2D, 3D and Hybrid QSAR Studies of Nostoclide Analogues as Inhibitors of the Photosystem II

Synthetic herbicides are widely used for weed control in crops. Continuous application of chemicals induces the adaptation of weeds, leading to the development of resistance. Therefore, research for novel synthetic herbicides plays an important role in crop protection and food production. Within this context, it is important to understand the relationship between the structures of a set of compounds endowed with herbicide activity and their biological response. In this work, a quantitative structure-activity relationship (QSAR) study of 34 nostoclide analogues was carried out in order to analyze their ability to inhibit the photosystem-II. Compounds were optimized using the DFT/B3LYP/Def2-TZVPP method. 2D and 3D (Lennard-Jones and electrostatic potential energies) molecular descriptors were calculated at the same theory level. QSAR models showed major importance of electrostatic ligand-acceptor interactions and indicated the presence of aromatic interaction in the benzyl group. A photoelectron transfer via hydrogen-bond was proposed to occur between His215 residue and the carbonyl group.

Keywords:
QSAR; DFT; herbicides; photosystem II; nostoclides

Authorship

Pedro O. M. de Carvalho

Instituto de Química, Universidade Estadual de Campinas, 13083-970 Campinas-SP, BrazilUniversidade Estadual de CampinasBrazilCampinas, SP, BrazilInstituto de Química, Universidade Estadual de Campinas, 13083-970 Campinas-SP, Brazil

Márcia M. C. Ferreira ^**e-mail: marcia@iqm.unicamp.br

http://orcid.org/0000-0001-5677-0299

*e-mail: marcia@iqm.unicamp.br

SCIMAGO INSTITUTIONS RANKINGS

Figures | Tables | Formulas

Figures (16)
Tables (6)
Formulas (6)

Molecule	R₁	R₂	R₃	R₄	R₅	PI^b b percentage of photosynthetic inhibitory activity; / %
m01	–H	–H	–H	–H	–H	29.2
m02	–CH₃	–H	–H	–H	–H	26.0
m02i	–H	–H	–H	–H	–CH₃	26.0
m03	–H	–CH₃	–H	–H	–H	16.7
m03i	–H	–H	–H	–CH₃	–H	16.7
m04	–H	–H	–CH₃	–H	–H	8.0
m05	–H	–H	–CH₂CH₃	–H	–H	50.7
m06	–H	–CN	–H	–H	–H	NE^c c not effective: PI < 5%.
m06i	–H	–H	–H	–CN	–H	NE^c c not effective: PI < 5%.
m07	–H	–H	–CN	–H	–H	NE^c c not effective: PI < 5%.
m08	–CF₃	–H	–H	–H	–H	44.9
m08i	–H	–H	–H	–H	–CF₃	44.9
m09	–H	–CF₃	–H	–H	–H	49.9
m09i	–H	–H	–H	–CF₃	–H	49.9
m10	–H	–H	–CF₃	–H	–H	55.5
m11	–H	–H	–Ph	–H	–H	NE^c c not effective: PI < 5%.
m12	–H	–N(CH₃)₂	–H	–H	–H	38.6
m12i	–H	–H	–H	–N(CH₃)₂	–H	38.6
m13	–H	–H	–N(CH₃)₂	–H	–H	7.1
m14	–Cl	–H	–N(CH₃)₂	–H	–H	25.4
m14i	–H	–H	–N(CH₃)₂	–H	–Cl	25.4
m15	–H	–NO₂	–H	–H	–H	39.5
m15i	–H	–H	–H	–NO₂	–H	39.5
m16	–H	–H	–NO₂	–H	–H	57.8
m17	–OH	–H	–H	–H	–H	NE^c c not effective: PI < 5%.
m17i	–H	–H	–H	–H	–OH	NE^c c not effective: PI < 5%.
m18	–H	–OH	–H	–H	–H	6.9
m18i	–H	–H	–H	–OH	–H	6.9
m19	–H	–H	–OH	–H	–H	6.5
m20	–H	–OH	–OCH₃	–H	–H	NE^c c not effective: PI < 5%.
m20i	–H	–H	–OCH₃	–OH	–H	NE^c c not effective: PI < 5%.
m21	–H	–H	–OCH₃	–H	–H	15.0
m22	–OCH₃	–H	–H	–OCH₃	–H	43.5
m22i	–H	–OCH₃	–H	–H	–OCH₃	43.5
m23	–OCH₃	–H	–OCH₃	–H	–OCH₃	43.6
m24	–H	–OCH₃O–		–H	–H	NE^c c not effective: PI < 5%.
m24i	–H	–H	–OCH₃O–		–H	NE^c c not effective: PI < 5%.
m25	–F	–H	–H	–H	–H	49.5
m25i	–H	–H	–H	–H	–F	49.5
m26	–H	–F	–H	–H	–H	28.8
m26i	–H	–H	–H	–F	–H	28.8
m27	–H	–H	–F	–H	–H	25.6
m28	–F	–F	–F	–F	–F	6.6
m29	–Cl	–H	–H	–H	–H	30.8
m29i	–H	–H	–H	–H	–Cl	30.8
m30	–H	–Cl	–H	–H	–H	20.3
m30i	–H	–H	–H	–Cl	–H	20.3
m31	–H	–H	–Cl	–H	–H	5.8
m32	–Br	–H	–H	–H	–H	21.8
m32i	–H	–H	–H	–H	–Br	21.8
m33	–H	–Br	–H	–H	–H	22.6
m33i	–H	–H	–H	–Br	–H	22.6
m34	–H	–H	–Br	–H	–H	NE^c c not effective: PI < 5%.

Model	2D-QSAR	3D-QSAR	Hybrid QSAR
Number of descriptors	5	7	7
Latent variables	1	2	2
Explained variance / %	57.4	63.4	61.2
k_C^a a Slope of measured versus predicted regression line through the origin;	0.901	0.935	0.939
RMSEC^b b root mean square error of calibration; / %	9.11	7.55	7.33
^c c Pearson correlation coefficient of calibration;	0.717	0.813	0.824
R2 ^d d correlation coefficient of multiple determination;	0.717	0.813	0.824
RMSECV^e e root mean square error of cross-validation; / %	9.60	8.52	8.38
Q2 ^f f cross-validated correlation coefficient.	0.663	0.734	0.743

	CO+CM	wHO02	QMafYY	QMbYY	QMbZZ
wHO02	0.310
QMafYY	0.107	0.439
QMbYY	–0.679	–0.212	–0.232
QMbZZ	0.778	0.558	0.297	–0.794
PI^a a Percentage of photosynthetic inhibitory activity;	–0.683	–0.521	–0.439	0.705	–0.792
R. coef.^b b regression coefficients for descriptors in autoscaled form;	–0.238	–0.182	–0.153	0.246	–0.276
R. model^c c regression mathematical model.	PI = –498(CO + CM) – 283(wHO02) – 0.358(QMafYY) + 75.6(QMbYY) – 82.2(QMbZZ) – 335

	E1	E2	E3	E4	E5	L1	L2
E2	–0.622
E3	–0.340	0.391
E4	–0.333	0.407	0.991
E5	–0.0848	0.0630	0.165	0.190
L1	–0.293	–0.0468	0.213	0.239	0.487
L2	–0.335	0.526	0.381	0.382	0.403	–0.0209
PI^a a Percentage of photosynthetic inhibitory activity;	–0.429	0.567	0.571	0.550	–0.344	–0.400	0.407
R. coef.^b b regression coefficients for descriptors in autoscaled form;	–0.180	0.244	0.248	0.228	–0.343	–0.373	0.157
R. model^c c regression mathematical model.	PI = –0.109(E1) + 0.151(E2) + 0.142(E3) + 0.124(E4) – 0.143(E5) – 44.2(L1) + 4.75(L2) – 293

	Detail
CO+CM	sum of CHELPG atomic charges for ortho and meta carbons of the benzyl’s ring, i.e., C16, C17, C19, C20
HO+HM	sum of CHELPG atomic charges for ortho and meta hydrogens of the benzyl’s ring, i.e., H23, H24, H26, H27
wHO2	sum of squares of the wave functions’ molecular orbital coefficients for C02 carbon
QMafYY^a a Regression coefficients for descriptors in autoscaled form. CHELPG: charges from electrostatic potentials using a grid based method.	YY component of the electrostatic quadrupole moment from arylidene and furanone, i.e., atoms 1-13 and 28-34
QMbYY^a a Regression coefficients for descriptors in autoscaled form. CHELPG: charges from electrostatic potentials using a grid based method.	YY component of the electrostatic quadrupole moment from benzyl’s ring, i.e., atoms 15-20 and 23-27
QMbZZ^a a Regression coefficients for descriptors in autoscaled form. CHELPG: charges from electrostatic potentials using a grid based method.	ZZ component of the electrostatic quadrupole moment from benzyl’s ring, i.e., atoms 15-20 and 23-27