Design, Synthesis, Biological Activities and 3D-QSAR of New N,N'-Diacylhydrazines Containing 2,4-Dichlorophenoxy Moieties

A series of new N,N'-diacylhydrazine derivatives were designed and synthesized. Their structures were verified by 1H-NMR, MS and elemental analysis. The herbicidal activities and plant growth regulating activity of these N,N'-diacylhydrazines were evaluated. The herbicidal activity results showed that most of these N,N'-diacyl-hydrazines showed excellent in vivo activities against Echinochloa crus-galli, Digitaria sanguinalis, Brassica napus, Amaranthus retroflerus. Most of them exhibited higher herbicidal activities against dicotyledonous weeds than monocotyledonous weeds. To further investigate the structure-activity relationship, comparative molecular field analysis (CoMFA) was performed on the basis of herbicidal activity data. Both the steric and electronic field distributions of CoMFA are in good agreement in this work.

In view of these facts mentioned above, and also as a part of our continuing work [26] on the synthesis of bioactive lead compounds, the title compounds were designed by introducing the 2,4-dichlorophenoxy acetic acid pharmacophore into a diacylhydrazine scaffold. Our original strategy is depicted in Figure 2. Twenty-three novel diacylhydrazine derivatives were synthesized and characterized by 1 H-NMR, MS and elemental analysis. The herbicidal activity and plant growth regulate activity of the diacylhydrazine compounds were tested in vivo.

Synthesis and Characterization
The synthetic route of the title compounds is outlined in Scheme 1. First, the title compounds were synthesized under conventional conditions. 2,4-Dichlorophenol, K 2 CO 3 , DMF and ethyl 2-chloro-acetate were placed in a round-bottomed flask and the mixture was stirred overnight at room temperature. When the reaction was complete, the mixture was poured into water, and compound 2 was collected. Next, to a solution of compound 2 in ethanol excess 80% hydrazine hydrate was added. The mixture was refluxed for 5 h, cooled to room temperature and the precipitate formed was filtered off to afford the pure product 3. For the last step 3, Et 3 N and a substituted acyl chloride were mixed in THF and refluxed for 4 h to afford the crude solid. The target compounds were recrystallised from ethanol to afford the pure products. In order to optimize the reaction condition and reaction times, microwave irradiation was employed next. In the synthesis process of compound 2, the microwave method is used KI as catalysis, in order to accelerate reaction. The key intermediates 2, 3 can be obtained in this way in excellent yield (>95%) after short reaction times (Table 1).
In the 1 H-NMR spectra of title compounds, the CH 2 proton signals sppeared at δ 4.57~4.82 ppm. The NH proton was observed as two single peaks, although sometimes, it appeared as one broad peak. All the other alkyl or aryl groups showed the normal peak locations. All the mass spectra of the title compounds show the molecular ion peaks. Scheme 1. Synthetic route to the title compounds.

Herbicidal Activities
The herbicidal activities of these compounds were determined in vivo. The results for compounds 4a-x are summarized in Table 2. As shown in Table 2, all the compounds exhibit excellent post-emergence herbicidal activity against Amaranthus retroflexus and Brassica napus, except 4r. Most of these compounds alsoshow better herbicidal activities in post-emergence treatment than in pre-emergence treatment against Amaranthus retroflexus and Brassica napus. On the other hand, these compounds exhibited low activity against Echinochloa crus-galli, Digitaria sanguinalis under post-emergence conditions, but in pre-emergence conditions, 4p, 4q, 4r, 4u, 4v, 4w exhibited good herbicidal activity against Echinochloa crus-galli. Notably, alkyl substituted compounds or pyridine substituted compounds show surprisingly higher pre-emergence herbicidal activity against the monocotyledon Echinochloa crus-galli than the corresponding aryl substituted compounds. At 1,500 g ai/ha, the compounds 4a-x exhibited the same herbicidal activities as seen in our previous work [26].
The subsequent results in Table 3 showed that the compound 4a had good herbicidal activity against Echinochloa crus-galli, Digitaria sanguinalis, Brassica napus, Amaranthusretroflerus at a dosage of 750 g ai/ha. The herbicidal activities against dicotyledon weeds are comparable with the commercial herbicide 2,4-D. Surprisingly, cmpound 4a exhibited good activities against monocotyledon weeds, so a further bioassay was performed. At dosage of 350 g ai/ha, it can be found that the title compound has low effective inhibition against all tested monocotyledon weeds. The compound 4a displayed higher activities than that seen in the previous work for (N'-(2-(2,4-dichlorophenoxy)propanoyl)cyclopropanecarbohydrazide) [26] at the dosage of 350 g ai/ha. It can be found that the title compound 4a has the most effective inhibitory activity against all tested dicotyledon weeds, even at a dosage as low as 187.5 g ai/ha. Under pre-emergence conditions, compound 4a exhibited better activity against Amaranthus retroflerus and Brassica napus than that of the control 2,4-D.

Plant Growth Regulatory Activity
In order to elucidate why these compounds had excellent herbicidal activity, the activity of title compounds on cotyledon root of cucumber was determined. Surprisingly, as listed in Table 2, all the tested compounds showed 100% inhibition on the root cotyledon of cucumber. Such strong inhibition of plant root growth can possibly the reason behind the aforementioned excellent herbicidal activity.

CoMFA Analysis
The CoMFA method is widely used in drug design, because it allows for rapid prediction of QSAR of newly designed molecules [27]. The results of these computations are summarized in Table 4. The CoMFA contour models are very similar, suggesting that for this training set, using four components is acceptable. Experimental and activities predicted by CoMFA for all compounds are listed in Table 4. As shown, a predictive CoMFA model was established with the conventional correlation coefficient r 2 = 0.886 and the cross-validation coefficient q 2 = 0.57. The plots of the predicted versus the actual activity values for all the compounds are shown in Figure 3. It is shown in Figure 4 that the contributions of steric and electrostatic fields ("stdev × coeff") are 66.9% and 33.1% respectively.  With the CoMFA analysis, we obtained the isocontour diagrams of the steric and electrostatic field contributions ("stdev × coeff"), which is displayed in Figure 4. In Figure 4a, the steric field contours are represented with different colors: the green color at 3-position means a bulky group here would be favorable for higher herbicidal activity, while the yellow color means oppositely. As shown in Figure 4a, there is a green region located around the 3-position of the benzene ring, indicating that the bulky groups at this position will increase the herbicidal activity. This is in agreement with the actual experimental data: for example, compounds 4f and 4g have higher herbicidal activity with a bulky group in this position. In the same Figure 4b, the electrostatic contours are displayed in distinguishable colors: blue means an increase in the positive charge will lead to an increase in the activity, while the red contour defines in the opposite. So, the target compounds bearing an electron-withdrawing group at the 4-position of the benzene ring and an electron-donating group at the other positions displayed higher activity. These results provided useful information for further optimization of the compounds.

Instruments
Melting points were determined using an X-4 apparatus (Beijing Tech Instruments Co., Beijing, China) and are uncorrected. 1 H-NMR spectra were measured on a Bruker AC-P500 instrument (300 MHz, Bruker, Fallanden, Switzerland) using TMS as an internal standard and DMSO-d 6 as solvent. Mass spectra were recorded on a Thermo Finnigan LCQ Advantage LC/mass detector instrument (Thermo Finnigan, MA, USA). Elemental analyses were performed on a Vario EL elemental analyzer. An LWMC-250 domestic microwave oven (Jiang-Ling Company, Nanjing, China) was used to do microwave reactions. All reagents are commercially available analytical grade or were synthesized by us.

General Procedure
2,4-Dichlorophenol (5 mmol), KI (1 mmol), DMF (1 mL), ethyl 2-chloroacetate (5 mmol) and TBAB (0.5 mmol) were place in a dried round-bottomed flask and the mixture was irradiated by microwaves (200 W) for 5 min. On completion of the reaction, the mixture was cooled to room temperature and then added to ethanol (10 mL) with constant stirring. After filtering off the inorganic salts, the reaction mixture was added to 85% hydrazine hydrate (5 mmol) and subjected to microwave irradiation (500 W) for an additional 2 min. Then, it was cooled to room temperature, allowed to settle for 1 h, and the precipitates were filtered off and recrystallized from ethanol to afford the pure product 3. Then 3 (1 mmol) and substituted acyl chloride (1 mmol) were mixed in THF. The mixture was put into the microwave oven (400 W) and irradiated for 10 min to produce the crude solid, which on recrystallization with ethanol gave the pure product as shown in Scheme 1.

3D-QSAR Analysis
Molecular modeling was performed using SYBYL 6.91 software [27] (Tripos, Inc., St. Louis, MO, USA), and the CoMFA method was done according to our previous work [28]. The herbicidal activities of 24 compounds against Digitaria sanguinalis data (% I) at pre-emergence condition used to derive the CoMFA analyses model are listed in Table 5. The activity was expressed in terms of pIC 50 by the formula pIC 50 = log(I/((100 − I) × Mw)), where I is the percent inhibition and Mw is the molecular weight of the tested compounds. The compound 4f was used as a template to build the other molecular structures. Because these compounds share a common skeleton, 16 atoms marked with an asterisk were used for rms-fitting onto the corresponding atoms of the template structure ( Figures 5 and 6).    Each structure was fully geometry-optimized using a conjugate gradient procedure based on the TRIPOS force field and Gasteiger and Hückel charges. Because these compounds share a common skeleton, 10 atoms marked with an asterisk were used for rms-fitting onto the corresponding atoms of the template structure. CoMFA steric and electrostatic interaction fields were calculated at each lattice intersection on a regularly spaced grid of 2.0 Å. The grid pattern was generated automatically by the SYBYL/CoMFA routine, and an sp 3 carbon atom with a van der Waals radius of 1.52 Å and a +1.0 charge was used as the probe to calculate the steric (Lennard-Jones 6-12 potential) field energies and electrostatic (Coulombic potential) fields with a distance-dependent dielectric at each lattice point. Values of the steric and electrostatic fields were truncated at 30.0 kcal/mol. The CoMFA steric and electrostatic fields generated were scaled by the CoMFA-STD method in SYBYL. The electrostatic fields were ignored at the lattice points with maximal steric interactions. A partial least-squares (PLS) approach was used to derive the 3D-QSAR, in which the CoMFA descriptors were used as independent variables, and ED values were used as dependent variables. The cross-validation with the leave-one-out (LOO) option and the SAMPLS program, rather than column filtering, was carried out to obtain the optimal number of components to be used in the final analysis. After the optimal number of components was determined, a non-cross-validated analysis was performed without column filtering. The modeling capability (goodness of fit) was judged by the correlation coefficient squared, r 2 , and the prediction capability (goodness of prediction) was indicated by the cross-validated r 2 (q 2 ).

Herbicidal Activities Assay
Plant Material. Two dicotyledonous weeds, rape (Brassica napus) and amaranth pigweed (Amaranthus retroflexus), and two monocotyledonous crops, barnyardgrass (Echinochloa crus-galli) and hairy crabgrass (Digitaria sanguinalis), were used to test the herbicidal activities of the title compounds. The seeds of amaranth pigweed were reproduced outdoors and stored at room temperature. 2,4-dichlorophenoxyl acetic acid (2,4-D) was obtained as a commercial material.
Culture Method. The seeds were planted in 6 cm diameter plastic boxes containing artificial mixed soil. Before plant emergence, the boxes were covered with plastic film to retain moisture. Plants were grown in the green house. The fresh weight of the above ground tissues was measured 17 days after treatment. The inhibition percent was used to describe the control efficiency of the compounds.
Treatment. The dosage (activity ingredient) for each compound is corresponded to 1500 g/ha. The purified compounds were dissolved in 100 μL of N,N-dimethylformamide with 0.1 μL of Tween 20, and then prepared for spraying with a laboratory belt sprayer delivering a 750 L/ha spray volume. Compounds and 2,4-D were sprayed immediately after seed planting (preemergence treatment) or after the expansion of the first true leaf (postemergence treatment). The mixture of same amount of water, N,N-dimethylformamide, and Tween 20 was sprayed as the blank. Each treatment was triplicated. The activity numbers represented the percent displaying herbicidal damage as compared to the blank. The error of the experiments was 2%.

Plant Growth Regulatory Activity Assay
After dipping into distilled water for 1 h at 23 °C, the cucumber seeds (Jinke, no. 4, commercially available) were then sown into the soil with 0.7% agar on a covered porcelain enamel plate and incubated at 26 °C in a darkroom for 3 days. The same sized cotyledons were carefully selected for the subsequent biological assay. The analyte (3 mg) was resolved in N,N-dimethyl formamide (3 mL) and this solution was then diluted to 10% concentration with distilled water. A sample solution (0.3 mL) was sprayed over a 6 cm diametered filter paper and solvent was volatilized to dryness on air. The filter paper thus prepared was placed into a 6 cm diameter incubation vessel and soaked with 10 cm distilled water. Finally, 10 pieces of cotyledon of the same size were added into the incubation vessel. These cotyledons were incubated at 26 °C in a darkroom for 5 days. Then the rhizogenesis numbers of every 10 pieces of hypocotyls were measured. Each sample was repeated twice. In contrast, the distilled water was used as a blank experiment. The relative ratios of cucumber cotyledon rhizogenesis were calculated according to the following formula: Relative ratio % = (N S − N C )/N C × 100% where N S and N C are the numbers of cucumber cotyledon rhizogenesis of tested compound and control experiment, respectively.

Conclusions
In summary, a series of diacylhydrazine derivatives containing 2,4-dichlorophenoxy moieties were synthesized in good yields. The preliminary bioassays showed that some of the compounds had good herbicidal activity. Structure-activity relationship and comparative molecular field analysis (CoMFA) studies was done. The present findings provided a powerful complement to the SARs of herbicides, and warrant future investigation of the mechanism of action of these new analogues.