G Model
CCLET-2932; No. of Pages 4
X.-L. Yue et al. / Chinese Chemical Letters xxx (2014) xxx–xxx
3
The IR and 1H NMR spectral data of the other target compounds
and all the 1H NMR spectra can be found in the Supporting
information.
3.2. Herbicidal activity
All synthesized compounds 3–47 and bensulfuron, a commer-
cially available herbicide, were evaluated for their herbicidal
activity against AR and SV. Bensulfuron was employed as a positive
control. Herbicidal activity was evaluated by measuring the
lengths of roots and hypocotyls. The biological testing results
showed that all of compounds showed better herbicidal activity
against the roots of AR and SV. The results were expressed as
concentrations of IC50 and listed in Table 1.
As shown in Table 1, compounds 10, 25, 40, and 41 exhibit high
activities with IC50 below 50 mg/L against SV, and compound 25
(11.67 mg/L) showed slightly better activity than bensulfuron
(27.45 mg/L). For AR, compounds 20, 37, 39, 40, 41, 44 and
46 showed slightly weaker activity, and the IC50 are less than
100 mg/L. In terms of the structure–activity relationship, the
different position of the substituent on the benzene ring displayed
different activities. For AR, compounds where the substituent-F lies
in the para-position display the best herbicidal activity, such as 14
(4-F) > 13 (3-F), 12 (2-F) and 32 (4-F) > 31 (3-F), 30 (2-F).
However, the substituent-F in the meta-position led to the highest
activity against SV, for example, 4 (3-F) > 3 (2-F), 5 (4-F) and 40
(3-F) > 39 (2-F), 41 (4-F). The substituent –CF3 on the meta-
position and -2,4-di-F led to higher activity in both AR and SV, for
instance, 16 (3-CF3) > 15 (2-CF3), 17 (4-CF3) and 46 (2,4-di-F) > 45
(2,6-di-F), 47 (2,5-di-F). In addition, the compounds with optimal
activity are 41 (35 mg/L) for AR and 25 (11 mg/L) for SV.
2.2. Herbicidal activity testing
Herbicidal activity testing of compounds 3–47 against AR and
SV were evaluated according to the standard protocol [18]. All
compounds were formulated as 10000 mg/L emulsified concen-
trates by using dimethyl sulfoxide (DMSO) as the solvent and TW-
80 as an emulsification reagent. Then, they were diluted with
distilled water to the desired concentration (10, 25, 50, 100, and
200 mg/L). Twenty seedlings were each grown in a 9 cm Petri dish
containing two pieces of filter paper and 10 mL solution at
25 ꢁ 1 8C, relative humidity (RH) (60 ꢁ 5) % in a greenhouse. Distilled
water and bensulfuron, a commercially available herbicide, were
used as the controls. For the entire bioassay test, each treatment was
repeated twice. Herbicidal activity was evaluated by measuring the
lengths of the roots and hypocotyls. Then, Predictive Analytics
Software (PASW) 18.0 was used to perform the regression analysis in
order to obtain the IC50 values.
2.3. 3D-QSAR analysis
The 45 target compounds were divided into a training set and a
testing set which included 38 and 7 compounds, respectively. The
testing set compounds were randomly chosen. The IC50 values
have been changed into the minus logarithmic scale [pIC50] for the
QSAR study [19,20]. The three-dimensional structures of all
compounds were built using the SYBYL 7.3 software [21]. Partial
atomic charges were calculated by the Gasteiger–Hu¨ckel method,
and energy minimizations were performed using the Tripos force
3.3. Quantitative structure–activity relationship analysis
Compared to AR, the biological testing results showed that most
of the synthesized compounds showed higher herbicidal activity
against SV. Subsequently, we performed QSAR studies on the
biological testing data for SV. The accuracy of the prediction of the
QSAR model and reliability of the contour maps are directly
dependent on the structural alignment rule. So, the molecular
alignment is considered one of the most sensitive parameters in
QSAR analysis. In this study, compound 25 with the best activity
was selected as a reference molecule, and the common fragment
(shown in Fig. 1A) was used as a template for all compound
alignments, as shown in Fig. 1B.
The statistical results of the 3D-QSAR model showed that the
cross-validated correlation coefficients (q2) and the regression
coefficients (r2) for SV are 0.869 and 0.989, respectively. The
obtained results indicated that the CoMFA model has good
prediction capability. The predicted pIC50 values are generally in
good agreement with the experiment data, and the residuals are all
small as shown in Fig. S46 and Table S1 in Supporting information.
The contribution of the steric field and electrostatic field are
34.8% and 65.2% in the 3D-QSAR model, respectively. The steric and
electrostatic contribution contour maps of the model are displayed
in Fig. 2. The 3D contour maps showed that the changes of
molecular fields are associated with the differences of the
biological activity. The steric fields are in green and yellow. The
region of green contour suggests that more bulky substituents in
these positions will improve the biological activity, while the
field and the Powell conjugate gradient algorithm with
a
˚
convergence criterion of 0.05 kcal/(mol A) [22,23].
The steric and electrostatic interaction fields for CoMFA were
˚
calculated using the SYBYL default parameters: 2.0 A grid points
spacing, a sp3 carbon probe atom with +1 charge and a van der
˚
Waals radius of 1.52 A, and column filtering of 2.0 kcal/mol [24].
The descriptors calculated from the CoMFA analysis were used as
independent variables, and the experimental pIC50 values were
used as dependent variables in partial least squares (PLS) analysis
to derive 3D-QSAR model. Leave-one-out (LOO) cross-validated
and the SAMPLS program [25] were performed to obtain the
optimal number of components (n) and cross-validated coefficient
(q2). After the optimal number of components was determined, a
non-cross-validated analysis was performed without column
filtering to obtain regression coefficients (r2) which determine
the external predictive ability [26–28].
3. Results and discussion
3.1. Synthesis of N-fluorinated phenyl-N0-pyrimidyl urea derivatives
3–47
The synthetic route for N-fluorinated phenyl-N0-pyrimidyl urea
derivatives is depicted in Scheme 1. Fluorinated phenyl isocya-
nates were prepared by reacting BTC with respective amines in
toluene at 0 8C for 1 h and then stirred at 80 8C for 2 h [16,17]. The
intermediate 1, without further isolation, reacted with the
corresponding amine to produce the target compounds 3–47 with
yields of 46–79%. All the compounds were purified by crystalliza-
tion, and all the compound structures were assigned by IR and
1H NMR spectral data, which were reported in the experimental
section and Supporting information.
Fig. 1. Structure of the urea derivatives: (A) general structure for title compounds,
(B) 3D view of all the aligned molecules in training and testing sets.
Please cite this article in press as: X.-L. Yue, et al., N-Fluorinated phenyl-N0-pyrimidyl urea derivatives: Synthesis, biological evaluation
Yue, Xia-Li
