Original Article

Oncogene (2006) 25, 4525–4533. doi:10.1038/sj.onc.1209489; published online 27 March 2006

BH3 peptidomimetics potently activate apoptosis and demonstrate single agent efficacy in neuroblastoma

K C Goldsmith1, X Liu1, V Dam1, B T Morgan1, M Shabbout2, A Cnaan2,3, A Letai4, S J Korsmeyer5 and M D Hogarty1,3

  1. 1Division of Oncology, The Children's Hospital of Philadelphia, Philadelphia, PA, USA
  2. 2Division of Biostatistics and Epidemiology, The Children's Hospital of Philadelphia, Philadelphia, PA, USA
  3. 3Department of Pediatrics, University of Pennsylvania School of Medicine, Philadelphia, PA, USA
  4. 4Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA, USA
  5. 5Department of Cancer Immunology and AIDS, Dana-Farber Cancer Institute, Boston, MA, USA

Correspondence: Dr MD Hogarty, Division of Oncology, The Children's Hospital of Philadelphia, 9 North ARC, Suite 902C, 3615 Civic Center Boulevard, Philadelphia, PA 19104-4318, USA. E-mail: hogartym@email.chop.edu

Received 11 October 2005; Revised 1 February 2006; Accepted 2 February 2006; Published online 27 March 2006.



The major impediment to cure for many malignancies is the development of therapy resistance with resultant tumor progression. Genetic alterations leading to subversion of inherent apoptosis pathways are common themes in therapy resistance. Bcl-2 family proteins play a critical role in regulating mitochondrial apoptosis that governs chemotherapeutic effects, and defective engagement of these pathways contributes to treatment failure. We have studied the efficacy of BH3 peptidomimetics consisting of the minimal death, or BH3, domains of the proapoptotic BH3-only proteins Bid and Bad to induce apoptosis using neuroblastoma (NB) as a model system. We demonstrate that BH3 peptides, modified with an arginine homopolymer for membrane transduction (called r8-BidBH3 and r8-BadBH3, respectively), potently induce apoptosis in NB cells, including those with MYCN amplification. Cell death is caspase 9 dependent, consistent with a requirement for the intrinsic mitochondrial pathway. Substitutions at highly conserved residues within the r8-BidBH3 peptide abolish apoptotic efficacy supporting activity through specific BH domain interactions. Concomitant exposure to r8-BadBH3 and r8-BidBH3 at sublethal monotherapy doses revealed potent synergy consistent with a competitive displacement model, whereby BH3 peptides displace sequestered BH3 proteins to induce cell death. Further, BH3 peptides demonstrate antitumor efficacy in a xenograft model of NB in the absence of additional genotoxic or trophic stressors. These data provide proof of principle that targeted re-engagement of apoptosis pathways may be of therapeutic utility, and BH3-like compounds are attractive lead agents to re-establish therapy-induced apoptosis in refractory malignancies.


apoptosis, neuroblastoma, Bcl-2 homologues, programmed cell death, MYCN


BH, Bcl-2 homology domain; DISC, death-inducing signaling complex; PTD, peptide transduction domain



Cytotoxic drugs exert their antitumor effects through the activation of apoptosis, an evolutionarily conserved programmed cell death process, rather than directly through the disruption of vital metabolic functions (Dive et al., 1992). Diverse classes of anticancer agents induce discrete damage response programs that engage the apoptosis machinery, and apoptosis signaling cascades subsequently converge upon common downstream effector pathways that execute programmed cell death (Nicholson, 2000). It is not surprising, therefore, that defects in apoptotic mediators are commonly selected for in myriad solid and hematologic tumors, either de novo or during treatment, with the end result being the emergence of broad therapy resistance and treatment failure (Green and Reed, 1998). Apoptosis mediators whose deregulation is frequently selected for, and that contribute to the chemoresistant phenotype, include p53 pathway regulators (p53, p19/ARF, Mdm2) (Schmitt et al., 1999), BCL-2 family members (Bcl-2, Mcl1, BclXl, Bim and others) (Reed, 1995), caspases and their activators (Casp8, Apaf1, death receptor signaling components) (Teitz et al., 2000; Soengas et al., 2001) and inhibitors of apoptosis effectors (IAP proteins) (Hersey and Zhang, 2003). Fortunately, an improved understanding of PCD pathways, as well as the mechanisms employed by cancer cells to subvert them, has led to the emergence of novel classes of apoptosis-directed therapies that may find utility as anticancer therapeutics (Reed, 1999).

Apoptosis of cancer cells in response to cytotoxic agents requires an intact mitochondrial apoptosis pathway in most cell types, and the Bcl-2 family of proteins are central regulators of this process (Wei et al., 2001). Bcl-2 proteins contain up to four highly conserved domains ('Bcl-2 homology' or BH domains). Multidomain members predominantly exert their effects at the outer mitochondrial membrane and endoplasmic reticulum. They suppress (Bcl-2, Bcl-xl, Mcl-1) or promote (Bak, Bax) apoptosis by altering the threshold for release of apoptogenic factors from mitochondrial stores. Members containing only the BH3 domain (e.g., Bid, Bad, Noxa, Puma and others) are exclusively proapoptotic and function through interactions with the multidomain Bcl-2 members. Specific cellular stressors activate distinct BH3-only proteins to deliver a prodeath signal to the mitochondria. The prodeath BH3-only proteins may further be considered in two functionally distinct subsets (Letai et al., 2002; Kuwana et al., 2005). Some (e.g., Bid and Bim) are capable of directly activating Bak or Bax to induce their oligomerization and programmed cell death. Others (e.g., Bad, Puma, Noxa) do not bind to Bak or Bax, but bind with high affinity to prosurvival Bcl-2 proteins. Although this latter class does not directly activate Bax or Bak, they may enable apoptosis by preventing the sequestration of Bid-like death activators through competitive displacement. Thus, interactions among the pro- and antiapoptotic Bcl-2 family proteins, dictated by their abundance and relative affinities, are critical determinants of apoptosis sensitivity, and their deregulated expression is a frequent cause of enhanced cancer cell survival (Korsmeyer, 1992).

Efforts to abrogate the cancer cell survival bias engendered by alterations in these death effectors are a major focus in experimental therapeutics. One approach utilizes prodeath BH3 minimal death domains to re-establish mitochondrial sensitivity (Cosulich et al., 1997; Holinger et al., 1999; Moreau et al., 2003). The proapoptotic effects of BH3 proteins have been recapitulated by small peptides (20–25 amino acids) encompassing their unique BH3 domains (Letai et al., 2002). BH3 peptidomimetics representing select BH3-only death domains have been shown to induce cytochrome c release from isolated mitochondria, coincident with Bak and Bax oligomerization. These effects could be abolished using Bak-/- mitochondria or by Bcl-2 overexpression (Letai et al., 2002). Thus, the delivery of a minimal death domain peptide may be useful to engage apoptosis in tumor cells through Bcl-2 family targeted effects, either by activating proapoptotic members or by counteracting antiapoptotic members, consistent with genetic models of apoptosis.

To assess the relative potency of BH3 domain peptides, and as proof of principle of their potential therapeutic utility against refractory malignancies, we investigated their effects against neuroblastoma (NB)-derived cell lines in vitro and in vivo. Neuroblastoma is a solid tumor arising from undifferentiated neural crest-derived neuroblasts in the sympathetic nervous system, and a major cause of childhood cancer mortality (Matthay et al., 1999). The emergence of therapy resistance is largely responsible as many tumors demonstrate a complete or near-complete response to chemotherapy before progressing as chemoresistant tumor, despite maximal multimodality therapy (Matthay et al., 1999). Improvements in survival for this tumor are likely to require innovative treatment approaches targeting the pathways altered during malignant evolution, including those mediating programmed cell death. High-risk NBs may demonstrate alterations in caspase 8 (Teitz et al., 2000) and/or TP53 (Keshelava et al., 2001; Tweddle et al., 2001). As BH3-only proteins engage mitochondrial apoptosis downstream of these defects, we hypothesized that BH3 peptidomimetics might prove an effective therapeutic against this tumor. We demonstrate that BH3 peptides effectively induce apoptosis in highly refractory NB cell lines (including those with MYCN amplification, which correlates with highly aggressive tumor behavior), synergize when given in combination with other BH3-only peptides and induce programmed cell death both in vitro and in vivo via the genetically defined intrinsic, or mitochondrial, death pathway.



r8-modified BH3 peptides maintain their alpha-helicity

BH3 domains exist as amphipathic alpha-helices within native BH3-only proteins. The r8-BidBH3 and r8-BadBH3 peptides include the BH3 domains of human Bid or Bad modified with a D-isomer homopolymer of eight arginine residues (Wadia and Dowdy, 2002) followed by a glycine linker to function as a peptide transduction domain (PTD). To control for nonspecific membrane disruptive effects of the arginine PTD sequence, a control peptide was synthesized identical to the r8-BidBH3 peptide except for substitutions at two highly conserved residues predicted to abolish Bid functional interactions (called r8-BidBH3alt; Table 1). To ensure that these peptides were capable of adopting a near-native conformation despite these alterations, we determined their potential alpha-helicity. Circular dichroism (CD) spectra were obtained for each peptide and demonstrated that significant alpha-helicity was maintained despite their truncated form and the PTD modification (Table 1). The similar degree of alpha-helicity in the r8-BidBH3alt peptide with respect to the native r8-BidBH3 peptide ensured that the control peptide did not differ from the test peptides in conformation alone.

r8-BH3 peptides induce apoptosis in neuroblastoma-derived cell lines in vitro

Cells were exposed to r8-modified BH3 peptides in media with 10% fetal bovine serum (FBS) during logarithmic growth phase. Significant cytotoxicity (by morphologic alterations consistent with apoptosis as well as viability assay) was demonstrable in all cell lines studied with the r8-BidBH3 or r8-BadBH3 at 20 or 50 muM concentrations (Figure 1 and data not shown). The four highly aggressive MYCN amplified cell lines were most sensitive to the active BH3 peptides with demonstrable cytotoxicity at 10 muM and >80% reduction in viable cell mass after 2 days of exposure at 20 muM (Figure 1). The three MYCN non-amplified cell lines were moderately less sensitive by comparison, and cytotoxicity was readily apparent at concentrations of 20 muM r8-BidBH3 or r8-BadBH3 in two of the three cell lines. Of note, the only caspase 8-expressing cell line investigated, SK-N-AS, was the least sensitive (<40% decrease in cell mass at 20 muM peptide exposure). Still, morphologic evidence for apoptosis was demonstrated even at 20 muM of BH3 peptide exposure for this cell line (and confirmed with Annexin-V and caspase activity assays; Figures 2a and 3c).

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

(a) Cytotoxicity assay: three MYCN single-copy and four MYCN amplified neuroblastoma-derived cell lines were studied. Peptides were added to subconfluent cells at increasing concentrations (x axis) and assessed by MTT at 48 h. Values plotted are the viable cell number (biomass value by MTT assay) normalized to vehicle-treated cells (y axis). (filled square): r8-BidBH3; (filled circle): r8-BadBH3; (square): r8-BidBH3alt. *Caspase-8-expressing cell line (remainder are Casp8 null). Standard error at all data points is <5%. (b) Peptides were added to seeded bone marrow stromal cells to assess for effects on a nonmalignant, non-transformed cell line.

Full figure and legend (32K)

Figure 2.
Figure 2 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

(a) Assessment of BH3 peptide-induced apoptosis in neuroblastoma-derived cell lines using Annexin-V-EGFP fluorescence. Cells were exposed to the indicated peptide and concentration (x axis), vehicle or no treatment for 8 h. Apoptosis is calculated as the percent of Annexin-V-EGFP surface fluorescence-positive cells to total cells counted (<200 cells in all cases). Results of a single representative experiment are shown. Dark gray bars: r8-BidBH3; hatched bars: r8-BadBH3; open bars: r8-BidBH3alt. (b) Morphology of SK-N-AS cells exposed to r8-BidBH3 (50 muM) and caspase inhibitors or vehicle, at an early time point (8 h). Morphologic signs of apoptosis (cells are refractile with evident blebbing, increased nuclear condensation and apoptotic body formation) are demonstrated when pre-exposed to vehicle or caspase 8 inhibitor alone. In contrast, cells remained viable (with minimal toxic vacuolization but absent apoptotic body formation) when pre-exposed to either caspase 9 inhibitor or general caspase inhibitor through 48 h of exposure (data not shown).

Full figure and legend (159K)

Figure 3.
Figure 3 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Synergistic activity of BH3 peptides. (a) Fluorescence micrographs of SK-N-AS cells exposed to r8-BidBH3, r8-BadBH3 or both peptides at sublethal monotherapy concentrations. Cell number was assessed using the Hoechst 33342 cell-permeable nuclear counterstain and Annexin-V-EGFP was used to detect membrane alterations associated with apoptosis. Representative images for each condition are shown. (b) NGP or SK-N-AS cells were exposed to vehicle or peptides at the indicated concentrations (x axis). Percent apoptosis is shown as the ratio of Annexin-V-EGFP-positive cells to total cells counted (<200 cells in all cases). (c) Synergistic activity of r8-BidBH3 and r8-BadBH3 is demonstrated with caspase activation assays using sublethal concentrations of r8-BH3 peptides and combinations (x axis).

Full figure and legend (145K)

In comparison, no cytotoxic effect was apparent with the control r8-BidBH3alt peptide through 50 muM concentrations, although there was a modest reduction in viable cell number at higher concentrations (Figure 1). This may be attributable to membrane disruptive effects from the transduction sequence on cell proliferation, as this motif inserts through lipid matrices and might therefore interfere with membrane functions at higher concentrations. In support of this, no apoptosis was demonstrable at concentrations up to 50 muM with the control peptide using AnnexinV-EGFP (enhanced green fluorescent protein) or caspase activation assays to assess cell death (Figures 2a and 3c). To assess for toxicity of peptides on nonmalignant 'normal' cells, human bone marrow stromal cells (similar in behavior to tissue fibroblasts) were exposed to r8-modified BH3 peptides under identical tissue culture conditions. Minimal to no cytotoxic effects were seen at concentrations lethal to the NB cells (Figure 1b).

To demonstrate that NB cytotoxicity was owing to peptide-mediated engagement of programmed cell death, we performed immunofluorescence microscopy with Annexin-V-EGFP, which labels externalized membrane phosphatidylserine occurring during apoptotic cell death. Both r8-BidBH3 and r8-BadBH3 demonstrated a dose-dependent induction of apoptosis in all four cell lines studied (Figure 2a). Of note, SK-N-AS was least sensitive to BH3 peptides at lower concentrations (10 muM), consistent with viability assay results, but this relative insensitivity could be overcome at greater than or equal to20 muM of r8-BidBH3 (r8-BadBH3 activity was only seen at 50 muM). In no cell line was significant apoptosis induced by the r8-BidBH3alt control peptide (even at the 50 muM concentration) or vehicle, supporting a targeted effect for the r8-BidBH3 peptide.

To confirm that apoptosis was proceeding through the mitochondrial pathway, we exposed SK-N-AS cells to r8-BidBH3 peptide following incubation with caspase inhibitors. SK-N-AS cells express caspase 8 as well as the required apoptosome components, Apaf1 and caspase 9 (data not shown). Cells were exposed to 50 muM of either z-vad-fmk (general caspase inhibitor), z-lehd-fmk (caspase 9 inhibitor), z-ietd-fmk (caspase 8 inhibitor) or vehicle before co-exposure with 50 muM r8-BidBH3 and cells were assessed morphologically. SK-N-AS cells pre-exposed to vehicle exhibited marked apoptosis at early time points when treated with r8-BidBH3 (Figure 2b). Cells became denser and more refractile (with apoptotic body formation) before losing substrate anchorage, depleting the well of viable cells in <12 h (data not shown). Incubation with the caspase 8 inhibitor, z-ietd-fmk, did not inhibit this process. In contrast, both the pancaspase inhibitor z-vad-fmk and the selective caspase 9 inhibitor z-lehd-fmk were capable of significantly inhibiting r8-BidBH3-induced cell death. Viable cells persisted through the 48 h of observation and wells remained near-confluent under both these conditions. No evidence of cytotoxicity was seen in cells exposed to any of the caspase inhibitors alone (data not shown). These data support r8-BidBH3-induced apoptosis as a caspase-dependent process requiring caspase 9 participation, but not caspase 8, consistent with a mechanism of action through formation of an active apoptosome, and independent of upstream signaling via caspase 8 activation.

r8-BidBH3 and r8-BadBH3 act synergistically to induce apoptosis in vitro

We tested competitive displacement of Bid with Bad by co-exposing SK-N-AS and NGP cells with sublethal monotherapy doses of r8-BidBH3 and r8-BadBH3 peptides, alone and in combination. SK-N-AS showed no apoptosis in response to 5 muM r8-BidBH3 alone and minimal effect at 10 muM (<10% apoptotic cells) (Figure 3). Likewise, r8-BadBH3 did not induce apoptosis at 10 muM when given alone. However, the combination of sublethal monotherapy concentrations of r8-BadBH3 (10 muM) with r8-BidBH3 (5 or 10 muM) induced apoptosis in 25 and 51% of SK-N-AS cells, respectively. Similarly, the NGP cell line was not sensitive to 5 muM r8-BidBH3 and only moderately sensitive at 10 muM exposure. r8-BadBH3 had an intermediate effect at 10 muM. Co-exposure to r8-BidBH3 and r8-BadBH3 demonstrated synergy with apoptosis induced in 64 and 80% of cells, respectively, at the 5 and 10 muM r8-BidBH3 concentration. These effects are markedly more than would be predicted based on additive BH3 peptide stoichiometry alone and instead support the competitive displacement supposition whereby select prosurvival Bcl-2 docking sites can be blocked by Bad peptide allowing Bid access to prodeath Bak or Bax, thereby potentiating its efficacy.

To provide confirmatory evidence for synergistic apoptosis engagement, we determined caspase activation following peptide exposure. SK-N-AS cells were exposed to vehicle or r8-BidBH3, r8-BadBH3 or r8-BidBH3alt at a range of concentrations over a 2-h period (Figure 3c). Despite the inability of either 10 muM r8-BidBH3 or 10 muM r8-BadBH3 to activate caspases above baseline when used alone (by comparison with controls), the combination of both peptides at 10 muM led to a potent induction of caspases consistent with synergy. The in vitro caspase activation curves using NGP were similar to those shown in Figure 3C for SK-NAS (also with no caspase activation following treatment with the r8-BIDBH3alt peptide) as also demonstrated with Annexin-V-EGFP staining in Figure 2.

r8-BH3 peptides induce apoptosis in an in vivo xenograft model of neuroblastoma

Athymic nu/nu mice bearing NGP or SK-N-AS tumors (both are highly tumorigenic) were treated with 50 mul of 100 muM r8-BidBH3, r8-BadBH3, r8-BidBH3alt, vehicle or both 50 muM r8-BidBH3 and 50 muM r8-BadBH3 combined, via intratumoral injection on days 0, 2 and 4 only. This design assesses the ability of these peptides to induce cell death in tumor cells growing within a tissue architecture that recapitulates prosurvival cell–cell contacts (homotypic and heterotypic) available in situ (Sutherland, 1988; Jacks and Weinberg, 2002). We assessed a minimalist dosing scheme to provide a more rigorous test of peptide potency, alone and in combination.

Table 2 shows the mean tumor volume and 95% confidence interval at each time point for all treatment groups and Figure 4 shows the volume over time for each tumor. SK-N-AS xenografts demonstrated significant BH3 peptide sensitivity in vivo. No tumors were 'cured' with this brief therapeutic exposure, although four tumors had measurable regression (reduced tumor volume for greater than or equal totwo consecutive measurements): three in the combined r8-BidBH3/r8-BadBH3 group (three of six, 50%) and one in the r8-BidBH3 group (one of six, 17%). No tumors receiving vehicle, r8-BadBH3 alone or r8-BidBH3alt showed regression. Overall, there was a statistically significant antitumor response following brief exposures to r8-BidBH3 alone and r8-BidBH3/r8-BadBH3 combined (P=0.02 and 0.04, respectively), when compared to vehicle. In addition, r8-BadBH3 peptide monotherapy demonstrated a trend toward an antitumor response (P=0.06). The functionally deficient r8-BidBH3alt did not show antitumor efficacy (P=0.13).

Figure 4.
Figure 4 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Neuroblastoma xenograft growth curves. (a) SK-N-AS. (b) NGP cell lines. N=6 mice bearing established xenografts in each treatment group; day 0 is the day of treatment initiation (d0, 2 and 4 via intratumoral injection). P-value is shown for each treatment group and calculated by comparison to vehicle-treated tumors.

Full figure and legend (52K)

NGP xenografts grow more slowly by comparison, despite having genomic amplification of the MYCN proto-oncogene, and there was more variation in tumor size over time in all groups. A significant proportion of xenografted tumors (12 of 30; 40%) had measurable regression by day 5 of treatment: four receiving r8-BidBH3/r8-BadBH3 (67%), three receiving r8-BidBH3 alone (50%), two each receiving r8-BadBH3 alone or r8-BidBH3alt (33%) and one receiving vehicle (17%). In addition, five tumors demonstrated lack of tumor growth or 'dormancy' through >28 days of observation: two each in the r8-BidBH3/r8-BadBH3 and r8-BadBH3 groups and one in the r8-BidBH3 group (all tumors in the vehicle or r8-BidBH3alt groups demonstrated continuous growth over time). Comparing tumor growth in peptide-treated xenografts versus vehicle, no statistically significant impact of therapy was demonstrable with NGP xenografts owing to the increased variability in tumor kinetics in this model and the limited mouse numbers. There was a statistically significant difference in tumor growth comparing r8-BidBH3/r8-BadBH3-treated xenografts with those receiving r8-BidBH3alt control peptides (P<0.01).



Cancer is a complex disease defined by a heterogeneous collective of genetic lesions that maintain tumorigenesis. Multiple pressures favor the acquisition of apoptosis suppression in this process. Indeed, antineoplastics exert their effects through the activation of programmed cell death, so selective pressure for apoptosis suppression is intensified during therapy. Cancer cells are more dependent on subversion of apoptosis than normal cells, providing a therapeutic index for agents that restore sensitivity to apoptotic stimuli, and many such agents are in development (Reed, 2003). These include several that target Bcl-2 family interactions using small molecule inhibitors, native or synthetic mimetics, and antisense compounds. We have investigated minimal death domains from representative BH3-only proteins for their capacity to induce apoptosis in highly malignant NB-derived cell lines both in vitro and in vivo.

Neuroblastoma presents a therapeutic challenge with survival <40% by 5 years in high-risk disease (Matthay et al., 1999). The emergence of therapy resistance is the most significant obstacle to cure, and subversion of programmed cell death pathways plays a predominant role. Tumors frequently demonstrate caspase 8 silencing and defective death receptor-induced apoptosis (Teitz et al., 2000). Thus, a subset of NBs harbor intrinsic defects in apical caspase activation that impede death signaling upstream of the mitochondria. Additionally, a subset of NBs have acquired TP53 dysfunction at the time of relapse, posited to occur through aberrant cytoplasmic p53 sequestration (Ostermeyer et al., 1996; Nikolaev et al., 2003), TWIST1-mediated repression (Valsesia-Wittmann et al., 2004) or TP53 mutation following genotoxic therapy (Keshelava et al., 2001; Tweddle et al., 2001). This functional deficit results in an inability to transactivate Noxa and Puma, the BH3 domain proteins largely responsible for the execution of p53-mediated apoptosis (Villunger et al., 2003). Therefore, an inability to deliver an appropriate BH3 protein burden to mitochondrial effectors contributes to chemoresistance in NBs.

Synthetic BH3 domain peptides of proapoptotic Bid and Bad deliver a prodeath signal to mitochondrial death effectors. Unlike their full-length endogenous counterparts, these peptides do not require activation (native Bid requires proteolytic processing and N-myristoylation), nor are they sequestered (Bad is phosphorylated and sequestered by cytosolic 14-3-3 proteins). Instead, they are untethered from their regulatory domains to interact with Bcl-2 homologues. Therefore, it has been hypothesized that BH3 peptides may circumvent the cancer cell survival bias engendered by tumor-specific alterations in death effector machinery. We have shown that r8-BidBH3 and r8-BadBH3 specifically and potently engage apoptosis in NB-derived cell lines as single agents without concomitant genotoxic or trophic stressors. Apoptosis was induced in all cell lines tested including all four with MYCN amplification, a harbinger of therapy-resistant disease. Normal bone marrow stromal cells were largely resistant to cell death at the concentrations studied. The inefficacy of r8-BidBH3alt strongly supports that potency was not the result of off-target disruption of membrane function, but results from specific protein–protein interactions mediated by the BH3 helices themselves. Programmed cell death required caspase 9, but not caspase 8, consistent with with a requirement for apoptosome formation and death induction downstream of the DISC. Interestingly, only one of the seven cell lines studied herein expresses caspase 8 (SK-N-AS), which had reduced BH3 peptide sensitivity in vitro. It is plausible that SK-N-AS has defects in apoptosis pathways downstream of caspase 8, counteracting the effects of an exogenously delivered BH3 burden. We are currently investigating the relative expression levels of Bcl-2 members and IAP proteins to correlate with BH3 peptide responses in tumor cells.

Our in vivo studies similarly demonstrate single agent activity. Particularly interesting is the response of the SK-N-AS cell line that traditionally is highly refractory to in vivo growth inhibition. Brief exposure of established tumors to peptide over 5 days led to an antitumor effect without overt toxicities. Using NGP xenografts, no tumor inhibition was achieved using our a priori defined end points, yet secondary end points suggest an in vivo antitumor effect as well. Of particular interest, in no control or vehicle-treated tumor was a growth plateau or dormancy-like tumor state achieved. Yet five of 18 (28%) NGP xenografts treated with active peptides (either r8-BidBH3, r8-BadBH3 or both) remained latent with no notable growth over the >28 days of tumor observation, despite a >95% rate of tumor establishment with exponential growth in this model, and represented a dramatic response in our experience.

Although r8-BidBH3 was more potent than r8-BadBH3, likely attributable to its capacity for direct Bak or Bax activation (Kuwana et al., 2005), r8-BadBH3 demonstrated efficacy as well. We propose that r8-BadBH3, which has not been shown to directly engage Bak or Bax, induces apoptosis through displacement of endogenously activated BH3 proteins that are being actively sequestered by Bcl-2 homologues in cancer cells. This hypothesis is supported by the demonstration that r8-BadBH3 at low concentrations markedly sensitizes NB cells to programmed cell death induced by sublethal concentrations of r8-BidBH3, and that r8-BadBH3 alone at higher concentrations is capable of activating programmed cell death in these cells. Further support derives from studies of a MYC-induced leukemia model with conditional Bcl-2 expression (Letai et al., 2004) in which BadBH3 induces cytochrome c release from mitochondria derived from Bcl-2-dependent leukemic blasts but not from pro-B-lineage FL5.12 mitochondria. This suggests that Bcl-2 is sequestering activated BH3 proteins in this model as well.

Further investigation is warranted to determine whether alternative BH3 peptides (from native BH3 proteins, such as Puma or Noxa, or synthetic peptides with improved affinity profiles) may have increased potency over prototypical Bid- or Bad-derived peptides by extending their affinities for prosurvival multidomain members (Chen et al., 2005; Kuwana et al., 2005). BH3 therapeutics need not restore the precise death insult abrogated in a cancer cell, but instead may drive a cell past its apoptotic threshold by delivering an analogous or even redundant prodeath BH3 signal. Although peptides represent an intriguing pharmaceutical class, they also have potential drawbacks including limited serum stability and cellular penetrance. Newer means of peptide delivery and stabilization through chemical modification may prove these obstacles surmountable. Second-generation BH3 peptides have been engineered using olefin chemistry to create a hydrocarbon staple to stabilize the amphipathic alpha-helix while maintaining binding specificity (Walensky et al., 2004). These peptides demonstrate improved kinetics, serum stability and potency in an in vivo model of leukemia (Walensky et al., 2004). These studies serve as proof of principle that targeted peptide mimetics can engage apoptotic pathways in cancer cells that serve as promising lead compounds and encourage the further development of such targeted prodeath therapeutics.


Materials and methods

Cell lines

Neuroblastoma cell lines with MYCN amplification (IMR5 (Tumilowicz et al., 1970), SMS-KAN (Reynolds et al., 1986), NGP (Schwab et al., 1983) and CHP-134 (Schlesinger et al., 1976)) and without (NB-69 (Feder and Gilbert, 1983), SK-N-SH (Biedler and Spengler, 1976) and SK-N-AS (Schmechel et al., 1978)) were used. Bone marrow stromal cells (Cooperman et al., 2004) were used as normal control cells. Cells were grown in RPMI media 1640 (Life Technologies) supplemented with 10% FBS, 2 mM L-glutamine, 1% OPI, 100 U/ml of penicillin and 100 mug/ml gentamicin. Tissue culture was at 37°C in a humidified atmosphere of 5% CO2 as described previously (Tajiri et al., 2003).

Peptide synthesis

Peptides were synthesized by the University of Pennsylvania Protein Chemistry Laboratory using solid-phase Fmoc chemistry and purified by HPLC (<95% purity) (sequences in Table 1). To enhance protease resistance and improve stability, the arginine (r8) membrane transduction sequence was synthesized in D-conformation, and the N-terminus and C-terminus were blocked in all peptides by an acetyl and amide group, respectively. Stock solutions were 6 mM in DMSO, and all working dilutions contained less than or equal to0.8% DMSO. As a vehicle control, 1% DMSO in PBS was used.

Circular dichroism

r8-modified BH3 domain peptides were dissolved at a final concentration of 40 muM in 50 mM potassium phosphate buffer, with and without 30% 2,2,2-trifluoroethanol (TFE), in a 1 mm quartz cuvette. Circular dichroism spectra were obtained on a JASCO J-810 spectrapolorimeter, and were recorded from 190 to 300 nm wavelength with a step resolution of 1 nm and a scan speed of 50 nm/min. Temperature was maintained at 25°C, and six spectra were measured and averaged for each peptide studied. The raw data in millidegrees were converted to molar ellipticity by the following equation: theta=(thetaobs)105/(NL(protein (muM))), where thetaobs=measured signal in millidegrees, N=number of amino-acid residues, L=path length of the cell and (protein)=protein concentration in muM. The molar ellipticity at 222 nm in the presence and absence of TFE was used to define the percent alpha-helicity.

Cytotoxicity assay

Cells were seeded into 96-well microplates at approx2 times 104 cells per well and treated with peptide (10, 20 or 50 muM of r8-BidBH3, r8-BadBH3, r8-BidBH3alt or 1% DMSO as vehicle). The MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay was performed according to the manufacturer's protocol on days 0, 2 and 5 (Sigma-Aldrich, St Louis, MO, USA). Cell biomass, proportional to viable cell number, was obtained measuring MTT absorbance on a Bio-Rad Benchmark Plus microplate reader at 570 nm. All experiments were performed with at least three wells per condition and replicated at least once.

Apoptosis assays

Adherent cells were grown on a multichamber slide (Lab-Tek II; Nalge Nunc International) and exposed to peptide or vehicle for 5 h. The peptide was then removed and replaced with fresh RPMI 1640 media with 10% FBS. After 24 h, apoptotic cells were stained with Annexin-V-EGFP (BD Biosciences Clontech, Palo Alto, CA, USA) and counterstained with the nuclear fluorochrome Hoechst 33342 (Molecular Probes Inc.). Death was quantified by flourescence microscopy. Cells were counted with the Hoechst setting (excitation 350 nm, emission 460 nm) and apoptotic cells quantified according to Annexin-V-EGFP membrane positivity (FITC filter set). The percentage of apoptotic cells was determined from >200 cells for each test condition.

To assess caspase activation, approximately 2 times 104 cells were plated per well in triplicate in 96-well microplates and exposed to peptide for 2, 4, 6 and 8 h. Caspase activity was then detected using the Homogeneous Caspase Assay (Roche Applied Science, Mannheim, Germany) according to the manufacturer's protocol. In brief, the activity of caspases 2, 3, 6, 7, 8, 9 and 10 was determined measuring rhodamine fluorescence emitted when the quenching substrate bound to it is cleaved by activated caspases, and read with a Wallac 1420 VICTOR2 luminometer (Perkin-Elmer) (excitation 480 nm, emission 510 nm). The average of three wells for each experimental condition was determined.

To determine the apoptotic pathway of primary importance, studies were performed with the SK-N-AS cell line (caspase 8, caspase 9 and apaf1 expressing) and various inhibitors of caspase activity. Cells were plated in serum replete media and exposed to the cell-permeable caspase inhibitors (R&D Systems, Minneapolis, MN, USA) z-vad-fmk (general caspase inhibitor), z-ietd-fmk (caspase 8 inhibitor), z-lehd-fmk (caspase 9 inhibitor) (all at 50 muM) or 1% DMSO (vehicle) for 1 h before addition of 50 muM r8-BidBH3 peptide. Cells were incubated and assessed for morphologic changes indicative of apoptosis. Images were obtained with a digital camera and microscope apparatus (Canon Powershot G5 and Zeiss Axiovert 40C).

Mouse xenograft model

Six-week-old female athymic nu/nu mice from the Jackson Laboratory received subcutaneous flank injections of 106 NB cells (NGP or SK-N-AS) suspended in 50 muL of matrigel matrix (BD Biosciences). Tumor take is >95% with this model. When established tumors reached approx150 mm3, the mice were randomized for treatment (N=6 mice each) with 50 mul intratumoral injections of 100 muM r8-BidBH3, r8-BadBH3 or r8-BidBH3alt; 50 mul of 1% DMSO in PBS (vehicle); or 50 mul of r8-BidBH3 (50 muM) combined with r8-BadBH3 (50 muM). Treatments were given on days 0, 2 and 4. Tumor growth was assessed 2–3 times per week by caliper measurement by a single technician with volume calculated using the formula (lwh)*0.52 mm3. Animals were euthanized before symptoms of tumor burden (generally at a tumor volume of approx3000 mm3). All studies were approved by the Institutional Animal Care and Utilization Committee (IACUC) at The Children's Hospital of Philadelphia.


Mixed effects models with repeated measurements were used to analyse xenograft data for the treatment groups over time. We modeled the covariance structure within mice, using the autoregressive order one structure. Quadratic models were fitted for 'time' and 'tumor volume' and modeled as a polynomial function of time for the best fit. In these models, we examined time2 by treatment–group interaction. The significance of each of the independent variables was evaluated by its F-statistic overall. Contrast statements were used to make pairwise comparisons among the five treatment groups over time.



  1. Biedler JL, Spengler BA. (1976). J Natl Cancer Inst 57: 683–695. | PubMed | ChemPort |
  2. Chen L, Willis SN, Wei A, Smith BJ, Fletcher JI, Hinds MG et al. (2005). Mol Cell 17: 393–403. | Article | PubMed | ISI | ChemPort |
  3. Cooperman J, Neely R, Teachey DT, Grupp S, Choi JK. (2004). Stem Cells 22: 1111–1120. | Article | PubMed | ISI |
  4. Cosulich SC, Worrall V, Hedge PJ, Green S, Clarke PR. (1997). Curr Biol 7: 913–920. | Article | PubMed | ISI | ChemPort |
  5. Dive C, Evans CA, Whetton AD. (1992). Semin Cancer Biol 3: 417–427. | PubMed | ChemPort |
  6. Feder MK, Gilbert F. (1983). J Natl Cancer Inst 70: 1051–1056. | PubMed | ChemPort |
  7. Green DR, Reed JC. (1998). Science 281: 1309–1312. | Article | PubMed | ISI | ChemPort |
  8. Hersey P, Zhang XD. (2003). J Cell Physiol 196: 9–18. | Article | PubMed | ISI | ChemPort |
  9. Holinger EP, Chittenden T, Lutz RJ. (1999). J Biol Chem 274: 13298–13304. | Article | PubMed | ISI | ChemPort |
  10. Jacks T, Weinberg RA. (2002). Cell 111: 923–925. | Article | PubMed | ISI | ChemPort |
  11. Keshelava N, Zuo JJ, Chen P, Waidyaratne SN, Luna MC, Gomer CJ et al. (2001). Cancer Res 61: 6185–6193. | PubMed | ISI | ChemPort |
  12. Korsmeyer SJ. (1992). Cancer Surv 15: 105–118. | PubMed | ISI | ChemPort |
  13. Kuwana T, Bouchier-Hayes L, Chipuk JE, Bonzon C, Sullivan BA, Green DR et al. (2005). Mol Cell 17: 525–535. | Article | PubMed | ISI | ChemPort |
  14. Letai A, Bassik MC, Walensky LD, Sorcinelli MD, Weiler S, Korsmeyer SJ. (2002). Cancer Cell 2: 183–192. | Article | PubMed | ISI | ChemPort |
  15. Letai A, Sorcinelli MD, Beard C, Korsmeyer SJ. (2004). Cancer Cell 6: 241–249. | Article | PubMed | ISI | ChemPort |
  16. Matthay KK, Villablanca JG, Seeger RC, Stram DO, Harris RE, Ramsay NK et al. (1999). New Engl J Med 341: 1165–1173. | Article | PubMed | ISI | ChemPort |
  17. Moreau C, Cartron PF, Hunt A, Meflah K, Green DR, Evan G et al. (2003). J Biol Chem 278: 19426–19435. | Article | PubMed | ISI | ChemPort |
  18. Nicholson DW. (2000). Nature 407: 810–816. | Article | PubMed | ISI | ChemPort |
  19. Nikolaev AY, Li M, Puskas N, Qin J, Gu W. (2003). Cell 112: 29–40. | Article | PubMed | ISI | ChemPort |
  20. Ostermeyer AG, Runko E, Winkfield B, Ahn B, Moll UM. (1996). Proc Natl Acad Sci USA 93: 15190–15194. | Article | PubMed | ChemPort |
  21. Reed JC. (1995). Curr Opin Oncol 7: 541–546. | PubMed | ChemPort |
  22. Reed JC. (1999). J Clin Oncol 17: 2941–2953. | PubMed | ISI | ChemPort |
  23. Reed JC. (2003). Cancer Cell 3: 17–22. | Article | PubMed | ISI | ChemPort |
  24. Reynolds CP, Biedler JL, Spengler BA, Reynolds DA, Ross RA, Frenkel EP. (1986). J Natl Cancer Inst 76: 375–387. | PubMed | ChemPort |
  25. Schlesinger HR, Gerson JM, Moorhead PS, Maguire H, Hummeler K. (1976). Cancer Res 36: 3094–3100. | PubMed | ISI | ChemPort |
  26. Schmechel D, Marangos PJ, Brightman M. (1978). Nature 276: 834–836. | Article | PubMed | ISI | ChemPort |
  27. Schmitt CA, McCurrach ME, de_Stanchina E, Wallace_Brodeur RR, Lowe SW. (1999). Gene Dev 13: 2670–2677. | Article | PubMed | ISI | ChemPort |
  28. Schwab M, Alitalo K, Klempnauer KH, Varmus HE, Bishop JM, Gilbert F et al. (1983). Nature 305: 245–248. | Article | PubMed | ISI | ChemPort |
  29. Soengas MS, Capodieci P, Polsky D, Mora J, Esteller M, Opitz-Araya X et al. (2001). Nature 409: 207–211. | Article | PubMed | ISI | ChemPort |
  30. Sutherland RM. (1988). Science 240: 177–184. | PubMed | ISI | ChemPort |
  31. Tajiri T, Liu X, Thompson PM, Tanaka S, Suita S, Zhao H et al. (2003). Clin Cancer Res 9: 3345–3355. | PubMed | ISI | ChemPort |
  32. Teitz T, Wei T, Valentine MB, Vanin EF, Grenet J, Valentine VA et al. (2000). Nat Med 6: 529–535. | Article | PubMed | ISI | ChemPort |
  33. Tumilowicz JJ, Nichols WW, Cholon JJ, Greene AE. (1970). Cancer Res 30: 2110–2118. | PubMed | ISI | ChemPort |
  34. Tweddle DA, Malcolm AJ, Bown N, Pearson AD, Lunec J. (2001). Cancer Res 61: 8–13. | PubMed | ISI | ChemPort |
  35. Valsesia-Wittmann S, Magdeleine M, Dupasquier S, Garin E, Jallas AC, Combaret V et al. (2004). Cancer Cell 6: 625–630. | Article | PubMed | ISI | ChemPort |
  36. Villunger A, Michalak EM, Coultas L, Mullauer F, Bock G, Ausserlechner MJ et al. (2003). Science 302: 1036–1038. | Article | PubMed | ISI | ChemPort |
  37. Wadia JS, Dowdy SF. (2002). Curr Opin Biotechnol 13: 52–56. | Article | PubMed | ISI | ChemPort |
  38. Walensky LD, Kung AL, Escher I, Malia TJ, Barbuto S, Wright RD et al. (2004). Science 305: 1466–1470. | Article | PubMed | ISI | ChemPort |
  39. Wei MC, Zong WX, Cheng EH, Lindsten T, Panoutsakopoulou V, Ross AJ et al. (2001). Science 292: 727–730. | Article | PubMed | ISI | ChemPort |


This work is dedicated to the memory of Stanley J Korsmeyer. We thank Al Knudson, Garrett Brodeur, Audrey Evans and John Maris for manuscript review and commentary, Michael Rosenblatt for assistance with circular dichroism analyses and Rosalind Barr and Agueda Lin for technical assistance. This work was supported by NIH CA97323, a Career Development Award in the Biomedical Sciences from the Burroughs Wellcome Fund, and the Richard and Sheila Sanford Chair in Pediatric Oncology (M.D.H.); and The Caitlin Robb Foundation (K.C.G.).