Báo cáo khoa học: Synthetic constrained peptide selectively binds and antagonizes death receptor 5 ppt

Synthetic constrained peptide selectively binds and antagonizes death receptor 5 Johanna Vrielink 1, *, Mariette S. Heins 1, *, Rita Setroikromo 1 , Eva Szegezdi 2 , Margaret M. Mullally 1 , Afshin Samali 2 and Wim J. Quax 1 1 Department of Pharmaceutical Biology, University of Groningen, the Netherlands 2 Department of Biochemistry, Cell Stress and Apoptosis Research Group, National Centre for Biomedical Engineering Science, National University of Ireland, Galway, Ireland Introduction Tumour necrosis factor-related apoptosis-inducing ligand (TRAIL ⁄ Apo2L), a member of the tumour necrosis factor (TNF) ligand family, is well known for its ability to induce apoptosis in many cancer cells but not in most untransformed cells [1]. This quality makes TRAIL an interesting and promising target for cancer therapy and the main focus of TRAIL research has therefore been on its role in cancer. Less attention has been paid to the role of TRAIL in neurodegenerative diseases. Normally, mature neurones will last the lifespan Keywords apoptosis; DR5; phage display; R2C16; TRAIL Correspondence W. J. Quax, Department of Pharmaceutical Biology, University of Groningen, Antonius Deusinglaan 1, 9713 AV, Groningen, the Netherlands Fax: +31 50 363 3000 Tel: +31 50 363 2558 E-mail: w.j.quax@rug.nl *These authors contributed equally to this work (Received 12 October 2009, revised 17 December 2009, accepted 25 January 2010) doi:10.1111/j.1742-4658.2010.07590.x Apoptosis or programmed cell death is an inherent part of the development and homeostasis of multicellular organisms. Dysregulation of apoptosis is implicated in the pathogenesis of diseases such as cancer, neurodegenera- tive diseases and autoimmune disorders. Tumour necrosis factor-related apoptosis-inducing ligand (TRAIL) is able to induce apoptosis by binding death receptor (DR)4 (TRAIL-R1) and DR5 (TRAIL-R2), which makes TRAIL an interesting and promising therapeutic target. To identify peptides that specifically interact with DR5, a disulfide-constrained phage display peptide library was screened for binders towards this receptor. Phage-displayed peptides were identified that bind specifically to DR5 and not to DR4, nor any of the decoy receptors. We show that the synthesized peptide, YCKVILTHRCY, in both monomeric and dimeric forms, binds specifically to DR5 in such a way that TRAIL binding to DR5 is inhibited. Surface plasmon resonance studies showed higher affinity towards DR5 for the dimeric form then the monomeric form of the peptide, with apparent K d values of 40 nm versus 272 nm, respectively. Binding studied on cell lines by flow cytometry analyses showed concentration-dependent binding. Upon co-incubation with increasing concentrations of TRAIL, the peptide binding was reduced. Moreover, both the monomeric and dimeric forms of the peptide reduced TRAIL-induced cell death in Colo205 colon carcinoma cells. The peptide, YCKVILTHRCY, or its derivates, may be a useful investigative tool for dissecting signalling via DR5 relative to DR4 or could act as a lead peptide for the development of therapeutic agents in diseases with dysregulated TRAIL-signalling. Abbreviations DcR, decoy receptor; DR, death receptor; EAE, experimental autoimmune encephalomyelitis; FACS, fluorescence-activated cell sorting; HRP, horseradish peroxidase; OPG, osteoprotegerin; PE, phycoerythrin; pfu, plaque forming unit; RU, response unit; sTRAIL, soluble TRAIL; TMB, tetramethylbenzidine; TNF, tumour necrosis factor; TRAIL, TNF-related apoptosis-inducing ligand; TRAIL-R, TRAIL-receptor. FEBS Journal 277 (2010) 1653–1665 ª 2010 The Authors Journal compilation ª 2010 FEBS 1653 of the organism. However, in neurodegenerative diseases, the balance is shifted towards death and the major cause of neuronal loss is apoptosis. TRAIL is present in the central nervous system and can induce apoptosis in brain cells. Studies have shown the involvement of TRAIL in neurodegenerative diseases such as HIV-1-associated dementia, Alzheimer’s disease and multiple sclerosis [2–4]. Molecules that counteract the dysregulation of apoptosis could com- prise effective therapeutic agents in these degenerative disorders. TRAIL induces apoptosis by binding to the cyste- ine-rich extracellular domain of death receptors DR4 (TRAIL-R1) [5] and DR5 (TRAIL-R2) [6–8]. Binding of the trimeric ligand leads to clustering of the cytoplasmic death domains of the receptors and recruitment of signalling molecules to form the death- inducing signalling complex, which activates the caspase cascade and thus initiates apoptosis [9,10]. In addition to DR4 and DR5, TRAIL is capable of bind- ing to three decoy receptors (DcR): DcR1 ⁄ TRAIL-R3 [6,7], DcR2 ⁄ TRAIL-R4 [11] and osteoprotegerin (OPG) [12]. DcR1 lacks a functional death domain, DcR2 contains a truncated death domain and OPG is a soluble receptor; therefore, they cannot trigger a pro- apoptotic signal. Although the crystal structure of TRAIL in complex with DR5 is known, the exact mechanism of binding and signal initiation is still not completely understood [13–15]. The first step in signalling by members of the TNF family is considered to comprise ligand-induced trimerization of the receptor. However, the identifica- tion of a pre-ligand assembly domain suggested that receptors may already be pre-assembled as trimers before ligand binding [16–18]. Another intriguing fea- ture of DR5 is that it appears to be able to mediate distinctly different cell signals depending on the inter- action with different receptor agonists [19,20]. Further- more, depending on the cell type, TRAIL can signal apoptosis either via DR4 [21–23] or DR5, or both [24–26]. The reasons for the differences between DR4 versus DR5 signalling are not yet fully understood. To study the differences in mechanism of ligand binding and subsequent intracellular signalling through DR4 versus DR5, receptor-selective agonists and antagonists are necessary. Recently, we have described DR5 [25] and DR4 [27] selective agonistic variants of TRAIL. To identify an antagonist to address the differences in DR4 versus DR5 signalling, we now select for a pep- tide that binds specifically to DR5. Peptides and small proteins were demonstrated to bind their targets with high affinity and specificity and to have advantage over antibodies [28]. Phage display, a sophisticated technique that links genotype and phenotype, was used to select for such ligand-mimicking peptides. Earlier studies have shown that it is a practical method for identifying peptides with either agonistic [29–32] or antagonistic properties for various receptors [33–35]. In the present study, by screening a disulfide-con- strained phage display peptide library, we report the identification of a peptide that specifically interacts with DR5 and blocks binding of TRAIL to DR5. The identified DR5-binding phage-displayed peptides shows a strong consensus sequence and the monomeric and dimeric forms of one of these peptides, YCK- VILTHRCY, were further characterized. Both the monomeric and dimeric peptide show selective binding to DR5 in vitro. To confirm the binding specificity of the monomeric and dimeric peptides on the membrane of intact cells, we show binding towards Jurkat cells that can be competed with soluble TRAIL (sTRAIL). Finally, we demonstrate that the peptides can reduce TRAIL-induced apoptosis on Colo205 cells. Compared to the monomeric form, the dimeric form has higher affinity for DR5 and increased antagonistic activity. The identified peptide, or its derivatives, can be a use- ful tool for elucidating the mechanism of TRAIL signalling or the mechanism of controlling differential signalling through DR4 or DR5. In addition, this pep- tide may act as a lead peptide for the development of therapeutic agents in diseases with dysregulated TRAIL-signalling. Results Identification of DR5-binding phages To select for peptides able to bind to DR5 with high affinity, we used a cystein-constrained heptamer pep- tide phage library. After three rounds of selection (as described in the Experimental procedures), 25 indi- vidual clones were picked and sequenced (Table 1). The binding ability of the phages displaying these peptides to DR5 was analysed using ELISA. The wells were coated with DR5-Fc, the extracellular domain of DR5 fused to the Fc-portion of human IgG 1 , and the bound phages were detected with an horseradish per- oxidase (HRP)-antibody against the phage coat protein g8p. The background signal measured for a well with no receptor coated was subtracted. Fourteen out of 25 phages showed binding to DR5 (Table 1). These peptides share a highly homologous consensus C(K ⁄ I ⁄ L)V(Y ⁄ I ⁄ A)LT(Q ⁄ H ⁄ L)(K ⁄ R)C. Phage 77-R2C16 (CKVILTHRC) showed the highest affinity for DR5 and was selected for further investigation. Purified phage 77-R2C16 showed a dose-dependent binding to Inhibition of DR5 signalling J. Vrielink et al. 1654 FEBS Journal 277 (2010) 1653–1665 ª 2010 The Authors Journal compilation ª 2010 FEBS DR5. Even when using increased amounts, no binding to DR5 of control phage (i.e. a phage that displays g3p without a C7C peptide) was observed (Fig. 1A). Phage 77-R2C16 was also used to determine selectivity towards DR5 using ELISA. Wells were coated with different TNF family receptors (i.e. DR4-Fc, DR5-Fc, DcR1-Fc, DcR2-Fc, OPG-Fc, mouse OPG-Fc, mouse receptor activator of nuclear factor-jB-Fc and TNF- receptor 1-Fc) and binding of 77-R2C16 or control phage was measured. Phage 77-R2C16 exclusively binds to DR5 and not to any of the other receptors tested. The control phage demonstrated no binding to any of the receptors, confirming that binding of 77-R2C16 to DR5 was via the displayed peptide, and not via other regions of the phage (Fig. 1B). To assess where the peptide binds to the receptor ⁄ ligand inter- face of DR5, we tested whether sTRAIL competes for DR5-binding with phage 77-R2C16. Phage 77-R2C16 was added to the wells at a concentration of 1 · 10 10 plaque forming units (pfu)ÆmL )1 and the sTRAIL concentration was increased. With an increasing concentration of sTRAIL, the binding of the phage to the receptor DR5 decreased, suggesting that phage 77-R2C16 attaches to a binding patch on DR5 over- lapping with sTRAIL (Fig. 1C). Competition studies with synthetic constrained peptides Because the phage-displayed peptide may have differ- ent binding characteristics compared to the constrained peptide alone, the corresponding constrained peptide YCKVILTHRCY (peptide R2C16) was synthesized. Tyrosine residues were added to the ends of this hydrophobic peptide to increase its solubility. During the synthesis, next to the monomer, a dimeric peptide was also formed. This dimeric peptide was separated from the monomeric peptide by HPLC. The mass of the dimeric peptide and measurements using MALDI- TOF indicated that all cysteines in the dimeric peptide Table 1. Sequences of the 25 clones picked after three rounds of biopanning against DR5. The sequences are denoted by the single-letter amino acid code. Of these 25 clones, 14 clones showed binding to DR5 with ELISA (indicated by an asteriak). These 14 clones show a strong consensus sequence with valine at position 2, leucine at position 4, threonine at position 5 and a basic amino acid (arginine or lysine) at position 7. The random residues are shown in bold; fixed cysteines and the preceding alanine in are shown in normal text; and consensus residues are shaded grey. Clone Sequence 77-R2C16 ACKVILTHRC * 89-R2C2 ACKVILTHRC * 77-R2C5 ACKVALTLRC * 77-R2C12 ACKVALTLRC * 77-R2C15 ACKVALTLRC * 77-R2C18 ACKVALTLRC * 77-R2C20 ACKVALTLRC * 77-R2C8 ACLVYLTQRC * 77-R2C19 ACLVYLTQRC * 89-R2C5 ACLVYLTQRC * 77-R2C2 ACIVYLTQKC * 77-R2C3 ACIVYLTQKC * 77-R2C13 ACIVYLTQKC * 77-R2C1 ACILYLTQKC * 77-R2C4 ACKLAMTMKC 77-R2C9 ACKLAMTMKC 89-R2C4 ACKLAMTMKC 77-R2C6 ACFLVMSQRC 77-R2C10 ACLWFPREQC 77-R2C14 ACLWFPREQC 89-R2C3 ACMLPLYFPC 77-R2C11 ACELPRSPSC 77-R2C7 ACTVPAFPAC 89-R2C1 ACTNSAMADC 77-R2C17 ACKHEPTPNC Consensus ACKVYLTQRC LA HK II L J. Vrielink et al. Inhibition of DR5 signalling FEBS Journal 277 (2010) 1653–1665 ª 2010 The Authors Journal compilation ª 2010 FEBS 1655 were oxidized and that covalent bonds were formed between the two monomers. This suggests that the two monomers are linked to each other via disulfide bridges. To determine the orientation of the two monomeric peptides in relation to each other (i.e. par- allel or symmetrical), the dimeric peptide was digested with trypsin and measured using MALDI-TOF. The analysis showed that the dimeric peptide sample con- tained both parallel and antiparallel orientated mono- mers (data not shown). The monomeric and dimeric peptides were used in competitive studies with ELISA. By adding increasing concentrations of the peptide R2C16, a competition with the phage 77-R2C16 (1 · 10 10 pfuÆmL )1 ) for DR5-binding could be seen with ELISA. A known TNFa antagonist peptidomimetic, WP9QY [36], was used as a control peptide. When used at the same con- centrations, it did not compete with phage 77-R2C16 for binding to DR5 (Fig. 2A). This indicates that the peptide R2C16, in both monomeric and dimeric forms, is capable of binding to DR5. Competitive ELISA was also used to analyse the effect of an increasing concen- tration of the monomeric and dimeric forms of R2C16 on TRAIL DR5-binding. In this competition ELISA, the amount of bound sTRAIL was measured. The results obtained show that both forms of the R2C16 peptide could compete with sTRAIL for binding to DR5, not only confirming that the R2C16 peptide is indeed a DR5-binding peptide, but also suggesting that R2C16 and TRAIL bind to an overlapping area on DR5 (Fig. 2B). Binding studies with surface plasmon resonance Binding of the monomeric and dimeric form of peptide R2C16 to immobilized DR4-Fc and DR5-Fc receptor was assessed in real time by using surface plasmon resonance. Both forms of R2C16 bound in a dose-depen- dent manner to DR5 (Fig. 3A, B). It was observed that, after saturation was reached, injection of higher concentrations of peptides resulted in increasing response units (RUs), indicating the accumulation of the peptide. Furthermore, at higher concentrations (> 2000 nm monomer or > 120 nm dimer), some binding to DR4 was observed (Fig. 3C, D). Because of the hydrophobic nature of the peptide, we consider that the accumulation on DR5 and binding to DR4 is caused by aggregation of the peptides. Thus, for K d determination, we decided to use the lane coated with DR4-Fc as a control lane instead of an empty lane. The signal obtained at equilibrium (176 s after injec- tion) was plotted against the concentration of the pep- tide and apparent K d values were calculated from these A B C Fig. 1. ELISAs with phage 77-R2C16 and control phage. (A) Wells are coated with DR5-Ig. Phage 77-R2C16 (•) bound to DR5 in a dose- dependent manner; control phage ( ) did not show any binding. This indicates that binding to DR5 is not mediated by nonspecific binding of the phage particle. (B) Wells are coated with different receptors of the TNF-family. Phage 77-R2C16 showed specific binding to DR5, and control phage showed no binding to any of the receptors. This confirms that binding of the phage 77-R2C16 to DR5 is via the dis- played peptide, and not via other parts of the phage particle. (C) Com- petition ELISA of sTRAIL with phage 77-R2C16 (1 · 10 10 pfuÆmL )1 ) for binding to DR5. Increasing amounts of sTRAIL decreased the binding of phage 77-R2C16 to DR5, suggesting that phage 77-R2C16 and sTRAIL bind to an overlapping region on DR5. Inhibition of DR5 signalling J. Vrielink et al. 1656 FEBS Journal 277 (2010) 1653–1665 ª 2010 The Authors Journal compilation ª 2010 FEBS plots. The monomer has an apparent K d value of 272 nm (range 251–294 nm) and the dimer has a value of 40 nm (37–44 nm) (Fig. 4A, B). Binding studies towards Jurkat cells with fluorescence-activated cell sorting (FACS) analysis To further evaluate the binding of the monomeric and dimeric peptides, we characterized their binding towards Jurkat cells by flow cytometry. Jurkat cells are a widely used model of DR5 only cells. The con- strained peptide R2C16 was synthesized with a biotin label at the C-terminus (YCKVILTHRCY-K[biotin]). Again, both a monomeric and a dimeric form of the peptide were formed and they were separated from each other by HPLC. Both forms of biotin-R2C16 bound in a dose-dependent manner to DR5 (Fig. 5A1, B2). Compared to the control, the increasing amounts of biotinylated monomeric and dimeric peptide showed an increased fluorescence signal. At higher concentra- tions of peptide (> 23 nm biotinylated monomer or >12nm biotinylated dimer), the fluorescence signal suddenly and drastically dropped to almost control levels (data not shown). Again, this suggests the for- mation of aggregates as a result of the hydrophobic nature of the peptides, which would only be increased by the addition of the biotin label. To determine the specificity of this binding interac- tion, the Jurkat cells were co-incubated with the bioti- nylated peptides (5.71 nm of monomer and 1.43 nm of dimer) and increasing concentrations of sTRAIL (0.21 pm to 2.1 nm). Compared to the monomeric or dimeric biotinylated-R2C16 alone, analyses showed that the fluorescence signal can be gradually reduced by increasing the concentrations of sTRAIL (Fig. 5A2, A3, B2, B3). This indicates that the bioti- nylated peptides and sTRAIL bind at a similar posi- tion on the cells, most likely at an overlapping patch of DR5. Constrained R2C16 peptide inhibits TRAIL-induced apoptosis To assess the effect of R2C16 on DR5 apoptosis induction, Colo205 colon carcinoma cells were used. We have previously reported that Colo205 cells were sensitive to TRAIL-induced apoptosis and the TRAIL-death signal was primarily transmitted by DR5 in these cells [25]. Treatment of the cells with increasing concentrations of the monomeric R2C16 peptide caused no cell death, as measured by annexin V labelling of the dying cells (Fig. 6A). Treatment with the dimeric form of R2C16 lead to similar results, where only the highest concentration (3.6 lm) caused a small (7.1 ± 3.2%) increase of cell death (Fig. 6A). Next, the possibility of antagonistic action of R2C16 was tested. Colo205 cells were treated with increasing concentration of monomeric or dimeric R2C16 for 1 h before treatment with 20 ngÆmL )1 TRAIL for 2 h and annexin V staining was used to quantify cell death. Both forms of R2C16 were able to inhibit TRAIL- induced cell death, with the dimeric form being more efficient than the monomer. In addition, the dimeric A B Fig. 2. (A) Competition ELISAs of peptide R2C16, in both mono- meric and dimeric forms, and control peptide WP9QY with phage 77-R2C16 (1 · 10 10 pfuÆmL )1 ). Increasing amounts of monomer (•) and dimer ( ) competed with phage 77-R2C16 for binding to DR5-Ig, whereas peptide WP9QY ( ) did not. This showed that the monomeric form of the dimeric peptide can bind to DR5. (B) Competition ELISA of peptide R2C16, in both monomeric and dimeric forms, with sTRAIL (10 ngÆmL )1 ). The amount of sTRAIL binding was measured. Increasing amounts of monomer (•) and dimer ( ) competed with sTRAIL for binding to DR5-Ig, not only confirming that the R2C16 peptide is indeed a DR5-binding peptide, but also suggesting that R2C16 and TRAIL bind in an overlapping area on DR5. J. Vrielink et al. Inhibition of DR5 signalling FEBS Journal 277 (2010) 1653–1665 ª 2010 The Authors Journal compilation ª 2010 FEBS 1657 form of R2C16 acted as an antagonist best in the con- centration range 1–5 lgÆmL )1 (equal to 0.36–1.8 lm). Above this concentration, the efficiency of the peptide to inhibit TRAIL-induced cell death was reduced (Fig. 6B). This reduction may be caused by aggrega- tion of the peptide. Discussion TRAIL and its receptors (DR4, DR5, DcR1, DcR2 and OPG) are of high interest because of the potential of TRAIL to specifically induce apoptosis in cancer cells, as well as its involvement in many diseases with dysregulated apoptosis. Using phage display, we identi- fied peptides that share a homologous consensus, C(K ⁄ I ⁄ L)V(Y ⁄ I ⁄ A)LT(Q ⁄ H ⁄ L)(K⁄ R)C, and, when dis- played on a phage, bind to DR5. The phage displayed peptide that showed the highest affinity for DR5 was further characterized and shown to bind exclusively to DR5, with no affinity towards any of the other four TRAIL receptors, receptor activator of nuclear factor- jB or TNF-receptor I. The synthetic constrained peptide, YCKVILTHRCY, in both monomeric and dimeric forms, competed with the phage 77-R2C16 and with sTRAIL for binding to DR5 in a concentra- tion-dependent manner and retained DR5 selectivity. The dimeric form of the peptide displayed higher affin- ity for DR5 compared to the monomeric form with an apparent K d value of 40 nm versus 272 nm, respec- tively. This is in accordance with the reported observa- tions in the literature of higher affinities of dimeric and even multimeric peptides compared to their mono- meric form as a result of an avidity effect [30,37–40]. A similar phenomenon was seen for the biotin-labelled synthetic constrained peptide. Both the monomeric and the dimeric form were able to bind to Jurkat cells in a concentration-dependent manner. The dimeric form of the biotin-labelled peptide appeared to have a higher affinity for the Jurkat cells, although part of this effect can be attributed the double biotin label. The higher avidity for the dimeric peptide, which was observed in all binding studies, might also be a conse- quence of the reduced rigidity compared to the mono- meric peptide. This would allow the dimeric peptides to adapt an improved confirmation for DR5-binding. Co-incubation of the Jurkat cells with biotinylated AB CD Fig. 3. BIAcore curves of peptide R2C16, in both monomeric and dimeric forms, binding to immobilized DR5-Ig (A, B) or DR4-Ig (C, D). (A) Increasing concentrations of the monomeric peptide (4086, 2043, 1021, 511, 255, 128, 63.8, 31.9, 16.0 and 8.0 n M) were injected, demonstrating an increased signal. The curves shown are corrected for the signal obtained in the lane coated with DR4-Ig. (B) Increasing concentrations of dimeric peptide (179, 119, 89.5, 59.7, 44.8, 29.8, 22.4, 14.9, 11.2 and 7.5 n M) were injected, demonstrating an increased signal. The curves shown are corrected for the signal obtained in the lane coated with DR4-Ig. (C, D) Increasing concentrations of monomeric (C) or dimeric peptide (D) demonstrated some binding to DR4-Ig at higher concentrations. This binding is most likely caused by aggregation of the peptide as a result of its hydrophobic nature. Inhibition of DR5 signalling J. Vrielink et al. 1658 FEBS Journal 277 (2010) 1653–1665 ª 2010 The Authors Journal compilation ª 2010 FEBS peptide and sTRAIL showed a decrease in peptide binding with increasing sTRAIL concentrations. This suggests that the peptides bind at a similar patch on DR5. Furthermore, both the monomeric and the dimeric peptide acted as DR5 antagonists because they were able to inhibit TRAIL-mediated apoptosis in Colo205 cells, with the dimeric form of R2C16 demon- strating the most efficient antagonistic effect. This is the first DR5-specific antagonistic peptide described. In 2004, Kajiwara et al. [41] described synthetic pep- tides inhibiting TRAIL-induced cell death, although these peptides bound to TRAIL instead of DR5. More recently, Li et al. [32] described peptides binding to DR5, although none of these were antagonistic. We hypothesize that, at higher concentrations, the peptide might aggregate in an aqueous environment based on the fact that the amino acids in the sequences of pep- tide are rather hydrophobic. In addition, we observed that biotinylated was more prone to aggregation than the label-free peptide, which is in accordance with biotin being well known for its hydrophobic character- istics. One mechanism of DR5 that is not yet fully under- stood is the ability to mediate distinct cell signals when interacting with different receptor agonists. An agonistic DR5 monoclonal antibody could induce both caspase-dependent and caspase-independent cell death in Jurkat cells, whereas TRAIL could only trig- ger the caspase-dependent cell death [20]. Thomas et al. [19] found that TRAIL and some agonistic anti- bodies required the C-terminal tail of DR5 for recruitment of Fas-associated death domain, whereas other agonistic antibodies could function in the absence of this C-terminal tail. Thus, different recep- tor agonists can use distinct molecular mechanisms to activate signalling from the same receptor. It is postu- lated that the binding of different agonists to the extracellular domain causes different conformational changes in the intracellular domain, which may inter- act with different cytoplasmic adaptor proteins and trigger different cell signals [20]. To address these questions, the R2C16 antagonistic peptide described in the presrent study comprises a useful tool for eluci- dating the mechanisms of binding and signalling initi- ation of DR5. Accordingly, the manner in which the peptide is able to antagonize TRAIL should be fur- ther elucidated by studying the trimerization of DR5. In addition, the effects of the peptide on the different cell signals that it can trigger or block should be investigated. Up to now, the focus of TRAIL research has been mainly on its therapeutic value in cancer, as a result of the quality of TRAIL that leads to the induction of apoptosis in a broad range of cancer cells but not in most untransformed cells [1]. However, the involve- ment of TRAIL in neurodegenerative diseases has not received much attention, despite the mounting evidence emphasizing the role of TRAIL in these disorders. Recent data show that, although TRAIL is absent in normal brain, it is upregulated under pathological con- ditions such as Alzheimer’s disease. Human brain cells express all four TRAIL receptors and are sensitive to TRAIL-induced apoptosis. TRAIL is also suggested to be involved in the pathogenesis of HIV-1-associated dementia [2]. HIV infection triggers TRAIL expression in macrophages and these TRAIL expressing macro- phages can initiate neuronal injury. The involvement of TRAIL in Alzheimer’s disease was shown by A B Fig. 4. Dose–response curves of increasing concentration of pep- tide R2C16, (A) monomeric (•) or (B) dimeric ( ), binding to DR5-Ig measured with surface plasmon resonance. The amount of binding is depicted as the RU measured at 176 s. Graphs were fit with four-parameter sigmoid curves. Apparent K d values were calculated as the concentration of peptide that gives a signal of 50% of the maximum RU. J. Vrielink et al. Inhibition of DR5 signalling FEBS Journal 277 (2010) 1653–1665 ª 2010 The Authors Journal compilation ª 2010 FEBS 1659 neutralization of the TRAIL death pathway, which protected a human neuronal cell line from b-amyloid toxicity [2]. TRAIL plays a dual role in T cell-induced experimental autoimmune encephalomyelitis (EAE), an animal model of multiple sclerosis. Blockade of TRAIL within the periphery exacerbates EAE, whereas blockade of TRAIL in the central nervous system sup- presses EAE by inhibiting brain cell apoptosis [42]. Inhibition of TRAIL-induced apoptosis within the cen- tral nervous system may represent a possible therapeu- tic strategy for preventing neuronal damage in patients with neurodegenerative diseases. The R2C16 peptide, with its DR5 antagonistic activity and lipophilic prop- erties, has the potential to act as a lead peptide in studies aiming to block TRAIL and reducing its toxic- ity in neurodegenerative diseases. Overall, the peptides described in the present study, or their derivatives, may have various applications in the field of TRAIL-mediated signalling and diseases caused by dysregulated TRAIL signalling. The small size of this peptide offers the possibility of designing structurally mimetic nonpeptidic molecules. A1 B1 A2 B2 A3 B3 Fig. 5. Dose– response histograms of biotinylated peptide R2C16, (A1) monomeric or (B1) dimeric, on Jurkat cells. Increasing amounts of both monomeric (m1 = 0.57 n M, m2 = 2.86 n M, m3 = 11.42 nM) and dimeric (d1 = 0.14 n M, d2 = 1.43 nM, d3 = 5.71 nM) biotinylated peptide give rise to a right shift in the fluorescent PE signal compared to the control (C, filled grey). Competition his- tograms of 5.71 n M monomeric biotinylated peptide (m, tinted) (A2) and 1.43 n M dimeric biotinylated peptide (d, tinted) (B2) show a strong right shift compared to control (C, filled grey). Co-incubation of the biotiny- lated peptides with 2.1 n M sTRAIL [m + T (A2) and d + T (B2), respectively] show a decrease in the right shift that they were initially able to cause. The curves represent the relative binding signal binding of the (A3) monomer and (B3) dimer upon co- incubation with sTRAIL compared to the signal obtained when no sTRAIL was used. Inhibition of DR5 signalling J. Vrielink et al. 1660 FEBS Journal 277 (2010) 1653–1665 ª 2010 The Authors Journal compilation ª 2010 FEBS Experimental procedures TRAIL purification cDNA corresponding to soluble human TRAIL (C-terminal amino acids 114–281) was cloned into the NcoI and BamHI sites of a pET15b vector and transformed to and expressed in Escherichia coli BL21 (DE3). The trimeric soluble protein was purified as described previously [43]. Biopanning The Ph.D.–C7C phage display peptide library (New Eng- land Biolabs, Hitchin, UK), consisting of randomized hept- amer constrained peptides (1.2 · 10 9 individual clones), was used to identify peptides binding to DR5. The disulfide- constrained heptapeptides are expressed at the N-terminus of g3p, with the first cysteine preceded by an alanine residue, and the second cysteine followed by a short spacer (Gly-Gly-Gly-Ser). DR5-Fc, fusion of the ecto-domain of the receptor to the Fc-portion of human IgG 1 (R&D Sys- tems, Minneapolis, MN, USA), was used to coat Protein A magnetic dynabeads (Dynal, Hammerfest, Norway). To 6 lL of beads, 1 lg of DR5-Fc was added in 0.1 m NaH- CO 3 (pH 8.6) and incubated overnight at 4 °C. The beads were then blocked using 0.1 m NaHCO 3 (pH 8.6), 5mgÆmL )1 BSA and 0.02% NaN 3 . Phage library was added to the beads at 2 · 10 11 pfu and incubated for 45 min. Unbound phages were removed by washing ten times with washing buffer (NaCl ⁄ Tris containing 0.1% Tween 20). Bound phages were eluted with 1 mL of 0.2 m glycine ⁄ HCl (pH 2.2), 1 mgÆmL )1 BSA for no more then 10 min, and immediately neutralized with 150 l Lof1m Tris-HCl (pH 9.1). The eluted phages were amplified in E. coli ER2738 and titred according to the manufacturer’s instructions (New England Biolabs). Another two rounds of biopanning were then performed. The incubation time was decreased to 30 min in the second round and to 15 min in the third round. In both rounds, the concentration of Tween 20 in the washing buffer was increased to 0.5%, and 1 lm sTRAIL in NaCl ⁄ Tris was used for competitive elu- tion. Before the third round, a subtractive round was per- formed, by incubating the amplified phage of round two with human IgG Fc fragment (Rockland Immunochemi- cals, Inc., Gilbertsville, PA, USA) bound to protein A beads. To 20 lL of beads, 10 lg of Fc fragment was added, incubated and blocked as described above. Phages were incubated for 30 min with the beads, and unbound phages were subsequently used in the third positive panning round. After round three, 25 individual phage clones were picked from agar plates and amplified. The single-stranded DNA was isolated and sequenced according to the manufacturer’s instructions. Supernatants were screened using ELISA, as described below. Samples that showed a high binding signal were further purified using poly(ethylene glycol) precipita- tion according to the manufacturer’s instructions (New England Biolabs). ELISA with phage Maxisorp 96-wells plates (Nunc, Roskilde, Denmark) were coated for 1–2 h with 100 lLof1ngÆlL )1 receptor-Fc A B Fig. 6. Annexin V cell assays with peptide R2C16 using Colo205 cells. (A) Treatment of cells with increasing concentrations of the monomeric R2C16 peptide (•) caused no cell death. Treatment with the dimeric form of R2C16 ( ) gave similar results. Only the high- est concentration (8 l M) caused a small (7.1 ± 3.2%) increase of cell death, suggesting that the R2C16 peptide is not a DR5 agonist. (B) Reduction of TRAIL-induced cell death in Colo205 cells by pep- tide R2C16. Colo205 cells were treated with increasing concentra- tion of monomeric (•) or dimeric ( ) R2C16 before treatment with 20 ngÆmL )1 TRAIL and annexin V staining was used to quantify the dying cells. Treatment of Colo205 cells with only 20 ngÆmL )1 TRAIL induced 80% cell death. Both forms of R2C16 were able to reduce TRAIL-induced cell death, with the dimeric form being more efficient than the monomer. The dimeric form of R2C16 acted as an antagonist best in the concentration range 1–5 lgÆmL )1 (equal to 0.76–3.8 l M). Above this concentration, the efficiency of the peptide to inhibit TRAIL-induced cell death was reduced, most likely as a result of aggregation of the peptide. J. Vrielink et al. Inhibition of DR5 signalling FEBS Journal 277 (2010) 1653–1665 ª 2010 The Authors Journal compilation ª 2010 FEBS 1661 (R&D Systems) at 4 °C in 0.1 m NaHCO 3 (pH 8.6). The wells were blocked for 1–2 h with 200 lL of 2% BSA in 0.1 m NaHCO 3 and washed three times with NaCl ⁄ Tris, 0.5% Tween 20 (TBST). Phage supernatant, purified phage diluted in TBST with 0.5% BSA or a premix of different concentration of synthesized peptides with 1 · 10 10 pfuÆmL )1 of phage was added at 100 lL per well and incu- bated for 30 min. After washing the wells six times with TBST, 200 lL of 1 : 5000 diluted HRP ⁄ anti-M13 monoclo- nal conjugate (Amersham Pharmacia Biotech, Little Chal- font, UK) in TBST was added for 1 h. Wells were washed six times with TBST and bound phages were detected with 100 lL of tetramethylbenzidine (TMB; one-step Turbo TMB- ELISA) (Pierce Biotechnology, Rockford, IL, USA). The reaction was stopped with 100 lL of approximately 1.8 m H 2 SO 4 . The signal was read at 450 nm in a multiscan Ascent plate reader (Thermo Labsystems, Helsinki, Finland). Curves were fitted with four-parameter sigmoid curves. Competition ELISA with peptides and sTRAIL Maxisorp plates were coated for 1–2 h with 100 lLof 1ngÆlL )1 DR5-Fc at 4 °C in 0.1 m NaHCO 3 (pH 8.6). The wells were blocked for 1–2 h with 200 lL of 2% BSA in 0.1 m NaHCO 3 (pH 8.6) and washed three times with TBST. Different concentrations of synthesized peptides were premixed with 10 ngÆmL )1 sTRAIL in TBST and 100 lL of this premix was added to each well, and incu- bated for 30 min. After washing the wells six times with TBST, 200 lL of 1 : 200 diluted anti-human TRAIL sera (AF375; R&D Systems) in TBST was added to the wells for 1 h, washed six times with TBST and 200 lLof 1 : 25 000 diluted HRP conjugated swine anti-goat (Bio- Source International, Camarillo, CA, USA) was added for 1 h. After washing wells six times with TBST, 100 lLof TMB was added to measure the amount of bound sTRAIL. The reaction was stopped with 100 lL of approximately 1.8 m H 2 SO 4 . The signal was read at 450 nm. Curves were fitted with four-parameter sigmoid curves. Synthetic peptides The constrained peptide (YCKVILTHRCY), in both mono- meric and dimeric forms, was synthesized by Pepscan (Lelys- tad, the Netherlands). Both peptides had a free amine at the N-terminal and a free acid at the C-terminal. The lyophilized peptides were solubilized in acetonitril ⁄ water (1 : 1) to give a stock concentration of 20 mgÆmL )1 . A control peptide (YCWSQYLCY) was purchased from Bachem AG (Buben- dorf, Switzerland). The lyophilized peptide was solubilized in 50% acetic acid to give a stock concentration of 10 mgÆmL )1 . The stock of each peptide was stored at )20 °C. The constrained peptide with biotin label (YCK- VILTHRCY-K[biotin]), in both monomeric and dimeric forms, was also synthesized by Pepscan Systems, again with a free amine at the N-terminal and the biotin label at the C-terminus. The biotin label was coupled to the C-terminal tyrosine, resulting in two biotin labels for the dimeric peptide. The lyophilized peptides were solubilized in acetonitril ⁄ water (1 : 1) to give a stock concentration of 20 mgÆmL )1 . The stock of each peptide was stored at )20 °C. Interaction studies by surface plasmon resonance To evaluate the binding of the peptides to DR5, BIAcore 2000 (BIAcore AB, Uppsala, Sweden) was used. All reagents used were also purchased from BIAcore AB. Immobilization of the receptor DR4-Fc and DR5-Fc on the sensor surface of a CM5-chip was performed in accor- dance with a standard amine coupling procedure at a flow rate of 10 lLÆmin )1 . To activate carboxyl groups on the sensor surface, 70 lL of a solution containing 0.2 m 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochlo- ride and 0.05 m N-hydroxysuccinimide was injected. Recep- tors (6.7 lgÆmL )1 in 10 mm NaAc buffer, pH 5.0) were flowed over the chip surface until a surface density of approximately 3500 RU was reached. Remaining active groups were blocked by injecting 70 lL of 1.0 m ethanol- amine-HCl (pH 8.5). Assays were performed at 25 °C with NaCl ⁄ Po containing 0.005% (v ⁄ v) P20 surfactant as running buffer and a flow rate of 70 lLÆmin )1 . Peptides were injected at variable concentrations for 3 min followed by 4 min of running buffer. The surface was regenerated after each binding step by removing all bound peptide by injecting 35 lLof10mm glycine (pH 2.0). Sensorgrams were evaluated with BIAevaluation software, version 4.1 (Biacore, GE Healthcare, Chalfont St Giles, UK). The lane coated with DR4-Fc was used as a reference. The signal obtained at equilibrium (176 s after start injection) was plotted against the concentration of the peptide and fitted with four-parameter sigmoid curves. From these curves, the apparent K d values were calculated. Interaction studies on Jurkat cells by FACS analysis Jurkat cells were maintained in RPMI 1640 medium + GlutaMAX-I supplemented with 10% fetal bovine serum, 100 UÆmL )1 penicillin and 100 lgÆmL )1 strep- tomycin (all from Gibco, Gaithersburg, MD, USA) at 37 °C in 5% CO 2 in a humidified environment. Cells were harvested at 10 6 cells per sample and washed in ice-cold NaCl ⁄ Pi ⁄ 2% fetal bovine serum to prevent peptide reduc- tion by the reducing agent glutathione present in the RPMI medium. Increasing concentrations of biotin-labelled pep- tides were added to the cells and incubated for 1 h on ice. The cells were washed twice in ice-cold NaCl ⁄ Pi ⁄ 2% fetal bovine serum. The cells were incubated for an additional 1 h with streptavidin-phycoerythrin (PE) (BD Pharmingen, Inhibition of DR5 signalling J. Vrielink et al. 1662 FEBS Journal 277 (2010) 1653–1665 ª 2010 The Authors Journal compilation ª 2010 FEBS [...]... inducing ligand death receptor DR5 prevents beta-amyloid neurotoxicity Neuropsychopharmacology, 32, 872–880 5 Pan G, O’Rourke K, Chinnaiyan AM, Gentz R, Ebner R, Ni J & Dixit VM (1997) The receptor for the cytotoxic ligand TRAIL Science, 276, 111–113 6 Pan G, Ni J, Wei YF, Yu G, Gentz R & Dixit VM (1997) An antagonist decoy receptor and a death domain-containing receptor for TRAIL Science, 277, 8 15 818 7... LeBlanc HN & Ashkenazi A (2003) Apo2L ⁄ TRAIL and its death and decoy receptors Cell Death Differ, 10, 66– 75 11 Pan G, Ni J, Yu G, Wei YF & Dixit VM (1998) TRUNDD, a new member of the TRAIL receptor family that antagonizes TRAIL signalling FEBS Lett, 424, 41– 45 FEBS Journal 277 (2010) 1 653 –16 65 ª 2010 The Authors Journal compilation ª 2010 FEBS 1663 Inhibition of DR5 signalling J Vrielink et al 12 Emery JG,... A & Ruberti G (1999) Identification and characterization of a ligand-independent oligomerization domain in the extracellular region of the CD 95 death receptor J Biol Chem, 274, 38241–38 250 17 Chan FK, Chun HJ, Zheng L, Siegel RM, Bui KL & Lenardo MJ (2000) A domain in TNF receptors that mediates ligand-independent receptor assembly and signaling Science, 288, 2 351 –2 354 18 Siegel RM, Frederiksen JK, Zacharias... signaling and dominant inhibition by pathogenic mutations Science, 288, 2 354 –2 357 19 Thomas LR, Johnson RL, Reed JC & Thorburn A (2004) The C-terminal tails of tumor necrosis factorrelated apoptosis-inducing ligand (TRAIL) and Fas receptors have opposing functions in Fas-associated death domain (FADD) recruitment and can regulate agonist-specific mechanisms of receptor activation J Biol Chem, 279, 52 479 52 486... corresponding to ligand-binding region of death receptors, DR5, Fas, and TNFR, specifically inhibit cell death mediated by the death ligands, respectively Biochim Biophys Acta, 1699, 131–137 42 Aktas O, Smorodchenko A, Brocke S, Infante-Duarte C, Topphoff US, Vogt J, Prozorovski T, Meier S, Osmanova V, Pohl E et al (20 05) Neuronal damage in autoimmune neuroinflammation mediated by the death ligand TRAIL Neuron,... ligand 2 ⁄ tumor necrosis factor-related apoptosis-inducing ligand reveal a greater contribution of death receptor (DR) 5 than DR4 to apoptosis signaling J Biol Chem, 280, 22 05 2212 25 van der Sloot AM, Tur V, Szegezdi E, Mullally MM, Cool RH, Samali A, Serrano L & Quax WJ (2006) Designed tumor necrosis factor-related apoptosis-inducing ligand variants initiating apoptosis exclusively via the DR5 receptor. .. angiogenesis EMBO J, 19, 152 5– 153 3 England BP, Balasubramanian P, Uings I, Bethell S, Chen MJ, Schatz PJ, Yin Q, Chen YF, Whitehorn EA, Tsavaler A et al (2000) A potent dimeric peptide antagonist of interleukin -5 that binds two interleukin -5 receptor alpha chains Proc Natl Acad Sci USA, 97, 6862–6867 Takasaki W, Kajino Y, Kajino K, Murali R & Greene MI (1997) Structure-based design and characterization... of DR5 signalling 39 O’Leary PD & Hughes RA (2003) Design of potent peptide mimetics of brain-derived neurotrophic factor J Biol Chem, 278, 257 38– 257 44 40 Aggarwal S, Harden JL & Denmeade SR (2006) Synthesis and screening of a random dimeric peptide library using the one-bead-one-dimer combinatorial approach Bioconjug Chem, 17, 3 35 340 41 Kajiwara K, Saito A, Ogata S & Tanihara M (2004) Synthetic peptides... Eppendorf tubes and collected by centrifugation at 3000 g for 5 min The cell pellets were resuspended in 50 lL of calcium buffer [10 mm Hepes, pH 7 .5 (set with NaOH), 140 mm NaCl and 2 .5 mm CaCl2] containing 3 lL of annexin V (IQ Corporation, Groningen, the Netherlands) and incubated for 15 min on ice The staining was terminated by diluting the annexin V solution to 300 lL with calcium buffer and the samples... thrombopoietin receptor as potent as the natural cytokine Science, 276, 1696–1699 31 Sato A & Sone S (2003) A peptide mimetic of human interferon (IFN)-beta Biochem J, 371, 603–608 32 Li B, Russell SJ, Compaan DM, Totpal K, Marsters SA, Ashkenazi A, Cochran AG, Hymowitz SG & Sidhu SS (2006) Activation of the proapoptotic death receptor DR5 by oligomeric peptide and antibody agonists J Mol Biol 361, 52 2 53 6 33 . Synthetic constrained peptide selectively binds and antagonizes death receptor 5 Johanna Vrielink 1, *, Mariette S (2004) Synthetic peptides corresponding to ligand-binding region of death receptors, DR5, Fas, and TNFR, specif- ically inhibit cell death mediated by the death