Overexpression of Parathyroid Hormone-Related Protein Inhibits Pancreatic β-Cell Death In Vivo and In Vitro

Two independent RIP-PTHrP transgenic mouse lines (1799 and 1807) having similar phenotypes were generated as described previously (8). In all of the experiments, neonatal and adult transgenic mice were compared with their corresponding normal littermates. The age of the mice varied with the experiments and is specified in each experiment. Genotyping of the mice was performed by tail DNA PCR as described (20). All studies were performed with the approval of, and in accordance with, guidelines established by the University of Pittsburgh School of Medicine Institutional Animal Care and Use Committee.

Adult mice (3–6 months old) were injected intraperitoneally with STZ (Sigma, St. Louis, MO) at a single dose of 100 mg/kg. STZ was prepared fresh for each use in citrate buffer (10 mmol/l sodium citrate, 0.9% saline, pH 4.0–4.5) at a stock STZ concentration of 12.5 mg/ml. Initially, pancreata from normal mice were removed 6, 12, or 24 h after STZ treatment, fixed overnight in Bouin’s fixative, and immunostained with terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) to determine the extent of cell death. Based on these pilot studies, the 12-h time point was chosen to compare cell death in normal and transgenic mouse pancreata.

Neonatal mice (7–10 days) were injected in preliminary experiments with 100 or 150 mg STZ/kg; blood glucose was measured at 1, 2, or 3 days after treatment or pancreata were removed 6 h after STZ injection. Subsequently, in neonatal mouse studies, STZ was administered intraperitoneally at a dose of 150 mg/kg. Glucose concentrations were measured 24 h after this dose, and pancreata were harvested 6 hours after STZ injections to compare β-cell death between normal and transgenic neonatal mice. Blood glucose was measured by a Precision Q.I.D. portable glucometer (Medisense, Bedford, MA).

Mouse islets were isolated as previously described (27), with some modifications. Briefly, 3 ml of 1.7 mg/ml Collagenase P (Roche Molecular Biochemicals, Indianapolis, IN) in Hanks’ buffered saline solution (HBSS) was injected into the pancreatic duct. Subsequently, pancreata were removed and digested at 37°C for 17 min, before sieving through a 500-μm wire mesh to separate undigested tissue. The digested pancreas was pelleted and rinsed with HBSS, and islets were separated by density gradient in Histopaque (Sigma). After several washes with HBSS, islets were hand-picked under a microscope. Islets either were used immediately after isolation to obtain RNA or protein or were left in complete medium (RPMI medium with 10% fetal bovine serum, 5.5 mmol/l glucose, and 1% penicillin and streptomycin) for 1–2 days before islet cell cultures were prepared.

Cultures of single islet cells were prepared by handpicking 400 islet equivalents (IE) (1 IE = 200-μm-diameter islet) into microfuge tubes, rinsing with PBS, and trypsinizing with 400 μl trypsin-EDTA for 8–10 min at 37°C. Islet cells were dispersed by pipetting twice during digestion. One milliliter of complete medium was added to stop the trypsinization, and cells were pelleted by spinning at 2,000 rpm for 2 min before plating on 12-mm glass coverslips placed in 24-well plates. Initially, cells from ∼100 IE were plated on each coverslip in a small volume of 50 μl complete medium and incubated at 37°C for 2 h, to allow cells to attach to the glass surface. Subsequently, 800 μl medium was added and cells were allowed to grow for 2–4 days. Thereafter, cells were either treated with 5 mmol/l STZ for 6 h or were kept in complete medium or RPMI without serum and glucose for 16–24 h before rinsing with PBS and finally fixing in freshly prepared 2% paraformaldehyde for 30 min at room temperature. Normal islet cultures were treated with 10⁻⁶ mol/l peptides (1-36, 38-94, or 107-139) during the period of serum starvation.

Pancreata removed from transgenic and normal mice were fixed in Bouin’s solution, embedded in paraffin, sectioned, and immunostained after deparaffinization and rehydration. Cell death was detected by enzymatic in situ labeling of DNA strand breaks using the TUNEL method. TUNEL staining was performed using the In Situ Cell Death Detection Kit (Roche Molecular Biochemicals, Mannheim, Germany) according to the manufacturer’s protocol. Proteinase K (Invitrogen Life Technologies, Rockville, MD) and the substrate NBT/BCIP (Vector Laboratories, Burlingame, CA) were obtained from the indicated manufacturers. After TUNEL, sections were stained with a mixture of antibodies against islet hormones, including rabbit anti-pancreatic polypeptide antibody (1:20 dilution), rabbit anti-glucagon antibody (1:50 dilution) (both from Biogenex, San Ramon, CA), and rabbit anti-somatostatin antibody (1:200 dilution) (Novocastra Laboratories, Newcastle, U.K.) overnight at 4°C, to identify islets in the section. Visualization of staining was achieved using antibody coupled to horseradish peroxidase enzyme and the substrate diaminobenzidine tetrahydrochloride (Biogenex). Costaining with insulin and propidium iodide (PI) (Sigma) was also used to detect β-cell death. Immunofluorescent staining of pancreatic sections or islet cell cultures was carried out with a guinea pig anti-porcine insulin antibody (Zymed, San Francisco, CA) at a 1:15 dilution for 30 min at room temperature, followed by a fluorescein isothiocyanate-conjugated rabbit anti-guinea pig IgG secondary antibody (Zymed) at a 1:100 dilution for 1 h at room temperature in the dark. Finally, samples were incubated for 10 min at 37°C with 2 μg/ml PI and 100 μg/ml RNase I (Ambion, Austin, TX) made in PBS, washed several times with water, and coverslipped using the Prolong antifade kit (Molecular Probes, Eugene, OR). Immunostaining for Glut-2 was done using an affinity-purified goat anti-human Glut-2 antibody at a 1:500 dilution (Santa Cruz Biotechnology, Santa Cruz, CA) as described (28).

Total RNA was isolated from islets of mice aged 3 to 8 months, using Trizol reagent (Life Technologies, Grand Island, NY) according to the manufacturer’s instructions. Semiquantitative RT-PCR was performed as described in detail (28). The following pairs of murine gene-specific primers were used: Actin (Ambion), Glut-2, Bcl-2, Bcl-X_L, Bax, and Bad (Table 1). The size of the resulting product, the annealing temperature used, and the number of cycles previously determined to be in the nonsaturating part of the linear amplification portion of the PCR products are listed in Table 1 for each set of primers. Films were scanned and band densitometry was quantitated using the computer program ImageJ from the National Institutes of Health.

Whole islet extracts were made in freshly prepared lysis buffer (5% SDS, 80 mmol/l Tris-HCl, pH 6.8, 5 mmol/l EDTA, 0.5 mmol/l phenylmethylsulfonyl fluoride) at a ratio of 20 μl/100 IE. Islets were sonicated, the supernatant containing the cell lysate was separated by centrifugation, and protein concentrations were measured using the MicroBCA assay (Pierce). Forty micrograms of protein from each sample was added to loading buffer and analyzed using 10% SDS-polyacrylamide gels. Proteins were transferred from the gels to Immobilon-P membrane (Millipore, Bedford, MA) using standard techniques. Blots were incubated with primary antibodies against Bcl-2 (polyclonal rabbit anti-rat/mouse Bcl-2 antibody at 1:2,000 dilution) (PharMingen International, San Diego, CA), Bcl-X_L (mouse monoclonal antibody at 1:200 dilution) (Santa Cruz Biotechnology), Glut-2 (at a 1:200 dilution), and actin (rabbit polyclonal antibody at 1:400 dilution) (Sigma). For Bcl-2 and actin, the same secondary antibody, peroxidase-conjugated donkey anti-rabbit IgG, was used at a 1:8,000 dilution (Jackson ImmunoResearch Laboratories, West Grove, PA); for Glut-2, a peroxidase-conjugated donkey anti-goat IgG was used at a 1:8,000 dilution (Santa Cruz Biotechnology); and for Bcl-X_L, peroxidase-linked sheep anti-mouse IgG was used at a 1:5,000 dilution (Amersham, Piscataway, NJ). Chemiluminescence was detected using the enhanced chemiluminescence system (Amersham).

Data are expressed as means ± SE. Unpaired two-tailed Student’s t test was used to determine statistical significance. Differences were considered significant at P < 0.05.

To examine whether β-cells of RIP-PTHrP mice are resistant to cell death, we studied β-cell responses to in vivo injection of STZ, a β-cell-specific cytotoxic agent. In pilot studies, normal mice injected with STZ at a single dose of 100 mg/kg were killed 6, 12, or 24 h postinjection. Pancreatic sections from these mice were stained for TUNEL to determine the extent of β-cell death. Minimal TUNEL staining was observed at the 6-h time point in normal mice, indicating that 6 h is too early to assess cell death following STZ administration at that dose. Markedly increased numbers of TUNEL-positive nuclei were observed by 12 h, and by 24 h there was a massive loss of β-cells (data not shown), making analysis of cell survival at this late time point impossible to assess. The 12-h time point was therefore chosen for subsequent measurement and comparison of β-cell death between normal and transgenic mouse pancreata.

Figure 1A shows TUNEL staining of a representative section from normal and transgenic mouse pancreata following STZ administration. Numerous TUNEL-positive nuclei, representing dying cells, were seen in the islets of normal mouse pancreata. In contrast, only a few TUNEL-positive nuclei were observed in the RIP-PTHrP mouse pancreata. This was confirmed by blinded quantitation of the percentage of TUNEL-positive β-cell nuclei from 14 normal and 15 transgenic pancreata. As shown in Fig. 1B, direct quantitation showed a sixfold decrease in β-cell death in transgenic (1.8 ± 0.3%) versus normal (11.2 ± 3.1%) mouse pancreata following exposure to STZ.

Cell death was also quantitated by costaining pancreatic sections from these mice with insulin and PI. Condensed, bright-red pyknotic nuclei representing dead β-cells were more prevalent in the islets of normal mice relative to their transgenic siblings (Fig. 2A). Quantitation of the percentage of condensed β-cell nuclei in a subset of five normal and five transgenic mouse pancreata again revealed markedly (6.4-fold) lower β-cell death in transgenic mice (2.4 ± 0.7%) compared with normal siblings (15.4 ± 4.2%) following STZ administration (Fig. 2B). These two studies independently demonstrate that compared with their normal littermates, adult RIP-PTHrP mice are markedly resistant to the cytotoxic effects of STZ.

There are several possible explanations as to why the RIP-PTHrP mice might be resistant to the cytotoxic and diabetogenic (25) effects of STZ. One explanation for the apparent resistance to STZ in adult RIP-PTHrP mice could be that they have a two- to fourfold increase in overall β-cell mass and β-cell number (8,25). Thus, RIP-PTHrP mice could in theory lose the same number of β-cells as their normal littermates and yet show a decreased percentage of cell death due to their initial advantage of an amplified β-cell mass. A twofold increase in islet mass could also result in sufficient β-cells surviving after STZ treatment to keep RIP-PTHrP mice from becoming hyperglycemic. In this scenario, individual RIP-PTHrP β-cells need not be STZ resistant. Therefore, to directly determine the effect of STZ on normal and RIP-PTHrP mice matched for islet mass and β-cell number, we took advantage of our previous findings (25). We have shown earlier that neonatal RIP-PTHrP mice at 1 week of age have islet mass and plasma glucose values very similar to those of their normal littermates (25). Therefore, litters aged 7–10 days were injected with 150 mg/kg STZ. As shown in Fig. 3A, blood glucose levels obtained 24 h after STZ injection were strikingly higher in normal mice (178.8 ± 17.6 mg/dl) compared with transgenic mice (90.5 ± 12.2 mg/dl; P = 0.0003), in spite of their islet mass being identical at that age. To directly examine and compare the cytotoxic effects of STZ in islet mass-matched normal and RIP-PTHrP neonatal mice, β-cell death was quantitated by measuring the percentage of condensed nuclei in pancreatic sections costained with insulin and PI. As observed in older animals, neonatal transgenic mice had a significantly (P = 0.004) lower rate of β-cell death (3 ± 0.3%) compared with normal siblings (4.3 ± 0.3%) following treatment with STZ (Fig. 3B). These studies confirm that individual β-cells of neonatal RIP-PTHrP mice are indeed resistant to the diabetogenic and cytotoxic effects of STZ, in a setting in which islet mass is equivalent to that of normal mice.

Another possible explanation for the resistance to STZ in RIP-PTHrP mice could be that they have a reduced ability to transport STZ from the extracellular to the intracellular compartment compared with β-cells from normal mice. It is well established that STZ is transported into β-cells through the Glut-2 transporter (29), which is also responsible for the transport of glucose. Thus, a decrease in the activity of Glut-2 would lead to diminished sensitivity to STZ as well as to glucose. A priori, this is an unlikely explanation for STZ resistance in RIP-PTHrP mice, since we have previously shown by islet perifusion studies that transgenic and normal islets have identical insulin secretory responses to glucose in vitro (8), suggesting that the Glut-2 function in these islets is equivalent.

To directly compare levels of Glut-2 expression, Glut-2 mRNA was quantitated in islets of normal and transgenic mouse pancreata using semiquantitative RT-PCR (Fig. 4A). As shown in Fig. 4B, no significant difference was observed in the level of Glut-2 mRNA when islets from six normal and six transgenic mice were examined by semiquantitative RT-PCR followed by densitometric scanning.

This result was further confirmed at the protein level by immunohistochemical staining and quantitatively by Western blot analysis using a Glut-2 antibody. Immunohistochemistry revealed a similar pattern and intensity of staining in normal and transgenic mouse pancreata (data not shown). Western blot analysis indicated similar levels of Glut-2 protein in islets of normal and transgenic mice (Fig. 4C and D). The identical responses of normal and transgenic mouse islets to glucose in vitro, as well as similar levels of mRNA and protein, indicate that the resistance to STZ in RIP-PTHrP mice is unlikely to be due to reduced transport of STZ.

Studies described thus far strongly suggest that overexpression of PTHrP in β-cells sustains their survival when challenged by STZ in vivo. To unequivocally demonstrate the ability of PTHrP overexpression to inhibit β-cell death, we examined the effect on β-cell survival in vitro of two independent inducers of cell death: STZ and nutrient (serum/glucose) deprivation. Serum starvation and glucose deprivation are well known inducers of cell death in β-cells (30). β-Cells from normal and transgenic mice were cultured either under conditions of glucose and serum deprivation or in the presence of STZ in vitro and examined for cell death by co-staining for insulin and PI. Many pyknotic β-cell nuclei were observed in cell cultures of normal islets deprived of serum and glucose (Fig. 5A, panel III) compared with occasional pyknotic nuclei seen in cell cultures of transgenic islets (Fig. 5A, panel IV). Blinded quantitation of 7–11 islet cell cultures revealed 20 ± 4.9% pyknotic nuclei in β-cells from normal mice compared with only 3.9 ± 1.1% pyknotic nuclei in β-cells from transgenic mice grown in the absence of serum and glucose (Fig. 5B). In contrast, the numbers of pyknotic nuclei in normal (1.9 ± 0.6%) and RIP-PTHrP (1.1 ± 0.4%) β-cells were low and indistinguishable when grown in serum- and glucose-replete medium (Fig. 5B). Quantitation of pyknotic β-cell nuclei in STZ-treated cultures also showed a significant reduction in the percentage of cell death in cultures of transgenic islets (6.32 ± 0.68%) compared with normal islets (10.91 ± 1.44%), as shown in Fig. 5C. These studies clearly indicate that β-cells of RIP-PTHrP mice are resistant to cell death induced by both STZ and serum and glucose deprivation in vitro, and that this protective effect is independent of islet mass.

The results thus far clearly demonstrate that PTHrP overexpression in β-cells makes them resistant to cell death. To determine whether the protective effect of PTHrP requires chronic transgenic expression of PTHrP or whether short-term exposure to PTHrP peptides per se may have a protective effect on β-cells, we starved normal islet cell cultures of serum and glucose in the presence or absence of PTHrP peptides. We have previously demonstrated that PTHrP is posttranslationally processed to give rise to several peptides, including an NH₂-terminal peptide (1-36), a midregion peptide (38-94), and a COOH-terminal peptide (107-139) (31). As shown in Fig. 6, islet cultures treated with the NH₂-terminal PTHrP peptide showed a statistically significant 40% reduction in β-cell death compared with control cultures. Neither the midregion nor the COOH-terminal peptides had any significant effect on β-cell death. These results suggest that NH₂-terminal PTHrP may be responsible at least in part for the protective effects displayed by overexpression of the PTHrP transgene (1–141) in the RIP-PTHrP mice.

PTHrP is known to inhibit or delay cell death in many other cell types (32–37), including chondrocytes (32,33) and breast cancer cells (37), where it has been shown to specifically upregulate the expression of the anti-apoptotic gene Bcl-2 (37,38). To examine whether PTHrP might alter expression of members of the Bcl gene family in β-cells, mRNA expression of four major members of the Bcl family was measured by semiquantitative RT-PCR from freshly isolated islets. Steady-state Bcl-2 mRNA was about threefold higher in islets from both lines of RIP-PTHrP transgenic mice compared with normal mice (Fig. 7A), and the increase was significant for each RIP-PTHrP mouse line when 5–6 individual samples were quantitated by densitometric scanning (Fig. 7B). Importantly, the increase in Bcl-2 mRNA appeared to be specific, since there was no change in the mRNA levels of three other members of the Bcl family, the anti-apoptotic gene Bcl-X_L and the pro-apoptotic members Bad and Bax (Fig. 7C and D).

Bcl-2 and Bcl-X_L proteins were quantitated by Western blot analysis in islets isolated from normal and transgenic mice. Surprisingly, and in contrast to the Bcl-2 mRNA findings, there was no significant difference in the level of Bcl-2 protein in the islets of normal versus transgenic mice (Fig. 8A and B). The levels of Bcl-X_L protein were also similar between normal and transgenic mouse islets (Fig. 8A and C).

The results presented in this study clearly indicate that overexpression of PTHrP in pancreatic β-cells protects them from STZ-induced cell death in vivo and in vitro, as well as from cell death induced by nutrient deprivation in vitro. Our previous studies have shown that RIP-PTHrP mice display a progressive, life-long increase in overall islet mass, and an increase in the size and number of individual islets (8,25). This results from an increased number of β-cells, but occurs without a concomitant increase in the rate of β-cell proliferation or individual cell size (25). Taken together, our previous and present results provide compelling, although indirect, evidence that PTHrP overexpression enhances islet mass in transgenic mice through a subtle inhibition or deceleration of the normal rate of β-cell death in vivo.

The purpose of this study was to directly document whether a reduction in the rate of cell death actually occurs in RIP-PTHrP islets. We used the β-cell-specific cytotoxic and diabetogenic agent STZ to induce β-cell death in normal and transgenic littermates in vivo so that we could measure and compare β-cell death in these mice. We observed a sixfold reduction in β-cell death following STZ administration in transgenic mice compared with normal mice. It could be argued that the twofold increase in islet mass in adult RIP-PTHrP mice (8) provides sufficient extra mass to “buffer” STZ-induced β-cell death. To address this issue, we demonstrated that β-cells of neonatal transgenic mice, which have an islet mass equivalent to that of normal mice, are also resistant to the cytotoxic effects of STZ. This confirms that the resistance to STZ-induced cytotoxicity is not purely a result of enhanced islet mass.

Another potential explanation for the higher tolerance of RIP-PTHrP mice to the cytotoxic effects of STZ could be a decrease in the transport of STZ into PTHrP-overexpressing β-cells. Previous glucose-stimulated insulin secretion (GSIS) data (8), as well as similar Glut-2 mRNA and protein levels in islets from transgenic and normal mice in the current study, imply that Glut-2 levels and activity are probably similar in the β-cells of these mice. Collectively, these findings support the idea that PTHrP overexpression directly induces resistance within β-cells to STZ-mediated cell death in vivo.

To unequivocally and directly demonstrate the ability of PTHrP overexpression to inhibit β-cell death, islet cell cultures from normal and transgenic mice were examined in vitro for their ability to survive under conditions of STZ cytotoxicity or nutrient deprivation. The rationale for using a second, different, cell death-inducing stimulus—serum and glucose deprivation—in addition to STZ was twofold. First, if a subtle but undetectable decrease in the sensitivity and transport of STZ into β-cells of RIP-PTHrP mice were the cause of their resistance to STZ-induced cell death, then using nutrient deprivation as a stimulus would bypass that problem. Second, nutrient deprivation induces a more physiologic form of cell death, widely believed to occur through the mitochondrial apoptotic pathway (39). Thus, if β-cells of RIP-PTHrP mice were shown to be resistant to this form of cell death, a more general protective effect of PTHrP overexpression on β-cell survival would be indicated. As hypothesized, β-cells from RIP-PTHrP mice demonstrated strikingly and significantly reduced cell death compared with islet cultures from normal littermates under conditions of both STZ treatment and serum starvation. These studies indicate that overexpression of PTHrP has the ability to inhibit β-cell death mediated via at least two different stimuli, STZ induction in vivo and in vitro and nutrient-deprivation in isolated islet cell cultures in vitro.

Pancreatic β-cells can now be added to the growing list of cell types protected from cell death induced by a variety of different stimuli through PTHrP overexpression (32–37). Interestingly, it should be noted that the anti-cell death effect of PTHrP is cell-type specific. Whereas PTHrP is protective in some cell types, it has been shown to induce cell death in other cell types (40). The cellular and signaling mechanisms through which PTHrP inhibits death in these various cell types remain unclear. In some cases, nuclear localization of PTHrP, mediated by a nuclear localization signal, is required for anti-apoptotic action (32,36,37), whereas in other cases it appears to act through an autocrine/paracrine mechanism (34,35). Protection of β-cell death by the exogenous addition of the NH₂-terminal fragment of PTHrP suggests that this region of PTHrP is important in regulating β-cell survival and that it probably acts through an autocrine/paracrine mechanism.

To begin to address the cellular mechanism through which PTHrP might inhibit β-cell death, expression of members of the Bcl family of genes was examined in islets isolated from normal and transgenic mice, initially at the mRNA and subsequently at the protein level. The rationale for considering the Bcl family of genes was as follows. First, this pathway is the major pathway involved in apoptosis mediated by serum deprivation (39). Second, the Bcl gene family is also involved in other pathways of cell death, including necrosis (41), which may be the pathway of cell death mediated by STZ (42). Third, inhibition of apoptosis by PTHrP in breast cancer cells and chondrocytes is thought to be mediated through upregulation of Bcl-2 mRNA and protein (37,38). In the current study, semiquantitative RT-PCR demonstrated an increase in Bcl-2 mRNA in both lines of transgenic mice. This upregulation of Bcl-2 mRNA appeared to be specific, since there was no change in the level of Bcl-X_L, Bad, or Bax mRNA. However, no change was seen in Bcl-2 at the protein level in RIP-PTHrP islets. Although these studies leave the increase in Bcl-2 mRNA unexplained, one must reasonably conclude that upregulation of Bcl-2 protein is not critical for PTHrP-mediated cell survival in β-cells. This is in contrast to the anti-apoptotic action of PTHrP in chondrocytes and breast cancer cells, in which Bcl-2 protein upregulation has been observed (37,38). However, these results are in accord with those of Allison et al. (43), who demonstrated that transgenic mice overexpressing Bcl-2 in pancreatic β-cells are not resistant to high doses of STZ in vivo, unlike the RIP-PTHrP mice. Thus, cellular pathways other than Bcl-2 probably mediate the anti-apoptotic effect of PTHrP overexpression in β-cells.

One such potential mechanism through which PTHrP may inhibit β-cell death is through the regulation of l-type calcium channels. Addition of NH₂-terminal PTHrP peptide has been shown to inhibit cell death in cerebellar granule neurons by downregulating the activity of this channel (34,35). Given the neuroendocrine nature of pancreatic β-cells, the electrophysiological similarities they share with neuronal cells, and the fact that NH₂-terminal PTHrP inhibits cell death in an autocrine/paracrine manner in both neuronal and pancreatic β-cells, regulation of β-cell death and survival through the voltage-sensitive calcium channels is an attractive possibility.

In summary, we have clearly shown that overexpression of PTHrP in β-cells has a protective effect on their survival. Dissecting the signaling pathways and molecular mechanisms through which PTHrP overexpression inhibits β-cell death will likely provide important information from the perspective of both PTHrP biology and understanding β-cell death and survival. Whether the protective effect of PTHrP overexpression on β-cells will prove applicable to islets in the setting of islet transplantation or type 1 diabetes is an interesting question to be addressed in the future. The recognition that PTHrP overexpression enhances β-cell survival may provide potential therapeutic targets for pharmaceutical agents aimed at improving the survival of β-cells in diabetes.

This work was supported by National Institutes of Health Grant DK 47168 and American Diabetes Association Grant 701825.

We are grateful to Dr. Simon Watkins and personnel in the imaging core facility for the use of their equipment and assistance. We would also like to thank Dr. Daniel Johnson for providing us with a Bcl-2-overexpressing cell line used as a positive control in the Western blots. We thank Vasumathi Reddy, John Rossi, and Sophia Masters for superb technical assistance. Finally, we are grateful to Dr. Robert O’Doherty for his insightful comments on the manuscript.