First report on the synthesis and structural studies of trans-Phakellistatin 18: a rotamer of marine natural product phakellistatin 18

Abstract Phakellistatin peptides from marine organisms are the sources of proline-rich cyclic peptides with reported significant antitumor activities. Phakellistatin 18 (1), reported from marine sponge Phakellia fusca, contains three proline-peptide linkages in cis form. We attempted the total synthesis of natural product 1 through solution-phase macrocyclization approach, as a result, the synthetic cyclic peptide 2 was obtained as a rotamer of natural product having all three proline residues in trans-conformation. Here, we describe the synthesis, structural, and cytotoxicity studies of trans-Phakellistatin 18 (2), and its analog [Ala1,3,6]-Phakellistatin 18 (3). Detailed NMR studies were carried out to characterize the synthesized peptides, and anti-cancer screening was performed by using MTT assay. The synthetic trans-Phakellistatin 18 (2) (IC50=67.5 ± 2.938 µM) showed comparable cytotoxicity against HepG2 cancer cell line with standard drug doxorubicin (IC50=63.88 ± 6.48 µM). Here, the first synthetic and structural studies on trans-Phakellistatin 18 (2), and its anticancer screening against HepG2 cell line was reported. Graphical Abstract


Introduction
Natural products are the great source of anticancer agents (Tan et al. 2006). Various structural modifications are reported to be helpful in improving the anticancer potential of natural products (Mahal et al. 2017;Mahal et al. 2019;Duan et al. 2021). Proline-rich cyclic peptides are among such interesting class of naturally occurring secondary metabolites having enormous biological potentials (Pujals and Giralt 2008;Joo 2012;Fang et al. 2016;Jing and Jin 2020;Zhang et al. 2021). This class of cyclic peptides are obtained mainly from marine organisms (Zheng et al. 2011;Gogineni and Hamann 2018;Zheng et al. 2018;Macedo et al. 2021), and some higher plants (Morita et al. 1997;Fisher et al. 2020;Dahiya and Dahiya 2021;Daly and Wilson 2021). Generally cyclic peptides are reported to have higher selectivity, more stability, and greater potency than the linear peptides due to restricted mobility and less degree of freedom (Fung and Hruby 2005;Roxin and Zheng 2012;Yang et al. 2018;Bajaj 2019;Liu et al. 2021). The presence of proline or any cyclic amino acid residues embedded in a cyclic framework make their conformational mobility more restricted owing to the selective yet high potency of these molecules (Vermeer et al. 2012;Yang et al. 2018). Phakellistatin peptides obtained from marine organisms are one of the greatest source of proline-rich cyclic peptides with reported significant antitumor activities (Zheng et al. 2011;Zhang et al. 2021), which makes it the center of attention for many research groups (Meli et al. 2017). Twenty-two members of phakellistatin class have been reported so far, while many members of this class are reported as a mixture of proline rotamers (Meli et al. 2017;Kwon et al. 2018).
Phakellistatin peptides are obtained in very limited quantity from natural sources. To increase the yields of these secondary metabolites, alternative sources of these important secondary metabolites have been sought through synthesis based on detailed studies on their structures and biological activities. It has been reported from previous studies that biological activities of synthesized analogues were different from isolated one, which is yet a great challenge to be solved, most probably due to their differences in conformations or due to incomplete characterization of the structures of these compounds or from the existence of cis-trans stereoisomerism at equilibrium (Napolitano et al. 2005;Meli et al. 2017;Fisher et al. 2020).
Phakellistatin 18 (1) is a proline-rich cycloheptapeptide isolated from marine sponge Phakellia fusca. It has three proline residues having cis geometry. Phakellistatin 18 (1), obtained from natural source, was screened against P388 and BEL-7402 cancer cell lines, but no cytotoxic activity was observed (Zhang et al. 2010). Wu et al. (2018) reported the synthetic peptide similar in structure and geometry with the natural product peptide phakellistatin 18 (1) ), but there is no detailed discussion or evidence on the geometry of proline residues supported by DEPT experiment and 2D-NMR was provided. The synthetic product was also reported to have weak inhibitory activity against BEL-7402 (IC 50 ¼85.5 mM) and A549 (IC 50 ¼72.4 mM) cancer cell lines . This difference in activity is may be due to presence of cis-trans isomerism in equilibria.
In the current study, we attempted the total synthesis of natural product 1 via solution-phase macrocyclization approach, and further investigated the structural and the biological aspects of synthetic peptide 2. The linear peptide precursor was used as a starting material for the synthesis of cyclic peptide due to the fact that it is reported that solution-phase macrocyclization gives higher yields (Malesevic et al. 2004;Prior et al. 2018). The studies were further extended towards the anticancer screening of synthetic peptide 2 on HepG2, HeLa, MCF-7, MDA-MB-231, and PC3 cell lines in order to explore its anti-cancer activity on some other cancer cell lines. Alanine substituted analog [Ala 1,3,6 ]-Phakellistatin 18 (3) was also synthesized to study the effect of proline residues on the biological activity as literature suggested that proline residues have significant impact on the anti-cancer potential of cyclic peptides (Chiangjong et al. 2020;Zhang et al. 2021).

Results and discussion
The synthesis of cyclic peptides natural product 1, and its analog 3 was carried out in a combination of solid-and solution-phase methodologies (Malesevic et al. 2004;Shaheen et al. 2012;Shaheen et al. 2016;Ali et al. 2020) (Scheme S1, supplementary material). The linear precursors (2a, and 3a) were constructed using Fmoc assisted solid-phase peptide synthetic methodology (Fu et al. 2018) on Wang resin as a solid support. Each coupling was performed in the presence of DIC as an activating agent and OxymaPure as a racemization suppressor. The first coupling on resin was carried out in the presence of catalytic amount of DMAP. The resin was capped with acetic anhydride in the presence of pyridine after first coupling. Following each coupling, the Fmoc group was removed with 20% solution of 4-methylpipridine in DMF. The progress of each coupling and decoupling step was monitored with the help of Kaiser test. Later obtaining desired sequence, the peptides were cleaved from resin by using TFA cocktail with the composition of TFA/TIPS/Phenol/H 2 O (88:2:5:5 v/v), and precipitated with cold diethyl ether. The crude linear peptides were purified by RP-HPLC and cyclized using solution-phase methodology in the presence of DIPEA and HATU (Malesevic et al. 2004).
It is reported that in proline-rich peptides, cis-trans isomerism exist and both geometric isomers exist in dynamic equilibria with each other. In the absence of any structural restrictions, the proline-rich peptides tend to exist in more stable trans conformation as a major rotamer along with small detectable quantity of cis rotamer as minor conformer (Napolitano et al. 2003). Similarly, we have obtained linear peptide 2a after purification as a mixture of cis-and trans-rotamers, in which trans-rotamer was present as a major geometric isomer (supplementary material, Table S1) with some minor peaks of cis-rotamer in 13 C-NMR (supplementary material, Figure S6).
Cyclic product 2 was further purified by using RP-HPLC, and single product with molecular mass corresponding to our targeted peptide, phakellistatin 18. While the percent purity was established by UPLC profile (supplementary material, Figure S2). Cyclic product 2 showed molecular ion peak [M þ H] þ at m/z 828.4652 corresponding to molecular formula C 45 H 62 N 7 O 8 þ , and is established by HRESI-MS. Analysis of reported 13 C-NMR data of natural product showed that three prolines exist in cis conformation, while in synthetic product 2, the proline-peptide bonds between Ile 7 -Pro 1 , Tyr 2 -Pro 3 , and Phe 5 -Pro 6 exist in trans conformation with DdC b-C c 4.24, 4.58, and 4.84, respectively. Detailed comparison of NMR of reported natural product phakellistatin 18 (1) (Zhang et al. 2010) with synthesized peptide 2 is summarized in Table S2 (supplementary material). 1 H-1 H-NOESY showed that H d of all three prolines exhibited correlation with the H a of their preceding amino acid moieties, confirming the trans geometry at each proline residue of our synthetic peptide 2 (supplementary material, Figure S17).
From this, we have concluded that the synthetic peptide 2 has different geometry at all three proline residues, as compared to natural product phakellistatin 18. In order to analyze the effect of these three proline residues on the bioactivity of this peptide, new cyclic analog, [Ala 1,3,6 ]-Phakellistatin 18 (3) was prepared by substituting all three proline residues of peptide 2 by L-Ala, as described earlier for the natural product (Scheme S1, supplementary material). The crude [Ala 1,3,6 ]-linear precursor (3a) was purified by using recycling RP-HPLC, and its purity was established by UPLC (supplementary material, Figure S2). The pure peptide 3a showed molecular ion peak [M þ H] Table S4). Phakellistatin 18 (1) was previously isolated from natural source, and was evaluated for its biological activity against P388 and BEL-7402 cancer cell lines, but no cytotoxic activity was observed (Zhang et al. 2010). Wu et al. (2018) reported its synthesis, and evaluated the biological potential of their synthetic peptide against BEL-7402 (IC 50 ¼85.5 mM) and A549 (IC 50 ¼72.4 mM) cancer cell lines ). In the current study, the synthetic trans-Phakellistatin 18 (2), its cyclic alanine substituted analog  Table S5). The synthetic trans-Phakellistatin 18 (2) was found to be active, and showed cytotoxicity against HepG2 cell line with an IC 50 value of 67.5 ± 2.938 mM. This peptide 2 inhibited the HepG2 cell line in comparison with the standard drug doxorubicin (IC 50 ¼ 63.88 ± 6.48 mM). While the linear precursor 2a showed no cytotoxicity. When we have replaced all proline residues of peptide 2 with L-alanine, we found that the activity is affected as can be seen in linear precursor 3a which showed IC 50 of 95.83 ± 7.319 mM, and cyclic peptide 3 showed IC 50 of 92.57 ± 3.576 mM (supplementary material, Figure  S32). This concludes that the proline residues have effect on the activity of this peptide molecule on HepG2 cell line-. Cytotoxicity of these peptides was also evaluated against HeLa, MCF-7, MDA-MB-231, and PC3 cell lines using MTT protocol. These peptides were found to be inactive on other cell lines used in this study, showing the selective activity of these peptides towards HepG2 cell line. This study revealed that trans-phakellistatin 18 (2), and its alanine substituted analog (3) exhibited the potential to serve as a chemotherapeutic prototype agent.

General experimental procedures
All amino acids, Wang resin (loading capacity 1.2 meq/g) were purchased from Chem. Impex (USA). All the solvents for purification were HPLC grads and purchased from sigma Aldrich. All the chemicals used in the experimental portion were 95-99% pure. Purification of all the synthesized peptides was done using RP-HPLC preparative and recycling HPLC (LC-908W), Japan Analytical Industry Co., Ltd., Jaigel ODS-MAT 80 (C18) column with ACN: H 2 O: TFA (60: 40: 0.08) solvent system. Percentage purity of all peptides was determined on UPLC using the same solvent system on Agilent 1260 Infinity Diode Array, C-4 reversed-phase analytical column, 5 lm, 150 Â 4.6 mm. International Bruker Avance NMR spectrometer (600 MHz) was used to record NMR of samples. QSTAR XL MS/MS SYSTEMS (AB Science, USA) was used to record ESI-MS. Thermoscientific evolution 300 spectrophotometer was used to record UV (Ultraviolet) absorption spectra, while JASCO 302-A Infrared Spectrometer was used to record IR (Infrared absorption spectra). Optical rotations were measured on a JASCO DIP 360 polarimeter at the sodium D line.
% Inhibition ¼ 100 -Abs of test compound À Abs of blank ð Þ = Abs of control À Abs of blank ð Þ IC 50 (mM) of the compounds that showed 50% decrease in cell viability was evaluated by using EZ-FIT Enzyme Kinetics Program. The activity of the peptides was compared with a well-known anti-cancer drug; doxorubicin, which was used as a standard.

Conclusion
The synthesis of natural product phakellistatin 18 (1) by solution-phase macrocyclization approach afforded a natural product rotamer 2. In contrast to natural product conformation with all three proline residues in cis form, the synthetic peptide 2 obtained in current study, was observed to have trans conformation at all proline-peptide bond linkages We found that peptide 2 exhibited cytotoxicity against HepG2 cell line, which is comparable to standard drug doxorubicin. Synthetic analog 3, and its linear precursor 3a were found to be weakly active. These peptides showed no cytotoxicity towards the other cell lines used in this study.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
Authors acknowledge the financial support from Higher Education Commission, Pakistan. This work was supported by a grants 5738/Sindh/NRPU/R&D/HEC/2016 and 8169/Sindh/NRPU/R&D/ HEC/2017 from the Higher Education Commission, Pakistan.