Synthesis and antiproliferative activity of some 3-(pyrid-2-yl)-pyrazolines †

The synthesis and antiproliferative activity of eleven 3-(pyrid-2-yl)-pyrazolines in two cancer cell lines are reported. X-ray crystallography was obtained of the lead compound 8i which was screened in the NCI 60 human tumour cell line and displayed sub-micromolar activity. Cell cycle analysis, in vitro tubulin assay and confocal microscopy are also reported and suggest that the lead compound disrupts microtubule formation.

We have previously reported the synthesis and physicochemical properties of the simple pyrazoline 8a (R 1 ¼ H, R 2 ¼ Me) and the corresponding oxidised pyrazole derivative, which are "turn on" uorescent sensors for Cd 2+ and Zn 2+ , respectively. 19 Our interests in pyrazolines derived from 2-acetylpyridine 3 originated from possible chelation of ions in these systems by the pyridine nitrogen, the basic pyrazoline nitrogen (N2) and co-ordinating groups (R 2 , Scheme 1). Compounds of this structure have been reported as ligands for gold and display antiproliferative activities. 20 In addition, the 3-pyridyl motif would afford compounds with improved pharmacological properties. N1-Substituted 3-(pyrid-2-yl)-pyrazolines display various biological activities, including antitubercular, 21 antimycobacterial, 22 antifungal, 23 anti-inammatory 24,25 and antiproliferative 20,24 activities. We therefore wished to investigate novel compounds of this class and examine their antiproliferative activity. Scheme 1 Synthesis of 3-(pyrid-2-yl)-pyrazolines 8 and 9 (see Table 1 for R 2 groups and yields).

Results and discussion
Continuing our research interests in pyrazolines derived from 2acetylpyridine 3, we synthesised various derivatives following the sequence shown in Scheme 1. Claisen-Schmidt condensation of 2-acetylpyridine 3 with an aromatic aldehyde gave the corresponding chalcone 4 (97%) or 5 (85%) in excellent yields. Treatment with an excess of hydrazine hydrate gave the N-H pyrazoline derivative 6 or 7 which could be characterised by 1 H NMR. Attempts to purify these compounds by silica gel column chromatography however led to decomposition, as was previously observed with similar substrates, 22,23,26 possibly due to autoxidation. 27 Instead, the N-H pyrazolines 6 and 7 were treated directly with the desired electrophile to afford the Nsubstituted pyrazolines 8 and 9 which could be puried, in a one-pot transformation in good yields.
The N-Me and N-Ph substituted pyrazolines 8a,b and 9b were synthesised by reaction of the chalcone 4 or 5 with the commercially available substituted hydrazine NH 2 NHMe or NH 2 NHPh, affording the corresponding products 8a (72%), 19 8b (49%) or 9b (34%) directly ((iv), Scheme 1). Low yields with arylhydrazines have been previously reported, 26 and similar transformations suggest that the reaction proceeds via initial 1,2-addition followed by cyclisation and not 1,4-conjugate addition and subsequent cyclisation, which could generate different isomers. 19,28 Compounds were examined for antiproliferative activity in vitro using the MTS assay 29 with human colon carcinoma (HT29) and highly metastatic human breast carcinoma (MDA-MB-231) cell lines. The results are reported as the concentration required to inhibit 50% cell proliferation (IC 50 ) and are shown in Table 1.
The unsubstituted chalcone 4 has been reported to display modest antiproliferative activity in various cancer cell lines. 30 Likewise, the 3,4,5-trimethoxy derivative 5 showed good activity in both the cell lines examined in this study (entry 1, Table 1). As can be seen in Table 1, the R 2 substituent had a great effect on the antiproliferative activity of the pyrazolines. In the C5-phenyl pyrazoline series 8, the N-Me derivative 8a displayed weak activity, which improved two-fold with the N-Ph analogue 8b (entries 2 and 3). The introduction of an N-acetyl group (8c), which provides additional Lewis basic sites, resulted in partial loss of activity although it did exhibit some selectivity for the HT29 cell line (entry 4). The triuoroacetyl analogue 8d, which has a similar steric requirement as the acetyl but is a much less effective Lewis base, resulted in complete loss of antiproliferative activity in both cell lines (entry 5).
The carbothioamide 1 shown in Fig. 1 displayed potent antiproliferative activity 4 and similar compounds also show promise. 5,6 The related derivative 8e was previously reported to exhibit activity as a gold complex with HeLa and A549 cancer cell lines. 20 As the free ligand, however, the carbothioamide 8e gave poor inhibition of cell growth in the HT29 and MDA-MB-231 cell lines, while the monomethyl analogue 8f exhibited much improved activities for both cell lines (entries 6 and 7, Table 1).
X-ray crystallography of the carbothioamide 8e indicated a putative intramolecular H-bond to the basic pyrazoline nitrogen (shown in Fig. 2) and an intermolecular H-bond with the pyridine nitrogen (shown in the ESI †). Likewise, the monomethyl derivative 8f could also possess intramolecular co-ordination to the remaining N-H. Although these interactions are present in the solid state, in aqueous solution competing intermolecular H-bonding to the solvent will undoubtedly occur, affording different conformations such as that observed in co-ordination with gold. 20 To investigate further we attempted to generate the dimethyl analogue 8g, as this would remove the interaction altogether. Attempts to obtain pyrazoline 8g either by methylation of the carbothioamide 8e or 8f, or by reaction of the N-H pyrazoline 6 with dimethylthiocarbamoyl chloride, were however unsuccessful as previously observed in similar substrates. 26 The effect of an N-benzoyl substituent (8h) was also investigated and exhibited no observable antiproliferative activity in the two cell lines examined (entry 8, Table 1). Following our previous studies using 3,4,5-trimethoxyphenyl substituents, 31 which have been shown to play crucial roles in activity, [32][33][34][35][36] we investigated this substitution pattern in the N-benzoyl group.
Remarkably, the corresponding 3,4,5-trimethoxybenzoyl derivative 8i displayed potent activity and selectivity for the highly metastatic human breast carcinoma (MDA-MB-231, entry 9). The X-ray crystal structure of pyrazoline 8i was also obtained and is shown in Fig. 2. As with pyrazoline 8e, it shows that in the solid state the amide adopted the E-conformation. Continuing our interest in the 3,4,5-trimethoxyphenyl motif, we investigated the effect of this substitution pattern in three examples in the pyrazoline series 9. In all three examples, the compounds exhibited increased potency for the MDA-MB-231 cell line, as previously observed with derivative 8i which also possesses the 3,4,5-trimethoxyaryl motif. The N-Ph derivative 9b was more active in both cell lines than the corresponding C5-phenyl analogue 8b (entries 10 and 3, Table 1). Surprisingly, the N-Ac derivative 9c showed no activity in HT29 (>500 mM), yet exhibited improved activity and selectivity for MDA-MB-231 (52.5 mM) compared to pyrazoline 8c (entries 11 and 4).  Incorporating the 3,4,5-trimethoxy-substitution in both the C5aryl and N1-benzoyl positions of the pyrazoline ring (9i) did not further improve activity compared to compound 8i (entries 12 and 9). On the basis of these preliminary investigations, compound 8i was selected for further evaluation by the National Cancer Institute (NCI) and the results are shown in Table 2.
Pyrazoline 8i (NSC 761258) displayed promising GI 50 values across the NCI 60 human tumour cell line panel and exhibited sub-micromolar activity in a range of cancer cell lines, including the multidrug resistant ovarian cell line NCI/ADR-RES (0.519 mM). Also of interest is that one of the most susceptible ovarian cancer cell lines, OVCAR-3, is particularly sensitive to tubulin disruptors, 37 suggesting that pyrazoline 8i may be inhibiting cell proliferation through the disruption of tubulin. Although COMPARE analysis of the NCI 60 cancer cell line data did not afford strong correlations with compounds of known biological activity, drugs which inhibit the polymerization of tubulin did feature (see ESI †).
To investigate potential interaction with tubulin dynamics, cell cycle analysis was performed with HT29 cells with pyrazoline 8i and colchicine, a known inhibitor of microtubule polymerisation. Cell cycle analysis was performed three times and the results shown in Fig. 3 are representative of the complete study (shown in the ESI †). The presence of 100 nM of colchicine led to cell cycle arrest in the G2/M phase, shown in Fig. 3B. In the presence of pyrazoline 8i at 1 mM (Fig. 3C) and 5 mM (Fig. 3D), 50% and 95% respectively of the cells also arrested in G2/M phase, suggesting that pyrazoline 8i affects microtubule polymerisation.
To examine this interaction an in vitro tubulin polymerisation assay was performed in the presence of pyrazoline 8i (20 mM) or paclitaxel (5 mM), which is known to polymerise microtubule formation. As can be seen in Fig. 4, the presence of paclitaxel led to an increased optical density (OD) compared to the control experiment, due to increased microtubule formation. The presence of pyrazoline 8i however gave a lower OD reading than the control, implying that pyrazoline 8i inhibits microtubule polymerisation and is consistent with the cell cycle analysis.
Confocal microscopy was performed on HT29 cells to visualise the effects of pyrazoline 8i in microtubule formation, using Dm1a tubulin antibodies (Fig. 5). The effects of 100 nM colchicine, which inhibits microtubule polymerisation, can be seen in Fig. 5B, which contains much less tubulin compared to the control experiment (Fig. 5A). Likewise, compared to the control there is less tubulin present in cells treated with 5 mM pyrazoline 8i (Fig. 5C), which again supports the data obtained from the cell cycle analysis and in vitro tubulin polymerisation assay that pyrazoline 8i disrupts microtubule formation. The results are consistent with previous reports of similar compounds which were also shown to interfere with microtubule assembly. 18

Conclusions
In conclusion, we report the two-step synthesis of some 3-(pyrid-2-yl)-pyrazolines and their antiproliferative activities. The lead pyrazoline 8i was selected from this preliminary study and   further investigated against the NCI 60 cancer cell line (NSC 761258) and exhibits sub-micromolar activity in a number of cell lines. The mode of action was explored by cell cycle analysis, in vitro microtubule assembly assay and confocal microscopy with a-tubulin antibody Dm1a which suggest that the novel compound 8i inhibits microtubule polymerisation. The promising antiproliferative activity of 3-(pyrid-2-yl)-pyrazoline 8i and the ease of its synthesis provide the potential for further development, which is currently under investigation and will be reported in due course.