Cyclic Octamer Peptoids: Simplified Isosters of Bioactive Fungal Cyclodepsipeptides

Cyclic peptoids have recently emerged as an important class of bioactive scaffolds with unique conformational properties and excellent metabolic stabilities. In this paper, we describe the design and synthesis of novel cyclic octamer peptoids as simplified isosters of mycotoxin depsipeptides bassianolide, verticilide A1, PF1022A and PF1022B. We also examine their complexing abilities in the presence of sodium tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (TFPB) salt and explore their general insecticidal activity. Finally, we discuss the possible relationship between structural features of free and Na+-complexed cyclic octamer peptoids and bioactivities in light of conformational isomerism, a crucial factor affecting cyclic peptoids’ biomimetic potentials.


Introduction
While the vast inventory of natural products finds its limits in the finiteness of biosynthetic pathways [1], the synthesis of non-natural analogues has no boundaries [2]. For their distinct conformational properties [3], their excellent biostabilities [4] and straightforward modular construction [5], peptoids (i.e., oligomers of N-substituted glycines) [6] represent a formidable class of biomimetic compounds with striking bio-and pharmacological activities [7,8].
Among peptoid oligomers, cyclic peptoids hold a special status for their intrinsic ability to adopt compact folded conformations [9,10] and exhibit conspicuous therapeutic properties [9,11]. Their excellent biomimetic properties have been confirmed by a recent study in which abiotic hexamer cyclopeptoids (mimicking depsipeptide mycotoxins of the enniatins class) evoked potent cytotoxic activities against cancer cell lines [12].
Peptoid congeners 5-8 ( Figure 1) were obtained replacing the α-amino/hydroxy acids methines and intraannular N/O atoms present in natural cyclodepsipepides with nitrogen atoms and methylene groups, respectively. N-methyl substituents of L-leucine/alanine residues were removed in order to have a sequence of N-alkyl glycines. The relative position of the carbonyl groups was left unaltered to preserve possible metal chelating properties, essential for biological activities [14,[20][21][22][23].
Molecules 2018, 23  The depsipeptide/peptoid core switch yields congeners fully compatible with the solid-phase synthesis. Moreover, starting materials are inexpensive and the availability of hundreds of primary amines allows extensive structure-activity investigations.

Chemistry and Conformational Studies
The syntheses of cyclic octamer peptoids 5-8 began with the solid-phase "submonomer" construction of linear precursors 9-12 (Table 1) [5]. N,N'-diisopropylcarbodiimide-induced bromoacetic acid condensations were alternated with amine substitutions until formation of the desired linear oligomeric amides. Detachment from the acid-labile solid support in the presence of hexafluoroisopropanol yielded octamers in acceptable to good yields and decent purities (Table 1).  The depsipeptide/peptoid core switch yields congeners fully compatible with the solid-phase synthesis. Moreover, starting materials are inexpensive and the availability of hundreds of primary amines allows extensive structure-activity investigations.

Chemistry and Conformational Studies
The syntheses of cyclic octamer peptoids 5-8 began with the solid-phase "submonomer" construction of linear precursors 9-12 (Table 1) [5]. N,N'-diisopropylcarbodiimide-induced bromoacetic acid condensations were alternated with amine substitutions until formation of the desired linear oligomeric amides. Detachment from the acid-labile solid support in the presence of hexafluoroisopropanol yielded octamers in acceptable to good yields and decent purities (Table 1). Head-to-tail cyclisations of crude linear peptoids 9-12 were performed in high dilution conditions (3.0 mM concentration) in the presence of HATU as the coupling agent [26]. Cyclic compounds were purified from traces of unreacted/oligomerized linear peptoids through precipitation from hot acetonitrile (in the case of 5 and 7) or reverse-phase column chromatography (oligomers 6 and 8). Cyclic oligoamides 5-8 were isolated as white amorphous solids (purity > 98%, HPLC analysis).
2D homonuclear (COSY) and heteronuclear (HMQC, HMBC) experiments allowed assignment of the 1 H/ 13 C resonances for cyclic peptoids 5-7, suggesting formation of C 2 -symmetric ccttcctt species for 5 and 7 (half of the signals were present in the NMR spectra, as previously reported for octamer peptoids [29][30][31][32][33][34]) and a C 1 -symmetric species for 6. Peptoid bond geometries have been documented by 1 H NMR chemical shifts values. In particular, relatively low values for Cα-H N-side chains' or small ∆δ for diastereotopic Cα-H 2 testified for trans peptoid junctions; higher values of chemical shifts or larger ∆δ were indicative of cis amide bonds [12]. The NMR data identified two possible alternative structures for symmetric 5 and 7. In these congeners two of the four isobutyl/isopropyl and benzyl side chains were located on cis amide bonds and the other two were on trans peptoid junctions (5a/5b and 7a/7b, Figure 3, see also Figure S2 of Supplementary Material).
Accurate analysis of the 1 H NMR resonances showed no symmetry elements for compound 6 (to the best of our knowledge, this is the first case of symmetry breaking in the case of a conformationally stable, symmetrically substituted, cyclic octamer peptoid) (see experimental part and Figure S2 of Supplementary Material). The presence of two N-benzyl and two N-isobutyl side chains on cis amide bonds, plus two N-isobutyl and two N-methyl side chains on trans peptoid junctions testified for the asymmetric arrangement of cyclic oligomer 6.
or larger Δδ were indicative of cis amide bonds [12]. The NMR data identified two possible alternative structures for symmetric 5 and 7. In these congeners two of the four isobutyl/isopropyl and benzyl side chains were located on cis amide bonds and the other two were on trans peptoid junctions (5a/5b and 7a/7b, Figure 3, see also Figure S2 of Supplementary Material).
Accurate analysis of the 1 H NMR resonances showed no symmetry elements for compound 6 (to the best of our knowledge, this is the first case of symmetry breaking in the case of a conformationally stable, symmetrically substituted, cyclic octamer peptoid) (see experimental part and Figure S2 of Supplementary Material). The presence of two N-benzyl and two N-isobutyl side chains on cis amide bonds, plus two N-isobutyl and two N-methyl side chains on trans peptoid junctions testified for the asymmetric arrangement of cyclic oligomer 6. C2-symmetric ccttcctt-conformational diastereoisomers 5a/b and 7a/b were modelled by DFT (Density Functional Theory) studies and minimum energy structures were located in order to find the plausible most stable conformations. Free energies in CHCl3 have been reported in Figure 4 (see S.M. for computational details). Isomers 5a and 7b were found to be more stable than the corresponding conformational diastereoisomers 5b and 7a of 2.2 and 0.8 kcal/mol, respectively. The energy differences are mainly due to steric interactions among the N-substituents and the ring. More in detail, as reported in Figure 4, 5b shows isobutyl-ring distances shorter than 5a, whereas interactions among the phenyl groups and the ring is responsible for the higher energy of 7a with respect to 7b. Furthermore, the higher energy difference between 5a and 5b (2.2 kcal/mol) could explain why the NMR spectrum of 5 shows only one conformational isomer, whereas the 1 H NMR spectrum of 7 shows a small amount of the other isomer ( Figure 2c). These data are also concordant with the VT-experiments (the coalescent temperature and therefore the conformational interconversion barrier, is higher for cyclopeptoid 5). C 2 -symmetric ccttcctt-conformational diastereoisomers 5a/b and 7a/b were modelled by DFT (Density Functional Theory) studies and minimum energy structures were located in order to find the plausible most stable conformations. Free energies in CHCl 3 have been reported in Figure 4 (see S.M. for computational details). Isomers 5a and 7b were found to be more stable than the corresponding conformational diastereoisomers 5b and 7a of 2.2 and 0.8 kcal/mol, respectively. The energy differences are mainly due to steric interactions among the N-substituents and the ring. More in detail, as reported in Figure 4, 5b shows isobutyl-ring distances shorter than 5a, whereas interactions among the phenyl groups and the ring is responsible for the higher energy of 7a with respect to 7b. Furthermore, the higher energy difference between 5a and 5b (2.2 kcal/mol) could explain why the NMR spectrum of 5 shows only one conformational isomer, whereas the 1 H NMR spectrum of 7 shows a small amount of the other isomer ( Figure 2c). These data are also concordant with the VT-experiments (the coalescent temperature and therefore the conformational interconversion barrier, is higher for cyclopeptoid 5).  It is interesting to note that the removal of two phenyl groups in compounds 6 (where two Nmethyl groups take the place of two N-benzyl side chains in 7) causes a profound effect on the intrinsic stability of the cyclopeptoid framework and breaks the symmetry of the octameric species.
Another important aspect regarding the conformational isomerism of rigid cyclic octamer peptoids is related to their conformational chirality [10].
The presence of a C1-or C2-symmetry axis and the relatively high barrier for conformational inversion of the cyclooligomeric species (ΔG ‡ ≥ 10 kcal/mol) makes cyclic peptoid 5-7 chiral and It is interesting to note that the removal of two phenyl groups in compounds 6 (where two N-methyl groups take the place of two N-benzyl side chains in 7) causes a profound effect on the intrinsic stability of the cyclopeptoid framework and breaks the symmetry of the octameric species.
Another important aspect regarding the conformational isomerism of rigid cyclic octamer peptoids is related to their conformational chirality [10].
The presence of a C 1 -or C 2 -symmetry axis and the relatively high barrier for conformational inversion of the cyclooligomeric species (∆G ‡ ≥ 10 kcal/mol) makes cyclic peptoid 5-7 chiral and observable as conformational enantiomers in solution by NMR spectroscopy (which, at room temperature, reveals species with lifetimes exceeding 10 -2 s) [10].

Complexation Studies
The biological action of CDPs is often associated to the formation of metal complexes [18][19][20]. We therefore tested the metal chelating attitudes of cyclic peptoids 5-8 in the presence of sodium tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (NaTFPB, [12]) in CDCl 3 solutions ( 1 H NMR analysis, Figure 6). In all the performed experiments the presence of two equivalents of Na + cation induced formation of highly symmetric metalated species in slow equilibrium with the corresponding free hosts on the NMR time scale. Addition of one equivalent of NaTFPB showed the formation of a mixture of multiple complexes in slow equilibrium on the NMR time scale. Two equivalent of cationic guest established the formation of single detectable, highly symmetric, complexes. The relatively low chemical shift values observed for the NAla  [36].
The host/guest ratio (calculated integrating the signals of the host/guest complex respect to those of dissolved guest, NaTFPB) remained constant during all the titration experiments and documented a 1:2 macrocycle/Na + ratio for all the synthesized complexes. The apparent association constants were therefore calculated for the formation of the [5-   Formation of conformational enantiomers in rigid octamer cyclopeptoids as free hosts or complexes represents a crucial factor to predict/interpret their biological action. It should be noted that the presence of enantiomorphous species is not detrimental to the biological activity only when the biological target is not chiroselective (as in the case of the cell membrane) [39]. Things change dramatically when the target of biological action is a chirospecific receptor (proteins, carbohydrates or nucleic acids).
In the case of enniatin A and B, both enantiomers are bioactive [40] (cyclic hexadepsipeptides, in fact, interfere with the cations' transport through the membrane). Studies on the molecular action of octamer depsipetides indicate specific molecular targets for these molecules. It is known for example, that PF1022A acts on the latrophin-like receptor and on the Ca 2+ -activated K + channel [20]. Verticilide A1 has been shown to bind selectively to the insect ryanodine receptor, a major target for modern insecticides [15] and it efficiently inhibits the acyl-CoA cholesterol acyltransferases ACAT1 and ACAT2 [41]. Emodepside, a broad spectrum anthelmintic cyclooctadespispetide derived from PF1022A [21], has multiple molecular targets. On the basis of these considerations, it is clear that the presence of stable free and complexed conformationally stable diastereomers and enantiomers is detrimental for the potential bioactivity.
At this point, conscious of the intrinsic drawbacks due to the conformational and stereochemical dispersion of the synthesized cyclic peptoids, we decided to perform a test on silkworm larvae in order to evaluate general insecticidal activity.

Insecticidal Activity against Silkworm Larvae
Bassianolide (1) was reported to exhibit acute lethal toxicity against silkworm larvae [14]. We tested insecticidal activity of 5, a structural congener of bassianolide and analogues 6-8. Each peptoid in dimethyl sulfoxide (DMSO) was injected into open vessels of 4th-instar larvae. The administration of 5-8 caused death of a few larvae at a dose of 300 nmol/larva after 72-120 h ( Table 2). Their toxicities Formation of conformational enantiomers in rigid octamer cyclopeptoids as free hosts or complexes represents a crucial factor to predict/interpret their biological action. It should be noted that the presence of enantiomorphous species is not detrimental to the biological activity only when the biological target is not chiroselective (as in the case of the cell membrane) [39]. Things change dramatically when the target of biological action is a chirospecific receptor (proteins, carbohydrates or nucleic acids).
In the case of enniatin A and B, both enantiomers are bioactive [40] (cyclic hexadepsipeptides, in fact, interfere with the cations' transport through the membrane). Studies on the molecular action of octamer depsipetides indicate specific molecular targets for these molecules. It is known for example, that PF1022A acts on the latrophin-like receptor and on the Ca 2+ -activated K + channel [20]. Verticilide A1 has been shown to bind selectively to the insect ryanodine receptor, a major target for modern insecticides [15] and it efficiently inhibits the acyl-CoA cholesterol acyltransferases ACAT1 and ACAT2 [41]. Emodepside, a broad spectrum anthelmintic cyclooctadespispetide derived from PF1022A [21], has multiple molecular targets. On the basis of these considerations, it is clear that the presence of stable free and complexed conformationally stable diastereomers and enantiomers is detrimental for the potential bioactivity.
At this point, conscious of the intrinsic drawbacks due to the conformational and stereochemical dispersion of the synthesized cyclic peptoids, we decided to perform a test on silkworm larvae in order to evaluate general insecticidal activity.

Insecticidal Activity against Silkworm Larvae
Bassianolide (1) was reported to exhibit acute lethal toxicity against silkworm larvae [14]. We tested insecticidal activity of 5, a structural congener of bassianolide and analogues 6-8. Each peptoid in dimethyl sulfoxide (DMSO) was injected into open vessels of 4th-instar larvae. The administration of 5-8 caused death of a few larvae at a dose of 300 nmol/larva after 72-120 h ( Table 2). Their toxicities were much weaker than that of bassianolide (1), which induced atonic symptom within 0.5 h and 100% mortality in a week at a dose of 5 µg (=5.5 nmol)/larva [14]. Table 2. Toxicity of 5-8 against silkworms by haemolymph injection a . Anti−proliferative potentials of the 5-8 were tested on A375 (human melanoma) cancer cell line. The cells were incubated for 72 h with increasing concentration of compounds (10-25-50 µM) and cell viability was determined by MTT proliferation assay. The data indicated that the assayed compounds did not affect the cell vitality (see experimental part for general procedures).

General Methods
Starting materials and reagents purchased from commercial suppliers were generally used without purification unless otherwise mentioned. HPLC analyses were performed on a JASCO LC-NET II/ADC equipped with a JASCO Model PU-2089 Plus Pump and a JASCO MD-2010 Plus UV-vis multiple wavelength detector set at 220 nm. The column used was a C 18 reversed-phase analytical column (Waters, Bondapak, 10 µm, 125 Å, 3.9 mm × 300 mm) run with linear gradients of ACN (0.1% TFA) into H 2 O (0.1% TFA) over 30 min, at a flow rate of 1.0 mL/min for the analytical runs. ESI-MS analysis in positive ion mode was performed using a Finnigan LCQ Deca ion trap mass spectrometer (ThermoFinnigan, San Josè, CA, USA) and the mass spectra were acquired and processed using the Xcalibur software provided by Thermo Finnigan. Samples were dissolved in 1:1 CH 3 OH/H 2 O, 0.1% formic acid and infused in the ESI source by using a syringe pump; the flow rate was 5 µL/min. The capillary voltage was set at 4.0 V, the spray voltage at 5 kV and the tube lens offset at −40 V. The capillary temperature was 220 • C. Data were acquired in MS1 and MSn scanning modes. Zoom scan was used in these experiments. High-resolution mass spectra (HRMS) were recorded on a Bruker Solarix XR Fourier transform ion cyclotron resonance mass spectrometer (FTICR-MS) equipped with a 7T magnet, using electrospray ionization (ESI). Yields refer to chromatographically and spectroscopically ( 1 H-and 13 C NMR) pure materials. NMR spectra were recorded on a Bruker DRX 600 ( 1 H at 600.13 MHz, 13 C at 150.90 MHz), Bruker DRX 400 ( 1 H at 400.13 MHz, 13 C at 100.03 MHz), Bruker DRX 300 ( 1 H at 300.13 MHz, 13 C at 75.03 MHz). Chemical shifts (δ) are reported in ppm relative to the residual solvent peak (CHCl 3 , δ = 7.26; 13 CDCl 3 , δ = 77.00; C 2 DHCl 4 , TCDE, δ = 5.80) and the multiplicity of each signal is designated by the following abbreviations: s, singlet; d, doublet; dd, double doublet; t, triplet; sept, septet; m, multiplet; br, broad. 2D NMR experiments such as COSY, ROESY, HSQC and HMBC were performed for the full assignment of each signal. Coupling constants (J) are quoted in Hertz.

General Procedure for the Solid-Phase Synthesis of Linear Peptoids 9-12
The 2-chlorotrityl chloride resin (α-dichlorobenzhydryl-polystyrene cross-linked with 1% DVB; 100-200 mesh; 1.63 mmol g -1 , 0.400 g, 0.652 mmol) was swelled in dry CH 2 Cl 2 (4 mL) for 45 min and washed twice with dry CH 2 Cl 2 (4 mL). The first submonomer was attached onto the resin by adding bromoacetic acid (0.136 g, 0.978 mmol) in dry CH 2 Cl 2 (4 mL) and DIPEA (567 µL, 3.26 mmol) on a shaker platform for 60 min at room temperature, followed by washing with CH 2 Cl 2 (3 × 1 min) and then again with DMF (3 × 1 min). A solution of the chosen amine (1.6 M in dry DMF, 4 mL) was added to the bromoacetylated resin. The mixture was left on a shaker platform for 40 min at room temperature, then the resin was washed with DMF (3 × 1 min), CH 2 Cl 2 (3 × 1 min) and then again with DMF (3 × 1 min). Subsequent bromoacetylation reactions were accomplished by reacting the aminated oligomer with a solution of bromoacetic acid (0.910 g, 6.52 mmol) and DIC (1.11 mL, 7.17 mmol) in dry DMF (4 mL) for 60 min at room temperature. The completion of the acylation reactions was verified by using the chloranil test. The filtrated resin was washed with DMF (3 × 1 min), CH 2 Cl 2 (3 × 1 min), DMF (3 × 1 min) and treated again with the proper amine under the same conditions reported above. This cycle of reactions was iterated until the target linear oligomer was obtained. The cleavage was performed treating the resin, previously washed with CH 2 Cl 2 (3 × 1 min), three times with a solution of HFIP in CH 2 Cl 2 (20% v/v, 4.0 mL each time) on a shaker platform at room temperature for 30 min each time. The resin was then filtered away and the combined filtrates were concentrated in vacuo. 1 mg of the final products were dissolved in 60 µL of acetonitrile (0.1% TFA) and 60 µL of HPLC grade water (0.1% TFA) and analysed by RP-HPLC; purity ≥ 71%; conditions: 5 → 100% A in 30 min for the all oligomers (A, 0.1% TFA in acetonitrile, B, 0.1% TFA in water); flow: 1.0 mL min −1 , 220 nm]. The linear oligomers (isolated as amorphous solids) were subjected to ESI mass spectrometry (see Table 1) and, subsequently, to the cyclization reactions without further purification.
where T c is the coalescence temperature, ∆ν is the difference in Hertz between the two coupled signals and J is the coupling constant between the two signals [42]. For compounds [5][6][7][8]  To a 5.0 mM solution of cyclic peptoid 7 in CDCl 3 (0.5 mL), 1.0 equivalent of Pirkle's alcohol ((R)-1-(9-anthryl)-2,2,2-trifluoroethanol) was added. After the addition, the mixture was mixed for 1 min and the 1 H NMR spectrum was recorded. Further 1.0 equivalent of Pirkle's alcohol was added in order to increase the protons resonances' splitting. NMR spectra were recorded on a Bruker DRX 600 ( 1 H at 600.13 MHz).

Assay of Insecticidal Activity against Silkworm Larvae
Third-instar larvae of silkworm Bombyx mori were purchased from Kougensha Co, Ltd. (Matsumoto, Japan) and cultured on an artificial diet SilkMate 2S obtained from Nosan Corporation. Each compound dissolved or suspended in DMSO was injected into open vessels of 4th-instar larvae (mean body weight ca 0.9 g) resulting in the dose of 300 nmol/larva. A total of 10 larvae were tested and the number of dead larvae was counted.

Conclusions
Understanding the conformational features of cyclooligoamides represents a necessary step to exploit their biological properties. In particular, the comprehension of their stereochemical traits is crucial, especially when the generation of rigid structures induces the formation of conformational enantiomorphs.
The isosteric transformation of bioactive cyclooctadepsipeptides in cyclooctapeptoids represents a simple way to explore the molecular space conjunct to macrocyclic oligoamides. Combined NMR and computational techniques allowed us to assign the structural identity of the formed species. The intrinsic conformational stability of cyclooctamer peptoids, in their free and complexed form, clarified their intricate topological features and evidenced possible detrimental effects on the binding interactions with target biomolecules and their scarce biological activities on silkworm larvae.
Differently from conformationally heterogeneous cyclohexapeptoids [10,12], cyclic octamer peptoids show limited capacity to mimic the natural counterparts. Although they are perfectly suited for solid-phase synthesis and combinatorial approaches to drug discovery, they show undesirable conformational and stereochemical complication that can decrease their adaptability to the molecular target.
Additional efforts are currently underway in order to design libraries of diverse and more malleable compounds that can exhibit increased biological properties and higher conformational mobility.