Engaging Ly‐6A/Sca‐1 triggers lipid raft‐dependent and ‐independent responses in CD4+ T‐cell lines

Abstract Introduction The lymphocyte antigen 6 (Ly‐6) supergene family encodes proteins of 12–14 kda in molecular mass that are either secreted or anchored to the plasma membrane through a glycosyl‐phosphatidylinisotol (GPI) lipid anchor at the carboxy‐terminus. The lipidated GPI‐anchor allows localization of Ly‐6 proteins to the 10–100 nm cholesterol‐rich nano‐domains on the membrane, also known as lipid rafts. Ly‐6A/Sca‐1, a member of Ly‐6 gene family is known to transduce signals despite the absence of transmembrane and cytoplasmic domains. It is hypothesized that the localization of Ly‐6A/Sca‐1 with in lipid rafts allows this protein to transduce signals to the cell interior. Methods and Results In this study, we found that cross‐linking mouse Ly‐6A/Sca‐1 protein with a monoclonal antibody results in functionally distinct responses that occur simultaneously. Ly‐6A/Sca‐1 triggered a cell stimulatory response as gauged by cytokine production with a concurrent inhibitory response as indicated by growth inhibition and apoptosis. While production of interleukin 2 (IL‐2) cytokine by CD4+ T cell line in response to cross‐linking Ly‐6A/Sca‐1 was dependent on the integrity of lipid rafts, the observed cell death occurred independently of it. Growth inhibited CD4+ T cells showed up‐regulated expression of the inhibitory cell cycle protein p27kip but not of p53. In addition, Ly‐6A/Sca‐1 induced translocation of cytochrome C to the cytoplasm along with activated caspase 3 and caspase 9, thereby suggesting an intrinsic apoptotic cell death mechanism. Conclusions We conclude that opposing responses with differential dependence on the integrity of lipid rafts are triggered by engaging Ly‐6A/Sca‐1 protein on the membrane of transformed CD4+ T cells.


Introduction
Mouse Ly-6A, also known as Sca-1 (Stem cell antigen 1), is a GPI-anchored membrane protein and the prototypic member of the lymphocyte antigen 6 (Ly-6) supergene family. These proteins were first described as allo-antigens expressed on activated murine lymphocytes [1]. The Ly-6 proteins are expressed in a wide range of organisms ranging from C. elegans to humans, and across tissue types as variable as stem cells, lymphocytes, neurons, and muscle cells [2,3]. A number of Ly-6 proteins, including Ly-6A/Sca-1, have cellcell adhesion properties in a variety of cell types [4][5][6][7][8].
Cross-linking of Ly-6 proteins with anti-Ly-6 monoclonal antibodies alone is sufficient to induce cell activation in transformed T cells [9,10], but additional co-stimulation is required to activate primary mouse CD4 þ T lymphocytes [9,10]. Expression of Ly-6A/Sca-1 regulates signaling through the antigen receptor on CD4 þ T cells and their cytokine responses [11][12][13]. The Ly-6 gene locus also influences susceptibility to mouse adeno virus in murine models, West Nile virus, HIV-1, and several other DNA and RNA viruses [14][15][16][17]. While various members of Ly-6 family are recognized for their role in cytokine responses by T cells, the full spectrum of responses, and the contribution of lipid rafts to signaling initiated by engaging Ly-6A/Sca-1 is unknown.
Ly-6A/Sca-1 signals to the cell interior despite the absence of a transmembrane and cytoplasmic tail. Inclusion of the lipid anchored Ly-6A/Sca-1 protein in the lipid rafts on the plasma membrane raises the possibility that this tail-less protein may possibly co-opt these signaling platforms to transduce signals. Lipid rafts are dynamic nano-domains on the plasma membrane that play an essential role in signal transduction by providing a platform to assemble signaling receptors, enzymes, and adaptor proteins [18]. We report here that engaging Ly-6A/Sca-1 protein on transformed murine T cells signals for cytokine response, growth inhibition, and apoptosis. While the interleukin 2 (IL-2) cytokine response is dependent on the integrity of the lipid rafts, the apoptotic cell death triggered by Ly-6A/Sca-1 is lipid raft independent. High expression of Ly-6A/Sca-1 observed on transformed cells, and its growth inhibition and apoptosis triggered in immortalized T cell lines by engaging this protein, suggests its promise as a potential tumor antigen target.

Cellular proliferation-MTS assay
Cell proliferation was measured using the CellTiter 96 1 Aqueous One Solution Cell Proliferation Assay (Promega Corp., Madison, WI) as per the manufacturer's instructions. Briefly, YH16.33 (5 Â 10 3 per well) cells were seeded in a 96-well plate with 100 ml of fresh RPMI 1640-GlutaxMAX TM cell culture media. The cells were either cultured for 4, 8, 24, or 48 h in media alone that served as negative control or media containing Ly-6A/Sca-1 monoclonal antibody at 4 mg/ml concentration. A total of 20 ml of CellTiter 96 1 AQueous One Solution Reagent (Promega Corp., Madison, WI) was added to each well, and then the plate was incubated at 378C in humidified 5% CO 2 incubator for 1 h. The absorbance was read at 490 nm using a 96-well plate reader.

Cytokine assays
To quantify IL-2 in anti-Ly-6A and anti-CD3e treated YH16.33 cells, the top 100 ml of supernatants was harvested at both 24 h time points and then frozen at À208C for ELISA analysis. IL-2 assay kit was used (BD Bioscience, San Jose, CA). Briefly, 24 h before the cytokine assay was carried out, ELISA plate (Costar, USA) was coated with 50 ml/well capture antibody diluted in 1:250 in coating buffer (carbonate/bicarbonate, pH 9.0) and stored at 48C overnight as per instructions from the vendor (BD Biosciences). After overnight incubation, the unbound capture antibody from the 96-well plate was removed by dumping the contents, and each well was subsequently washed with 100 ul of Phosphate with 0.05% Tween 20 (Sigma-Aldrich, St. Louis, MO). This was then followed by a blocking step; 100 ul of blocking buffer (PBS without Tween-20 þ10% FCS) and incubated at room temperature for 30 min. The assay supernatants and standards were then added to the appropriate wells and incubated for 90 min at room temperature. After this incubation, the contents of the plate were dumped and washed five times with wash buffer. A total of 50 ml of detector reagent (biotinylated anti-IL-2 detection antibody at 1:1000 and Avidin HRP enzyme [BD Biosciences] at 1:250 dilution) for the cytokine being detected was added to the plate and incubated for 30 min at 48C with the plate covered with aluminum foil. After 30 min, the detector was dumped and the wells were washed seven times with wash buffer with each wash done for about 30 s. Assay was developed with substrate and chromogen solution (KPL, Milford, MA) in a 1:1 ratio and 100 ml of the mixture was added to each well. To serve as a blank, two untreated wells contained substrate. The assay was developed in dark for about 30 min. The plate was then read at 405 nm using Spectramax 190 plate reader (Molecular Devices, Sunnyvale, CA). The data for each treatment was done in triplicate. In the final analysis of data, an average of the three values was taken and the graphs that were plotted shows the average values.

Flow cytometry
Active caspase-3 staining The detection of active caspase-3 was carried out using The CaspGLOW TM Fluorescein Active Caspase-3 Staining Kit (BioVision, Mountain View, CA) per the manufacturer's instructions. Briefly, YH16.33 cells were incubated in media alone (Non-treated negative control) or with 4 mg/ml Ly-6A monoclonal antibody (mAb) for 4, 8, 24, or 48 h. Afterwards, 300 ml of the cell culture was incubated with 1 ml of fluorescein isothiocyanate (FITC)-DEVD-FMK (BioVision) marker for 1 h at 378C in humidified 5% CO 2 incubator. The cells were then centrifuged at 3000 rpm for 5 min, the supernatant removed, and the cells were re-suspend in 0.5 ml of Wash Buffer. The cells were then centrifuged again, then resuspended in 300 ml of Wash Buffer, and then analyzed by FACSCalibur flow cytometer (BD Biosciences, Palo Alto, CA) using the FL-1 channel.

Flow cytometry and cytochrome C staining
The intracellular cytochrome staining was carried out according to a published study [23]. Briefly, YH.33 cells were treated with 100 ml of a digitonin solution (50 mg/ml in PBS with 100 mM KCl) for 5 min on ice. The cells were then fixed with 4% paraformaldehyde solution (Electron Microscopic Sciences, Hatfield, PA) made in PBS for 20 min. Paraformaldehyde was washed three times, and treated with a blocking buffer consisting of 3% BSA, 0.05% saponin, and made in PBS for 1 h. The cells incubated with a FITC-tagged monoclonal antibody against cytochrome C (clone 6H2.B4) (BioLegend, San Diego, CA), at 1:200 monoclonal dilution and were then analyzed using Flow Cytometry.

Flow cytometry and staining with Di-4-ANEPPDHQ
Lymph node cells were stained with fluorescent membrane dye, di-4-ANEPPDHQ (Invitrogen-Life Technologies, Grand Island, NY) at 0.5 mM final concentration for 20 min at room temperature as previously reported [22]. Labeled cells were analyzed by FACSCalibur flowcytometer (BD Biosciences, East Rutherford, NJ) using 488 nm excitation lasers. Emission from the labeled cells was recorded at wavelengths 570 nm (FL2 channel), 630 nm (FL3 channel). Folowing formula was used to assess generalized polarization (GP) values ¼ I 570 ÀI 630 /I 570 þ I 630; I ¼ mean fluorescence intensity.

Statistical analyses
All Statistical analysis, except where indicated in the legend, was done using one-way ANOVA testing. Either One-Way ANOVA or non-parametric Krusal-Wallis ANOVA was initially used to examine the overall variance, followed by either Tukey HSD or MCTP testing respectively in RStudio to determine where differences in the data lie. Results were considered significant if the p-value was less than 0.05.

Ethics statement
No animals or human subjects were used or involved in our study.

Results
Engaging Ly-6A/Sca-1 with anti-Ly-6A/Sca-1 mAb promotes cytokine production in YH16.33 T cells dependent on the integrity of lipid raft-based membrane order Engaging Ly-6A/Sca-1 with anti-Ly-6A/Sca-1 mAbs is known to activate CD4 þ T cell lines resulting in gene transcription and translation of multiple cytokines [9,10,19,20]. We examined the role of lipid raft-based membrane order in CD4 þ T cell response generated after engaging Ly-6A/Sca-1 protein with anti-Ly-6A mAB (8G12). Membrane order was disrupted by inserting an oxyseterol, 7-Keto Cholesterol (7-KC) in the membrane of a CD4 þ T cell line as reported previously [22]. Exposure to 7-KC resulted in loss of lipid raft-based membrane order in a concentration-dependent manner, (Supplementary Fig. S1) similar to what was observed with primary mouse CD4 þ T cells reported previously [22]. YH16.33 T cells exposed to different concentrations of 7-KC, prior to engaging Ly-6A/ Sca-1 protein with a Ly-6A/Sca-1 specific monoclonal antibody (mAb) 8G12, produced significantly lower IL-2 (Fig. 1A). The amount of IL-2 produced with highest concentration of the anti-Ly-6A/Sca-1 antibody was inversely proportional to the 7-KC concentration with >80% inhibition observed with highest 58 mM concentration used (Fig. 1A). Similarly, in response to stimulation through the (T cell receptor) TCRab/CD3 complex, 7-KC treated YH16.33 cells produced lower levels of IL-2 than either the untreated or vehicle (mbCD) treated cells (Fig. 1B). 7-KC treated cells expressed similar levels of Ly-6A/Sca-1, lymphocyte function antigen 1 (LFA-1), TCRb, and CD3e proteins on the cell surface as the untreated controls ( Supplementary Fig. S2) and therefore does not account for the observed altered IL-2 responses through the antigen receptor complex and Ly-6A/Sca-1 protein.
Anti-Ly-6A/Sca-1 mAb induces apoptotic cell death in YH16.33 T cells independent of lipid raft-based membrane order While the activated YH16.33 cells produced IL-2, a signature cytokine for T cell response, visual examination of cells suggested cell death and growth inhibitory effects of the antibody which prompted further investigation. To determine and quantify the effect of cross linking Ly-6A/Sca-1 proteins on cell death and cellular proliferation, YH16.33 cells were treated with different amounts of anti-Ly-6A/Sca-1 antibody at varying time periods before assessing their survival and cellular proliferation. Anti-Ly-6A/Sca-1-treated cells were assessed for apoptosis by staining with Annexin and 41% apoptotic (Annexin V þ PI À ) cells, respectively ( Fig. 2A). In contrast, apoptotic cell death in the YH16.33 control cultures treated with anti-CD3e antibody did not deviate significantly from the untreated YH16.33 cells (less than 8.9%) (Fig. 2B). Viability of YH16.33 cells at the beginning of the cell culture ranged from 93% to 97%.
Ligand induced crosslinking of ''death receptors,'' such as Fas, results in their aggregation and formation of Death-Inducing Signaling Complex (DISC) which in turn are recruited to lipid rafts [24,25]. Ly-6A/Sca-1 protein, because of its GPI-anchor, is housed in cholesterol and saturated lipid-rich lipid rafts as well on the plasma membrane [26,27]. Therefore, we tested the role of lipid rafts in Ly-6A/ Sca-1-mediated cell death. Cells were treated with 7ketocholesterol to disrupt the integrity of raft nano-domains as per previously published reports [22,28,29] and then incubated with anti-Ly6A mAb for 24 h. Such treatment caused an increase in apoptosis in anti-Ly-6A/Sca-1 mAb concentration-dependent manner but independent of the treatment with 7-KC. As seen in Figure 2C, even the lowest concentration of the Ly-6A/Sca-1 mAb induced apoptosis in YH16.33 T cell line four times above the background despite 7-KC treatment. An average of 31.6% (n ¼ 3) cells were scored apoptotic at 0.5 mg of anti-Ly-6A/Sca-1 antibody. To investigate the specificity of this response, we tested for the effects of cross-linking the TCRab/CD3 complex with anti-CD3e mAb. As shown in Figure 2C, apoptotic cells ranged from 4% to 10% of the population over the same range of antibody concentration making it indistinguishable from the background. We did not observe statistically significant differences between 7-KC-treated and untreated groups (p > 0.05). Our data suggest that apoptosis induced by engaging Ly-6A/Sca-1 is specific, and occurs independent of the integrity of the lipid rafts. In contrast, as shown in Figure 1, IL-2 response generated by YH16.33 under similar conditions is dependent on lipid raft integrity.
The growth inhibitory effects of anti-Ly-6A/Sca-1 was tested on other CD4 þ T cell lines and normal lymph node T cells. Similar effects were observed in KQ23.23.7 and D10. Growth inhibitory responses by Ly-6A/Sca-1 protein G4, two other independently-derived CD4 þ T cell lines ( Fig. 3A and B). A representative experiment (n ¼ 3) is shown ( Fig. 3A and B). However the extent of growth inhibition differed between the two cell lines. While KQ23.37.7 showed similar growth inhibition as the YH16.33 cell line, the D10.G4 with about 40-50% lower expression of Ly-6A/Sca-1 (data not shown) showed reduced growth inhibition (Fig. 3B). In contrast, anti-Ly-6A/Sca-1 antibody did not show growth inhibitory response on primary lymph node T cells (Fig. 3C). Taken together, our results suggest that growth inhibitory effects of engaging Ly-6A/Sca-1 protein are specific to transformed CD4 þ T cells and not on primary T cells from the lymph node.
Engaging Ly-6A/Sca-1 induces expression of p27 kip without affecting p53 expression Growth inhibition in a variety of cell types involves upregulation of p27 kip , a cell cycle inhibitory protein [30]. We next sought to examine p27 kip expression in YH16.33 cells after their Ly-6A/Sca-1 protein was engaged. Figure 4A shows that p27 kip protein was present at about equal level in YH16.33 cells treated with anti-CD3e or left untreated. Expression of p27 kip was increased by more than 10-fold when YH16 cells were treated with anti-Ly-6A/Sca-1 antibody. In these experiments the expression of p53 was not observed in any of the above treatments (Fig. 4B). In addition, we tested expression of p53 in YH16.33 cells treated with anti-Ly-6A/Sca-1 and Nutlin 3a, as the latter is known to inhibit degradation of p53 by blocking the binding of ubiquitin ligase mouse double minute 2 (MDM2) to p53 protein [31]. We were unable to observe p53 protein in lysates of YH16.33 cells treated with a combination of anti-Ly-6A/Sca-1 and Nutlin 3a by Western blots (Fig. 4B). The absence of detection was not due to the inactivity of the anti-p53 antibody used in our experiments as we were able to detect p53 in lysates from 293T cells. These results indicate that cross-linking of Ly-6A/Sca-1 protein results in upregulation of a cell cycle inhibitory protein, p27 kip . Additionally, in its steady state-level, p53 protein was undetectable in YH16.33 cells.
Ly-6A/Sca-1 mAb activates caspase-3 To further confirm that pro-apoptotic signals are induced by engaging Ly-6A/Sca-1 proteins, we examined the activation of caspase-3 using both Western blot and flow cytometric analyses. The activation of a family of cysteine-protease caspase(s) is one of the crucial events in apoptosis. One particular caspase, caspase-3, is a vital executioner of apoptosis as it is responsible for the proteolytic cleavage of endogenous proteins that are vital for signal transduction and structural maintenance of the cell [24]. Caspase-3 is activated by proteolytic processing of its inactive zymogen form into its p17 and p12 catalytically active fragments [25]. As seen in Figure 5A, Western blot analysis detected the cleaved (active) fragments of caspase-3 as early as 4 h after Ly-6A/Sca-1 mAb treatment. The active caspase-3 proteins fragments detected increased in the Ly-6A/Sca-1 mAb treatment group over the time course tested. In contrast, active caspase-3 fragments from the control group were not seen in early time points (data not shown). To corroborate the Western blot analysis, we also performed flow cytometry on YH16.33 cells stained with a FITC-labeled DEVD-FMK (FITC-DEVD-FMK) marker that binds irreversibly to active caspase-3. After 8 h of Ly-6A/Sca-1 mAb treatment, the percentage of YH16.33 cells that positively stained for the FITC-DEVD-FMK marker was 22.74% (Fig. 5B). While only 7.24% positively stained cells were seen in the control group at the same time period. This disparity in the percentage of FITC-DEVD-FMK þ cells between the control and Ly-6A/Sca-1 mAb group was also observed in the 24 and 48 h treatment time periods (9.7-32.3%, and 8.52-22.24%, respectively). As a result, from the data in Figure 5A and B, we conclude that the pro-apoptotic signal that is generated by cross linking Ly-6A/Sca-1 proteins can rapidly activate caspase-3. We next examined the role of caspase 8 in apoptotic cell death induced by anti-Ly-6A/Sca-1. Caspase 8 is an initiator caspase that typically is recruited by the death receptors to the membrane to trigger cell-extrinsic cell death, a pathway distinct from cell-intrinsic death pathway triggered by the release of cytochrome C from the mitochondria. We did not

454
detect, in our Western blots, the active fragment of caspase 8 with a specific rabbit monoclonal antibody directed against caspase 8 in YH16.33 cells exposed to anti-Ly-6A/Sca-1 antibody (Fig. 5E), nor did we observe changes in the levels of Bax, a pro-apoptotic protein, when compared to the controls (Fig. 5E). In contrast, the Ly-6A/Sca-1 engagement triggers the release of cytochrome C as assessed by intracellular staining with anti-cytochrome C antibody followed by flow cytometeric analysis (Fig. 5C) and generates active caspase 9 (Fig. 5D). These data indicated that apoptotic cell death through cell intrinsic pathway is triggered in the YH16.33 T cell line when Ly-6A/Sca-1 is engaged.

Discussion
In the present study, we report that engaging Ly-6A/Sca-1, a GPI-anchored protein, on clonal CD4 þ T cells triggers cytokine response, growth inhibition, and apoptosis through distinct mechanisms. While production of IL-2 by immortalized CD4 þ T cells in response to engaging Ly-6A/Sca-1 protein was dependent on the lipid raft-based membrane order, the apoptotic cell death was independent of lipid raft integrity. In addition, cells undergoing apoptotic cell death generated signatures associated with cell-intrinsic apoptosis. Cell death triggered after engaging Ly-6A/Sca-1 protein with a monoclonal antibody showed signatures of apoptosis. This inference is based on assessing the asymmetry of plasma membrane of anti-Ly-6A/Sca-1 treated cells by staining with Annexin V-FITC followed by flow cytometeric analysis (Fig. 2) and by detecting intracellular activated caspase-3 ( Fig. 5A and B). It is generally accepted that the apoptotic cascade proceeds through two distinct pathways [reviewed in 32]. The ''extrinsic pathway'' of apoptosis is initiated when the death receptors on the plasma membrane are engaged by their ligands in either the soluble or cell surface bound forms [32], resulting in the formation of deathinducing signaling complex (DISC) [24]. Activation induced cell death in T cells and exposure to chemotherapeutic compounds are known to induce apoptotic cell death in T cell lines by recruiting death receptors and initiator caspases to DISC [24,33,34]. Disruption of lipid rafts interferes with apoptosis induced through DISC where the Fas receptor and a number of initiation caspases (Caspase-8, 9, and 10) play a central role [35,36]. In contrast, anti-Ly-6A/Sca-1 triggered apoptosis was not inhibited when integrity of lipid rafts was disrupted with 7-KC (Fig. 2), therefore suggesting a different mechanism underlying cell death initiated by engaging Ly-6A/Sca-1 protein. Consistent with this data are our Western blot experiments, where we did not detect active caspase 8, an initiator caspase recruited by the death receptor (Fig. 5E). Together these observations suggest that the pro-apoptotic Ly-6A/Sca-1 signal does not utilize initiator caspases associated with extrinsic apoptotic cell death pathway. Detection of intracellular cytochrome C (Fig. 5C) suggests that by engaging Ly-6A/Sca-1, the mitochondrial membrane loses permeability resulting in release of cytochrome C that in turn triggers the activation of caspase 3 and 9 (Fig. 5A, B, and D), the executioners of apoptosis. These data support the idea that Ly-6A/Sca-1 triggers an ''intrinsic apoptotic pathway,'' however further experiments are required to investigate how a tail-less Ly-6A/Sca-1 protein on the outer leaflet of the membrane connects to the mitochondria present in the cell interior. Either direct signaling by Ly-6A/ Sca-1 protein into mitochondria or an indirect mechanism, where signaling through Ly-6A/Sca-1 results in production of a death factor that in turn initiates intrinsic cell death pathway are two possibilities for investigation in the future.
Direct engagement of Ly-6A/Sca-1 proteins with crosslinking antibodies activated immortalized CD4 þ T cell lines as assessed by their cytokine secretion [9,10]. How a tail-less GPI-anchored, Ly-6A/Sca-1 protein communicated with the cell interior has remained unclear. Published reports suggest that Ly-6A/Sca-1 requires a transmembrane protein complex for its signaling [20,37]. Surface expression of a transmembrane protein complex, the TCRab/CD3 [20] and cytoplasmic domain of one of its component, TCR-z, [reviewed in 37] are critical for responses through GPIanchored proteins (Ly-6A/Sca-1 and Thy-1) in CD4 þ T cells. However, physical associations between the antigen receptor and its components with Ly-6A/Sca-1 proteins have been hard to establish over decades of experimentation (unpublished results). GPI-anchored proteins, such as Ly-6A/Sca-1 are localized to lipid rafts in a number of cell types, including T lymphocytes [38][39][40]. Co-opting membrane domains like lipid rafts where signaling receptors (like TCR/CD3) and other signaling proteins assemble and can potentially interact with GPI-anchored proteins for its signaling remains to be fully explored. This idea is consistent with the reports that lipid rafts can coalesce to form signaling platforms to organize and recruit proteins that are involved in various cell signaling conduits, such as the apoptotic pathway [35] as well as cell activation [41]. Additionally, a co-immunoprecipitation study has indicated that GPIanchored proteins (like Ly-6A/Sca-1) associate with p56 lck and p59 fyn , the two key protein tyrosine kinases located in lipid rafts [42]. We have proposed that instead of their direct association, the co-immunoprecipitation experiments perhaps suggest their indirect association through their localization in lipid rafts [27]. Our experiments reported here show that YH16.33 T cells with compromised lipid raftbased membrane order produce reduced levels of IL-2 in response to anti-Ly-6A/Sca-1 mAbs (Fig. 1). Further experimentation is required to identify a transmembrane protein(s), (TCRab/CD3 or others) housed in lipid rafts that possibly can connect Ly-6A/Sca-1 present on the outer leaflet of the membrane to the signaling kinases (e.g., p56 lck and p59 fyn ) tethered to the inner leaflet, and therefore aid in, a better understanding of the signaling pathway triggered through the tail-less, GPI-anchored proteins.
The growth inhibitory effect of anti-Ly-6A/Sca-1 antibody has been described previously [37,43]. These inhibitory effects are not due to down-regulation of the protein expression. Activated CD4 þ T cells and transformed CD4 þ T cells express high level of Ly-6A protein [2]. While the growth inhibitory effect of anti-Ly-6A/Sca-1 are known, but cell cycle proteins involved in this process were not reported. We speculated that the p53 protein was involved in the growth inhibition and cell death observed in YH16.33 cells after engaging Ly-6A/Sca-1 protein on its plasma membrane. In transformed cells, p53 is often mutated, ubiquinated by MDM2 for degradation, thus preventing it from performing its normal function (cell cycle arrest) resulting in uncontrollable cell growth [44,45]. The observation of growth inhibition and apoptotic cell death in YH16.33 cells after engagement of Ly-6A/Sca-1 proteins prompted us to examine the expression of p53 in YH16.33 cells in the presence and absence of anti-Ly-6A/Sca-1 mAb. In some of these experiments Nutlin-3a, an inhibitor of MDM2 E3 ligase was included, given its role in inhibiting the interaction between p53 and MDM2 [31]. We did not detect p53 protein in YH16.33 cells at its steady-state level either in the absence of or presence of anti-Ly-6A mAb (Fig. 4B). p53 protein was undetectable in YH16.33 cells even after pharmacological treatement of YH16.33 cells with Nutlin 3a at 10 mg/ml concentration (Fig. 4B). Absence of detectable p53 in YH16.33 lysates strongly suggest that Ly-6A/Sca-1 specific growth inhibition observed in YH16.33 cells occur in the absence of p53 tumor suppressor gene. However, we observed an up-regulated expression of p27 kip , a cell cycle inhibitor protein instead (Fig. 4A). P27 kip regulates activity of cyclin-dependent kinase in response to anti-mitotic signaling [46][47][48]. High expression of p27 kip is associated with differentiation phenotype of a variety of cell types and low expression with embryonic stem cell phenotype in human embryonic stem cells by regulating expression of Brachyury and Twist [49]. Further experimentation will be required to identify downstream targets of p27 kip in T cell lines and link between growth inhibition and cell death by apoptosis, if any. The growth inhibitory response observed after engaging Ly-6A/Sca-1 protein by cross-linking with a monoclonal antibody appears to mirror the growth inhibitory role of this protein on immune and non-immune cells. CD4 þ T cells overexpressing Ly-6A/Sca-1 protein show reduced proliferation in response to a specific antigen [13] and CD4 þ T cells lacking the expression of Ly-6A/Sca-1 protein show modest hyper-proliferation [12]. These findings are consistent with a role of Ly-6A/Sca-1 protein in regulating cell signaling and proliferation through the antigen receptor in primary CD4 þ T cells. Ly-6A/Sca-1 induced expression of a p27 kip suggests a role of this cell cycle inhibitory protein in the growth inhibition and provides rationale to investigate this further.
While expression of Ly-6A/Sca-1 on primary T cells influences signaling through the antigen receptor, direct engagement of Ly-6A/Sca-1 alone does not generate cell death and growth inhibition responses (Unpublished data and Fig. 4). In contrast, directly engaging Ly-6A/Sca-1 on transformed CD4 þ T cell lines show measurable cellular responses, thus suggesting that Ly-6A/Sca-1 signaling pathways are likely to be wired differently between primary and immortalized CD4 þ T cells. Ly-6A/Sca-1 protein is highly expressed on transformed cells and its expression co-relates to the tumorigenic potential of these cells [50][51][52]. Prostate stem cell antigen (PSCA), a member of the Ly-6 family is overexpressed in prostate cancer cells [50]. In addition, the Ly-6A/Sca-1 protein appears to regulate tumorigenesis [53,54]. The high expression level and the growth inhibitory/cell death inducing properties of Ly-6A/Sca-1 in transformed cells provide an opportunity to target Ly-6A/ Sca-1 for cancer therapy. Figure S2. Expression of CD4 þ T cell molecules on the surface of either 7-KC treated or untreated cells by Flow cytometry is shown. YH16.33 cells treated with 7-KC for 15 min at RT or left untreated were stained with either PE or FITC conjugated anti-Ly-6A/Sca-1, anti-CD3e, anti-TCReb, and anti-LFA-1 monoclonal antibodies and analyzed by flow cytometer. Histograms shown represents the two experiments. Figure S3. Growth inhibition in anti-Ly-6A/Sca-1 antibody treated cells. YH16.33 or control MVB2 cells were treated with either anti-Ly-6A/Sca-1 (8G12) or anti-CD3e (145-2C11) for 48 hours and cell cultures were assessed for growth and survival by MTS assay. Absorbance at 490 nm was measured in triplicate samples. The data represent the mean AE S.D. of 7 independent experiments. Groups denoted with different letters are statistically different (p < 0.05).