT cells in spl (A) and liv (B) is indicated by the number in the FACS plots. Mean percentage and absolute number of iNKT cells are shown in A and B. C, Mean percentage of GrBhi and (D) active caspase 3+iNKT cells (CD1d tet+) from day 3 after PBS-57challenge. E, Ag-challenged iNKT cell proliferation on day 3. n = 3–5

mice. FACS plot shows the S-phase DNA as Downloaded from shown by BrdU labeling. Mean percentage of BrdU incorporation in unstimulated and PBS-57 stimulated BM, thy, spl, and liv iNKT cells of Wt and Spi6 KO mice. pp , 0.05; ppp , 0.001. liv, liver; spl, spleen; thy, thymus. http://www.jimmunol.org/

(Fig. 2B), there was about a 2-fold increase in the percentage of level of the CD1d molecule, which presents Ag to iNKT, was the GrBhi Spi6 KO iNKT cells during the clonal burst on day 3 after same as Wt on Spi6 KO bone marrow (BM) cells, thymocytes, and Ag challenge (Fig. 3C). We next examined the activity of caspase splenocytes (Supplemental Fig. 1D). Therefore, impaired Ag- 3, which is a key mediator of apoptosis activated by GrB (16), presentation is unlikely to be responsible for the diminished ex- by guest on September 28, 2021 using fluorogenic substrates and flow cytometry. The activity of pansion of Spi6 KO iNKT cells. We conclude that an increase in caspase 3 was ∼2-fold higher in iNKT cells from Spi6 KO mice GrB-mediated apoptosis was the cause of the reduced clonal burst compared with Wt mice (Fig. 3D). Therefore, increased apoptosis of Ag-stimulated Spi6 KO iNKT cells. is a likely cause of the decreased clonal burst size of Spi6 KO iNKT cells. Clonal burst of Spi6 KO iNKTs is not impaired after TCR- To rule out the possibility of defective proliferation as a reason independent activation for impaired iNKT cell expansion, we performed BrdU labeling to Tofurther investigate the stimulation requirements for GrB-mediated measure proliferation on day 3 after challenge with PBS-57. The loss of iNKTs, we infected Spi6 KO mice with LCMV to examine proportion of proliferating iNKT cells in Spi6 KO mice was no dif- bystander activation (27–30). As others have reported (28, 29), we ferent to Wt mice (Fig. 3E). Consistent with its specificity for observed a decrease in the percentage of iNKTs cells in the spleens of the TCR of iNKTs (9), PBS-57 did not drive the expansion of Wt mice after acute infection with LCMV (Fig. 4A,4B). However, either CD8 T cells or NK cells, which in turn were not affected in when we examined absolute cell number in multiple mice, we ob- challenged Spi6 KO mice (Supplemental Fig. 2) . In addition, the served an increase on day 8 of not only LCMV-specific CD8 T cells

FIGURE 4. Response of Spi6 KO iNKT cells to acute LCMVinfection in vivo. Wt and Spi6 KO mice were infected with LCMV. Splenocytes were har- vested at indicated time and stained for iNKT and CD8 T cells using specific Abs. A,Percentageof iNKT and CD8 population is shown in FACS plot on days 0, 3, and 8. B, Mean percentage and absolute number of iNKTand CD8 T cells are shown. Data shown as mean percentage (6 SEM) and ab- solute number. n = 5 mice. ppp , 0.05; pppp , 0.001. The Journal of Immunology 881

FIGURE 5. Cytotoxic granules of Spi6 KO iNKT cells. CIM images (original magnification 340) of FACS purified iNKT cells from day 3 of PBS-57 challenge. Left panels show nuclear dye DAPI (blue), middle panels show GrB staining (red), and right panels show merged blue and red images. Arrows indicate cytoplasmic GrB staining. Data are representative of at least two experi- ments. Downloaded from but also of bystander-activated iNKTs (Fig. 4D,4E). We think the apoptosis. To do this, we examined the effect of GrB deficiency on decrease in the percentage of splenic iNKTs is likely due to dilution iNKT cells in Spi6 KO3GrB cluster deficient (GrB KO) mice (19, by LCMV-specific CD8 T cells. As we reported previously (24, 25), 24). In Spi6 KO3GrB KO mice, the size of the clonal burst of we observed a severe reduction in the size of the CD8+ T cell clonal iNKT cells after PBS-57 challenge was rescued to the Wt level burst in Spi6 KO mice on day 8 after LCMV infection (Fig. 4C,4E). (Fig. 6A,6B). The proportion of iNKT cells undergoing apoptosis However, therewas no difference in the number ofiNKT cells in Spi6 in Spi6 KO3GrB KO mice, as evidenced by the expression of http://www.jimmunol.org/ KO mice after LCMV infection (Fig. 4D). Thus, TCR-independent activated caspase 3 was correspondingly reduced (Fig. 6C). In ad- stimulation of NKT cells did not result in the reduced survival of Spi6 dition, the levels of iNKT cells (Supplemental Fig. 3) were also KO iNKTs. corrected in single GrB KO mice compared with Spi6 KO mice. Thus, the absence of GrB corrected the deficit in iNKT cell survival Leakage of GrB from cytotoxic granules of Spi6 KO iNKT cells caused by Spi6 deficiency and returned the level of iNKT cells to Wt Spi6 not only protects CTL from GrB-mediated apoptosis but levels. Therefore, we conclude that protection from GrB is at least in also ensures the integrity of cytotoxic granules (24). To determine part a physiological mechanism by which Spi6 ensures the survival whether Spi6 also protects iNKT cells in this way, iNKT cells of iNKT cells after Ag-stimulation. were purified from Spi6 KO mice by FACS after PBS-57 chal- by guest on September 28, 2021 lenge. Spi6 KO iNKT cells were then stained for GrB and exam- Discussion ined by confocal immunofluorescence microscopy (CIM). We iNKT cells share many effector properties with peptide/MHC- observed the characteristic punctate pattern of Ab staining indic- specific classical T cells. However, compared with both CD8+ and ative of distribution of GrB to cytotoxic granules in Wt iNKT cells CD4+ effector T cells, the role of apoptosis in the homeostasis of (24) (Fig. 5). In Spi6 KO iNKT cells, we observed a more diffuse activated iNKT cells is poorly understood. We provide evidence for pattern of Ab staining indicative of increased distribution of GrB a physiological role for Spi6 for protecting iNKT cells from self- to the cytoplasm. Therefore, Spi6 prevents the leakage of GrB into inflicted apoptosis caused by GrB. Our findings support a new mech- the cytoplasm of iNKT cells. anism in which the size of the clonal burst of iNKT cells is deter- mined by protection from GrB-mediated apoptosis. Clonal burst deficit of Spi6 KO iNKT cells is rescued by GrB Wehavedemonstrated previously thatSpi6 protects classical CTL deficiency from GrB- mediated apoptosis and thereby determines the size of the We determined whether inhibition of GrB was a physiological clonal burst postinfection with either LCMVor Listeria monocyto- mechanism by which Spi6 protected Ag-stimulated iNKT cells from (24). Proteinase inhibitor 9 (PI9) is a human homolog of Spi6

FIGURE 6. Effect of GrB deficiency on the clonal burst of Spi6 KO iNKT cells. Wt, Spi6 KO, and Spi6 KO3GrB KO mice were injected i.p. with 2 mg PBS-57 and iNKT cells harvested on day 3. n = 3–5 mice. A,Re- presentative FACS plots showing percentage of CD1d tet+ iNKT cells in spleen. B, Mean percentage and absolute number of iNKT cells. C, Mean percentage of active cas- pase 3+ iNKT cells (CD1d tet+). pp , 0.05; ppp , 0.001; n.s., p . 0.5. 882 Spi6 AND iNKT CELLS that, as well as being a potent inhibitor of intracellular GrB, is also and after Ag stimulation, whereas, GrB exclusively induces apopto- expressed in cytotoxic lymphocytes (31). A previous study demon- sis after TCR stimulation. The PI9 human homolog of Spi6 is a strated GrB-mediated activation-induced cell death of human NK potent inhibitor of intracellular GrB and is also expressed in cyto- cells but did not directly demonstrate a physiological requirement toxic lymphocytes (31). Therefore, the expansion of iNKT cells in for PI9 in cell survival (32). Using Spi6 KO mice, we are able to clinically relevant diseases may also be controlled by the protection directly demonstrate a physiological requirement of GrB inhibition from self-inflicted damage by natural endogenous inhibitors of GrB. by Spi6 in determining the viability of a cytotoxic cell of the innate immune system. Acknowledgments iNKTcellsperformimportantimmunoregulatoryfunctionthrough We thank Dr. Paul Savage (Brigham Young University, Provo, UT) and Na- the rapid production of cytokines, such as IFN-g and IL-4, after tional Institutes of Health Tetramer Facility for providing glycolipid PBS-57 recognizing glycolipid Ag (2–7). Through this activity, iNKT cells and PBS-57 loaded/CD1d tetramer, respectively. We also thank Susan are thought to provide an important link between innate Ag recog- Byrne for helping with cell sorting and members of the PA-R laboratory. nition and the recruitment of Th1 (IFN-g) and Th2 (IL-4) arms of adaptive immunity. Therefore, the dramatic impairment in the ability of Spi6 KO mice to produce IFN-g and IL-4 after PBS-57 challenge Disclosures (Fig. 2F) strongly suggests that Spi6 is required for iNKT cell func- The authors have no financial conflicts of interest. tion by ensuring their viability. We show that Spi6 is a physiological inhibitor of GrB in iNKT References cells (Figs. 2, 3, 5). As we observed for CTL, Spi6 ensured the Downloaded from 1. Godfrey, D. I., H. R. MacDonald, M. Kronenberg, M. J. Smyth, and L. Van Kaer. integrity of cytotoxic granules and prevented GrB leakage into the 2004. NKT cells: what’s in a name? Nat. Rev. Immunol. 4: 231–237. cytoplasm of iNKT cells (Fig. 5). Interestingly, classical Wt CTLs 2. Bendelac, A., P. B. Savage, and L. Teyton. 2007. The biology of NKT cells. Annu. Rev. Immunol. 25: 297–336. appear to selectively leak GrB, but not granzyme A or perforin, 3. Borg, N. A., K. S. Wun, L. Kjer-Nielsen, M. C. Wilce, D. G. Pellicci, R. Koh, from their granules (24, 33). Given that Spi6 is specific for GrB, G. S. Besra, M. Bharadwaj, D. I. Godfrey, J. McCluskey, and J. Rossjohn. 2007. this may explain the nonredundant role for Spi6 in protecting CTL CD1d-lipid- recognition by the semi-invariant NKT T-cell receptor. Na- ture 448: 44–49. and iNKT cells from self-inflicted damage caused by the break- 4. Brigl, M., and M. B. Brenner. 2004. CD1: antigen presentation and T cell fun- http://www.jimmunol.org/ down of cytotoxic granules. However, how the increased activity ction. Annu. Rev. Immunol. 22: 817–890. of cytoplasmic GrB causes the breakdown of cytotoxic granules 5. Kronenberg, M. 2005. Toward an understanding of NKT cell biology: progress and paradoxes. Annu. Rev. Immunol. 23: 877–900. remains to be determined. 6. Parekh, V. V., M. T. Wilson, and L. Van Kaer. 2005. iNKT-cell responses to Spi6 is not a physiological inhibitor of Fas-induced apopto- glycolipids. Crit. Rev. Immunol. 25: 183–213. sis (22) and so the increase in apoptosis of Spi6 KO iNKT cells 7. Taniguchi, M., M. Harada, S. Kojo, T. Nakayama, and H. Wakao. 2003. The reg- ulatory role of Valpha14 NKT cells in innate and acquired immune response. is likely to be exclusively through increased GrB levels and ac- Annu. Rev. Immunol. 21: 483–513. tivity. This is confirmed by the observations that both the de- 8. Hayakawa, Y., D. I. Godfrey, and M. J. Smyth. 2004. Alpha-galactosylceramide: creased clonal burst of iNKT cells in Spi6 KO mice is rescued potential immunomodulatory activity and future application. Curr. Med. Chem.

3 11: 241–252. by guest on September 28, 2021 by GrB deficiency in Spi6 KO GrB KO mice (Fig. 6). Therefore, 9. Liu, Y., R. D. Goff, D. Zhou, J. Mattner, B. A. Sullivan, A. Khurana, C. Cantu, the GrB pathway of iNKT cell homeostasis is therefore likely to III, E. V. Ravkov, C. C. Ibegbu, J. D. Altman, et al. 2006. A modified alpha- be distinct to the Fas–FasL pathway of iNKT activation-induced galactosyl for staining and stimulating natural killer T cells. J. Immu- nol. Methods 312: 34–39. cell death (15). 10. Eberl, G., and H. R. MacDonald. 1998. Rapid death and regeneration of NKT iNKT cell can be activated in a TCR-independent fashion by cells in anti-CD3epsilon- or IL-12-treated mice: a major role for bone marrow in microbes, such as viruses, through inflammatory cytokines, such as NKT cell homeostasis. Immunity 9: 345–353. 11. Crowe, N. Y., A. P. Uldrich, K. Kyparissoudis, K. J. Hammond, Y. Hayakawa, type I IFN, IL-12, IL-18, and IFN-g (34). However, our findings S. Sidobre, R. Keating, M. Kronenberg, M. J. Smyth, and D. I. Godfrey. 2003. support the view that stimulation through the TCR is needed to Glycolipid antigen drives rapid expansion and sustained production by induce GrB-mediated death of iNKT cells. This is because we did NK T cells. J. Immunol. 171: 4020–4027. 12. Harada, M., K. Seino, H. Wakao, S. Sakata, Y. Ishizuka, T. Ito, S. Kojo, T. Nakayama, not observe any decrease in the level of iNKT cells after TCR- and M. Taniguchi. 2004. Down-regulation of the invariant Valpha14 antigen receptor independent stimulation with LCMV (Fig. 4). It was only after direct in NKT cells upon activation. Int. Immunol. 16: 241–247. stimulation through the TCR by PBS-57 did we observe an increase 13. Wilson, M. T., C. Johansson, D. Olivares-Villago´mez, A. K. Singh, A. K. Stanic, C. R. Wang, S. Joyce, M. J. Wick, and L. Van Kaer. 2003. The response of in GrB-mediated apoptosis of Spi6 KO iNKT cells (Figs. 2, 3, 6). Our natural killer T cells to glycolipid is characterized by surface receptor results suggest that TCR-mediated mobilization of GrB is required down-modulation and expansion. Proc. Natl. Acad. Sci. USA 100: 10913–10918. for death through misdirected GrB. Although it is not clear how GrB 14. Uldrich, A. P., N. Y. Crowe, K. Kyparissoudis, D. G. Pellicci, Y. Zhan, A. M. Lew, P. Bouillet, A. Strasser, M. J. Smyth, and D. I. Godfrey. 2005. NKT cell stimulation reaches the cytoplasm of iNKT cells, our CIM studies indicate that with glycolipid antigen in vivo: costimulation-dependent expansion, Bim-dependent the breakdown of cytotoxic granules on TCR activation is probably contraction, and hyporesponsiveness to further antigenic challenge. J. Immunol. 175: involved (Fig. 5). In contrast, classical NK cells, which constitutively 3092–3101. 15. Leite-de-Moraes, M. C., A. Herbelin, C. Gouarin, Y. Koezuka, E. Schneider, and express GrB, appear to undergo apoptosis in the absence of overt Ag M. Dy. 2000. Fas/ interactions promote activation-induced cell death stimulation (32). of NK T lymphocytes. J. Immunol. 165: 4367–4371. 16. Russell, J. H., and T. J. Ley. 2002. Lymphocyte-mediated cytotoxicity. Annu. Spi6 was not required for the development or survival on iNKT Rev. Immunol. 20: 323–370. cells in the absence of Ag stimulation. This is in contrast to the recent 17. Ka¨gi, D., B. Ledermann, K. Bu¨rki, P. Seiler, B. Odermatt, K. J. Olsen, E. R. Podack, finding that the Id2 transcriptional regulator determines the level of R. M. Zinkernagel, and H. Hengartner. 1994. Cytotoxicity mediated by T cells and natural killer cells is greatly impaired in perforin-deficient mice. Nature 369: 31–37. unstimulated iNKT cells by ensuringsurvival inthe liver (35). Spi6 is 18. Podack, E. R., J. D. Young, and Z. A. Cohn. 1985. Isolation and biochemical and transcriptional target of Id2 and so it is possible that Id2 (36) also functional characterization of perforin 1 from cytolytic T-cell granules. Proc. protects iNKT cells from activation-induced apoptosis. Deficiency Natl. Acad. Sci. USA 82: 8629–8633. 19. Heusel, J. W., R. L. Wesselschmidt, S. Shresta, J. H. Russell, and T. J. Ley. 1994. in the proapoptotic member of the Bcl-2 family, Bim could correct Cytotoxic lymphocytes require granzyme B for the rapid induction of DNA the defect in functional maturation and survival of Id2-deficient fragmentation and apoptosis in allogeneic target cells. Cell 76: 977–987. iNKT cells (35). In addition, Bim deficiency can correct activation 20. Ashton-Rickardt, P. G. 2005. The granule pathway of programmed cell death. Crit. Rev. Immunol. 25: 161–182. induced apoptosis of iNKT cells (11). Therefore, it would seem that 21. Remold-O’Donnell, E. 1993. The ovalbumin family of serpin . FEBS Bim is involved in the apoptosis of iNKT cells during development Lett. 315: 105–108. The Journal of Immunology 883

22. Phillips, T., J. T. Opferman, R. Shah, N. Liu, C. J. Froelich, and P. G. Ashton- 30. Roberts, T. J., Y. Lin, P. M. Spence, L. Van Kaer, and R. R. Brutkiewicz. 2004. Rickardt. 2004. A role for the granzyme B inhibitor serine protease inhibitor 6 in CD1d1-dependent control of the magnitude of an acute antiviral immune re- CD8+ memory cell homeostasis. J. Immunol. 173: 3801–3809. sponse. J. Immunol. 172: 3454–3461. 23. Sun, J., L. Ooms, C. H. Bird, V. R. Sutton, J. A. Trapani, and P. I. Bird. 1997. A 31. Sun, J., C. H. Bird, V. Sutton, L. McDonald, P. B. Coughlin, T. A. De Jong, new family of 10 murine ovalbumin serpins includes two homologs of proteinase J. A. Trapani, and P. I. Bird. 1996. A cytosolic granzyme B inhibitor related to inhibitor 8 and two homologs of the granzyme B inhibitor (proteinase in- the viral apoptotic regulator cytokine response modifier A is present in cytotoxic hibitor 9). J. Biol. Chem. 272: 15434–15441. lymphocytes. J. Biol. Chem. 271: 27802–27809. 24. Zhang, M., S. M. Park, Y. Wang, R. Shah, N. Liu, A. E. Murmann, C. R. Wang, 32. Ida, H., T. Nakashima, N. L. Kedersha, S. Yamasaki, M. Huang, Y. Izumi, M. E. Peter, and P. G. Ashton-Rickardt. 2006. Serine protease inhibitor 6 protects T.Miyashita, T.Origuchi, A. Kawakami, K.Migita, et al. 2003. Granzyme B leakage- cytotoxic T cells from self-inflicted injury by ensuring the integrity of cytotoxic induced cell death: a new type of activation-induced natural killer cell death. Eur. J. granules. Immunity 24: 451–461. Immunol. 33: 3284–3292. 25. Zhang, M., S. Byrne, N. Liu, Y. Wang, A. Oxenius, and P. G. Ashton-Rickardt. 33. Bide`re, N., M. Briet, A. Du¨rrbach, C. Dumont, J. Feldmann, B. Charpentier, G. de 2007. Differential survival of cytotoxic T cells and memory cell precursors. Saint-Basile, and A. Senik. 2002. Selective inhibition of dipeptidyl peptidase I, not J. Immunol. 178: 3483–3491. 26. Coquet, J. M., K. Kyparissoudis, D. G. Pellicci, G. Besra, S. P. Berzins, M. J. Smyth, caspases, prevents the partial processing of procaspase-3 in CD3-activated human and D. I. Godfrey. 2007. IL-21 is produced by NKT cells and modulates NKT cell CD8(+) T lymphocytes. J. Biol. Chem. 277: 32339–32347. activation and cytokine production. J. Immunol. 178: 2827–2834. 34. Tyznik, A. J., E. Tupin, N. A. Nagarajan, M. J. Her, C. A. Benedict, and M. Kronenberg. 27. Diana, J., and A. Lehuen. 2009. NKT cells: friend or foe during viral infections? 2008. Cutting edge: the mechanism of invariant NKT cell responses to viral danger Eur. J. Immunol. 39: 3283–3291. signals. J. Immunol. 181: 4452–4456. 28. Hobbs,J.A.,S.Cho,T.J.Roberts,V.Sriram,J.Zhang,M.Xu,andR.R.Brutkiewicz. 35. Monticelli, L. A., Y. Yang, J. Knell, L. M. D’Cruz, M. A. Cannarile, I. Engel, 2001. Selective loss of natural killer T cells by apoptosis following infection with M. Kronenberg, and A. W. Goldrath. 2009. Transcriptional regulator Id2 controls lymphocytic choriomeningitis virus. J. Virol. 75: 10746–10754. survival of hepatic NKT cells. Proc. Natl. Acad. Sci. USA 106: 19461–19466. 29. Lin, Y., T. J. Roberts, C. R. Wang, S. Cho, and R. R. Brutkiewicz. 2005. Long- 36. Cannarile, M. A., N. A. Lind, R. Rivera, A. D. Sheridan, K. A. Camfield, B. B. Wu, term loss of canonical NKT cells following an acute virus infection. Eur. J. K. P. Cheung, Z. Ding, and A. W. Goldrath. 2006. Transcriptional regulator Id2 Immunol. 35: 879–889. mediates CD8+ T cell immunity. Nat. Immunol. 7: 1317–1325. Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021