Cytotoxic Effects of Dynorphins through Nonopioid Intracellular Mechanisms

Experimental Cell Research 269, 54 – 63 (2001) doi:10.1006/excr.2001.5309, available online at http://www.idealibrary.com on

Cytotoxic Effects of Dynorphins through Nonopioid Intracellular Mechanisms Koichi Tan-No, 1 Gvido Cebers, Tatjana Yakovleva, Bee Hoon Goh, Irina Gileva, 2 Kyrill Reznikov, Miguel Aguilar-Santelises,* Kurt F. Hauser,† Lars Terenius, and Georgy Bakalkin 3 Section of Alcohol and Drug Addiction Research, Department of Clinical Neuroscience, and *Department of Hematology, Karolinska Institute, S-171 76 Stockholm, Sweden; and †Department of Anatomy and Neurobiology, University of Kentucky College of Medicine, 800 Rose Street, Lexington, Kentucky 40536-0298

tumor cells, which express prodynorphin at high levels. © 2001 Academic Press Key Words: big dynorphin; cell death; apoptosis; p53.

Dynorphin A, a prodynorphin-derived peptide, is able to induce neurological dysfunction and neuronal death. To study dynorphin cytotoxicity in vitro, prodynorphin-derived peptides were added into the culture medium of nonneuronal and neuronal cells or delivered into these cells by lipofection or electroporation. Cells were unaffected by extracellular exposure when peptides were added to the medium. In contrast, the number of viable cells was significantly reduced when dynorphin A or “big dynorphin,” consisting of dynorphins A and B, was transfected into cells. Big dynorphin was more potent than dynorphin A, whereas dynorphin B; dynorphin B-29; [Arg 11,13]dynorphin A(1–13)-Gly-NH-(CH 2) 5-NH 2, a selective ␬-opioid receptor agonist; and poly-L-lysine, a basic peptide more positively charged than big dynorphin, failed to affect cell viability. The opioid antagonist naloxone did not prevent big dynorphin cytotoxicity. Thus, the toxic effects were structure selective but not mediated through opioid receptors. When big dynorphin was delivered into cells by lipofection, it became localized predominantly in the cytoplasm and not in the nuclei. Big dynorphin appeared to induce toxicity through an apoptotic mechanism that may involve synergistic interactions with the p53 tumor-suppressor protein. It is proposed that big dynorphin induces cell death by virtue of its net positive charge and clusters of basic amino acids that mimic (and thereby perhaps interfere with) basic domains involved in protein–protein interactions. These effects may be relevant for a pathophysiological role of dynorphins in the brain and spinal cord and for control of death of

INTRODUCTION

Dynorphin A, dynorphin B, and ␣-neoendorphin are the endogenous ligands for ␬-opioid receptors. Many actions of dynorphin A, however, have been found to be insensitive to the general opioid antagonist naloxone or the ␬-receptor-selective antagonist nor-binaltorphimine and are shared by the N-terminally truncated dynorphin A [2–17], which does not interact with opioid receptors. For instance, dynorphin A and dynorphin A [2–17] similarly affect motor function, electroencephalogram, and hippocampal unit activity [1]. Nonopioid effects of dynorphins may be relevant for several pathophysiological processes and the levels of dynorphins are increased at the sites of spinal cord injuries [2–5] and in the spinal cord with neuropathic pain [6 – 8]. Dynorphin A produces neurologic dysfunctions and hind-limb paralysis through a nonopioid mechanism when injected intrathecally into rats [9 –11]. Dynorphin A and dynorphin A [2–17] also cause depletion of cell bodies of inter-, sensory, and motoneurons in the lumbosacral cord associated with loss of motor activity [12] and kill neurons cultivated in vitro [13]. A long-lasting state of allodynia is induced by a single intrathecal injection of dynorphin A via a nonopioid mechanism [14 –16]. NMDA receptors have been suggested to be involved in the nonopioid effects of dynorphin A since NMDA receptor antagonists protect against hind-limb paralysis [8, 17–20] and loss of neuronal cell bodies [18]. Dynorphin A has been found to bind directly to these receptors [21, 22] and also to interact with several other membrane proteins, including acetylcholine [23] and vanniloid receptors [24], adenylyl cyclase [25], and Class II MHC molecules [26], as well as lipid vesicles [27, 28] and membrane nonopioid dynorphin A-binding

1

Current address: Department of Pharmacology, Tohoku Pharmaceutical University, 4-4-1 Komatsushima, Aoba-ku, Sendai 981-8558, Japan. 2 Current address: State Research Center of Virology and Biotechnology “Vector,” Institute of Molecular Biology, 633159 Koltsovo, Novosibirsk Region, Russian Federation. 3 To whom correspondence and reprint requests should be addressed at the Section of Alcohol and Drug Addiction Research, Department of Clinical Neuroscience, CMM L8-01, Karolinska Institute, S-171 76 Stockholm, Sweden. Fax: ⫹46 8 5177 61 80. E-mail: [email protected]. 0014-4827/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.

54

CYTOTOXIC EFFECTS OF DYNORPHINS

sites [29], with nano- or micromolar affinity. Thus, dynorphin A demonstrates a unique ability to interact with diverse proteins with high affinity and specificity. To gain insight into the mechanism of the cytotoxic effects of dynorphins we tested whether these effects are mediated via the p53 tumor-suppressor protein. p53 is a key transcriptional regulator of apoptosis and cell proliferation in the central nervous system [30]. In normal cells p53 is present in a latent form [30, 31]. The mechanisms of latency and activation are not understood but it is known that p53 can be activated in vitro for binding to target DNA sequences by basic fragments of its C-terminal domain [31]. Dynorphin A and a 32-amino-acid prodynorphin fragment big dynorphin (BD) 4 are probably among the most basic signal peptides that occur naturally [32, 33]. Initial in vitro experiments demonstrated that BD activated p53 more efficiently than the C-terminal fragment p53(361-382) [34]. This encouraged us to test whether BD would promote the ability of p53 to induce growth arrest and apoptosis. In the course of the study, we observed that BD is cytotoxic by itself because it reduces the number of viable cells when delivered into cells. These effects were sequence specific and may be a relevant control for proliferation and death of tumor cells, which express prodynorphin at high levels [35, 36], and for a pathophysiological role of prodynorphin in the central nervous system (CNS). MATERIALS AND METHODS Peptides and plasmids. BD and dynorphins A and B were obtained from Phoenix Pharmaceuticals, Inc. (Mountain View, CA) and other peptides were purchased from Bachem Feinchemikalien (Bubendorf, Switzerland). The CMV–wild-type p53 expression plasmid and PG-CAT and MG-CAT reporter plasmids were a gift from Dr. B. Vogelstein, Johns Hopkins Oncology Center (Baltimore, MD). A pCMV vector plasmid was used to adjust concentrations of transfected DNA. Cell culture. HeLa human cervical carcinoma and Saos-2 osteosarcoma cells were cultured in Iscove’s medium (Gibco BRL, Gaithersburg, MD) and neuroblastoma– glioma hybrid NG108-15 cells in DMEM/NUT-MIX, F-12 medium (Gibco BRL), all supplemented with 10% fetal bovine serum (Gibco BRL), penicillin, and streptomycin. Transfections and cell growth assays. Peptides were delivered into cells by lipofection with Lipofectamine (Gibco BRL) or by electroporation. Cells plated on 35-mm well plates (1.2 ⫻ 10 5 cells per well), 12-well plates with glass coverslips (1 ⫻ 10 4 cells per well), or 96-well microplates (3.5 ⫻ 10 3 cells per well) were transfected by 5 h incubation with a peptide, pCMV vector plasmid DNA, or CMV–wildtype p53 expression plasmid (1, 0.5, and 0.2 ␮g, respectively) and Lipofectamine (5.2, 2.6, and 0.4 ␮l, respectively) mixture. Naloxone was included in the transfection medium (1 nmol/ml in a 35-mm well) or added to the growth medium at 5 ⫻ 10 ⫺6 M concentration 5 h after transfection. The number of viable cells in 35-mm well plates was determined for pooled adherent and nonadherent cells by the 4 Abbreviations used: BD, big dynorphin; MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(sulfophenyl)2H-tetrazolium.

55

trypan blue exclusion assay. The viability of cells in the 96-well microplate was determined by the MTS (3-(4,5-dimethylthiazol-2-yl)5-(3-carboxymethoxyphenyl)-2-(sulfophenyl)-2H-tetrazolium, inner salt) assay [37]. Briefly, 20 ␮l of the combined MTS (1.9 mg/ml; Promega, Madison, WI) and phenazine methosulfate (0.04 mg/ml; Sigma) solution in PBS was added to the well, and after 2 h incubation absorption values at 492 nm were determined with a microplate reader. For electroporation, BD (5.27 nmol) was mixed with 1.1 ⫻ 10 5 HeLa or NG108-15 cells in 100 ␮l of RPMI 1640 medium (Gibco) and electroporated by a pulse at 125 ␮F, 230 V using the Gene Pulser II System (Bio-Rad). Cells were then diluted in 0.65 ml cultivation medium with 10% fetal bovine serum and antibiotics and aliquots were plated onto 96-well microplates, 100 ␮l per well, and cell viability was measured 24 h later by the MTS assay. Staining of apoptotic, p53-positive, and big dynorphin-transfected cells. HeLa and Saos-2 cells on glass coverslips in 12-well plates at 50 – 60% confluence were transfected with 2.4 ␮g of BD, 0.4 ␮g of CMV–wild-type p53 expression plasmid, or CMV vector plasmid and 2 ␮l of Lipofectamine in 0.4 ml of Opti-MEM per well, incubated in the DNA-intercalating dye propidium iodide (Sigma) (20 ␮g/ml, 10 min at 37°C), fixed with 3.7% paraformaldehyde, counterstained for 30 min with 0.4 ␮g/ml Hoechst 33258 dye (Hoechst, Frankfurt, Germany), and mounted on microscopic slides with Moviol 4-88 (Hoechst) as described previously [38]. Propidium iodide (fluoresces red) enters dead or dying cells with damaged membrane integrity, while chromatin condensation in small, round-shaped nuclei and fragmentation revealed by Hoechst staining (fluoresces blue) indicated apoptosis. The TUNEL technique was used for in situ nick DNA end-labeling with an in situ cell death detection kit (Boehringer Mannheim, Germany) according to the manufacturer’s instructions. The validity of the morphological identification of apoptosis was tested in nontransfected cells using the TUNEL technique (see also [38]); the concordance rate was 85%. The TUNEL technique cannot be used, however, for analysis of transfected cells due to the high incidence of strand breaks in transfected DNA. For immunofluorescence detection of p53-transfected cells, cells seeded on coverslips were fixed in 3.7% paraformaldehyde 48 h after transfection by lipofection, probed with a monoclonal anti-p53-antibody (clone DO-7; DAKO A/S, Copenhagen, Denmark), and exposed to the secondary rabbit anti-mouse IgG fluorescein-conjugated antibodies (DAKO A/S) in the presence of 0.4 ␮g/ml Hoechst 33258 dye. For staining of BD, cells transfected with the peptide for 5 h were washed with PBS, fixed in 3.7% paraformaldehyde for 1 h, incubated with immunoglobulins from rabbit anti-dynorphin B-antiserum [39] or from preimmune serum, and exposed to the secondary goat anti-rabbit IgG, F(ab⬘) 2 fragments conjugated with Alexa Fluor 488 (Molecular Probes) added in the presence of Hoechst 33258 dye. Immunoglobulins were purified with protein A–Sepharose CL-4B (Pharmacia Biotech, Uppsala, Sweden) according to the manufacturer’s recommendations. Quantification of the percentage of apoptotic cells. The percentage of apoptotic cells was assessed by morphological analysis with fluorescence microscopy after vital staining with propidium iodide and postfixation counterstaining with Hoechst 33258 dye [38]. Cells with round-shaped nuclei and chromatin condensation and fragmentation revealed by Hoechst staining indicated apoptosis. For each experimental observation, the number of apoptotic cells was scored on three coverslips with 300 cells on each slip. Apoptotic index (the number of apoptotic cells as percentage of the total number of counted cells) was calculated and presented as means ⫾ SEM. FACS analysis. Cells were harvested 48 h after transfection, washed with PBS, fixed with 70% ethanol at 4°C for 24 h, and stained with 5 ␮g/ml propidium iodide (PI). The samples were analyzed in a Becton–Dickinson FACScan system equipped with a 15mW, 448-nm air-cooled argon laser, using 10,000 cells for each determination and an acquisition rate lower than 400 cells/s. Staining with PI permitted us to recognize the DNA content of the cell popu-

56

TAN-NO ET AL.

TABLE 1 Effects of Dynorphins in Culture Medium on Growth of HeLa Cells a Time of treatment (h)

Dynorphin A

Dynorphin B

Big dynorphin

24 48 72

101 110 95

117 102 120

95 103 89

Cells were incubated with 100 ␮M peptides added to medium, and the number of viable cells was measured by the MTS assay in 96-well plates. The average number of viable cells, obtained in three experiments with six repeats in each, is shown as percentage of control untreated cells. a

lations studied and to quantify the cell cycle phases and the subG 0/G 1 fraction that corresponds to the percentage of apoptotic (having DNA fragmentation) cells. CAT assay. HeLa or Saos-2 cells (4 ⫻ 10 5 cells/6-cm plate) were transfected with 0.6 ␮g of the p53-responsive CAT reporter construct PG-CAT or a construct with mutant p53-binding sites, MG-CAT; 0.6 ␮g of CMV–wild-type p53 expression plasmid; CMV vector plasmid to adjust DNA concentration to 3 ␮g; peptide (17 ␮g); and 16 ␮l of Lipofectamine according to the manufacturer’s recommendations. CAT activity was measured [40] and quantified by Phosphoimager (Fuji-BAS-1000) analysis. To compare the levels of CAT expression under different experimental conditions, CMV-␤-galactosidase expression plasmid (0.6 ␮g of DNA/6-cm plate) was included as a reference plasmid and CAT activity was calculated as a ratio of the experimental values of CAT activity to the ␤-galactosidase activity. The assay was performed three times each in triplicate samples. Anti-p21 antibodies were obtained from Transduction Laboratories (Lexington, KY). Images. Slides were viewed under an epifluorescence microscope Zeiss Axioplan 2 imaging (Carl Zeiss AB, Oberkochen, Germany) and images captured by a CCD camera controlled by the Isis software (Metasystems GmbH, Altlussheim, Germany). Confocal images were collected by Leica TCS SP confocal laser scanning microscope (Leica Microsystem, Heidelberg, Germany). Images were prepared using the CorelDraw 8 software (Ottawa, ONT, Canada) and PhotoShop 6.0 (Adobe; San Jose, CA).

FIG. 1. The viability of HeLa cells transfected with prodynorphin-derived peptides. Cell viability was measured 72 h after transfection with peptide–Lipofectamine mixture by trypan blue exclusion assay and was taken as 100% in the mock-transfected cultures. Peptide concentrations are given in nanomoles. The data are given as means and SEM from three separate experiments. Student’s t test was used to compare cells transfected with peptides and mocktransfected (control) cells. **P ⬍ 0.01.

dependent and reached the maximum 45–55% reduction at a concentration of 1.8 nmol/ml of the transfection medium (Fig. 1). Cytotoxic effect of cotransfected dynorphins A and B (each 1.8 nmol) did not exceed 25% (data not shown) and did not differ from that of dynorphin A alone. BD, delivered into HeLa cells by lipofection, reduced by twofold the number of viable cells 24, 48, and 72 h after transfection (Figs. 1 and 2). The transfection procedure by itself under the conditions used decreased the number of viable cells by less than 20% 24 h after transfection compared with nontransfected cultures. Kinetics of inhibition of cell growth demonstrates that BD exerts its toxic

RESULTS

Cytotoxic effects of dynorphins. Dynorphin A, dynorphin B, and BD, a 32-amino-acid fragment of prodynorphin that consists of dynorphins A and B, added directly into the growth medium at 100 ␮M concentration, did not affect growth of HeLa cells during 72 h of incubation as assessed by the MTS (Table 1) and trypan blue (data not shown) assays. However, when peptides were delivered into HeLa cells by lipofection, dynorphin A, at concentrations of 1.8 and 3.5 nmol of peptide per milliliter of transfection medium, reduced the number of viable cells by 20 –25% after 72 h compared with mock-transfected cell cultures (Fig. 1). Dynorphin B and dynorphin B-29 used at the same concentrations produced no significant effects. Substantial reduction in the number of cells was observed with BD; the effect was generally concentration-

FIG. 2. Effects of big dynorphin on the viability of HeLa cells. Cell were transfected with peptide (0.3 nmol of peptide per milliliter of transfection medium)–Lipofectamine mixture and viability was measured by the MTS assay and was taken as 100% in nontransfected cultures. The data are given as means and SEM from three separate experiments. Student’s t test was used to compare cells transfected with peptides and mock transfected (control) cells. **P ⬍ 0.01.

CYTOTOXIC EFFECTS OF DYNORPHINS

FIG. 3. Effects of electroporation of big dynorphin on the viability of HeLa and NG108-15 cells. Cells were electroporated with big dynorphin (5.3 nmol) or without peptide (control) and plated into 96-well microplates. The MTS assay was done 24 h later. The data are given as mean and SEM. Student’s t test was used to compare electroporated with control cells. *P ⬍ 0.05, **P ⬍ 0.01 (n ⫽ 6).

effect by inducing cell death rather than by blocking cell cycle progression (Fig. 2); number of viable cells was reduced twofold 24 h after treatment with peptide, with no increase in this parameter in the mock-transfected culture. When electroporated into HeLa or NG108-15 cells, BD (5.3 nmol/100 ␮l of medium) also significantly reduced the number of viable cells in both cell lines (Fig. 3). Approximately the same percentage, about 50%, of HeLa cells were killed by lipofection or electroporation of BD, demonstrating that this effect does not depend on the delivery route. An analog of dynorphin A, the ␬ agonist [Arg 11,13]dynorphin A(1-13)-Gly-NH-(CH 2) 5-NH 2 (DAKLI), failed to reduce the number of viable cells when delivered into cells by lipofection (Fig. 4). This demonstrates that activation of ␬-opioid receptors, expression of which in these cells at very low levels cannot be ruled out, does not induce cytotoxicity. To our knowledge, there are no reports on cytotoxic effects of opioids mediated via ␬ receptors. In contrast, a neuroprotective action of dynorphin A is mediated through these receptors in a subpopulation of spinal cord neurons [13]. The general opioid antagonist naloxone, added to cell culture medium at 5 ⫻ 10 ⫺6 M concentration (Fig. 5) or to transfection medium (data not shown), did not prevent the effects of BD. Naloxone at this concentration blocks the effects of opioids in vitro on all three opioid receptor types, ␮, ␦, and ␬ [41– 44]. Accordingly to data with naloxone, nonopioid peptide des-Tyr-dynorphin A at concentration of 1.8 nmol/ml of transfection medium reduced the number of viable cells by 20% after 72 h of

57

FIG. 4. Effect of big dynorphin (0.3 nmol), DAKLI (0.5 nmol), and poly-L-lysine (1 ␮g ⫽ 0.3 nmol) transfected by lipofection on the viability of HeLa cells. Cells seeded in 96-well plates were transfected with peptides by lipofection and cell viability was measured by MTS assay 72 h after transfection. Cell viability in mock-transfected cultures (control) was taken as 100% and means and SEM from three separate experiments are shown. **P ⬍ 0.01 compared to control groups (Student’s t test was used).

incubation compared with mock-transfected cell cultures (data not shown) and, therefore, was equipotent to dynorphin A in this respect. Opioid receptors are located on the plasma membrane and opioids, added into cultivation medium, can bind to and activate them in in vitro experiments. In our study, no effect was observed when peptides were added into the medium, whereas toxicity was induced when peptides were de-

FIG. 5. Effect of the opioid antagonist naloxone on the big dynorphin-induced changes in HeLa cell viability. 5 ⫻ 10 ⫺6 M naloxone was added to growth medium 5 h after the initiation of transfection. Cell viability was measured by trypan blue exclusion assay 72 h after transfection. The data are given as means and SEM from three separate experiments. **P ⬍ 0.01 compared to control groups (Student’s t test was used).

58

TAN-NO ET AL.

FIG. 6. Hoechst 33258 staining demonstrating cell density and nuclei morphology in cultures incubated with transfection medium alone (A) or with big dynorphin–Lipofectamine mixture (B–D) 24 h after transfection. (B) Cultures incubated with big dynorphin show a considerably lower cell density and a higher number of condensed apoptotic nuclei compared to control cultures. (C, D) Nuclei displaying varying apoptotic morphology—irregularly shaped nuclei, condensation, and fragmentation of chromatin. Images in A and B were taken with a 20⫻ objective, whereas images in C and D were taken with a 100⫻ objective using an epifluorescence microscope equipped with a CCD camera. These images are representative of similar results obtained in three separate experiments.

livered into cells. No information on intracellular, functionally active opioid receptors is available to our knowledge. HeLa and Saos-2 cells do not express opioid receptors, at least at significant levels ([45] and our observations). These data altogether demonstrate that intracellular toxic effects of dynorphins are not mediated via opioid receptors. BD is a basic peptide. Another basic peptide, poly-Llysine, with a molecular size of up to 5 kDa, failed to reduce the number of viable cells (Fig. 4), suggesting that not only basic charge but some other properties of BD are important for toxicity. Apoptosis induced by big dynorphin. It was then investigated whether BD cytotoxicity was due to apoptosis and whether effects of BD may be mediated via p53-dependent pathways. HeLa cells with nonfunctional p53 and p53-deficient Saos-2 cells [31, 38, 40] were transfected with BD and the number of apoptotic cells was assessed using two types of analysis. First, in cultures transfected with BD, cells with chromatin con-

densation in small round-shaped nuclei, characteristic of early stages of apoptosis, and those with fragmented nuclei and apoptotic bodies, typical of the late stages of apoptosis, were registered (Figs. 6C and 6D). BD increased the proportion of apoptotic cells identified by morphological analysis by approximately fivefold (Fig. 7) and decreased the overall confluency of cells on coverslips (Figs. 6A and 6B). Second, cells were fixed, stained with propidium iodide, and then analyzed on FACS (Fig. 8). Loss of DNA is a characteristic of cells undergoing apoptosis and these cells can be distinguished from normal cells using FACS analysis. The sub-G 0/G 1 area represents this apoptotic population. Forty-eight hours after transfection with big dynorphin, 11.3% of cells were apoptotic, whereas only 2.3% of apoptotic cells were found in the mock-transfected cell cultures. The DNA staining pattern also suggested an accumulation of cells in S phase at the expense of G 1 and G 2 reduction in BD-transfected cultures (Fig. 8). The p53 tumor-suppressor protein is a latent tran-

CYTOTOXIC EFFECTS OF DYNORPHINS

59

FIG. 9. Effects of big dynorphin (BD) and dynorphin A (DA) on p53-dependent CAT transcription in HeLa cells. The PG-CAT and MG-CAT reporters contain consensus and mutant p53-binding sites, respectively. The relative units of CAT activity were normalized to the ␤-galactosidase activity with this ratio in mock-transfected cells (no p53-expression plasmid and no peptides; PG-CAT reporter) taken as a unit. Mean and SEM values are shown; n ⫽ 3.

FIG. 7. Morphological analysis of apoptosis in HeLa cells transfected with big dynorphin. The percentage of apoptotic cells in the total cell population is shown for the mock- (control) and big dynorphin-transfected cell cultures (left). The percentage of p53-positive apoptotic cells among p53-positive cells in cell cultures transfected with CMV–p53 expression plasmid with or without big dynorphin is shown at the right. The data are given as means and SEM (n ⫽ 6). **P ⬍ 0.01 compared to control groups. ##P ⬍ 0.01 compared to p53-positive apoptotic cells (Student’s t test was used).

scription factor and its activation in cells results in apoptosis or growth arrest. We have recently demonstrated that BD may efficiently 5- to 30-fold activate

FIG. 8. FACS analysis of apoptosis in HeLa cells, which were mock transfected (control) or transfected with big dynorphin. Cells were fixed in ethanol 48 h after transfection. DNA fragmentation was determined by FACS analysis after propidium iodide (PI) staining. The M1 marker shows the sub-G 0 cell fractions that correspond to apoptotic cells.

latent p53 for its binding to specific DNA sites in vitro [34]. We here tested whether BD would promote the ability of p53 to induce apoptosis. In HeLa cell cultures transfected with CMV–p53-expressing plasmid, the proportion of apoptotic cells was 21 times greater in the p53-positive cell population than that in the total cell population (Fig. 7). Cotransfection of CMV–p53 expression plasmid with BD resulted in a further 70% increase in the proportion of apoptotic cells in the p53positive cell population. Similar data were obtained with the p53-deficient Saos-2 cells (data not shown). Thus, BD-induced apoptosis was p53-independent but was also augmented in the presence of p53. Effects of BD on p53 transactivation. We examined the effects of BD on p53-dependent transcription in HeLa and Saos-2 cells using the experimental set-up described for p53 activation by the fragment of the C-terminal domain of p53 [40]. The p53-responsive CAT reporter construct PG-CAT or a construct with mutant p53-binding sites, MG-CAT, as well as CMV␤-galactosidase, CMV–p53, or CMV-vector plasmid mixed with peptide, was transfected by lipofection. BD and dynorphin A did not affect p53-induced CAT activity transcribed from the PG-CAT or MG-CAT constructs in HeLa (Fig. 9) and Saos-2 cells (data not shown). p53 expression induced a 5- to 10-fold increase in the p21 protein level. No effects of BD on the basal or p53-stimulated levels of p21 expression were registered (data not shown). Intracellular localization of big dynorphin delivered into HeLa cells. HeLa cells were incubated for 5 h in transfection medium with or without BD, fixed, and stained with rabbit anti-dynorphin B antibodies (Figs. 10A–10G). Preimmune serum did not stain the BDtransfected HeLa cells (not shown). Anti-dynorphin B immunoglobulins strongly stained cells transfected with BD (Figs. 10B, 10C, and 10E–10G) but not mocktransfected cells (Figs. 10A and 10D). Confocal laser scanning microscopy revealed that BD was localized throughout the cytoplasm but not in the cell nuclei (Figs. 10E–10G). Discrete, filamentous aggregates of BD were often observed throughout the cytoplasm and

60

TAN-NO ET AL.

FIG. 10. Visualization of big dynorphin delivery into HeLa cells by lipofection. Cells were incubated for 5 h with transfection medium (A, D) or big dynorphin (0.3 nmol)–Lipofectamine mixture (B, C, E–G), fixed, immunofluorescently stained using anti-dynorphin B antibodies (green fluorescence), counterstained with Hoechst 33258 dye (blue fluorescence), and examined using either epifluorescence (A–C) or confocal (D–G) microscopy. Note that the blue fluorescence from Hoechst 33258 is observed only with the epifluorescence microscope. (A, D) No big dynorphin staining (green fluorescence) could be observed in vehicle-treated cultures. (B) A group of big-dynorphin-transfected viable cells. (E) A group of big-dynorphin-transfected cells with a generally uniform distribution and aggregates of big dynorphin in the cytoplasm but not in the cell nuclei. (F) A pair of transfected cells connected with a brightly stained bridge-like structure observed in about 5% of the transfected cells. (C, G) Big dynorphin-transfected cells display signs of apoptotic cell death, blebbed cytoplasm, and condensed nuclei. Micrographs are representative of similar results obtained in four separate experiments with four different culture preparations are shown. Scale bars, 20 ␮m.

especially within the juxtanuclear area. In a small population of transfected cells, brightly stained bridgelike structures connecting pairs of transfected cells were observed (Fig. 10F). These structures were most probably remains of mitotic spindles connecting recently divided cells. Many BD-transfected cells displayed morphological features typically associated with apoptotic cell death, including blebbed cytoplasm and condensed nuclei (Figs. 10C and 10G). DISCUSSION

BD and dynorphin A significantly reduced the number of viable cells through nonopioid mechanisms when delivered into HeLa cells by lipofection or electroporation, and this effect was sequence-specific. The induction of apoptosis is the probable reason although the mechanism is not clear. BD synergistically augmented p53-induced apoptosis but BD did not affect the p53dependent transcription of reporter genes and the endogenous p21 WAF1/CIP1 gene in HeLa and Saos-2 cells. Also supporting evidence for a nontranscriptional mechanism is the fact that BD was located in the cytoplasm but not in the nuclei in transfected cells. Another basic peptide, the p53 C-terminal domain

(CTD) fragment fused to the Antennapedia homeobox domain to facilitate cellular uptake, has been found to suppress growth and induce apoptosis in cells expressing p53 via a nontranscriptional mechanism, mediated through a Fas/APO-1 signaling pathway and a Fas/ APO-1-specific protease FLICE mechanism [46]. A 50% reduction in the number of viable cells in BD-treated cell cultures was observed although only 1–10% of the cells were transfected with BD and 5–11% were apoptotic at each observation moment. The generalization of the BD effects may be due to killing of viable bystander cells by dead or dying cells. Bystander cell killing can be triggered experimentally by inducing apoptosis in single cells and may be based on the exchange of chemical cell death signals between nearby cells. We have recently demonstrated bystander cell killing by primary apoptotic cells in HeLa and Saos-2 cell monolayers [38]. Interestingly, that catalase, a peroxide scavenger, suppressed bystander killing. We further demonstrated that apoptotic cells generate hydrogen peroxide in very high, micromolar concentrations and that this oxidant penetrates cell membranes and causes death of viable neighbors [38]. Bystander killing through peroxides emitted by apo-

61

CYTOTOXIC EFFECTS OF DYNORPHINS

TABLE 2 Amino Acid Sequences of Dynorphins and Selected Basic Regions Involved in Regulatory Protein–Protein Interactions Peptide

Sequence

Big dynorphin a Dynorphin B-29 Fragment of the p53 C-terminal domain interacting with other proteins [31] MyoD domain interacting with M-Twist [52] M-Twist domain interacting with MyoD [52] ␤␥ binding domain of ␤-adrenergic receptor kinase [48] CaM-binding domain of CaM kinase I [50] CaM-binding domain of CaM kinase II [49]

YGGFLRRIRPKLKWDNQKRYGGFLRRDFKVVT YGGFLRRDFKVVTRSQEDPNAYYEELFDV GSRAHSSHLKSKKGQSTRHKK KRKTTNADRRKAATMRERRR QRVMANVRERQR WKKELRDAYREAQQLVQRVPKMKNKPRS KWKQAFNATAVVRHMRK NARRKLKGAILTTMLATR

a

The N-terminal dynorphin A and C-terminal dynorphin B sequences are underlined.

ptotic cells may therefore propagate tissue injury in different pathological situations and be relevant for a pathophysiological role of dynorphins. In addition to binding to opioid receptors, dynorphin A demonstrates unique ability to interact with diverse membrane proteins with nano- and micromolar affinity [21–26, 29]. One distinct structural feature of dynorphin A and BD as well is high positive charge (Table 2). BD is more basic (Table 2) and cytotoxic (see Fig. 1) than dynorphin A. Both peptides have higher positive charge than dynorphins B and B-29 and DAKLI, which are not toxic at all. However, the basic peptide poly-Llysine did not show any toxicity (see Fig. 4), suggesting that not only basic charge determines the ability of BD to kill cells. Amino acid sequences can be represented as sequences of hydrophobic and hydrophilic segments [47]. Alternating hydrophobic and hydrophilic segments in dynorphin A and BD predominantly consist of one or two amino acid residues (see sequences in Table 2). This feature along with the high content of Gly and Pro may give the dynorphins high flexibility, allowing them to adopt multiple compact conformations complementary to the diverse faces of protein targets. The p53 tumor-suppressor protein is a latent transcription factor and can be activated for the binding to target DNA sites by the basic fragment of it C-terminal domain, p53(361-383) [30, 31, 34]. We have recently demonstrated that BD, which is larger and has higher positive charge than p53(361-382) (Table 2), activates p53 more efficiently [34]. Shorter peptides such as dynorphins A and B were unable to activate p53. The net positive charge and overall peptide length seem to be the common structural feature of peptides activating p53, as well as of those inducing toxicity. Protein–protein interactions play an important regulatory role in multicomponent intracellular systems. Basic domains of many cytoplasmic and nuclear proteins are implicated in protein–protein interactions (Table 2; [48 –53]). The myogenic basic helix-loop-helix protein MyoD interacts with another basic helix-loop-

helix protein, Twist, via basic domains of the two proteins, which leads to inhibition of MyoD [52]. Basic domains involved in protein–protein interactions are characterized by common features in their primary structure with basic amino acids generally arranged in two or three clusters separated by short fragments enriched in Ser, Thr, Ala, and Gly, amino acids with small side chains (Table 2). BD shows a similar clustering of basic amino acids, suggesting that this peptide can interfere with functionally essential protein– protein interactions, which leads to changes in cell physiology, growth alterations, and cell death. Levels of dynorphin A increase markedly following spinal cord trauma and may contribute to secondary neurodegeneration. To better understand how dynorphins affect cell viability, we have recently performed structure–activity studies examining the effects of dynorphin A and its fragments on the survival of mouse spinal cord neurons coexpressing ␬-opioid and NMDA receptors in vitro [13]. Dynorphin A caused significant neuronal losses at 1 ␮M and higher concentrations. Exposure to dynorphin A fragments also caused a significant loss of neurons. The rank order of toxicity was dynorphin A(1-17) ⬎ dynorphin A(1-13) ⫽ dynorphin A(2-13) ⫽ dynorphin A(13-17). Thus, dynorphin A(13-17) was the minimal toxic fragment, while dynorphin A(1-11) and Leu-enkephalin did not cause neuronal losses. The NMDA receptor antagonist MK801 significantly attenuated toxic effects, suggesting that they are mediated by NMDA receptors [13]. Thus, toxicity resides in the carboxyl-terminal portion of dynorphin A. These findings suggest that dynorphin A and/or its metabolites may contribute to neurodegeneration during spinal cord or brain injury and that alterations in dynorphin A biosynthesis, metabolism, and/or degradation may be important in determining injury outcome. Comparison of these data with the results of the present investigation demonstrates different structure determinants for extra- and intracellular toxicity and suggests different mechanisms engaged. The short C-terminal fragment of dynorphin A,

62

TAN-NO ET AL.

dynorphin A(13-17), is sufficient to kill neurons, whereas the full-length 32-amino-acid BD seems to be required to induce cell death when delivered into cells. Structural determinants of BD cytotoxicity appear to include (a) high positive charge, (b) peptide length, (c) peptide flexibility, and (d) clustering of basic amino acids. We do not know yet whether the presence of the dynorphin A(13-17) segment, which is critical for toxic effects mediated via the NMDA receptors, is necessary for the intracellular effects of BD. Our preliminary data demonstrate that in addition to entering the secretory pathways, substantial amounts of prodynorphin dislocate from the endoplasmic reticulum to the cytoplasm where this protein is degraded (protein dislocation from the endoplasmic reticulum reviewed in [54]). Short peptides generated in the process of degradation may interfere with intracellular processes via binding to cytoplasmic proteins. A potential for such interference was demonstrated in the present study when dynorphins were delivered into the cells. Many human tumor cell lines and tumors naturally express prodynorphin lacking the signal peptide [35, 36], which therefore does not translocate into the endoplasmic reticulum but degrades in the cytosol, and degradation products may potentially be involved in the control of tumor cell fate. We thank Dr. B. Lusher for the CMV-␤-galactosidase plasmid, Dr. T. Shenk for the GST-C2 plasmid, and Dr. B. Vogelstein for the PG-CAT and MG-CAT and pC53-SN3 plasmids. This work was supported by grants from the Swedish Medical Research Council (3166) to L.T. and (12190) to G.B. and from the Swedish Cancer Society (3935) to G.B. and by a fellowship from the Royal Swedish Academy of Sciences to I.G.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

REFERENCES 1.

Walker, J. M., Moises, H. C., Coy, D. H., Baldrighi, G., and Akil, H. (1982). Nonopiate effects of dynorphin and des-Tyr-dynorphin. Science 218, 1136 –1138.

17.

2.

Cox, B. M., Molineaux, C. J., Jacobs, T. P., Rosenberger, J. G., and Faden, A. I. (1985). Effects of traumatic injury on dynorphin immunoreactivity in spinal cord. Neuropeptides 5, 571– 574.

18.

3.

Faden, A. I., Molineaux, C. J., Rosenberger, J. G., Jacobs, T. P., and Cox, B. M. (1985). Endogenous opioid immunoreactivity in rat spinal cord following traumatic injury. Ann. Neurol. 17, 386 –390.

4.

Faden, A. I., Molineaux, C. J., Rosenberger, J. G., Jacobs, T. P., and Cox, B. M. (1985). Increased dynorphin immunoreactivity in spinal cord after traumatic injury. Regul. Pept. 11, 35– 41.

5.

Yakovlev, A. G., and Faden, A. I. (1994). Sequential expression of c-fos protooncogene, TNF-alpha, and dynorphin genes in spinal cord following experimental traumatic injury. Mol. Chem. Neuropathol. 23, 179 –190.

6.

Malan, T. P., Ossipov, M. H., Gardell, L. R., Ibrahim, M., Bian, D., Lai, J., and Porreca, F. (2000). Extraterritorial neuropathic pain correlates with multisegmental elevation of spinal dynorphin in nerve-injured rats. Pain 86, 185–194.

19.

20.

21.

22.

Long, J. B., Martinez-Arizala, A., Petras, J. M., and Holaday, J. W. (1986). Endogenous opioids in spinal cord injury: A critical evaluation. Cent. Nerv. Syst. Trauma 3, 295–315. Shukla, V. K., and Lemaire, S. (1994). Non-opioid effects of dynorphins: Possible role of the NMDA receptor. Trends Pharmacol. Sci. 15, 420 – 424. Faden, A. I., and Jacobs, T. P. (1984). Dynorphin-related peptides cause motor dysfunction in the rat through a non-opiate action. Br. J. Pharmacol. 81, 271–276. Herman, B. H., and Goldstein, A. (1985). Antinociception and paralysis induced by intrathecal dynorphin A. J. Pharmacol. Exp. Ther. 232, 27–32. Long, J. B., Rigamonti, D. D., de Costa, B., Rice, K. C., and Martinez-Arizala, A. (1989). Dynorphin A-induced rat hindlimb paralysis and spinal cord injury are not altered by the kappa opioid antagonist nor-binaltorphimine. Brain Res. 497, 155– 162. Long, J. B., Petras, J. M., Mobley, W. C., and Holaday, J. W. (1988). Neurological dysfunction after intrathecal injection of dynorphin A (1-13) in the rat. II. Nonopioid mechanisms mediate loss of motor, sensory and autonomic function. J. Pharmacol. Exp. Ther. 246, 1167–1174. Hauser, K. F., Knapp, P. E., and Turbek, C. S. (2001). Structure–activity analysis of dynorphin A toxicity in spinal cord neurons: Intrinsic neurotoxicity of dynorphin A and its carboxyl-terminal, non-opioid metabolites. Exp. Neurol. 168, 78 – 87. Vanderah, T. W., Laughlin, T., Lashbrook, J. M., Nichols, M. L., Wilcox, G. L., Ossipov, M. H., Malan, T. P., and Porreca, F. (1996). Single intrathecal injections of dynorphin A or des-Tyrdynorphins produce long-lasting allodynia in rats: Blockade by MK-801 but not naloxone. Pain 68, 275–281. Laughlin, T. M., Vanderah, T. W., Lashbrook, J., Nichols, M. L., Ossipov, M., Malan, T. P., Porreca, F., and Wilcox, G. L. (1997). Spinally administered dynorphin A produces long-lasting allodynia: Involvement of NMDA but not opioid receptors. Pain 72, 253–260. Vanderah, T. W., Gardell, L. R., Burgess, S. E., Ibrahim, M., Dogrul, A., Zhong, C. M., Zhang, E. T., Malan, T. P., Jr., Ossipov, M. H., Lai, J., and Porreca, F. (2000). Dynorphin promotes abnormal pain and spinal opioid antinociceptive tolerance. J. Neurosci. 20, 7074 –7079. Bakshi, R., and Faden, A. I. (1990). Competitive and non-competitive NMDA antagonists limit dynorphin A-induced rat hindlimb paralysis. Brain Res. 507, 1–5. Skilling, S. R., Sun, X., Kurtz, H. J., and Larson, A. A. (1992). Selective potentiation of NMDA-induced activity and release of excitatory amino acids by dynorphin: Possible roles in paralysis and neurotoxicity. Brain Res. 575, 272–278. Bakshi, R., Ni, R. X., and Faden, A. I. (1992). N-Methyl-Daspartate (NMDA) and opioid receptors mediate dynorphininduced spinal cord injury: Behavioral and histological studies. Brain Res. 580, 255–264. Long, J. B., Rigamonti, D. D., Oleshansky, M. A., Wingfield, C. P., and Martinez-Arizala, A. (1994). Dynorphin A-induced rat spinal cord injury: Evidence for excitatory amino acid involvement in a pharmacological model of ischemic spinal cord injury. J. Pharmacol. Exp. Ther. 269, 358 –366. Chen, L., Gu, Y., and Huang, L.-Y. M. (1995). The opioid peptide dynorphin directly blocks NMDA receptor channels in the rat. J. Physiol. 482, 575–581. Tang, Q., Gandhoke, R., Burritt, A., Hruby, V. J., Porreca, F., and Lai, J. (1999). High-affinity interaction of (des-tyrosyl)dynorphin A(2-17) with NMDA receptors. J. Pharmacol. Exp. Ther. 291, 760 –765.

CYTOTOXIC EFFECTS OF DYNORPHINS 39.

23.

Itoh, H., Andoh, T., Watanabe, I., Sasaki, T., Kamiya, Y., and Okumura, F. (2000). Dynorphins directly inhibit neuronal nicotinic acetylcholine receptors in PC12 cells. Eur. J. Neurosci. 12, 1253–1262.

24.

Planells-Cases, R., Aracil, A., Merino, J. M., Gallar, J., PerezPaya, E., Belmonte, C., Gonzalez-Ros, J. M., and Ferrer-Montiel, A. V. (2000). Arginine-rich peptides are blockers of VR-1 channels with analgesic activity. FEBS Lett. 481, 131–136.

25.

Claye, L. H., Unterwald, E. M., Ho, A., and Kreek, M. J. (1996). Both dynorphin A(1-17) and [Des-Tyr1]dynorphin A(2-17) inhibit adenylyl cyclase activity in rat caudate putamen. J. Pharmacol. Exp. Ther. 277, 359 –396.

26.

Schmitt, L., Kratz, J. R., Davis, M. M., and McConnell, H. M. (1999). Catalysis of peptide dissociation from class II MHCpeptide complexes. Proc. Natl. Acad. Sci. USA 96, 6581– 6586.

27.

Erne, D., Sargent, D. F., and Schwyzer, R. (1985). Preferred conformation, orientation, and accumulation of dynorphin A-(113)-tridecapeptide on the surface of neutral lipid membranes. Biochemistry 24, 4261– 4263.

43.

28.

Wu, C. S., Lee, N. M., Loh, H. H., and Yang, J. T. (1986). Competitive binding of dynorphin-(1-13) and beta-endorphin to cerebroside sulfate in solution. J. Biol. Chem. 261, 3687–3691.

44.

29.

Yarygin, K. N., Zhang, X. H., and Lee, N. M. (1998). Non-opioid dynorphin binding site on secretory vesicles of a pituitaryderived cell line. Brain Res. 791, 99 –107.

45.

Amson, R., Lassalle, J. M., Halley, H., Prieur, S., Lethrosne, F., Roperch, J. P., Israeli, D., Gendron, M. C., Duyckaerts, C., Checler, F., Dausset, J., Cohen, D., Oren, M., and Telerman, A. (2000). Behavioral alterations associated with apoptosis and down-regulation of presenilin 1 in the brains of p53-deficient mice. Proc. Natl. Acad. Sci. USA 97, 5346 –5350.

46.

30.

31.

Hupp, T. R., Sparks, A., and Lane, D. P. (1995). Small peptides activate the latent sequence-specific DNA binding function of p53. Cell 83, 237–245.

32.

Cone, R. I., Weber, E., Barchas, J. D., and Goldstein, A. (1983). Regional distribution of dynorphin and neo-endorphin peptides in rat brain, spinal cord, and pituitary. J. Neurosci. 3, 2146 –2152.

33.

34.

Day, R., and Akil, H. (1989). The posttranslational processing of prodynorphin in the rat anterior pituitary. Endocrinology 124, 2392–2405.

40.

41.

42.

47.

48.

49.

Yakovleva, T., Pramanik, A., Kawasaki,T., Tan-No, K., Gileva, I., Lindegren, H., Langel, U., Ekstro¨m, T. J., Rigler, R., Terenius, L., and Bakalkin, G. (2001). p53 latency: C-terminal domain prevents binding of p53 core to target but not to nonspecific DNA sequences. J. Biol. Chem. 276, 15650 –15658.

50.

35.

Geijer, T., Bergh, J., and Terenius, L. (1991). Expression of preprodynorphin in human small cell lung carcinoma cell lines. Regul. Pept. 34, 181–188.

51.

36.

Geijer, T., Telkov, M., and Terenius, L. (1995). Characterization of human prodynorphin gene transcripts. Biochem. Biophys. Res. Commun. 215, 881– 888.

52.

37.

Cory, A. H., Owen, T. C., Barltrop, J. A., and Cory, J. G. (1991). Use of an aqueous soluble tetrazolium/formazan assay for cell growth assays in culture. Cancer Commun. 3, 207–212.

38.

Reznikov, K., Kolesnikova, L., Pramanik, A., Tan-No, K., Gileva, I., Yakovleva, T., Rigler, R., Terenius, L., and Bakalkin, G. (2000). Clustering of apoptotic cells via bystander killing by peroxides. FASEB J. 14, 1754 –1764.

Received February 16, 2001 Revised version received June 14, 2001 Published online August 6, 2001

53.

54.

63

Christensson-Nylander, I., Nyberg, F., Ragnarsson, U., and Terenius, L. (1985). A general procedure for analysis of proenkephalin B derived opioid peptides. Regul. Pept. 11, 65–76. Selivanova, G., Iotsova, V., Okan, I., Fritsche, M., Storm, M., Groner, B., Crafstorm, R. C., and Wiman, K. G. (1997). Restoration of the growth suppression function of mutant p53 by a synthetic peptide derived from p53 C-terminal domain. Nat. Med. 3, 632– 638. Brown, G. P., and Pasternak, G. W. (1998). 3H-Naloxone benzoylhydrazone binding in MOR-1-transfected Chinese hamster ovary cells: Evidence for G-protein-dependent antagonist binding. J. Pharmacol. Exp. Ther. 286, 376 –381. Fukuda, K., Kato, S., Shoda, T., Morikawa, H., Mima, H., and Mori, K. (1998). Partial agonistic activity of naloxone on the opioid receptors expressed from complementary deoxyribonucleic acids in Chinese hamster ovary cells. Anesth. Analg. 87, 450 – 455. George, S. R., Fan, T., Xie, Z., Tse, R., Tam, V., Varghese, G., and O’Dowd, B. F. (2000). Oligomerization of mu- and deltaopioid receptors. Generation of novel functional properties. J. Biol. Chem. 275, 26128 –26135. Remmers, A. E., Clark, M. J. Mansour, A., Akil, H., Woods, J. H., and Medzihradsky, F. (1999). Opioid efficacy in a C6 glioma cell line stably expressing the human kappa opioid receptor. J. Pharmacol. Exp. Ther. 288, 827– 833. Gioannini, T. L., Yao, Y. H., Hiller, J. M., Taylor, L. P., and Simon, E. J. (1993). Antisera against peptides derived from a purified ␮-opioid binding protein recognize the protein as well as mu-opioid receptors in brain regions and a cell line. Mol. Pharmacol. 44, 796 – 801. Kim, A. L., Raffo, A. J., Brandt-Rauf, P. W., Pincus, M. R., Monaco, R., Abarzua, P., and Fine, R. L. (1999). Conformational and molecular basis for induction of apoptosis by a p53 C-terminal peptide in human cancer cells. J. Biol. Chem. 274, 34924 –34931. Bakalkin, G., Rakhmaninova, A. B., Akparov, V. K., Volodin, A. A., Ovchinnikov, V. V., and Sarkisyan, R. A. (1991). Amino acid sequence pattern in the regulatory peptides. Int. J. Pept. Protein Res. 38, 505–510. Koch, W. J., Inglese, J., Stone, W. C., and Lefkowitz, R. J. (1993). The binding site for the beta gamma subunits of heterotrimeric G proteins on the beta-adrenergic receptor kinase. J. Biol. Chem. 268, 8256 – 8260. Meador, W. E., Means, A. R., and Quiocho, F. A. (1993). Modulation of calmodulin plasticity in molecular recognition on the basis of x-ray structures. Science 262, 1718 –1721. Yokokura, H., Picciotto, M. R., Nairn, A. C., and Hidaka, H. (1995). The regulatory region of calcium/calmodulin-dependent protein kinase I contains closely associated autoinhibitory and calmodulin-binding domains. J. Biol. Chem. 270, 23851–23859. Gschwendt, M., Johannes, F. J., Kittstein, W., and Marks, F. (1997). Regulation of protein kinase Cmu by basic peptides and heparin. Putative role of an acidic domain in the activation of the kinase. J. Biol. Chem. 272, 20742–20746. Hamamori, Y., Wu, H. Y., Sartorelli, V., and Kedes, L. (1997). The basic domain of myogenic basic helix-loop-helix (bHLH) proteins is the novel target for direct inhibition by another bHLH protein, Twist. Mol. Cell. Biol. 17, 6563– 6573. Souroujon, M. C., and Mochly-Rosen, D. (1998). Peptide modulators of protein–protein interactions in intracellular signaling. Nat. Biotech. 16, 919 –924. Kopito, R. R. (1997). ER quality control: The cytoplasmic connection. Cell 88, 427– 430.