DOI: 10.1002/cbic.201402513
Full Papers
Chemical Synthesis of A Pore-Forming Antimicrobial Protein, Caenopore-5, by Using Native Chemical Ligation at a Glu-Cys Site Karima Medini,[a] Paul W. R. Harris,[a, b] Kiel Hards,[c] Andrew J. Dingley,[d] Gregory M. Cook,[c] and Margaret A. Brimble*[a, b] AMP. The protein (82 amino acids) was successfully synthesised by using Boc solid-phase peptide synthesis and native chemical ligation. No g-linked by-product was observed despite the use of a C-terminal Glu-thioester. The folding of the synthetic protein was confirmed by 1H NMR spectroscopy and circular dichroism and compared with data recorded for recombinant caenopore-5. The permeabilisation activities of the protein and of shortened analogues were evaluated.
The 2014 report from the World Health Organization (WHO) on antimicrobial resistance revealed an alarming rise in antibiotic resistance all around the world. Unlike classical antibiotics, with the exception of a few species, no acquired resistance towards antimicrobial peptides (AMPs) has been reported. Therefore, AMPs represent leads for the development of novel antibiotics. Caenopore-5 is constitutively expressed in the intestine of the nematode Caenorhabditis elegans and is a pore-forming
Introduction Antimicrobial proteins are typically small amphipathic peptides that have positively charged regions that allow them to readily interact with the negative lipid head groups found in abundance in prokaryotic membranes.[1] Additionally, AMPs may also possess secondary functions such as the modulation of the immune response, inactivation of bacterial endotoxins and the induction of cell apoptosis.[2] One group of AMPs recently identified in the nematode Caenorhabditis elegans is the caenopores. There are 33 caenopores that collectively enable C. elegans to survive in its natural habitat.[3] The caenopores consist of a short N-terminal hydrophobic amino acid region representing a signal peptide and a single domain with homology to members of the saposin-like protein (SAPLIP) family.[4] The saposin (spp) genes are expressed primarily in the intestinal epithelial cells of C. elegans and the resultant gene products are secreted into the gut lumen, which provides the optimal environment for antimicrobial activity.[3] Caenopores act directly on microbes that have entered the worm orally, thereby pre-
venting the establishment of intestinal infections. Caenopore-5 (Cp-5) is essential for the survival of C. elegans, because an spp-5 knockdown (T08A9.9 gene, GenBank U40417) induced a range of abnormal features, including developmental retardation, a significant reduction in the number of eggs laid and the inability to accumulate fat. Additionally, spp-5 silenced worms struggled to survive on lawns of Escherichia coli and intact bacteria were able to colonise the intestinal tract, leading to the death of the nematode.[3] The slightly acidic composition of the intestine provides the optimal environment for the function of caenopores; for example, the optimum pH for Cp-5 has been reported as ~ 5.2.[3] Cp-5 is the best functionally characterised member of the caenopore family and structural data has confirmed that the protein adopts the canonical protein fold of SAPLIPs.[4a] The solution structure shows that Cp-5 comprises five a-helices connected by three disulfide bonds from six conserved cysteines.[4a] Recombinant Cp-5 has been shown to be active against both Gram-positive (Bacillus megaterium) and Gramnegative (E. coli) bacteria with minimum inhibitory concentration (MIC) values of 0.05 and 0.1 mm, respectively, and minimum bactericidal concentration (MBC) values of 0.1 and 0.2 mm, respectively.[3] As a first step to design analogues with more potent biological activity, a chemically robust synthesis of the native Cp-5 sequence needs to be developed. To realise this aim, we herein report the synthesis of Cp-5 by native chemical ligation (NCL) of two unprotected peptide segments. The full-length protein was folded using a glutathione redox-coupled reaction and the folded protein was compared with recombinantly produced Cp-5 by using one-dimentional 1H NMR and circular dichroism. Moreover, shortened peptides representing the five
[a] K. Medini, Dr. P. W. R. Harris, Prof. M. A. Brimble Maurice Wilkins Centre for Molecular Biodiscovery School of Biological Sciences, The University of Auckland 3A Symonds Street, Auckland 1010 (New Zealand) E-mail: [email protected] [b] Dr. P. W. R. Harris, Prof. M. A. Brimble School of Chemical Sciences, The University of Auckland 23 Symonds Street, Auckland 1010 (New Zealand) [c] K. Hards, G. M. Cook Department of Microbiology and Immunology School of Medical Sciences, University of Otago, 720 Cumberland Street, Dunedin 9054 (New Zealand) [d] Dr. A. J. Dingley Institute of Complex Systems 6 Structural Biochemistry, Forschungszentrum Jlich 52425 Jlich (Germany)
ChemBioChem 2015, 16, 328 – 336
328
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers individual a-helices in Cp-5 were synthesised to evaluate their structure–activity relationship (SAR) profile.
Second ligation strategy The second strategy (Scheme 2) we tried was devised to use only two fragments, consisting of 35 (fragment 1, Gly1-Glu35) and 47 (fragment 2, Cys36-Pro82) amino acid residues in length (Scheme 1), and exploited the native N-terminal cysteinyl residue (Cys36), thereby avoiding the introduction of nonnative cysteines or derivatives. Because of the sensitivity of thioesters to piperidine, which is used for 9-fluorenylmethyloxycarbonyl (Fmoc) SPPS, Boc SPPS was preferred to synthesise the C-terminal thioester fragment 1 namely Gly1-Glu35COSCH2CH2-Ala-OH.[9] On a scale of 0.1 mmol, we obtained a 40 % yield of fragment 1, which was characterised by HPLC and LC–MS (obs = 802.4 Da [M+5 H]5 + ; calcd = 802.7 Da).
Results and Discussion Synthetic design Caenopore-5 is a protein consisting of 82 amino acid residues (Scheme 1), which renders the direct linear synthesis of the full-length protein difficult using SPPS. NCL is a powerful technique to ligate peptides in order to prepare longer proteins.[5]
Scheme 1. Amino acid sequence of Cp-5,[3] with the ligation sites of the first and second strategies marked with arrows.
This method requires a peptide fragment with a thioester on the C terminus, which undergoes a chemoselective reaction with a second fragment bearing a cysteine residue on the N terminus, and results in the formation of a native amide bond.
First ligation strategy The first strategy we tried used two ligation sites, which would give plausibly sized polypeptide fragments for solid-phase synthesis. The sequence of Cp-5 reveals two possible ligation sites, which could be assembled from three fragments comprising 20–33 amino acids each. Ser20-Ala21 and Tyr53-Val54 were identified as possible ligation sites; however these sites require N-terminal cysteinyl residues (as shown in Scheme 1) for native chemical ligation. Therefore, Ala21 and Val54 were substituted by the introduction of non-native cysteine[6] and penicillamine (Pen),[7] respectively, as surrogates for NCL, which requires a subsequent desulfurization step to regenerate the native sequence.[8] Although we could synthesise the requisite peptide fragments by Boc SPPS, namely the 1-20 thioester, Thz-21-53-thioester (Thz = thiazolidine) and Pen-54-82 fragments and the ligation was successful, the Pen ligation was slow (five days) and significant hydrolysis at the tyrosine thioester was observed. Moreover, the desulfurization of Pen54 to Val54 on the ligated 1–82 peptides was slow and side products were observed, reducing the yield of the reaction to 10 %. The final yield was only 3 %. ChemBioChem 2015, 16, 328 – 336
www.chembiochem.org
Scheme 2. The second ligation strategy for the chemical synthesis of caenopore-5. The cartoon representation of Cp-5 was generated using PyMOL from the solution structure (PDB ID: 2JS9).[4a] The disulfide bonds are shown as sticks. a) Boc-Ala-Pam, DIC, CH2Cl2, 1 h; b) N-terminal Boc removal; c) TrtSCH2CH2CO2H, HATU, iPr2EtN, DMF, 15 min, d) TFA/TIPS/H2O (95:2.5:2.5, v/v/v) 2 1 min; e) Boc-Glu-(OcHex)-OH, HATU, iPr2EtN, DMF, 1 h; f) Boc-AAOH HATU, iPr2EtN, DMF, 5 min; g) HF/p-cresol (20:1, v/v), 1 h at 0 8C; h) 0.2 m Na2HPO4/6 m Gn·HCl, 100 mm MPAA, 20 mm TCEP, pH 7.0.
329
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers
Scheme 3. Modified pathway proposed by Botti et al. for the formation of a- and g- linked peptides product during NCL between Glu-thioester and A) MPAA or a B) Cys-peptide 2.[12]
Figure 1. LC–MS analysis of MPAA exchange using a short C-terminal Glu-thioester peptide (R = CH2-Ala-OH; reaction conditions: MPAA (100 mm), Gd·HCl (6 m), Na2HPO4 (0.2 m), TCEP·HCl (20 mm), at RT and pH 7).
a microwave synthesiser.[10] Although the desired peptide was detected by LC–MS, an elution peak corresponding to an aspartyl piperidide peptide [M+67 Da] was observed (50 %),
The synthesis of fragment 2 (47 amino acid residues), which does not contain a sensitive thioester, was first attempted by Fmoc SPPS on a 2-chlorotrityl chloride resin with the aid of ChemBioChem 2015, 16, 328 – 336
www.chembiochem.org
330
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers 0.2 m Na2HPO4, 20 mm TCEP·HCl at pH 7.0), and the results are shown in Figure 1 A. The reaction was monitored by LC–MS using a slow gradient (0.1 % B/min) at room temperature to reveal the formation of any by-products, and there was no evidence to suggest that the cyclic intermediate that would lead to a g-linked by-product was formed. Rather, a single species was observed after 3 h (Figure 1), namely IKKDFDAE-MPAA thioester. Because the ligation of the peptide aryl thioester with N-terminal cysteinyl fragment 2 is fast, it was concluded that formation of a g-linked product would not be significant in our key ligation step. NCL of fragments 1 and 2: Ligation of fragments 1 and 2 (Scheme 2) was next undertaken by adding 100 mm MPAA[14]
probably owing to the Asp73-Lys72 motif (despite the use of the less basic piperizine for Fmoc deprotection).[11] We therefore used Boc SPPS to synthesise fragment 2 using HATU and iPr2EtN as the activating agent and base, respectively. The 47 residue peptide was obtained in good yield (45 %) with no aspartimide by-product and the purity was confirmed by integration of the HPLC chromatogram at 210 nm and LC–MS (obs = 881.6 Da [M+6 H]6 + ; calcd = 881.8 Da). Native chemical ligation
NCL at the Glu35-Cys36 site: The NCL site was chosen at position Glu35-Cys36 of the peptide chain. However, it is known that ligations between Asp-Cys and Glu-Cys residues can result in the migration of the ligation site onto the side chain, generating b- and g-linked by-products, respectively.[12] Scheme 3 shows the initial formation of a cyclic product from a peptide bearing a C-terminal Glu-thioester. In the presence of either MPAA (1) or an unprotected peptide containing a thiol at the N terminus (a Cys in our example, 2), two possible ring-opening reactions can occur. In the case of attack of MPAA or unprotected peptide 2 at the a-carbonyl group, the native bond is preserved. However, if MPAA or peptide 2 reacts at the g-carbonyl group of the cyclic anhydride, an unnatural glinked by-product is formed (Scheme 3). In a study conducted by Kent et al., the authors reported the formation of 3 % of the unnatural g amide bond at a Glu-Cys site, whereas 25 % of unnatural b amide bond formation took place at an Asp-Cys site during NCL.[13] In our present work, in order to study the formation of this side product during the ligation, a truncated moiety of the fragment 1 thioester was synthesised (IKKDFDAE-thioester). This short peptide was synthesised using in situ Boc SPPS, and the same procedure as fragment 1 (Scheme 3). The short peptide was reacted with MPAA under the same conditions that Figure 2. HPLC chromatogram of native chemical ligation at A) t = 0 and B) after completion at t = 10 h; the asterwere to be used for NCL of Cp-5 isk (*) represents the MPAA catalyst. C) After purification, the protein was folded. D) From the linear sequence, (100 mm MPAA, 6 m Gn·HCl, a shift of the retention time was observed from 17.30 to 16.65 min after folding. ChemBioChem 2015, 16, 328 – 336
www.chembiochem.org
331
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers and 20 mm TCEP to 1 mm peptide, at pH 7. The reaction was monitored by LC–MS and at t = 0 h we observed that some of fragment 1 reacted to form an MPPA-reactive thioester intermediate by way of a transthioesterification reaction (Figure 2 A). Although co-elution of the peptide fragments hampered the HPLC analysis, the ligated peptide 3 was the major product after 10 h (Figure 2 B), as evidenced by mass spectrometry (obs = 912.4 Da [M+10 H]10 + ; calcd = 912.6 Da). The ligated peptide 3 was purified by HPLC (Figure 2 C) to give 40 % yield. To obtain active Cp-5, which contains three disulfide bonds, the linear sequence was folded in buffer with 50 mm Tris·HCl, 150 mm NaCl and a redox system (final concentration: 10 mm reduced and 2 mm oxidised glutathione, pH 7.4) at 4 8C. The redox-coupled reaction enabled a facile thiol/disulfide exchange process and promoted the formation of the correctly folded protein. After 6 h, we observed a shift of the retention time from 17.30 min (unfolded) to 16.65 min (folded; Figure 2 D). After 12 h, the folding was deemed complete and the protein was purified by HPLC and lyophilised. The pure folded final product (Figure 3) was characterised by LC–MS (obs = 911.8 Da [M+10 H]10 + ; calcd = 912 Da], which corresponded to the loss of six hydrogen atoms as expected for the formation of three disulfide bonds (80 % yield).
Figure 4. HPLC chromatogram and MS of recombinant Cp-5.
fied using HPLC (4 mg of pure recombinant protein was obtained) and the mass was confirmed by LC–MS. The observed molecular mass of 9368 0.8 Da [M+H] + (Figure 4) was in agreement with the calculated average isotopic mass of 9369.7 Da.
Confirmation of the protein secondary structure and protein fold NMR spectroscopy has been shown to be a reliable technique to determine how well folded a protein is. One-dimensional 1 H NMR can provide rapid information on the folded state of a protein,[17] and proteins that are folded generally give a welldispersed set of resonances in the amide region of the 1D 1 H NMR spectrum.[18] Before the folding reaction of synthetic Cp-5, a 1D 1H NMR spectrum was recorded and gave a strong set of resonances clustered at approximately 8.3 ppm, confirming that Cp-5 was unstructured (Figure 5 A). After folding, the
Figure 3. HPLC chromatogram and MS of synthetic Cp-5.
Recombinant expression and purification of Cp-5 Recombinantly expressed Cp-5 was purified according to a slightly modified procedure from Mysliwy et al.[4a] The Cp-5 ORF (open reading frame) was cloned into the pProEx HTb vector[15] with an rTEV cleavage site rather than the original Factor X cleavage site,[16] and the protein was solubly expressed in E. coli using lysogeny broth medium (4 L). Thus, the protein was produced as a folded protein. After the first step of purification using immobilized metal affinity chromatography (IMAC), the His6-tag was successfully removed by overnight digestion with rTEV protease. The resulting protein amino-acid sequence was identical to the synthetic one except for the presence of an extra three N-terminal residues (Glu-AlaMet = GAM). These residues were generated recombinantly from the pProEx HTb vector linker between the rTEV recognition site and Cp-5. The recombinant protein was further puriChemBioChem 2015, 16, 328 – 336
www.chembiochem.org
Figure 5. A) The 1D 1H NMR spectra of Cp-5 before and after folding, recorded in sodium phosphate buffer (20 mm, pH 5.2) at 25 8C. B) Far-UV CD spectra of Cp-5. Grey line: synthetic Cp-5; black line: recombinant Cp-5.
332
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers It appears that the peptides of the five a-helices of Cp-5 are too short to exhibit activity, because they probably lack a combination of key residues required for membrane permeabilisation. It is possible that the peptides do interact transiently with the membrane, but do not permeabilise the membrane to release the dye.
spectrum showed a distinct set of well-dispersed resonances, indicating that synthetic Cp-5 was folded (Figure 5 A). As a comparison, the 1D 1H NMR spectrum of recombinant Cp-5 is also shown in Figure 5 A. The 1D 1H NMR spectroscopic data suggested that both proteins are folded. Chemical shift dispersion was also observed in the aliphatic region of the 1D 1H NMR spectrum of the synthetic protein (data not shown), further indicating that the protein has folded. We suspect that the slight difference between these two NMR spectra might be due to the presence of the tag (GAM) in the recombinant protein. Folded synthetic and recombinant Cp-5 were also investigated by far-UV/CD spectroscopy (180–300 nm).[19] The CD spectra (Figure 5 B) demonstrated that both the synthetic and recombinant Cp-5 exhibit the features expected of a-helical proteins, with the standard double negative ellipticity maxima at 208 and 222 nm, and a positive maximum near 194 nm.
Key residues involved in the membrane permeabilisation? The primary sequence of the N-terminal region of Cp-5 that showed the highest activity (1–35 amino acids) was compared with other SAPILIP family members (Figure 7). The proteins amaoebapore A (APA-1) and saposin-C belong to the same family as Cp-5 and by comparing them with the Cp-5 sequence, information about their functional properties can be obtained. The protein sequences of these SAPLIP members are
Table 1. Sequence of Cp-5 analogues.
Synthesis of shortened analogues
Peptide Sequence
In order to study the SAR profile of caenopore-5 with membranes, each of the 5 a-helices of Cp-5, comprising 15–22 residues each, were synthesised (Table 1). The fragments previously synthesised en route to synthetic native Cp-5 (fragments 1 and 2, namely Cp-5F and Cp-5G) were also used for this study. The CD spectra for each peptide established that these peptides were unstructured (despite the addition of trifluoroethanol, which induces a-helix formation).[20]
Cp-5A Cp-5B Cp-5C Cp-5D Cp-5E Cp-5F Cp-5G
H2N-1GRSALSCQMCELVVKKYEGSA21-COOH H2N-22DKDANVIKKDFDAECKKLFHT42-COOH H2N-43IPFGTRECDHYV54-COOH H2N-54NSKVDPIIHELEGG67-COOH H2N-69TAPKDVCTKLNECP82-COOH H2N-1GRSALSCQMCELVVKKYEGSADKDANVIKKDFDAE35-COOH H2N-36CKKLFHTIPFGTRECDHYVNSKVDPIIHELEGGTAPKDVCTKLNECP82-COOH
Permeabilisation activity The ability of these peptides to permeabilise inverted membrane vesicles of E. coli was measured by the requenching of the pHresponsive weak base, acridine orange, in NADH-energised vesicles. The peptides Cp-5 (recombinant) and the N-terminal region Gly1-Glu35-OH, namely Cp-5F were found to be effective at collapsing the pH gradient induced by NADH (Figure 6, Table 2). The effects of most other peptides were not greatly different from the vehicle control. This suggests that permeabilisation induced by Cp-5 recombinant and Cp-5F is sufficient to destabilise the proton component of the proton motive force. This may either lead to the uncoupling of respiration-driven ATP synthesis or lysis, in whole cells, depending on the extent of permeabilisation. ChemBioChem 2015, 16, 328 – 336
Figure 6. Measurement of the activity of A) the recombinant and B) the synthetic Cp-5, C) the C-terminal region and D) the N-terminal region (Cp-5G and Cp-5F, respectively), using fluorescence quenching experiments. The percentage of relative fluorescence intensity is shown as a function of time. The protonophore carbonyl cyanide mchlorophenyl hydrazone (CCCP) was used as a positive control and is shown in black; untreated cells are shown in light grey. R = the time that the peptide was injected (at different concentrations: 1, 2, 4 and 8 mm) to induce a reverse quenching.
www.chembiochem.org
333
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers and, although requiring longer fragments (35 and 47 amino acids in length), was shown to be a better strategy for the synthesis of Cp-5 than using a three fragment approach with two ligations. The synthetic protein was folded successfully and the synthesised material was found to have the same structural features as recombinant Cp-5, as determined by 1H NMR spectroscopy and circular dichroism experiments. Moreover, the Nterminal fragment (Gly1-Glu35-OH) exhibited cell permeability activity and is shorter than the native protein (35 residues compared to 82), hence future work will focus on the design of peptidomimetics from the N-terminal region (Cp-5F). Owing to its permeabilisation activity, analogues of Cp-5F may be suitable as cell-penetrating peptides (CPPs), which can traverse the lipophilic barrier of the cell membrane and deliver large, active biomolecules as well as small molecules (e.g., antibodies, contrast imaging agents, toxins and nanoparticle drug carriers including liposomes) into cells.
Table 2. Permeabilisation activity.[a] Peptide
Fluorescence at 8 mm [%]
Cp-5 recombinant Cp-5 synthetic folded Cp-5 synthetic reduced Cp-5A Cp-5B Cp-5C Cp-5D Cp-5E Cp-5F Cp-5G
35.2 16.4 10.2 0.0 15.3 8.5 7.5 10.5 48.1 20.9
[a] Peptides tested with their % fluorescence after injection (to induce a reverse quenching) after subtraction of the baseline value (untreated) and relative to the quenching effect of the positive control, CCCP (i.e., CCCP = 100 %).
Experimental Section All solvents and reagents were used as supplied. Dimethylformamide (DMF; AR grade) and acetonitrile (HPLC grade) were purchased from Scharlau (Barcelona, Spain). N,N’-diisopropylethylamine (DIPEA), ethanedithiol (EDT), 4-mercaptophenylacetic acid (MPAA), tris (2-carboxyethyl) hydrochloride (TCEP·HCl), triisopropylsilane (TIPS) diiosopropylcarbodiimide (DIC) and glutathione were purchased from Aldrich. Trifluoroacetic acid (TFA) was purchased from Oakwood Chemicals (West Columbia, SC). Anhydrous hydrogen fluoride was purchased from Matheson Trigas (Basking Ridge, NJ). 1-[bis-dimethylaminomethylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxid-hexafluorophosphate (HATU) was purchased from GL Biochem (Shanghai, China).
Figure 7. Sequence alignment of SAPLIP members: caenopore-5, amoebapore A (APA-1) and saposin-C, showing the N-terminal region (1–35 amino acids). The charged residues conserved between Cp-5 with at least another two SAPLIP members are boxed and indicated by arrows. The sequence of the shortened analogue Cp-5B is also shown (Asp22–Thr42).
diverse, yet there are a number of features that are conserved and these features may be functionally significant across the SAPLIP family. By looking at the N-terminal region, common charged residues are conserved between Cp-5, APA-1 and saposin-C, namely E18, K23, D22 and D29 (Figure 7, boxed and indicated by arrows). This suggests that the charges at these positions in the structures may play similar functional roles in membrane binding and permeabilisation properties. Moreover, de Alba and co-workers[21] reported that K22 in saposin-C was a key residue for the membrane-binding activity of the protein, so it appears that the equivalent lysine residue, K23, present in Cp-5 may also be involved in the initial binding of Cp-5 with the membrane. Additionally, the involvement of E18 is suggested to be important because the pKa of this residue is elevated (pKa = 5.2), well above the intrinsic pKa value for Glu residues (i.e., ~ 4).[22] Consequently, the surface charge of Cp-5 is more positive at lower pH values because of protonation of E18 (and possibly other acidic amino acids), thus favouring the interaction with the negative head-groups of the membrane phospholipids and therefore contributing to the activity of Cp5.
Syntheses were carried out on Boc-Pro-PAM-OH or Boc-Ala-PAMOH 4-(PAM = hydroxymethylphenylacetamidomethyl) starting material (Polypeptides; Strasbourg, France). Boc-amino acids were also purchased from Polypeptides with the following side-chain protection: Boc-Arg(Tos)-OH (TOS = p-toluenesulfonyl), Boc-Asp(cHex)-OH (cHex = cyclohexyl), Boc-Cys(4-MeBzl)-OH (MeBzl = 4methylbenzyl), Boc-Asn(Xan)-OH (Xan = xanthyl), Boc-Glu(cHex)-OH, Boc-His(Tos)-OH·DCHA (DCHA = dicyclohexylamine), Boc-Lys(2-Cl-Z)OH (Z = carboxybenzyl), Boc-Ser(Bzl)-OH (Bzl = benzyl), Boc-Thr(Bzl)OH and Boc-Tyr(2-Br-Z)-OH. HPLC and LC–MS: Peptides were purified by using a Dionex (Sunnyvale, California, USA) Ultimate 3000 system equipped with a Foxy Jr fraction collector and a Gemini C18 column (Phenomenex; 10 250 mm, 5 mm, 110 , flow rate of 5 mL min1). Purified peptides were eluted with an increasing gradient of acetonitrile containing 0.1 % TFA. The purity and molecular weight of the peptides were confirmed by LC–MS (Agilent Technologies; 1120 Compact LC equipped with a Hewlett Packard 1100 MSD mass spectrometer) using ESI in the positive mode. The fractions were analysed by using an Agilent Zorbax C3 (3.5 m; 3.0 150 mm) column at 0.3 mL min1 with a linear gradient of 5–65 % B over 21 min (i.e., 3 % increase in B per minute). The solvent system used was A (0.1 % formic acid in H2O) and B (0.1 % formic acid in CH3CN). Fractions containing the desired peptide were pooled and lyophilised.
Conclusion The 82 residue protein caenopore-5 was successfully synthesised by native chemical ligation of two smaller polypeptide fragments, which were prepared using Boc SPPS. The strategy adopted used a single NCL site at a native cysteine residue ChemBioChem 2015, 16, 328 – 336
www.chembiochem.org
Peptide synthesis Thioester fragment 1: Boc-Ala-PAM linker (0.2 mmol) was coupled to aminomethyl resin (synthesised as described;[23] loading
334
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers 1 mmol g1) with DIC (0.4 mmol) in CH2Cl2 (3 mL) for 1 h, drained and washed with CH2Cl2. The Kaiser test was negative. Fragment 1 a-thioester was synthesised on an S-trityl-associated mercaptopropionic acid alanine resin to produce activated C-terminal a-thioester upon cleavage with HF.[24] The S-trityl group was cleaved with a 2 2 min wash of triisopropylsilane (2.5 % v/v) and water (2.5 % v/v) in TFA. The first amino acid Boc-Glu-(ocHex) (0.4 mmol) was coupled using HATU/DIPEA in DMF for 1 h.
37 8C and 200 rpm, and then the culture was grown at 18 8C and 200 rpm overnight. Protein purification: The cells were harvested by centrifugation at 6000 g and resuspended in Tris·HCl (50 mm, pH 7.4), NaCl (150 mm), and phenylmethanesulfonylfluoride (PMSF, 1 mm) and lysed using a cell disrupter (Microfluidizer M-110P (Microfluidics, Westwood, USA)). The lysate was then centrifuged at 12 000 rpm at 4 8C to separate the insoluble material from the supernatant. Recombinant protein was initially purified by using immobilisedmetal affinity chromatography (IMAC; GE Healthcare). This step required loading the supernatant onto an IMAC column that had been pre-equilibrated in Na2HPO4 (50 mm), NaCl (300 mm) and imidazole (10 mm). The column was washed with Na2HPO4 (50 mm), NaCl (300 mm) and imidazole (40 mm) until no protein was detected in the flow through by using a Bio-Rad protein assay. The protein was subsequently eluted with Na2HPO4 (50 mm), NaCl (300 mm) and imidazole (250 mm) and the purity of the protein was assessed by SDS-PAGE. The N-terminal tag was removed by incubating the fusion protein in the presence of rTEV protease (1 mg mL1) and dialyzed overnight against a buffer (Tris·HCl (5 mm, pH 7.4; NaCl (150 mm)) to remove the imidazole; the digestion was monitored by SDS-PAGE. An apparently homogeneous product was formed following purification by reversed-phase (RP)HPLC using a Dionex UltiMate 3000 binary semi-preparative system (Thermo Scientific, Victoria, Australia) with a Gemini C18 column (10 250 mm, 5 mm, 110 ). From 4 L of LB culture, 4 mg of pure protein were obtained. The molecular mass of the purified protein was confirmed by LC–MS (obs = 937.8 Da [M+10 H]10 + ; calcd = 938.5 Da).
N-terminal cysteinyl fragment 2: Boc-Pro-PAM linker (0.2 mmol) was coupled to aminomethyl resin (0.1 g for 0.1 mmol scale, loading 1 mmol g1) with DIC (0.4 mmol) in CH2Cl2 (3 mL) for 1 h, drained and washed with CH2Cl2. For both fragments, SPPS was performed manually using the Boc in situ neutralization procedure.[25] Syntheses were carried out on a 0.1 mmol scale. All operations were performed manually in a glass reaction vessel (20 mL) with a Teflon-lined screw cap. The Na-Boc group was removed by treatment with TFA (100 %) for 2 1 min followed by a 30 s flow wash with DMF. Boc amino acids (0.4 mmol) were coupled by using HATU/DIPEA as activating agents. Coupling times were 5 min throughout the synthesis, without any double-coupling procedures. Following chain assembly, the crude peptides 1 and 2 were cleaved from the resin with simultaneous removal of the side-chain protecting groups by using HF/p-cresol (20:1, v/v) for 1 h at 0 8C. Following the evaporation of HF, the peptides were precipitated with cold diethyl ether, isolated by centrifugation, washed twice with cold diethyl ether, dissolved in acetonitrile/water (1:1 v/v) containing TFA (0.1 %), filtered and lyophilised. Both fragments were purified with a slow gradient as described.[26] The solvent system used was A = 0.1 % TFA in H2O and B = 0.1 % TFA in MeCN.
Protein folding: The protein was refolded overnight in a redox buffer (50 mm Tris-HCl, pH 7.4, 150 mm NaCl, 10 mm reduced and 2 mm oxidised glutathione) at 4 8C. Precipitated protein was removed by centrifugation at 6000 g for 15 min. The supernatant was loaded onto a C18 column, purified by RP-HPLC and recovered by lyophilisation (6.5 mg, yield 80 %). In order to determine if the protein remained folded, and was not subject to degradation after lyophilisation, samples of both proteins (synthetic and recombinant) were re-dissolved in sodium phosphate (20 mm) and re-analyzed by LC–MS and CD spectroscopy.
Native chemical ligation of Cp-5: MPAA (100 mm) and TCEP (20 mm) were dissolved in guanidine hydrochloride (6.0 m) and Na2HPO4 (200 mm), and the mixture was degassed with argon. The pH of the resulting solution was adjusted to 7.0 by addition of NaOH (10 m or 2 m). The two reactants, fragments 1 and 2 (5 mm final concentration) were then added to the reaction mixture, and the pH was readjusted to 7.0. The vial was capped under argon and the reaction mixture was stirred at room temperature. Aliquots of the reaction mixture (2 mL) were diluted four fold into an aqueous solution (5 % TFA v/v) for LC–MS analysis. After 10 h, the reaction reached completion and was quenched by the addition of HCl (5 m), isolated by solid-phase extraction and lyophilised to afford the crude ligation product that was purified by RP-HPLC to give the ligated peptide 3 (8 mg, 40 % yield). The mass of the product was confirmed by LC–MS, (obs = 912.4 Da [M+10 H]10 + ; calcd = 912.6 Da).
NMR spectroscopy: All NMR experiments were performed at 25 8C on a Bruker AV600 spectrometer equipped with a 5 mm z-gradient 1 H/15N/13C cryoprobe optimised for 1H detection. Homonuclear 1D 1 H NMR experiments were performed on a Cp-5 sample (0.1 mm) in sodium phosphate (20 mm) and H2O/D2O (9:1, v/v) at pH 5.2. The protein concentrations were quantified using an extinction coefficient (e280) = 3,355 m1 cm1, as determined by the method of Gill and von Hippel.[27] Proton chemical shifts were referenced to trimethylsilyl propionate (TSP, 1 mm). The 1D spectra were acquired with 2048 complex data points, and water suppression was achieved using the WATERGATE sequence.[28] Spectra were obtained with 1H spectral widths of 7500 Hz and processed and analysed by using Topspin 2.1 software (Bruker).
Cloning and expression of Cp-5: The ORF of Cp-5 was re-cloned into the pProEx HTb vector (Invitrogen) with an N-terminal (His)6-tag and rTEV cleavage site. Chemically competent E. coli cells BL21(DE3), (Invitrogen, California, USA) were transformed with the plasmid DNA pProEx HTb-spp5 construct for overexpression of Cp5. A pre-culture was prepared by inoculating a single colony into lysogeny broth (LB, 50 mL) with ampicillin (0.1 mg mL1) and incubated overnight at 37 8C with shaking at 200 rpm. The pre-culture was used to inoculate 4 L of LB (containing ampicillin; 0.1 mg mL1) at a dilution factor of 1/50. The growth of the cells was monitored by measuring the optical density at 600 nm. Once the OD600 was 0.5, isopropyl-b-d-1-thiogalactopyranoside (IPTG) was added to the culture to a final concentration of 1 mm to induce protein production. Cells were grown for a further 3 h at
ChemBioChem 2015, 16, 328 – 336
www.chembiochem.org
Circular dichroism: All CD measurements were performed on a PiStar spectrometer (Applied Photophysics; Leatherhead, UK). Protein spectral data are reported in terms of the mean residue ellipticity (q [deg cm2 dmol1]), calculated as follows [Eq. (1)]:
q ¼ S=ð10 c L nÞ
335
ð1Þ
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers Where S is the raw CD signal (mdeg), c is the protein concentration (M), L is the cuvette path length (cm) and n is the number of peptide bonds in the protein.
[2] a) Z. Wang, E. Choice, A. Kaspar, D. Hanson, S. Okada, S. C. Lyu, A. M. Krensky, C. Clayberger, J. Immunol. 2000, 165, 1486 – 1490; b) J. E. McInturff, S.-J. Wang, T. Machleidt, T. R. Lin, A. Oren, C. J. Hertz, S. R. Krutzik, S. Hart, K. Zeh, D. H. Anderson, R. L. Gallo, R. L. Modlin, J. Kim, J. Invest. Dermatol. 2005, 125, 256 – 263; c) A. Nakashima, A. Shiozaki, S. Myojo, M. Ito, M. Tatematsu, M. Sakai, Y. Takamori, K. Ogawa, K. Nagata, S. Saito, Am. J. Pathol. 2008, 173, 653 – 664. [3] T. Roeder, M. Stanisak, C. Gelhaus, I. Bruchhaus, J. Grotzinger, M. Leippe, Dev. Comp. Immunol. 2010, 34, 203 – 209. [4] a) J. Mysliwy, A. J. Dingley, M. Stanisak, S. Jung, I. Lorenzen, T. Roeder, M. Leippe, J. Grotzinger, Dev. Comp. Immunol. 2010, 34, 323 – 330; b) E. Liepinsh, M. Andersson, J. M. Ruysschaert, G. Otting, Nat. Struct. Biol 1997, 4, 793 – 795. [5] P. E. Dawson, S. B. Kent, Annu. Rev. Biochem. 2000, 69, 923 – 960. [6] L. Z. Yan, P. E. Dawson, J. Am. Chem. Soc. 2001, 123, 526 – 533. [7] a) J. Chen, Q. Wan, Y. Yuan, J. L. Zhu, S. J. Danishefsky, Angew. Chem. Int. Ed. 2008, 47, 8521 – 8524; Angew. Chem. 2008, 120, 8649 – 8652; b) C. Haase, H. Rohde, O. Seitz, Angew. Chem. Int. Ed. 2008, 47, 6807 – 6810; Angew. Chem. 2008, 120, 6912 – 6915. [8] Q. Wan, S. J. Danishefsky, Angew. Chem. Int. Ed. 2007, 46, 9248 – 9252; Angew. Chem. 2007, 119, 9408 – 9412. [9] P. W. Harris, M. A. Brimble, Biopolymers 2013, 100, 356 – 365. [10] a) E. Atherton, R. C. Sheppard, J. Chem. Soc. Chem. Comm 1985, 165 – 166; b) P. W. R. Harris, M. A. Brimble, Biopolymers 2010, 94, 542 – 550. [11] R. Subirs-Funosas, A. El-Faham, F. Albericio, Tetrahedron 2011, 67, 8595 – 8606. [12] M. Villain, H. Gaertner, P. Botti, Eur. J. Org. Chem. 2003, 3267 – 3272. [13] B. Dang, T. Kubota, K. Mandal, F. Bezanilla, S. B. Kent, J. Am. Chem. Soc. 2013, 135, 11911 – 11919. [14] E. C. B. Johnson, S. B. H. Kent, J. Am. Chem. Soc. 2006, 128, 6640 – 6646. [15] J. Dhakal, G. S. Brah, R. K. Agrawal, H. N. Pawar, D. Kaur, R. Verma, Indian J. Med. Microbiol. 2013, 31, 123 – 129. [16] N. Abdullah, H. A. Chase, Biotechnol. Bioeng. 2005, 92, 501 – 513. [17] T. Rehm, R. Huber, T. A. Holak, Structure 2002, 10, 1613 – 1618. [18] C. Scheich, D. Leitner, V. Sievert, M. Leidert, B. Schlegel, B. Simon, I. Letunic, K. Bussow, A. Diehl, BMC Struct. Biol. 2004, 4, 4. [19] S. M. Kelly, T. J. Jess, N. C. Price, Biochim. Biophys. Acta Proteins Proteomics 2005, 1751, 119 – 139. [20] H. H. Hauge, J. Nissen-Meyer, I. F. Nes, V. G. Eijsink, Eur. J. Biochem. 1998, 251, 565 – 572. [21] C. A. Hawkins, E. d. Alba, N. Tjandra, J. Mol. Biol. 2005, 346, 1381 – 1392. [22] W. Li, Personal communication, 2014. [23] P. W. R. Harris, S. H. Yang, M. A. Brimble, Tetrahedron Lett. 2011, 52, 6024 – 6026. [24] T. M. Hackeng, J. H. Griffin, P. E. Dawson, Proc. Natl. Acad. Sci. USA 1999, 96, 10068 – 10073. [25] M. Schnçlzer, P. Alewood, A. Jones, D. Alewood, S. B. Kent, Int. J. Pept. Protein Res. 1992, 40, 180 – 193. [26] P. W. R. Harris, D. J. Lee, M. A. Brimble, J. Pept. Sci. 2012, 18, 549 – 555. [27] S. C. Gill, P. H. von Hippel, Anal. Biochem. 1989, 182, 319 – 326. [28] M. Piotto, V. Saudek, V. Sklenar, J. Biomol. NMR 1992, 2, 661 – 665. [29] M. B. McNeil, H. G. Hampton, K. J. Hards, B. N. Watson, G. M. Cook, P. C. Fineran, FEBS Lett. 2014, 588, 414 – 421.
Each CD spectrum measurement represents the average of seven scans obtained with a 2 nm optical bandwidth. Baseline spectra were collected with buffer alone and then subtracted from the raw protein spectral data. The measurements were performed at protein concentrations of 2–5 mm in sodium phosphate buffer (20 mm, pH 5.2) at 25 8C in quartz cuvettes (1 mm; Hellma Analytics, Mllheim, Germany). Synthesis of shortened analogues: The syntheses of peptides Cp5A–Cp-5E (described in Table 1) were carried out on a 0.1 mmol scale on aminomethyl resin, which was pre-loaded with a Boc-AAPAM linker (AA corresponds to the C-terminal residue of each fragment). The Na-Boc group was removed by treatment with neat TFA followed by a flow wash with DMF. Boc amino acids were coupled by using HATU/iPr2EtN as the activating agent and base, respectively. Coupling times were 5 min throughout the synthesis, without the use of any double-coupling procedures. Following peptide chain assembly, the crude peptides were cleaved from the resin with simultaneous removal of side-chain protecting groups by using HF/p-cresol (20:1, v/v) for 30 min at 0 8C. Fluorescence quenching dependent on DpH: Proton translocation into inverted membrane vesicles was measured by the quenching of the fluorescent, pH-dependent probe acridine orange (AO), which was monitored by using a Cary Eclipse Fluorescence spectrophotometer. The assay buffer contained HEPES (10 mm, pH 6.0), KCl (100 mm), MgCl2 (5 mm), E. coli inverted membrane vesicles (0.085 mg mL1) and AO (2.5 mm). Inverted membrane vesicles were prepared as described by McNeil et al.,[29] with the exception that the cells were grown aerobically in conical flasks. The reactions were initiated with NADH (50 mm) and quenching was reversed by injection of the desired compound.[29] The excitation and emission wavelengths were 493 and 530 nm, respectively.
Acknowledgements The authors would like to thank Maurice Wilkins Centre for Molecular Biodiscovery for funding this work and Prof. Russel Snell for providing a picture of C. elegans. Keywords: antibiotics · antimicrobial protein · caenopore · native chemical ligation · protein design · protein folding [1] a) T. Kolter, F. Winau, U. E. Schaible, M. Leippe, K. Sandhoff, J. Biol. Chem. 2005, 280, 41125 – 41128; b) K. A. Brogden, Nat. Rev. Microbiol. 2005, 3, 238 – 250; c) R. I. Lehrer, Nat. Rev. Microbiol. 2004, 2, 727 – 738; d) M. Zasloff, Nature 2002, 415, 389 – 395.
ChemBioChem 2015, 16, 328 – 336
www.chembiochem.org
Received: September 7, 2014 Published online on November 25, 2014
336
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers
Chemical Synthesis of A Pore-Forming Antimicrobial Protein, Caenopore-5, by Using Native Chemical Ligation at a Glu-Cys Site Karima Medini,[a] Paul W. R. Harris,[a, b] Kiel Hards,[c] Andrew J. Dingley,[d] Gregory M. Cook,[c] and Margaret A. Brimble*[a, b] AMP. The protein (82 amino acids) was successfully synthesised by using Boc solid-phase peptide synthesis and native chemical ligation. No g-linked by-product was observed despite the use of a C-terminal Glu-thioester. The folding of the synthetic protein was confirmed by 1H NMR spectroscopy and circular dichroism and compared with data recorded for recombinant caenopore-5. The permeabilisation activities of the protein and of shortened analogues were evaluated.
The 2014 report from the World Health Organization (WHO) on antimicrobial resistance revealed an alarming rise in antibiotic resistance all around the world. Unlike classical antibiotics, with the exception of a few species, no acquired resistance towards antimicrobial peptides (AMPs) has been reported. Therefore, AMPs represent leads for the development of novel antibiotics. Caenopore-5 is constitutively expressed in the intestine of the nematode Caenorhabditis elegans and is a pore-forming
Introduction Antimicrobial proteins are typically small amphipathic peptides that have positively charged regions that allow them to readily interact with the negative lipid head groups found in abundance in prokaryotic membranes.[1] Additionally, AMPs may also possess secondary functions such as the modulation of the immune response, inactivation of bacterial endotoxins and the induction of cell apoptosis.[2] One group of AMPs recently identified in the nematode Caenorhabditis elegans is the caenopores. There are 33 caenopores that collectively enable C. elegans to survive in its natural habitat.[3] The caenopores consist of a short N-terminal hydrophobic amino acid region representing a signal peptide and a single domain with homology to members of the saposin-like protein (SAPLIP) family.[4] The saposin (spp) genes are expressed primarily in the intestinal epithelial cells of C. elegans and the resultant gene products are secreted into the gut lumen, which provides the optimal environment for antimicrobial activity.[3] Caenopores act directly on microbes that have entered the worm orally, thereby pre-
venting the establishment of intestinal infections. Caenopore-5 (Cp-5) is essential for the survival of C. elegans, because an spp-5 knockdown (T08A9.9 gene, GenBank U40417) induced a range of abnormal features, including developmental retardation, a significant reduction in the number of eggs laid and the inability to accumulate fat. Additionally, spp-5 silenced worms struggled to survive on lawns of Escherichia coli and intact bacteria were able to colonise the intestinal tract, leading to the death of the nematode.[3] The slightly acidic composition of the intestine provides the optimal environment for the function of caenopores; for example, the optimum pH for Cp-5 has been reported as ~ 5.2.[3] Cp-5 is the best functionally characterised member of the caenopore family and structural data has confirmed that the protein adopts the canonical protein fold of SAPLIPs.[4a] The solution structure shows that Cp-5 comprises five a-helices connected by three disulfide bonds from six conserved cysteines.[4a] Recombinant Cp-5 has been shown to be active against both Gram-positive (Bacillus megaterium) and Gramnegative (E. coli) bacteria with minimum inhibitory concentration (MIC) values of 0.05 and 0.1 mm, respectively, and minimum bactericidal concentration (MBC) values of 0.1 and 0.2 mm, respectively.[3] As a first step to design analogues with more potent biological activity, a chemically robust synthesis of the native Cp-5 sequence needs to be developed. To realise this aim, we herein report the synthesis of Cp-5 by native chemical ligation (NCL) of two unprotected peptide segments. The full-length protein was folded using a glutathione redox-coupled reaction and the folded protein was compared with recombinantly produced Cp-5 by using one-dimentional 1H NMR and circular dichroism. Moreover, shortened peptides representing the five
[a] K. Medini, Dr. P. W. R. Harris, Prof. M. A. Brimble Maurice Wilkins Centre for Molecular Biodiscovery School of Biological Sciences, The University of Auckland 3A Symonds Street, Auckland 1010 (New Zealand) E-mail: [email protected] [b] Dr. P. W. R. Harris, Prof. M. A. Brimble School of Chemical Sciences, The University of Auckland 23 Symonds Street, Auckland 1010 (New Zealand) [c] K. Hards, G. M. Cook Department of Microbiology and Immunology School of Medical Sciences, University of Otago, 720 Cumberland Street, Dunedin 9054 (New Zealand) [d] Dr. A. J. Dingley Institute of Complex Systems 6 Structural Biochemistry, Forschungszentrum Jlich 52425 Jlich (Germany)
ChemBioChem 2015, 16, 328 – 336
328
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers individual a-helices in Cp-5 were synthesised to evaluate their structure–activity relationship (SAR) profile.
Second ligation strategy The second strategy (Scheme 2) we tried was devised to use only two fragments, consisting of 35 (fragment 1, Gly1-Glu35) and 47 (fragment 2, Cys36-Pro82) amino acid residues in length (Scheme 1), and exploited the native N-terminal cysteinyl residue (Cys36), thereby avoiding the introduction of nonnative cysteines or derivatives. Because of the sensitivity of thioesters to piperidine, which is used for 9-fluorenylmethyloxycarbonyl (Fmoc) SPPS, Boc SPPS was preferred to synthesise the C-terminal thioester fragment 1 namely Gly1-Glu35COSCH2CH2-Ala-OH.[9] On a scale of 0.1 mmol, we obtained a 40 % yield of fragment 1, which was characterised by HPLC and LC–MS (obs = 802.4 Da [M+5 H]5 + ; calcd = 802.7 Da).
Results and Discussion Synthetic design Caenopore-5 is a protein consisting of 82 amino acid residues (Scheme 1), which renders the direct linear synthesis of the full-length protein difficult using SPPS. NCL is a powerful technique to ligate peptides in order to prepare longer proteins.[5]
Scheme 1. Amino acid sequence of Cp-5,[3] with the ligation sites of the first and second strategies marked with arrows.
This method requires a peptide fragment with a thioester on the C terminus, which undergoes a chemoselective reaction with a second fragment bearing a cysteine residue on the N terminus, and results in the formation of a native amide bond.
First ligation strategy The first strategy we tried used two ligation sites, which would give plausibly sized polypeptide fragments for solid-phase synthesis. The sequence of Cp-5 reveals two possible ligation sites, which could be assembled from three fragments comprising 20–33 amino acids each. Ser20-Ala21 and Tyr53-Val54 were identified as possible ligation sites; however these sites require N-terminal cysteinyl residues (as shown in Scheme 1) for native chemical ligation. Therefore, Ala21 and Val54 were substituted by the introduction of non-native cysteine[6] and penicillamine (Pen),[7] respectively, as surrogates for NCL, which requires a subsequent desulfurization step to regenerate the native sequence.[8] Although we could synthesise the requisite peptide fragments by Boc SPPS, namely the 1-20 thioester, Thz-21-53-thioester (Thz = thiazolidine) and Pen-54-82 fragments and the ligation was successful, the Pen ligation was slow (five days) and significant hydrolysis at the tyrosine thioester was observed. Moreover, the desulfurization of Pen54 to Val54 on the ligated 1–82 peptides was slow and side products were observed, reducing the yield of the reaction to 10 %. The final yield was only 3 %. ChemBioChem 2015, 16, 328 – 336
www.chembiochem.org
Scheme 2. The second ligation strategy for the chemical synthesis of caenopore-5. The cartoon representation of Cp-5 was generated using PyMOL from the solution structure (PDB ID: 2JS9).[4a] The disulfide bonds are shown as sticks. a) Boc-Ala-Pam, DIC, CH2Cl2, 1 h; b) N-terminal Boc removal; c) TrtSCH2CH2CO2H, HATU, iPr2EtN, DMF, 15 min, d) TFA/TIPS/H2O (95:2.5:2.5, v/v/v) 2 1 min; e) Boc-Glu-(OcHex)-OH, HATU, iPr2EtN, DMF, 1 h; f) Boc-AAOH HATU, iPr2EtN, DMF, 5 min; g) HF/p-cresol (20:1, v/v), 1 h at 0 8C; h) 0.2 m Na2HPO4/6 m Gn·HCl, 100 mm MPAA, 20 mm TCEP, pH 7.0.
329
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers
Scheme 3. Modified pathway proposed by Botti et al. for the formation of a- and g- linked peptides product during NCL between Glu-thioester and A) MPAA or a B) Cys-peptide 2.[12]
Figure 1. LC–MS analysis of MPAA exchange using a short C-terminal Glu-thioester peptide (R = CH2-Ala-OH; reaction conditions: MPAA (100 mm), Gd·HCl (6 m), Na2HPO4 (0.2 m), TCEP·HCl (20 mm), at RT and pH 7).
a microwave synthesiser.[10] Although the desired peptide was detected by LC–MS, an elution peak corresponding to an aspartyl piperidide peptide [M+67 Da] was observed (50 %),
The synthesis of fragment 2 (47 amino acid residues), which does not contain a sensitive thioester, was first attempted by Fmoc SPPS on a 2-chlorotrityl chloride resin with the aid of ChemBioChem 2015, 16, 328 – 336
www.chembiochem.org
330
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers 0.2 m Na2HPO4, 20 mm TCEP·HCl at pH 7.0), and the results are shown in Figure 1 A. The reaction was monitored by LC–MS using a slow gradient (0.1 % B/min) at room temperature to reveal the formation of any by-products, and there was no evidence to suggest that the cyclic intermediate that would lead to a g-linked by-product was formed. Rather, a single species was observed after 3 h (Figure 1), namely IKKDFDAE-MPAA thioester. Because the ligation of the peptide aryl thioester with N-terminal cysteinyl fragment 2 is fast, it was concluded that formation of a g-linked product would not be significant in our key ligation step. NCL of fragments 1 and 2: Ligation of fragments 1 and 2 (Scheme 2) was next undertaken by adding 100 mm MPAA[14]
probably owing to the Asp73-Lys72 motif (despite the use of the less basic piperizine for Fmoc deprotection).[11] We therefore used Boc SPPS to synthesise fragment 2 using HATU and iPr2EtN as the activating agent and base, respectively. The 47 residue peptide was obtained in good yield (45 %) with no aspartimide by-product and the purity was confirmed by integration of the HPLC chromatogram at 210 nm and LC–MS (obs = 881.6 Da [M+6 H]6 + ; calcd = 881.8 Da). Native chemical ligation
NCL at the Glu35-Cys36 site: The NCL site was chosen at position Glu35-Cys36 of the peptide chain. However, it is known that ligations between Asp-Cys and Glu-Cys residues can result in the migration of the ligation site onto the side chain, generating b- and g-linked by-products, respectively.[12] Scheme 3 shows the initial formation of a cyclic product from a peptide bearing a C-terminal Glu-thioester. In the presence of either MPAA (1) or an unprotected peptide containing a thiol at the N terminus (a Cys in our example, 2), two possible ring-opening reactions can occur. In the case of attack of MPAA or unprotected peptide 2 at the a-carbonyl group, the native bond is preserved. However, if MPAA or peptide 2 reacts at the g-carbonyl group of the cyclic anhydride, an unnatural glinked by-product is formed (Scheme 3). In a study conducted by Kent et al., the authors reported the formation of 3 % of the unnatural g amide bond at a Glu-Cys site, whereas 25 % of unnatural b amide bond formation took place at an Asp-Cys site during NCL.[13] In our present work, in order to study the formation of this side product during the ligation, a truncated moiety of the fragment 1 thioester was synthesised (IKKDFDAE-thioester). This short peptide was synthesised using in situ Boc SPPS, and the same procedure as fragment 1 (Scheme 3). The short peptide was reacted with MPAA under the same conditions that Figure 2. HPLC chromatogram of native chemical ligation at A) t = 0 and B) after completion at t = 10 h; the asterwere to be used for NCL of Cp-5 isk (*) represents the MPAA catalyst. C) After purification, the protein was folded. D) From the linear sequence, (100 mm MPAA, 6 m Gn·HCl, a shift of the retention time was observed from 17.30 to 16.65 min after folding. ChemBioChem 2015, 16, 328 – 336
www.chembiochem.org
331
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers and 20 mm TCEP to 1 mm peptide, at pH 7. The reaction was monitored by LC–MS and at t = 0 h we observed that some of fragment 1 reacted to form an MPPA-reactive thioester intermediate by way of a transthioesterification reaction (Figure 2 A). Although co-elution of the peptide fragments hampered the HPLC analysis, the ligated peptide 3 was the major product after 10 h (Figure 2 B), as evidenced by mass spectrometry (obs = 912.4 Da [M+10 H]10 + ; calcd = 912.6 Da). The ligated peptide 3 was purified by HPLC (Figure 2 C) to give 40 % yield. To obtain active Cp-5, which contains three disulfide bonds, the linear sequence was folded in buffer with 50 mm Tris·HCl, 150 mm NaCl and a redox system (final concentration: 10 mm reduced and 2 mm oxidised glutathione, pH 7.4) at 4 8C. The redox-coupled reaction enabled a facile thiol/disulfide exchange process and promoted the formation of the correctly folded protein. After 6 h, we observed a shift of the retention time from 17.30 min (unfolded) to 16.65 min (folded; Figure 2 D). After 12 h, the folding was deemed complete and the protein was purified by HPLC and lyophilised. The pure folded final product (Figure 3) was characterised by LC–MS (obs = 911.8 Da [M+10 H]10 + ; calcd = 912 Da], which corresponded to the loss of six hydrogen atoms as expected for the formation of three disulfide bonds (80 % yield).
Figure 4. HPLC chromatogram and MS of recombinant Cp-5.
fied using HPLC (4 mg of pure recombinant protein was obtained) and the mass was confirmed by LC–MS. The observed molecular mass of 9368 0.8 Da [M+H] + (Figure 4) was in agreement with the calculated average isotopic mass of 9369.7 Da.
Confirmation of the protein secondary structure and protein fold NMR spectroscopy has been shown to be a reliable technique to determine how well folded a protein is. One-dimensional 1 H NMR can provide rapid information on the folded state of a protein,[17] and proteins that are folded generally give a welldispersed set of resonances in the amide region of the 1D 1 H NMR spectrum.[18] Before the folding reaction of synthetic Cp-5, a 1D 1H NMR spectrum was recorded and gave a strong set of resonances clustered at approximately 8.3 ppm, confirming that Cp-5 was unstructured (Figure 5 A). After folding, the
Figure 3. HPLC chromatogram and MS of synthetic Cp-5.
Recombinant expression and purification of Cp-5 Recombinantly expressed Cp-5 was purified according to a slightly modified procedure from Mysliwy et al.[4a] The Cp-5 ORF (open reading frame) was cloned into the pProEx HTb vector[15] with an rTEV cleavage site rather than the original Factor X cleavage site,[16] and the protein was solubly expressed in E. coli using lysogeny broth medium (4 L). Thus, the protein was produced as a folded protein. After the first step of purification using immobilized metal affinity chromatography (IMAC), the His6-tag was successfully removed by overnight digestion with rTEV protease. The resulting protein amino-acid sequence was identical to the synthetic one except for the presence of an extra three N-terminal residues (Glu-AlaMet = GAM). These residues were generated recombinantly from the pProEx HTb vector linker between the rTEV recognition site and Cp-5. The recombinant protein was further puriChemBioChem 2015, 16, 328 – 336
www.chembiochem.org
Figure 5. A) The 1D 1H NMR spectra of Cp-5 before and after folding, recorded in sodium phosphate buffer (20 mm, pH 5.2) at 25 8C. B) Far-UV CD spectra of Cp-5. Grey line: synthetic Cp-5; black line: recombinant Cp-5.
332
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers It appears that the peptides of the five a-helices of Cp-5 are too short to exhibit activity, because they probably lack a combination of key residues required for membrane permeabilisation. It is possible that the peptides do interact transiently with the membrane, but do not permeabilise the membrane to release the dye.
spectrum showed a distinct set of well-dispersed resonances, indicating that synthetic Cp-5 was folded (Figure 5 A). As a comparison, the 1D 1H NMR spectrum of recombinant Cp-5 is also shown in Figure 5 A. The 1D 1H NMR spectroscopic data suggested that both proteins are folded. Chemical shift dispersion was also observed in the aliphatic region of the 1D 1H NMR spectrum of the synthetic protein (data not shown), further indicating that the protein has folded. We suspect that the slight difference between these two NMR spectra might be due to the presence of the tag (GAM) in the recombinant protein. Folded synthetic and recombinant Cp-5 were also investigated by far-UV/CD spectroscopy (180–300 nm).[19] The CD spectra (Figure 5 B) demonstrated that both the synthetic and recombinant Cp-5 exhibit the features expected of a-helical proteins, with the standard double negative ellipticity maxima at 208 and 222 nm, and a positive maximum near 194 nm.
Key residues involved in the membrane permeabilisation? The primary sequence of the N-terminal region of Cp-5 that showed the highest activity (1–35 amino acids) was compared with other SAPILIP family members (Figure 7). The proteins amaoebapore A (APA-1) and saposin-C belong to the same family as Cp-5 and by comparing them with the Cp-5 sequence, information about their functional properties can be obtained. The protein sequences of these SAPLIP members are
Table 1. Sequence of Cp-5 analogues.
Synthesis of shortened analogues
Peptide Sequence
In order to study the SAR profile of caenopore-5 with membranes, each of the 5 a-helices of Cp-5, comprising 15–22 residues each, were synthesised (Table 1). The fragments previously synthesised en route to synthetic native Cp-5 (fragments 1 and 2, namely Cp-5F and Cp-5G) were also used for this study. The CD spectra for each peptide established that these peptides were unstructured (despite the addition of trifluoroethanol, which induces a-helix formation).[20]
Cp-5A Cp-5B Cp-5C Cp-5D Cp-5E Cp-5F Cp-5G
H2N-1GRSALSCQMCELVVKKYEGSA21-COOH H2N-22DKDANVIKKDFDAECKKLFHT42-COOH H2N-43IPFGTRECDHYV54-COOH H2N-54NSKVDPIIHELEGG67-COOH H2N-69TAPKDVCTKLNECP82-COOH H2N-1GRSALSCQMCELVVKKYEGSADKDANVIKKDFDAE35-COOH H2N-36CKKLFHTIPFGTRECDHYVNSKVDPIIHELEGGTAPKDVCTKLNECP82-COOH
Permeabilisation activity The ability of these peptides to permeabilise inverted membrane vesicles of E. coli was measured by the requenching of the pHresponsive weak base, acridine orange, in NADH-energised vesicles. The peptides Cp-5 (recombinant) and the N-terminal region Gly1-Glu35-OH, namely Cp-5F were found to be effective at collapsing the pH gradient induced by NADH (Figure 6, Table 2). The effects of most other peptides were not greatly different from the vehicle control. This suggests that permeabilisation induced by Cp-5 recombinant and Cp-5F is sufficient to destabilise the proton component of the proton motive force. This may either lead to the uncoupling of respiration-driven ATP synthesis or lysis, in whole cells, depending on the extent of permeabilisation. ChemBioChem 2015, 16, 328 – 336
Figure 6. Measurement of the activity of A) the recombinant and B) the synthetic Cp-5, C) the C-terminal region and D) the N-terminal region (Cp-5G and Cp-5F, respectively), using fluorescence quenching experiments. The percentage of relative fluorescence intensity is shown as a function of time. The protonophore carbonyl cyanide mchlorophenyl hydrazone (CCCP) was used as a positive control and is shown in black; untreated cells are shown in light grey. R = the time that the peptide was injected (at different concentrations: 1, 2, 4 and 8 mm) to induce a reverse quenching.
www.chembiochem.org
333
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers and, although requiring longer fragments (35 and 47 amino acids in length), was shown to be a better strategy for the synthesis of Cp-5 than using a three fragment approach with two ligations. The synthetic protein was folded successfully and the synthesised material was found to have the same structural features as recombinant Cp-5, as determined by 1H NMR spectroscopy and circular dichroism experiments. Moreover, the Nterminal fragment (Gly1-Glu35-OH) exhibited cell permeability activity and is shorter than the native protein (35 residues compared to 82), hence future work will focus on the design of peptidomimetics from the N-terminal region (Cp-5F). Owing to its permeabilisation activity, analogues of Cp-5F may be suitable as cell-penetrating peptides (CPPs), which can traverse the lipophilic barrier of the cell membrane and deliver large, active biomolecules as well as small molecules (e.g., antibodies, contrast imaging agents, toxins and nanoparticle drug carriers including liposomes) into cells.
Table 2. Permeabilisation activity.[a] Peptide
Fluorescence at 8 mm [%]
Cp-5 recombinant Cp-5 synthetic folded Cp-5 synthetic reduced Cp-5A Cp-5B Cp-5C Cp-5D Cp-5E Cp-5F Cp-5G
35.2 16.4 10.2 0.0 15.3 8.5 7.5 10.5 48.1 20.9
[a] Peptides tested with their % fluorescence after injection (to induce a reverse quenching) after subtraction of the baseline value (untreated) and relative to the quenching effect of the positive control, CCCP (i.e., CCCP = 100 %).
Experimental Section All solvents and reagents were used as supplied. Dimethylformamide (DMF; AR grade) and acetonitrile (HPLC grade) were purchased from Scharlau (Barcelona, Spain). N,N’-diisopropylethylamine (DIPEA), ethanedithiol (EDT), 4-mercaptophenylacetic acid (MPAA), tris (2-carboxyethyl) hydrochloride (TCEP·HCl), triisopropylsilane (TIPS) diiosopropylcarbodiimide (DIC) and glutathione were purchased from Aldrich. Trifluoroacetic acid (TFA) was purchased from Oakwood Chemicals (West Columbia, SC). Anhydrous hydrogen fluoride was purchased from Matheson Trigas (Basking Ridge, NJ). 1-[bis-dimethylaminomethylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxid-hexafluorophosphate (HATU) was purchased from GL Biochem (Shanghai, China).
Figure 7. Sequence alignment of SAPLIP members: caenopore-5, amoebapore A (APA-1) and saposin-C, showing the N-terminal region (1–35 amino acids). The charged residues conserved between Cp-5 with at least another two SAPLIP members are boxed and indicated by arrows. The sequence of the shortened analogue Cp-5B is also shown (Asp22–Thr42).
diverse, yet there are a number of features that are conserved and these features may be functionally significant across the SAPLIP family. By looking at the N-terminal region, common charged residues are conserved between Cp-5, APA-1 and saposin-C, namely E18, K23, D22 and D29 (Figure 7, boxed and indicated by arrows). This suggests that the charges at these positions in the structures may play similar functional roles in membrane binding and permeabilisation properties. Moreover, de Alba and co-workers[21] reported that K22 in saposin-C was a key residue for the membrane-binding activity of the protein, so it appears that the equivalent lysine residue, K23, present in Cp-5 may also be involved in the initial binding of Cp-5 with the membrane. Additionally, the involvement of E18 is suggested to be important because the pKa of this residue is elevated (pKa = 5.2), well above the intrinsic pKa value for Glu residues (i.e., ~ 4).[22] Consequently, the surface charge of Cp-5 is more positive at lower pH values because of protonation of E18 (and possibly other acidic amino acids), thus favouring the interaction with the negative head-groups of the membrane phospholipids and therefore contributing to the activity of Cp5.
Syntheses were carried out on Boc-Pro-PAM-OH or Boc-Ala-PAMOH 4-(PAM = hydroxymethylphenylacetamidomethyl) starting material (Polypeptides; Strasbourg, France). Boc-amino acids were also purchased from Polypeptides with the following side-chain protection: Boc-Arg(Tos)-OH (TOS = p-toluenesulfonyl), Boc-Asp(cHex)-OH (cHex = cyclohexyl), Boc-Cys(4-MeBzl)-OH (MeBzl = 4methylbenzyl), Boc-Asn(Xan)-OH (Xan = xanthyl), Boc-Glu(cHex)-OH, Boc-His(Tos)-OH·DCHA (DCHA = dicyclohexylamine), Boc-Lys(2-Cl-Z)OH (Z = carboxybenzyl), Boc-Ser(Bzl)-OH (Bzl = benzyl), Boc-Thr(Bzl)OH and Boc-Tyr(2-Br-Z)-OH. HPLC and LC–MS: Peptides were purified by using a Dionex (Sunnyvale, California, USA) Ultimate 3000 system equipped with a Foxy Jr fraction collector and a Gemini C18 column (Phenomenex; 10 250 mm, 5 mm, 110 , flow rate of 5 mL min1). Purified peptides were eluted with an increasing gradient of acetonitrile containing 0.1 % TFA. The purity and molecular weight of the peptides were confirmed by LC–MS (Agilent Technologies; 1120 Compact LC equipped with a Hewlett Packard 1100 MSD mass spectrometer) using ESI in the positive mode. The fractions were analysed by using an Agilent Zorbax C3 (3.5 m; 3.0 150 mm) column at 0.3 mL min1 with a linear gradient of 5–65 % B over 21 min (i.e., 3 % increase in B per minute). The solvent system used was A (0.1 % formic acid in H2O) and B (0.1 % formic acid in CH3CN). Fractions containing the desired peptide were pooled and lyophilised.
Conclusion The 82 residue protein caenopore-5 was successfully synthesised by native chemical ligation of two smaller polypeptide fragments, which were prepared using Boc SPPS. The strategy adopted used a single NCL site at a native cysteine residue ChemBioChem 2015, 16, 328 – 336
www.chembiochem.org
Peptide synthesis Thioester fragment 1: Boc-Ala-PAM linker (0.2 mmol) was coupled to aminomethyl resin (synthesised as described;[23] loading
334
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers 1 mmol g1) with DIC (0.4 mmol) in CH2Cl2 (3 mL) for 1 h, drained and washed with CH2Cl2. The Kaiser test was negative. Fragment 1 a-thioester was synthesised on an S-trityl-associated mercaptopropionic acid alanine resin to produce activated C-terminal a-thioester upon cleavage with HF.[24] The S-trityl group was cleaved with a 2 2 min wash of triisopropylsilane (2.5 % v/v) and water (2.5 % v/v) in TFA. The first amino acid Boc-Glu-(ocHex) (0.4 mmol) was coupled using HATU/DIPEA in DMF for 1 h.
37 8C and 200 rpm, and then the culture was grown at 18 8C and 200 rpm overnight. Protein purification: The cells were harvested by centrifugation at 6000 g and resuspended in Tris·HCl (50 mm, pH 7.4), NaCl (150 mm), and phenylmethanesulfonylfluoride (PMSF, 1 mm) and lysed using a cell disrupter (Microfluidizer M-110P (Microfluidics, Westwood, USA)). The lysate was then centrifuged at 12 000 rpm at 4 8C to separate the insoluble material from the supernatant. Recombinant protein was initially purified by using immobilisedmetal affinity chromatography (IMAC; GE Healthcare). This step required loading the supernatant onto an IMAC column that had been pre-equilibrated in Na2HPO4 (50 mm), NaCl (300 mm) and imidazole (10 mm). The column was washed with Na2HPO4 (50 mm), NaCl (300 mm) and imidazole (40 mm) until no protein was detected in the flow through by using a Bio-Rad protein assay. The protein was subsequently eluted with Na2HPO4 (50 mm), NaCl (300 mm) and imidazole (250 mm) and the purity of the protein was assessed by SDS-PAGE. The N-terminal tag was removed by incubating the fusion protein in the presence of rTEV protease (1 mg mL1) and dialyzed overnight against a buffer (Tris·HCl (5 mm, pH 7.4; NaCl (150 mm)) to remove the imidazole; the digestion was monitored by SDS-PAGE. An apparently homogeneous product was formed following purification by reversed-phase (RP)HPLC using a Dionex UltiMate 3000 binary semi-preparative system (Thermo Scientific, Victoria, Australia) with a Gemini C18 column (10 250 mm, 5 mm, 110 ). From 4 L of LB culture, 4 mg of pure protein were obtained. The molecular mass of the purified protein was confirmed by LC–MS (obs = 937.8 Da [M+10 H]10 + ; calcd = 938.5 Da).
N-terminal cysteinyl fragment 2: Boc-Pro-PAM linker (0.2 mmol) was coupled to aminomethyl resin (0.1 g for 0.1 mmol scale, loading 1 mmol g1) with DIC (0.4 mmol) in CH2Cl2 (3 mL) for 1 h, drained and washed with CH2Cl2. For both fragments, SPPS was performed manually using the Boc in situ neutralization procedure.[25] Syntheses were carried out on a 0.1 mmol scale. All operations were performed manually in a glass reaction vessel (20 mL) with a Teflon-lined screw cap. The Na-Boc group was removed by treatment with TFA (100 %) for 2 1 min followed by a 30 s flow wash with DMF. Boc amino acids (0.4 mmol) were coupled by using HATU/DIPEA as activating agents. Coupling times were 5 min throughout the synthesis, without any double-coupling procedures. Following chain assembly, the crude peptides 1 and 2 were cleaved from the resin with simultaneous removal of the side-chain protecting groups by using HF/p-cresol (20:1, v/v) for 1 h at 0 8C. Following the evaporation of HF, the peptides were precipitated with cold diethyl ether, isolated by centrifugation, washed twice with cold diethyl ether, dissolved in acetonitrile/water (1:1 v/v) containing TFA (0.1 %), filtered and lyophilised. Both fragments were purified with a slow gradient as described.[26] The solvent system used was A = 0.1 % TFA in H2O and B = 0.1 % TFA in MeCN.
Protein folding: The protein was refolded overnight in a redox buffer (50 mm Tris-HCl, pH 7.4, 150 mm NaCl, 10 mm reduced and 2 mm oxidised glutathione) at 4 8C. Precipitated protein was removed by centrifugation at 6000 g for 15 min. The supernatant was loaded onto a C18 column, purified by RP-HPLC and recovered by lyophilisation (6.5 mg, yield 80 %). In order to determine if the protein remained folded, and was not subject to degradation after lyophilisation, samples of both proteins (synthetic and recombinant) were re-dissolved in sodium phosphate (20 mm) and re-analyzed by LC–MS and CD spectroscopy.
Native chemical ligation of Cp-5: MPAA (100 mm) and TCEP (20 mm) were dissolved in guanidine hydrochloride (6.0 m) and Na2HPO4 (200 mm), and the mixture was degassed with argon. The pH of the resulting solution was adjusted to 7.0 by addition of NaOH (10 m or 2 m). The two reactants, fragments 1 and 2 (5 mm final concentration) were then added to the reaction mixture, and the pH was readjusted to 7.0. The vial was capped under argon and the reaction mixture was stirred at room temperature. Aliquots of the reaction mixture (2 mL) were diluted four fold into an aqueous solution (5 % TFA v/v) for LC–MS analysis. After 10 h, the reaction reached completion and was quenched by the addition of HCl (5 m), isolated by solid-phase extraction and lyophilised to afford the crude ligation product that was purified by RP-HPLC to give the ligated peptide 3 (8 mg, 40 % yield). The mass of the product was confirmed by LC–MS, (obs = 912.4 Da [M+10 H]10 + ; calcd = 912.6 Da).
NMR spectroscopy: All NMR experiments were performed at 25 8C on a Bruker AV600 spectrometer equipped with a 5 mm z-gradient 1 H/15N/13C cryoprobe optimised for 1H detection. Homonuclear 1D 1 H NMR experiments were performed on a Cp-5 sample (0.1 mm) in sodium phosphate (20 mm) and H2O/D2O (9:1, v/v) at pH 5.2. The protein concentrations were quantified using an extinction coefficient (e280) = 3,355 m1 cm1, as determined by the method of Gill and von Hippel.[27] Proton chemical shifts were referenced to trimethylsilyl propionate (TSP, 1 mm). The 1D spectra were acquired with 2048 complex data points, and water suppression was achieved using the WATERGATE sequence.[28] Spectra were obtained with 1H spectral widths of 7500 Hz and processed and analysed by using Topspin 2.1 software (Bruker).
Cloning and expression of Cp-5: The ORF of Cp-5 was re-cloned into the pProEx HTb vector (Invitrogen) with an N-terminal (His)6-tag and rTEV cleavage site. Chemically competent E. coli cells BL21(DE3), (Invitrogen, California, USA) were transformed with the plasmid DNA pProEx HTb-spp5 construct for overexpression of Cp5. A pre-culture was prepared by inoculating a single colony into lysogeny broth (LB, 50 mL) with ampicillin (0.1 mg mL1) and incubated overnight at 37 8C with shaking at 200 rpm. The pre-culture was used to inoculate 4 L of LB (containing ampicillin; 0.1 mg mL1) at a dilution factor of 1/50. The growth of the cells was monitored by measuring the optical density at 600 nm. Once the OD600 was 0.5, isopropyl-b-d-1-thiogalactopyranoside (IPTG) was added to the culture to a final concentration of 1 mm to induce protein production. Cells were grown for a further 3 h at
ChemBioChem 2015, 16, 328 – 336
www.chembiochem.org
Circular dichroism: All CD measurements were performed on a PiStar spectrometer (Applied Photophysics; Leatherhead, UK). Protein spectral data are reported in terms of the mean residue ellipticity (q [deg cm2 dmol1]), calculated as follows [Eq. (1)]:
q ¼ S=ð10 c L nÞ
335
ð1Þ
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Papers Where S is the raw CD signal (mdeg), c is the protein concentration (M), L is the cuvette path length (cm) and n is the number of peptide bonds in the protein.
[2] a) Z. Wang, E. Choice, A. Kaspar, D. Hanson, S. Okada, S. C. Lyu, A. M. Krensky, C. Clayberger, J. Immunol. 2000, 165, 1486 – 1490; b) J. E. McInturff, S.-J. Wang, T. Machleidt, T. R. Lin, A. Oren, C. J. Hertz, S. R. Krutzik, S. Hart, K. Zeh, D. H. Anderson, R. L. Gallo, R. L. Modlin, J. Kim, J. Invest. Dermatol. 2005, 125, 256 – 263; c) A. Nakashima, A. Shiozaki, S. Myojo, M. Ito, M. Tatematsu, M. Sakai, Y. Takamori, K. Ogawa, K. Nagata, S. Saito, Am. J. Pathol. 2008, 173, 653 – 664. [3] T. Roeder, M. Stanisak, C. Gelhaus, I. Bruchhaus, J. Grotzinger, M. Leippe, Dev. Comp. Immunol. 2010, 34, 203 – 209. [4] a) J. Mysliwy, A. J. Dingley, M. Stanisak, S. Jung, I. Lorenzen, T. Roeder, M. Leippe, J. Grotzinger, Dev. Comp. Immunol. 2010, 34, 323 – 330; b) E. Liepinsh, M. Andersson, J. M. Ruysschaert, G. Otting, Nat. Struct. Biol 1997, 4, 793 – 795. [5] P. E. Dawson, S. B. Kent, Annu. Rev. Biochem. 2000, 69, 923 – 960. [6] L. Z. Yan, P. E. Dawson, J. Am. Chem. Soc. 2001, 123, 526 – 533. [7] a) J. Chen, Q. Wan, Y. Yuan, J. L. Zhu, S. J. Danishefsky, Angew. Chem. Int. Ed. 2008, 47, 8521 – 8524; Angew. Chem. 2008, 120, 8649 – 8652; b) C. Haase, H. Rohde, O. Seitz, Angew. Chem. Int. Ed. 2008, 47, 6807 – 6810; Angew. Chem. 2008, 120, 6912 – 6915. [8] Q. Wan, S. J. Danishefsky, Angew. Chem. Int. Ed. 2007, 46, 9248 – 9252; Angew. Chem. 2007, 119, 9408 – 9412. [9] P. W. Harris, M. A. Brimble, Biopolymers 2013, 100, 356 – 365. [10] a) E. Atherton, R. C. Sheppard, J. Chem. Soc. Chem. Comm 1985, 165 – 166; b) P. W. R. Harris, M. A. Brimble, Biopolymers 2010, 94, 542 – 550. [11] R. Subirs-Funosas, A. El-Faham, F. Albericio, Tetrahedron 2011, 67, 8595 – 8606. [12] M. Villain, H. Gaertner, P. Botti, Eur. J. Org. Chem. 2003, 3267 – 3272. [13] B. Dang, T. Kubota, K. Mandal, F. Bezanilla, S. B. Kent, J. Am. Chem. Soc. 2013, 135, 11911 – 11919. [14] E. C. B. Johnson, S. B. H. Kent, J. Am. Chem. Soc. 2006, 128, 6640 – 6646. [15] J. Dhakal, G. S. Brah, R. K. Agrawal, H. N. Pawar, D. Kaur, R. Verma, Indian J. Med. Microbiol. 2013, 31, 123 – 129. [16] N. Abdullah, H. A. Chase, Biotechnol. Bioeng. 2005, 92, 501 – 513. [17] T. Rehm, R. Huber, T. A. Holak, Structure 2002, 10, 1613 – 1618. [18] C. Scheich, D. Leitner, V. Sievert, M. Leidert, B. Schlegel, B. Simon, I. Letunic, K. Bussow, A. Diehl, BMC Struct. Biol. 2004, 4, 4. [19] S. M. Kelly, T. J. Jess, N. C. Price, Biochim. Biophys. Acta Proteins Proteomics 2005, 1751, 119 – 139. [20] H. H. Hauge, J. Nissen-Meyer, I. F. Nes, V. G. Eijsink, Eur. J. Biochem. 1998, 251, 565 – 572. [21] C. A. Hawkins, E. d. Alba, N. Tjandra, J. Mol. Biol. 2005, 346, 1381 – 1392. [22] W. Li, Personal communication, 2014. [23] P. W. R. Harris, S. H. Yang, M. A. Brimble, Tetrahedron Lett. 2011, 52, 6024 – 6026. [24] T. M. Hackeng, J. H. Griffin, P. E. Dawson, Proc. Natl. Acad. Sci. USA 1999, 96, 10068 – 10073. [25] M. Schnçlzer, P. Alewood, A. Jones, D. Alewood, S. B. Kent, Int. J. Pept. Protein Res. 1992, 40, 180 – 193. [26] P. W. R. Harris, D. J. Lee, M. A. Brimble, J. Pept. Sci. 2012, 18, 549 – 555. [27] S. C. Gill, P. H. von Hippel, Anal. Biochem. 1989, 182, 319 – 326. [28] M. Piotto, V. Saudek, V. Sklenar, J. Biomol. NMR 1992, 2, 661 – 665. [29] M. B. McNeil, H. G. Hampton, K. J. Hards, B. N. Watson, G. M. Cook, P. C. Fineran, FEBS Lett. 2014, 588, 414 – 421.
Each CD spectrum measurement represents the average of seven scans obtained with a 2 nm optical bandwidth. Baseline spectra were collected with buffer alone and then subtracted from the raw protein spectral data. The measurements were performed at protein concentrations of 2–5 mm in sodium phosphate buffer (20 mm, pH 5.2) at 25 8C in quartz cuvettes (1 mm; Hellma Analytics, Mllheim, Germany). Synthesis of shortened analogues: The syntheses of peptides Cp5A–Cp-5E (described in Table 1) were carried out on a 0.1 mmol scale on aminomethyl resin, which was pre-loaded with a Boc-AAPAM linker (AA corresponds to the C-terminal residue of each fragment). The Na-Boc group was removed by treatment with neat TFA followed by a flow wash with DMF. Boc amino acids were coupled by using HATU/iPr2EtN as the activating agent and base, respectively. Coupling times were 5 min throughout the synthesis, without the use of any double-coupling procedures. Following peptide chain assembly, the crude peptides were cleaved from the resin with simultaneous removal of side-chain protecting groups by using HF/p-cresol (20:1, v/v) for 30 min at 0 8C. Fluorescence quenching dependent on DpH: Proton translocation into inverted membrane vesicles was measured by the quenching of the fluorescent, pH-dependent probe acridine orange (AO), which was monitored by using a Cary Eclipse Fluorescence spectrophotometer. The assay buffer contained HEPES (10 mm, pH 6.0), KCl (100 mm), MgCl2 (5 mm), E. coli inverted membrane vesicles (0.085 mg mL1) and AO (2.5 mm). Inverted membrane vesicles were prepared as described by McNeil et al.,[29] with the exception that the cells were grown aerobically in conical flasks. The reactions were initiated with NADH (50 mm) and quenching was reversed by injection of the desired compound.[29] The excitation and emission wavelengths were 493 and 530 nm, respectively.
Acknowledgements The authors would like to thank Maurice Wilkins Centre for Molecular Biodiscovery for funding this work and Prof. Russel Snell for providing a picture of C. elegans. Keywords: antibiotics · antimicrobial protein · caenopore · native chemical ligation · protein design · protein folding [1] a) T. Kolter, F. Winau, U. E. Schaible, M. Leippe, K. Sandhoff, J. Biol. Chem. 2005, 280, 41125 – 41128; b) K. A. Brogden, Nat. Rev. Microbiol. 2005, 3, 238 – 250; c) R. I. Lehrer, Nat. Rev. Microbiol. 2004, 2, 727 – 738; d) M. Zasloff, Nature 2002, 415, 389 – 395.
ChemBioChem 2015, 16, 328 – 336
www.chembiochem.org
Received: September 7, 2014 Published online on November 25, 2014
336
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
