Conoidecyclics A-C from marine macroalga Turbinaria conoides: Newly described natural macrolides with prospective bioactive properties Author Chakraborty, Kajal * & Marine Bioprospecting Section of Marine Biotechnology Division, Central Marine Fisheries Research Institute, Ernakulam North, P. B. No. 1603, Cochin, Kerala State, & * Author Dhara, Shubhajit text Phytochemistry 2021 112909 2021-11-30 191 1 14 http://dx.doi.org/10.1016/j.phytochem.2021.112909 journal article 10.1016/j.phytochem.2021.112909 1873-3700 8258118 2.4. Bioactive potential of the conoidecyclics isolated from T. conoides Conoidecyclic A exhibited dual attenuation property against inducible inflammatory enzymes COX-2 and 5-LOX (IC 50 1.75 and 4.24 mM, respectively). The anti-inflammatory activities of conoidecyclic A were higher than those displayed by conoidecyclic B (IC 50 > 1.9 mM) and conoidecyclic C (IC 50 5-LOX 5.07 mM) ( Table 2 ). The anti-inflammatory selectivity index (SI) was greater for conoidecyclic A (1.79) than those displayed by conoidecyclic B and C (1.65–1.68) as well as synthetic anti-inflammatory agents (ibuprofen, 0.44 and sodium salicylate, 0.73) ( Table 2 ). The lesser selectivity ratio of synthetic anti-inflammatory agents specified the selective inhibition towards COX-1, leading to several side effects ( Laneuville et al., 1994 ). Therefore, it could possibly be concluded that conoidecyclic A, with higher SI and greater specificity towards COX-2 was noteworthy towards the development of selective anti-inflammatory therapeutic lead ( Spangler, 1996 ). Conoidecyclic A exhibited significantly greater attenuation properties against ACE-I and PTP-1B (IC 50 1.23 and 1.39 mM, respectively) as compared to other studied conoidecyclics (IC 50 > 1.80 mM) ( Table 2 ). The radical scavenging activities (IC 50DPPH 1.20 and IC 50ABTS 1.48 mM) exhibited by conoidecyclic A were greater compared to those displayed by conoidecyclic B (IC 50 1.35–1.54 mM), conoidecyclic C (IC 50 1.54–1.81 mM) and commercially available standards (IC 50 1.46–1.69 mM). Table 2 Bioactivities of conoidecyclics A-C isolated from the organic extract of T. conoides and their molecular descriptors.

Bioactivities a (IC; mM) 50	Conoidecyclic A	Conoidecyclic B	Conoidecyclic C	Standard	Standard
Antioxidant activities a
DPPH. scavenging activity	1.20d ± 0.05	1.35c±0.02	1.54a±0.02	1.46b ± 0.04 P	1.18d ± 0.03Q
ABTS+. scavenging	1.48d ± 0.01	1.54c±0.02	1.81a±0.04	1.69b ± 0.05P	1.28e±0.02Q
Anti-inflammatory activities a
COX-1 inhibition	3.13 c ±0.04	3.19 c ±0.04	3.35b ± 0.06	0.19d ± 0.01Ib	12.06 a ±0.05S
COX-2 inhibition	1.75 c±0.04	1.93b ± 0.02	1.99b ± 0.03	0.43d ± 0.02Ib	16.56 a±0.05S
Selectivity index b	1.79 a±0.03	1.65b ± 0.04	1.68b ± 0.02	0.44d±0.03Ib	0.73 c±0.02S
5-LOX inhibition	4.24 e±0.06	4.88 c±0.04	5.07b ± 0.09	4.51d ± 0.11Ib	10.93 a±0.12S
Anti-hypertensive activity a
ACE-I inhibition	1.23c±0.02	1.89b ± 0.04	2.23a±0.02	0.53d ± 0.02CP	–
Anti-diabetic activity a
PTP-1B inhibition	1.39 c ±0.03	2.33b ± 0.02	3.13 a ±0.04	1.13d ± 0.06SV	–
Molecular descriptors c	Conoidecyclic A	Conoidecyclic B	Conoidecyclic C	α -tocopherol (BHT)	Ibuprofen (Captopril)
Electronic
tPSA	96.22	122.52	122.52	29.46 (20.23)	37.30 (57.61)
PI (× 10 24 cm3)	51	61	55	53.54 (27.64)	23.96 (21.00)
Steric
MR (cm3/mol)	129.78	154.66	140.80	135.6 (69.73)	60.44 (54.85)
MV(cm3)	447.4	534.9	522.0	462.70 (237.50)	200.10 (170.7)
P (cm3)	1092.7	1316.5	1198.4	1123.00 (556.00)	499.30 (463.4)
Hydrophobic
Log POW	3.13	3.83	2.61	9.98 (5.74)	3.75 (0.24)

Different superscripts (P–V) indicate the standards used for different activities; P- α -Tocopherol; Q-Butylated hydroxy toluene; Ib-Ibuprofen; S-Sodium salicylate; CP- Captopril; SV- Sodium metavandate. a Bioactivities were expressed as IC values (mM). Samples were analyzed in triplicate (n = 3) and expressed as mean ± standard deviation. Means followed by the 50 different superscripts (a-e) within the same row indicated significant differences ( P < 0.05). b Selectivity index has been calculated as the ratio of anti COX-1(IC) to that of anti COX-2(IC). 50 50 c Structure-activity relationship analysis was carried out by using different molecular descriptors of the purified compounds as described in the text. tPSA, topological polar surface area; MV, molar volume; P, parachor; MR, molar refractivity; Log POW, logarithmic scale of the octanol-water partition coefficient; PI, polarizability. 2.5. Structure-activity correlation study analysis of conoidecyclic analogues isolated from T. conoides The steric factors of the studied compounds might play pivotal roles towards their potential bioactivities. Notably, the electronic properties of conoidecyclic B and C were higher than those of conoidecyclic A ( Table 2 ), even though the bioactivities of the latter were greater. This could be explained by the comparatively lesser steric bulkiness of conoidecyclic A (P 1092.7 cm 3 , MV 447.4 cm 3 ) than those recorded for conoidecyclic B (P 1316.5 cm 3 , MV 534.9 cm 3 ) and C (P 1198.4, MV 522.0 cm 3 ) owing to the presence of bulkier side chain in the latter. These inferences were appropriately corroborated by the efficiency of conoidecyclic A towards the conformationally favorable interaction with the active binding sites of the target enzymes. Notably, the hydrophobicity of conoidecyclic A and B (log POW 3.13–3.83) were found to reside within the permissible limit of hydrophobic-lipophilic threshold ( Lipinski, 2004 ), which could attribute to their prospective bioactive properties. An earlier report of literature inferred that the effective permeability in the cellular network (through inter membrane barrier) along with the radical scavenging activities of the pharmacophore agents might result in their potential bioactivities ( Ishige et al., 2001 ). 2.6. ADME and other physicochemical parametrs Swiss ADME tools were used ( Daina et al., 2017 ) for the estimation of different physicochemical parameters, drug-likeness, solubilities and ADME behaviors of the isolated compounds (conoidecyclics A-C). Based upon the specific physicochemical parameters, the qualitative prediction was performed, and only conoidecyclic A passed the filter of Lipinski’ s rule without any violation, whereas conoidecyclic B and C had one violation (MW > 500) ( Table 3 ). Therefore, conoidecyclic A could possess greater oral bioavailability. Notably, all the three compounds could pass the filter of Veber rule without any violations ( Table 3 ) ( Daina et al., 2017 ). In addition to that, predicted bioavailability score for the studied compounds were comparable (0.55) with that of ibuprofen (0.55), which apparently recognized at least 10% oral bioavailability in rat and permeability towards Caco-2 cell lines ( Daina et al., 2017 ) ( Table 3 ). For the rapid estimation of drug-likeness, the bioavailability radar plot was adopted, and six physicochemical parameters (size, lipophilicity, polarity, flexibility, solubility and saturation) were taken into account (Fig. S34). The optimum range for each parameter was shown by a pink area. Evidently, conoidecyclic B and C displayed a deviation including larger size (MW > 500), even though no eccentricity was apparent for conoidecyclic A, and all the six values (comparing to ibuprofen, conoidecyclic A exhibited lesser flexibility) led to optimal physicochemical attributes leading to an acceptable oral bioavailability. Solubility is considered to be one of the vital parameters for drug development activities, and also related to absorption ( Daina et al., 2017 ). Estimated aqueous solubility (Log S; ESOL and SILICOS-IT) for three isolated compounds based on molecular structure were in a range of moderately soluble to soluble ( Table 3 ). Logarithm value of skin permeability coefficient Kp (regarding pharmacokinetics) was calculated for the isolated compounds including the standard (ibuprofen), whereas more negative value of Kp inferred lesser skin permeability. The latter could be linearly correlated with the molecular size, lipophilicity, and it was observed that conoidecyclic A exhibited a Kp value (- 6.51 cm /s) closer to that exhibited by conoidecyclic B and C (- 6.56 and - 7.25 cm /s, respectively) ( Table 3 ). 2.7. Kinetic properties of ACE-I, PTP-1B and 5- LOX inhibition Kinetic studies were performed to determine the mode of inhibition of conoidecyclics A-C, and the inhibition constants ( Ki ) were determined by Lineweaver-Burk and Dixon plots. Conoidecyclics were found to inhibit ACE-I, PTP-1B and 5- LOX enzymes, in a non-competitive fashion as determined by the Lineweaver-Burk plot (Fig. S35). Increase of substrate concentrations could result in non-intersect series of line on Y-axis in the Lineweaver-Burk plot (Fig. S35) but intersected on the negative X axis ( Ki ) in the Dixon plots ( Fig. 6 ). Conoidecyclic A exhibited lesser inhibition constant towards inhibition of ACE-I (1.1 mM), PTP-1B (1.2 mM) and 5- LOX (4.0 mM) than those displayed by other studied metabolites ( Fig. 6 ). An inverse relation of Vmax with various concentrations of conoidecyclics A-C inferred the non-competitive inhibition of the target enzymes ( Blat, 2010 ). Among the studied metabolites, conoidecyclic A exhibited lesser apparent Vmax (0.31–0.14, 0.29–0.17 and 0.32–0.17 ΔA min 1 for ACE-I, PTP-1B and 5- LOX inhibition, respectively) ( Table S2 ) than other studied macrolides, which implied that the former could efficiently bind with targeted enzyme to diminish the reaction velocity. Table 3 Physicochemical and pharmacokinetic parameters of conoidecyclics A-C.

Parameters	Conoidecyclic A	Conoidecyclic B	Conoidecyclic C
Physicochemical properties
Formula	C27H44O6	C32H52O8	C29H46O8
Molecular	464.63 g/mol	564.75 g/mol	522.67 g/mol
weight
Num. heavy	33	40	37
atoms
Fraction Csp 3	0.74	0.75	0.72
Num. rotatable	1	6	4
bonds
Num. H-bond	6	8	8
acceptors
Num. H-bond	3	3	3
donors
Molar	132.11	157.43	143.01
refractivity
TPSA a	96.22 Å2	122.52 Å2	122.52 Å2
Lipophilicity
Log Pb o/w	3.95	4.31	4.38
(iLOGP)
Log Po/w	3.70	4.49	3.15
(XLOGP3)
Log Po/w	4.52	5.09	4.06
(WLOGP)
Consensus Log	4.05	4.63	3.86
Po/w
Water solubility
Log S c (ESOL)	4.99,	5.77, moderately	4.80, moderately
moderately	soluble	soluble
soluble
Log S (SILICOS-	3.01, soluble	3.82, soluble	3.04, soluble
IT)
Drug-likeness
Lipinski d	Yes; 0 violation	Yes; 1 violation:	Yes; 1 violation:
MW > 500	MW > 500
Ghose e	No; 2 violations:	No; 3 violations:	No; 3 violations:
MR > 130,	MW > 480, MR >	MW > 480, MR >
#atoms > 70	130, #atoms > 70	130, #atoms > 70
Veber f	Yes	Yes	Yes
Bioavailability	0.55	0.55	0.55
score
Pharmacokinetics
Log Kg (skin p	6.51 cm/s	6.56 cm/s	7.25 cm/s
permeation)

a TPSA = Topological polar surface area. b Log P= The partition coefficient between n-octanol and water. o/w c Log S = The decimal logarithm of the molar solubility in water. d Lipinski criteria range are, lipophilicity (Log P) ≤ 5, MW ≤ 500, H-bond o/w donors ≤ 5 and H-bond acceptors ≤ 10. e Ghose filter criteria range are, Log Pin 0.4 to +5.6 range, MR from 40 to o/w 130, MW from 180 to 480, No. of atoms from 20 to 70. f Veber rule criteria range are, RB ≤ 10 and TPSA ≤140 Å 2 . g Log K= skin permeability coefficient. p 2.8. In silico molecular modeling analysis of conoidecyclics isolated from T. conoides The macrocyclic derivatives (conoidecyclics A-C) were subjected to in-silico molecular modeling studies against pro-inflammatory 5- LOX and COX-2 enzymes, and the results were obtained with the help of RMSD data. Conoidecyclic A, on molecular modeling with COX-2 and ACE-I exhibited three hydrogen bonding interactions with the enzyme active site, whereas five hydrogen bonds were apparent between the ligand and the active site amino acyl residues of PTP-1B ( Fig. 3 , Table 4 ). In comparison, conoidecyclic B and C exhibited lesser number of hydrogen bonding interactions with the active site of targeted enzymes ( Figs. 4–5 , Table 4 ). Likewise, conoidecyclic A recorded least binding energy (13.34, 14.51, 13.87 and 11.27 kcal mol 1 with 5- LOX , COX-2, PTP-1B and ACE-I respectively) and docking score (~ 12 to 15 kcal mol 1 ) than those displayed by conoidecyclics B and C ( Table 4 ). Likewise, the constant of enzyme inhibition, Ki upon interaction with COX-2 and 5- LOX were lesser for conoidecyclic A (23.20 and 33.23 pM, respectively) followed by those of conoidecyclic B and C ( Table 4 ). The lowest docking score as well as binding energy of conoidecyclic A described its greater attenuation potential against 5- LOX and COX-2 enzymes, which were reported to produce inflammatory prostaglandins (PGE 2 , PGG 2 , PGI 2 , PGF 2 etc), thromboxane (TXA 2 ) and leukotrienes (such as LTB 4 ) causing the development of inflammation ( Hanna and Hafez, 2018 ).