Synthesis and Antiproliferative Activity of Novel Heterocyclic Indole-Trimethoxyphenyl Conjugates

The synthesis and biological evaluation of a series of novel heterocyclic indole derivatives is described. The consolidation of the combretastatin and bisindolylmaleimide templates towards the inclusion of a novel heterocyclic ring proffered a versatile pharmacophore with which to pursue chemical diversification. Given literature precedent, maleimide was initially investigated in this role and the bioactivity assessed by measurement of NCI-60 cell panel growth. Subsequently, a range of 5-aminopyrazoles was designed and developed to explore the specific effect of heterocycle hydrogen bonding on cell growth. The unique electronic nature of the 5-aminopyrazole moiety allowed for regiospecific monosubstitution on different sites of the ring, such as thiourea substitution at the N(1) position for derivative 45 or trifluoroacetylation on the 5-amino position for 43. Further derivatisation led to the ultimate development of bicyclic pyrazolotriazinedione 41 and pyrimidine 42 systems. The antiproliferative activities of these 3,4-diaryl-5-aminopyrazoles were assessed using the NCI-60 cell screen, disclosing the discovery of distinct selectivity profiles towards a number of cell lines, such as SNB-75 CNS cancer, UO-31 and CAKI-1 renal cancer cells. A series of DNA topological assays discounted the interaction with topoisomerase II as a putative mechanism of action.


Introduction
The 3,4,5-trimethoxyphenyl moiety is a commonly-used fragment in anticancer drug discovery due to its provenance from natural products and their derivatives. One such example is Combretastatin A4 (CA-4, 1) which was first isolated from the African willow tree, Combretum caffrum, by Pettit and co-workers ( Figure 1) and is a highly potent tubulin and cell growth inhibitor. The water-soluble prodrug of CA-4, fosbretabulin is currently in clinical trials [1][2][3].
One of the primary mechanisms by which CA-4 1 or other 3,4,5-trimethoxyphenyl derivatives such as podophyllotoxin (2) can mediate anticancer activity is by binding to tubulin, thus preventing microtubule formation and arresting the cell cycle [4]. An additional chemotherapeutic effect of CA-4 1 is its ability to disrupt established vasculature within solid tumours, while simultaneously rendering normal vascular networks intact [5]. Examining the structural motif of CA-4, it is established that the 3,4,5-trimethoxyphenyl ring is an essential moiety for cytotoxicity [6]. For example, the 3,4,5-trimethoxyphenyl ring has been found in many other natural products with activities against tubulin, such as steganacin (3, Figure 2) [7]. In addition, variants of the moiety are also to be found in other anticancer agents, such as the topoisomerase II (topo II) inhibitors etoposide (4) and azatoxin (5), which incorporates an indole in design [8,9]. Significantly, at lower concentrations, azatoxin (5) is found to exhibit considerable tubulin inhibition, while at higher doses, topo II activity is primarily observed [9]. Alterations to the 3,4,5-trimethoxyphenyl functional group of CA-4 1, such as their complete removal or absence, result in loss of potency [10,11].
A number of subsequent chemical elaborations have primarily been focused on modifications to the ethene double bond and the 2-methoxyphenol subunit. For example, in many instances, the ethene double bond of CA-4 1 has been modified towards the inclusion of a heterocyclic functionality, such as a pyrazole in the case of 6 or an imidazole for 7 ( Figure 3) and, more recently, tetrazoles, cyclopropylamides and bicyclic thiazolopyrimidinones [12][13][14][15][16]. In other combretastatin analogues, both the ethene double bond and 2-methoxyphenol ring have been replaced with heterocycles, often resulting in the mediation of highly promising chemotherapeutic activity [17].  In terms of kinase activity, Peifer and co-workers described the synthesis of moguntinone 8 (Figure 4), containing maleimide and indole moieties, which was found to be an extraordinarily Examining the structural motif of CA-4, it is established that the 3,4,5-trimethoxyphenyl ring is an essential moiety for cytotoxicity [6]. For example, the 3,4,5-trimethoxyphenyl ring has been found in many other natural products with activities against tubulin, such as steganacin (3, Figure 2) [7]. In addition, variants of the moiety are also to be found in other anticancer agents, such as the topoisomerase II (topo II) inhibitors etoposide (4) and azatoxin (5), which incorporates an indole in design [8,9]. Significantly, at lower concentrations, azatoxin (5) is found to exhibit considerable tubulin inhibition, while at higher doses, topo II activity is primarily observed [9]. Alterations to the 3,4,5-trimethoxyphenyl functional group of CA-4 1, such as their complete removal or absence, result in loss of potency [10,11].
A number of subsequent chemical elaborations have primarily been focused on modifications to the ethene double bond and the 2-methoxyphenol subunit. For example, in many instances, the ethene double bond of CA-4 1 has been modified towards the inclusion of a heterocyclic functionality, such as a pyrazole in the case of 6 or an imidazole for 7 ( Figure 3) and, more recently, tetrazoles, cyclopropylamides and bicyclic thiazolopyrimidinones [12][13][14][15][16]. In other combretastatin analogues, both the ethene double bond and 2-methoxyphenol ring have been replaced with heterocycles, often resulting in the mediation of highly promising chemotherapeutic activity [17]. Examining the structural motif of CA-4, it is established that the 3,4,5-trimethoxyphenyl ring is an essential moiety for cytotoxicity [6]. For example, the 3,4,5-trimethoxyphenyl ring has been found in many other natural products with activities against tubulin, such as steganacin (3, Figure 2) [7]. In addition, variants of the moiety are also to be found in other anticancer agents, such as the topoisomerase II (topo II) inhibitors etoposide (4) and azatoxin (5), which incorporates an indole in design [8,9]. Significantly, at lower concentrations, azatoxin (5) is found to exhibit considerable tubulin inhibition, while at higher doses, topo II activity is primarily observed [9]. Alterations to the 3,4,5-trimethoxyphenyl functional group of CA-4 1, such as their complete removal or absence, result in loss of potency [10,11].
A number of subsequent chemical elaborations have primarily been focused on modifications to the ethene double bond and the 2-methoxyphenol subunit. For example, in many instances, the ethene double bond of CA-4 1 has been modified towards the inclusion of a heterocyclic functionality, such as a pyrazole in the case of 6 or an imidazole for 7 ( Figure 3) and, more recently, tetrazoles, cyclopropylamides and bicyclic thiazolopyrimidinones [12][13][14][15][16]. In other combretastatin analogues, both the ethene double bond and 2-methoxyphenol ring have been replaced with heterocycles, often resulting in the mediation of highly promising chemotherapeutic activity [17].  In terms of kinase activity, Peifer and co-workers described the synthesis of moguntinone 8 (Figure 4), containing maleimide and indole moieties, which was found to be an extraordinarily  Examining the structural motif of CA-4, it is established that the 3,4,5-trimethoxyphenyl ring is an essential moiety for cytotoxicity [6]. For example, the 3,4,5-trimethoxyphenyl ring has been found in many other natural products with activities against tubulin, such as steganacin (3, Figure 2) [7]. In addition, variants of the moiety are also to be found in other anticancer agents, such as the topoisomerase II (topo II) inhibitors etoposide (4) and azatoxin (5), which incorporates an indole in design [8,9]. Significantly, at lower concentrations, azatoxin (5) is found to exhibit considerable tubulin inhibition, while at higher doses, topo II activity is primarily observed [9]. Alterations to the 3,4,5-trimethoxyphenyl functional group of CA-4 1, such as their complete removal or absence, result in loss of potency [10,11].
A number of subsequent chemical elaborations have primarily been focused on modifications to the ethene double bond and the 2-methoxyphenol subunit. For example, in many instances, the ethene double bond of CA-4 1 has been modified towards the inclusion of a heterocyclic functionality, such as a pyrazole in the case of 6 or an imidazole for 7 ( Figure 3) and, more recently, tetrazoles, cyclopropylamides and bicyclic thiazolopyrimidinones [12][13][14][15][16]. In other combretastatin analogues, both the ethene double bond and 2-methoxyphenol ring have been replaced with heterocycles, often resulting in the mediation of highly promising chemotherapeutic activity [17].  In terms of kinase activity, Peifer and co-workers described the synthesis of moguntinone 8 (Figure 4), containing maleimide and indole moieties, which was found to be an extraordinarily In terms of kinase activity, Peifer and co-workers described the synthesis of moguntinone 8 (Figure 4), containing maleimide and indole moieties, which was found to be an extraordinarily potent kinase inhibitor of VEGF-R2 (IC 50 = 2.5 nM) and VEGF-R3 (IC 50 = 5 nM) [18]. Modification of the maleimide moiety of 8 towards regioisomeric lactam derivatives produced strikingly different biological profiles. Though pyrrol-2-one 9 was found to potently inhibit VEGF-R2 kinase (IC 50 = 31 nM), its regioisomeric analogue 10 displayed significantly weaker activity (IC 50 = 11 µM), thus illustrating the importance of the orientation of the pyrrolone within the active site for the effective exploitation of H-bonding interactions [19].
Elevated levels of angiogenic factors, such as basic fibroblast growth factor, vascular endothelial growth factor (VEGF) and matrix metalloproteinases (MMPs), have been implicated in the promotion of angiogenesis [20,21]. A correlation has previously been demonstrated between high expression of VEGF, which promotes dense formation of tumour microvessels and a poor prognosis in different types of human tumour [22,23].
Recent work on analogous derivatives by Dannhardt et al., such as maleimides 11 and 12 containing 7-azaindolyl and 6-azaindolyl functionalities, respectively, has illustrated potent inhibition of angiogenesis and cell proliferation ( Figure 4) [24,25]. In examining the molecular targets associated with the candidates 11 and 12, significant activity was observed for both against VEGFR-2 and FLT-3 protein kinases. Furthermore, it was found that although 6-azaindolyl 12 was a potent inhibitor of GSK-3β, 7-azaindolyl 11 is completely inactive against the same enzyme. In addition, this group has described selective treatment of colorectal cancer using this template [24,25]. potent kinase inhibitor of VEGF-R2 (IC50 = 2.5 nM) and VEGF-R3 (IC50 = 5 nM) [18]. Modification of the maleimide moiety of 8 towards regioisomeric lactam derivatives produced strikingly different biological profiles. Though pyrrol-2-one 9 was found to potently inhibit VEGF-R2 kinase (IC50 = 31 nM), its regioisomeric analogue 10 displayed significantly weaker activity (IC50 = 11 µM), thus illustrating the importance of the orientation of the pyrrolone within the active site for the effective exploitation of H-bonding interactions [19]. Elevated levels of angiogenic factors, such as basic fibroblast growth factor, vascular endothelial growth factor (VEGF) and matrix metalloproteinases (MMPs), have been implicated in the promotion of angiogenesis [20,21]. A correlation has previously been demonstrated between high expression of VEGF, which promotes dense formation of tumour microvessels and a poor prognosis in different types of human tumour [22,23].
Recent work on analogous derivatives by Dannhardt et al., such as maleimides 11 and 12 containing 7-azaindolyl and 6-azaindolyl functionalities, respectively, has illustrated potent inhibition of angiogenesis and cell proliferation ( Figure 4) [24,25]. In examining the molecular targets associated with the candidates 11 and 12, significant activity was observed for both against VEGFR-2 and FLT-3 protein kinases. Furthermore, it was found that although 6-azaindolyl 12 was a potent inhibitor of GSK-3β, 7-azaindolyl 11 is completely inactive against the same enzyme. In addition, this group has described selective treatment of colorectal cancer using this template [24,25]. Therefore, it is evident that structural diversification of this template can lead to potent anticancer and kinase inhibitory activity. From Structures 2-12, it is obvious that this pharmacophore incorporating an indole, trimethoxyphenyl and a linking headgroup has direct relevance to the disruption of cancer progression. Rationalising the enhancement of existing H-bonding interactions or potential exploitation of new contacts, we embarked on the synthesis of derivatives of the 3,4-diaryl-1-hydroxymaleimide scaffold 13, in order to probe the effects of oxygen insertion and indole N-substitution. Subsequently, replacement of the maleimide/pyrrolone moieties with a closely analogous 5-aminopyrazole ring (14 along with a regioisomeric series of general structure 15; Figure 5) was undertaken to validate the approach and explore new chemical space [26,27].
In this work, we report the synthesis, full characterization and initial screening of effects on cancer cell growth to investigate the anticancer activity of this novel template. We also examine the interaction of these derivatives with topoisomerase II as a potential mechanism of action via a series of DNA topological assays, given the inherent activity of azatoxin and etoposide.  Therefore, it is evident that structural diversification of this template can lead to potent anticancer and kinase inhibitory activity. From Structures 2-12, it is obvious that this pharmacophore incorporating an indole, trimethoxyphenyl and a linking headgroup has direct relevance to the disruption of cancer progression. Rationalising the enhancement of existing H-bonding interactions or potential exploitation of new contacts, we embarked on the synthesis of derivatives of the 3,4-diaryl-1-hydroxymaleimide scaffold 13, in order to probe the effects of oxygen insertion and indole N-substitution. Subsequently, replacement of the maleimide/pyrrolone moieties with a closely analogous 5-aminopyrazole ring (14 along with a regioisomeric series of general structure 15; Figure 5) was undertaken to validate the approach and explore new chemical space [26,27].
In this work, we report the synthesis, full characterization and initial screening of effects on cancer cell growth to investigate the anticancer activity of this novel template. We also examine the interaction of these derivatives with topoisomerase II as a potential mechanism of action via a series of DNA topological assays, given the inherent activity of azatoxin and etoposide. Figure 5) was undertaken to validate the approach and explore new chemical space [26,27].
In this work, we report the synthesis, full characterization and initial screening of effects on cancer cell growth to investigate the anticancer activity of this novel template. We also examine the interaction of these derivatives with topoisomerase II as a potential mechanism of action via a series of DNA topological assays, given the inherent activity of azatoxin and etoposide.

Synthesis of the 3,4-Diaryl-1-Hydroxymaleimide Scaffold 13
The synthesis of a series of 3,4-diaryl-1-hydroxymaleimides was initially undertaken in order to investigate the H-bonding framework required to affect cell growth. Synthesis begins from indole (Scheme 1, 16), which may be alkylated under standard conditions (19 and 22) to investigate the effect of N-H capping and incorporate an element of solubility. This substitution has the possibility to probe two effects: the influence of the indole N-H on binding and a malleable nitrile functional group to modify solubility. Indole potassium glyoxalate salt formation is undertaken in two steps with acylation using oxalyl chloride followed by treatment with base (18, 21 and 24). Difficulties of translation to 7-azaindole necessitated an alternate method, with aluminium trichloride followed by ethyl chlorooxoacetate and then base hydrolysis yielding the glyoxalate (Scheme 2, 27).

Synthesis of the 3,4-Diaryl-1-Hydroxymaleimide Scaffold 13
The synthesis of a series of 3,4-diaryl-1-hydroxymaleimides was initially undertaken in order to investigate the H-bonding framework required to affect cell growth. Synthesis begins from indole (Scheme 1, 16), which may be alkylated under standard conditions (19 and 22) to investigate the effect of N-H capping and incorporate an element of solubility. This substitution has the possibility to probe two effects: the influence of the indole N-H on binding and a malleable nitrile functional group to modify solubility. Indole potassium glyoxalate salt formation is undertaken in two steps with acylation using oxalyl chloride followed by treatment with base (18, 21 and 24). Difficulties of translation to 7-azaindole necessitated an alternate method, with aluminium trichloride followed by ethyl chlorooxoacetate and then base hydrolysis yielding the glyoxalate (Scheme 2, 27). The synthesis of the maleic anhydrides 28-31 was facilitated by a modified Perkin condensation with trimethoxyphenyl acetic acid and acetic anhydride in low to moderate yield. Conversion of this anhydride to the 1-hydroxymaleimides 32-35 proceeded smoothly and in excellent yield (Scheme 3).

Synthesis of the 3,4-Diaryl-1-Hydroxymaleimide Scaffold 13
The synthesis of a series of 3,4-diaryl-1-hydroxymaleimides was initially undertaken in order to investigate the H-bonding framework required to affect cell growth. Synthesis begins from indole (Scheme 1, 16), which may be alkylated under standard conditions (19 and 22) to investigate the effect of N-H capping and incorporate an element of solubility. This substitution has the possibility to probe two effects: the influence of the indole N-H on binding and a malleable nitrile functional group to modify solubility. Indole potassium glyoxalate salt formation is undertaken in two steps with acylation using oxalyl chloride followed by treatment with base (18, 21 and 24). Difficulties of translation to 7-azaindole necessitated an alternate method, with aluminium trichloride followed by ethyl chlorooxoacetate and then base hydrolysis yielding the glyoxalate (Scheme 2, 27). The synthesis of the maleic anhydrides 28-31 was facilitated by a modified Perkin condensation with trimethoxyphenyl acetic acid and acetic anhydride in low to moderate yield. Conversion of this anhydride to the 1-hydroxymaleimides 32-35 proceeded smoothly and in excellent yield (Scheme 3).  The synthesis of the maleic anhydrides 28-31 was facilitated by a modified Perkin condensation with trimethoxyphenyl acetic acid and acetic anhydride in low to moderate yield. Conversion of this anhydride to the 1-hydroxymaleimides 32-35 proceeded smoothly and in excellent yield (Scheme 3). The synthesis of the maleic anhydrides 28-31 was facilitated by a modified Perkin condensation with trimethoxyphenyl acetic acid and acetic anhydride in low to moderate yield. Conversion of this anhydride to the 1-hydroxymaleimides 32-35 proceeded smoothly and in excellent yield (Scheme 3).

Antiproliferative Activity of 3,4-Diaryl-1-Hydroxymaleimide Derivatives
Assessment of antiproliferative activity was undertaken using the NCI-60 cell screen, where compounds are tested at a single dose (10 µM) against the panel of 60 cancer cells (see Section 2.6 for details). It is evident from the mean growth percentages ( Table 1) that there is little inhibition of growth, and on inspection of the individual cell lines, UO-31 appears to be the most susceptible with some limited effect on HOP-92. Substitution of the indole nitrogen or replacement of the indole with 7-azaindole disappointingly led to little difference in activity. Overall, there is minimal effect on cell growth, which led to a revision of our strategy to incorporate greater functionality in the headgroup as seen in Templates 14 and 15.

Antiproliferative Activity of 3,4-Diaryl-1-Hydroxymaleimide Derivatives
Assessment of antiproliferative activity was undertaken using the NCI-60 cell screen, where compounds are tested at a single dose (10 µM) against the panel of 60 cancer cells (see Section 2.6 for details). It is evident from the mean growth percentages ( Table 1) that there is little inhibition of growth, and on inspection of the individual cell lines, UO-31 appears to be the most susceptible with some limited effect on HOP-92. Substitution of the indole nitrogen or replacement of the indole with 7-azaindole disappointingly led to little difference in activity. Overall, there is minimal effect on cell growth, which led to a revision of our strategy to incorporate greater functionality in the headgroup as seen in Templates 14 and 15.

Synthetic Routes towards 3,4-Diaryl-5-Aminopyrazoles of General Structure 15
Synthesis of novel 3,4-diaryl-5-aminopyrazoles 15 to probe the influence of headgroup on cytotoxicity proved relatively straightforward. Initially, β-ketonitrile 39 was fashioned via reaction of acetonitrile 37 with 7-azaindolyl-3-acyl chloride (38, Scheme 4). Reaction of β-ketonitrile 39 with hydrazine hydrate, employing hydrochloric acid as catalyst, in absolute ethanol at reflux temperature led to aminopyrazole 40 (Scheme 5). The versatile nature of 5-aminopyrazole 40 easily lends itself towards further chemical elaboration of the headgroup space. Our initial plan set out to probe the chemical space by use of hydrophilic and lipophilic components and monosubstitution in tandem with ring formation. Following treatment with bidentate electrophiles N-chlorocarbonyl isocyanate and hexafluoroacetylacetone, bicyclic systems 41 and 42 were formed, respectively (Scheme 6).

Antiproliferative Activity of 3,4-Diaryl-1-Hydroxymaleimide Derivatives
Assessment of antiproliferative activity was undertaken using the NCI-60 cell screen, where compounds are tested at a single dose (10 µM) against the panel of 60 cancer cells (see Section 2.6 for details). It is evident from the mean growth percentages ( Table 1) that there is little inhibition of growth, and on inspection of the individual cell lines, UO-31 appears to be the most susceptible with some limited effect on HOP-92. Substitution of the indole nitrogen or replacement of the indole with 7-azaindole disappointingly led to little difference in activity. Overall, there is minimal effect on cell growth, which led to a revision of our strategy to incorporate greater functionality in the headgroup as seen in Templates 14 and 15.

Synthetic Routes towards 3,4-Diaryl-5-Aminopyrazoles of General Structure 15
Synthesis of novel 3,4-diaryl-5-aminopyrazoles 15 to probe the influence of headgroup on cytotoxicity proved relatively straightforward. Initially, β-ketonitrile 39 was fashioned via reaction of acetonitrile 37 with 7-azaindolyl-3-acyl chloride (38, Scheme 4). Reaction of β-ketonitrile 39 with hydrazine hydrate, employing hydrochloric acid as catalyst, in absolute ethanol at reflux temperature led to aminopyrazole 40 (Scheme 5). The versatile nature of 5-aminopyrazole 40 easily lends itself towards further chemical elaboration of the headgroup space. Our initial plan set out to probe the chemical space by use of hydrophilic and lipophilic components and monosubstitution in tandem with ring formation. Following treatment with bidentate electrophiles N-chlorocarbonyl isocyanate and hexafluoroacetylacetone, bicyclic systems 41 and 42 were formed, respectively (Scheme 6). The versatile nature of 5-aminopyrazole 40 easily lends itself towards further chemical elaboration of the headgroup space. Our initial plan set out to probe the chemical space by use of hydrophilic and lipophilic components and monosubstitution in tandem with ring formation. Following treatment with bidentate electrophiles N-chlorocarbonyl isocyanate and hexafluoroacetylacetone, bicyclic systems 41 and 42 were formed, respectively (Scheme 6).
Structural confirmation of the barbiturate moiety of triazinedione 41 is afforded by the two broad singlets arising from the NH protons at 11.59 and 11.74 ppm in the 1 H NMR spectrum. A second cyclic headgroup was furnished by condensation of aminopyrazole 40 with hexafluoroacetylacetone in an acidic medium gave highly conjugated pyrazolo[1,5-a]pyrimidine 42.
An intriguing pattern of regioselectivity was observed upon reaction of 5-aminopyrazole 40 with a number of monodentate electrophiles. Reaction with trifluoroacetic anhydride afforded trifluoroacetamide 43 (Scheme 6). The presence of two broad singlets at 11.37 and 13.30 ppm in the 1 H NMR spectrum confirm substitution on the exocyclic amino functionality rather than at the N(1) nitrogen.

X-Ray Crystal Structure Analysis of Substituted 3,4-Diaryl-5-Aminopyrazole Derivatives
Precedence for the difference in regioselectivity observed for the substitution of aminopyrazoles such as 40 exists in the literature [28]. For example, in the development of a series of novel protein kinase inhibitors, Nie et al. described the substitution of substituted 5-aminopyrazoles with ethoxycarbonyl isothiocyanate, with the regioselectivity of the reaction dependant on both the conditions employed and the nature of the ring substituent at the C(4) position [29].

X-ray Crystal Structure Analysis of Substituted 3,4-Diaryl-5-Aminopyrazole Derivatives
Precedence for the difference in regioselectivity observed for the substitution of aminopyrazoles such as 40 exists in the literature [28]. For example, in the development of a series of novel protein kinase inhibitors, Nie et al. described the substitution of substituted 5-aminopyrazoles with ethoxycarbonyl isothiocyanate, with the regioselectivity of the reaction dependant on both the conditions employed and the nature of the ring substituent at the C(4) position [29].
Therefore, in order to confirm the existence of both monosubstitution and the bicyclic templates, X-ray crystallographic studies were undertaken on a select panel of aminopyrazoles. As can be seen in Figure 6, acetyl aminopyrazole 54 and thiourea 55 demonstrate selective monosubstitution at the N(1) position of the pyrazole ring.
In contrast, pyrazolo[1,5-a]pyrimidine 53 is consistent with the structure proposed, confirming the reactivity of both the exocyclic amino functionality and the nitrogen at the N(1) position of the parent aminopyrazole. Therefore, in order to confirm the existence of both monosubstitution and the bicyclic templates, X-ray crystallographic studies were undertaken on a select panel of aminopyrazoles. As can be seen in Figure 6, acetyl aminopyrazole 54 and thiourea 55 demonstrate selective monosubstitution at the N(1) position of the pyrazole ring.
In contrast, pyrazolo[1,5-a]pyrimidine 53 is consistent with the structure proposed, confirming the reactivity of both the exocyclic amino functionality and the nitrogen at the N(1) position of the parent aminopyrazole. Comparing three analogous series of derivatized aminopyrazoles, it is evident that the pattern of regioselectivity with regard to reaction with monodentate electrophiles is not entirely dependent on the position or nature of the aryl substituents. In the case of the reaction with methyl isothiocyanate and acetic anhydride, substitution occurred at the N(1) position of the aminopyrazole ring in each case, i.e., the kinetic product of the reaction was more favourable than the thermodynamic product.
However, when treated under similar conditions with trifluoroacetic anhydride, the most reactive of the three monodentate electrophiles used, substitution occurred on the exocyclic amine of the aminopyrazole (such as with derivative 43), i.e., the thermodynamic product was the more favourable of the two. In this scenario, it is possible that the kinetic product of the reaction was more unstable than the analogous thiourea and acetyl derivatives, due to the greater electron withdrawing ability of the trifluoroacetyl moiety. Hence, substitution on the amino group may result in the more Comparing three analogous series of derivatized aminopyrazoles, it is evident that the pattern of regioselectivity with regard to reaction with monodentate electrophiles is not entirely dependent on the position or nature of the aryl substituents. In the case of the reaction with methyl isothiocyanate and acetic anhydride, substitution occurred at the N(1) position of the aminopyrazole ring in each case, i.e., the kinetic product of the reaction was more favourable than the thermodynamic product. However, when treated under similar conditions with trifluoroacetic anhydride, the most reactive of the three monodentate electrophiles used, substitution occurred on the exocyclic amine of the aminopyrazole (such as with derivative 43), i.e., the thermodynamic product was the more favourable of the two. In this scenario, it is possible that the kinetic product of the reaction was more unstable than the analogous thiourea and acetyl derivatives, due to the greater electron withdrawing ability of the trifluoroacetyl moiety. Hence, substitution on the amino group may result in the more stable and favoured product.

Antiproliferative Activity of 3,4-Diaryl-5-Aminopyrazole Derivatives
Assessment of antiproliferative activity elicited by the three 5-aminopyrazole series was undertaken using the NCI-60 cell screen, where compounds are tested at a single dose (10 µM) against the panel of 60 cells. Using this platform, distinctive patterns corresponding to differential sensitivities of cell lines to particular compounds can be correlated to the mechanism of cell growth inhibition, and so, it is an important probe in the investigation of mechanistic action in cancer chemotherapy [30].
Examining the first series of 5-aminopyrazoles (40-45), parent 40 displays appreciable selectivity for HOP-92 lung, SNB-75 CNS and UO-31 and CAKI-1 renal cancer cell lines, all of which display inhibition of growth greater than 50% after 48 h of incubation ( Table 1). Extension of aminopyrazole 40 towards bicyclic systems 41 and 42 subsequently resulted in a decrease in overall growth inhibition. In contrast, however, it was found that monosubstitution in the case of acetyl Derivative 44 and thiourea 45 led to both an overall increase in potency against the NCI-60 panel and also an increase in selectivity for UO-31, CAKI-1 renal and SNB-75 CNS cancer cells. For example, an average growth of 32.2% was recorded for thiourea 45 across these three cell lines, compared with a mean growth of 81.1% against the remainder of the panel.
Aminopyrazole 46 and thiourea 50 from the second series (Derivatives 46-50) also display considerable selectivity for the SNB-75, UO-31 and CAKI-1 cell lines, though acetylated derivative 49 is, somewhat surprisingly, relatively inactive by comparison (Table 1). Although conversion of aminopyrazole 46 towards bicyclic Analogues 47 and 48 resulted in an increase in mean growth in both cases, greater selectivity was achieved towards the HOP-92 cell line. This was particularly evident for pyrimidine 48, which completely arrests growth in this tumour cell line and even exhibits a degree of cytotoxicity (−3.1% growth after 48 h). Utilising the COMPARE algorithm, a high correlation was obtained between the growth inhibition data for three different sets of regioisomers (where one is an identical result): aminopyrazoles 40 and 46 (0.791), pyrimidines 42 and 48 (0.632) and thioureas 45 and 50 (0.805). Given the close agreement found, there exists the likelihood of a common mechanism of action, which provides the scope for further SAR study.
The final series of 3,4-diaryl-5-aminopyrazoles (51-55), containing an indolyl functionality rather than a 7-azaindolyl subunit, displayed the most potent overall antiproliferative activity across the derivatives tested. In keeping with the trend displayed in the previous two series, parent aminopyrazole 51 and monosubstituted Derivatives 54 and 55 are more potent than their bicyclic Analogues 52 and 53. In particular, thiourea 55 almost completely arrests growth in HOP-92 lung cancer cells (0.3% growth after 48 h), whilst also exhibiting excellent selectivity for SNB-75 CNS and UO-31 and CAKI-1 renal cancer cells. Once more, this pattern of selectivity compares favourably to that seen for analogous thioureas 45 and 50. Of great interest also is the fact that aminopyrazole 51, acetylated Derivative 54 and thiourea 55 exhibit a highly similar level of growth inhibition against the renal cancer cell lines UO-31 and CAKI-1 (Table 1).

Topoisomerase II Decatenation Assay
Consequent to screening the four series for antiproliferative activity, interaction with topoisomerase II of a subset was investigated as a potential mechanism of action. Topo II helps to modulate DNA processes via transient double-strand breaks in the DNA helix [31,32]. With higher expression of topo II in proliferating cells than inactive cells, it is a clinical target for the induction of cytotoxic effects in many cancer types, and inhibitors include the podophyllotoxin (2) derivative, etoposide 4.
Candidates 45-50 were screened for their ability to inhibit topoisomerase II-mediated circular kinetoplast DNA (kDNA) decatenation. The standard topo II inhibitor, ellipticine, was used as a reference compound in this study. From the assay, it was apparent that none of the derivatives displayed any activity against topo II (see the Supporting Information). It is possible that the lack of a core planar region in these derivatives prevents intercalation with DNA from occurring, thus eliminating the possibility of topo II inhibition as a function of their antiproliferative activity.
The topo II decatenation assay protocol involved initial incubation of each inhibitor candidate (100 µM) along with a stock solution containing water, ATP, assay buffer, kDNA obtained from the mitochondrial DNA of Crithidia fasciculate, and topo II, at 37 • C for 1 h. Following the addition of stop buffer, agarose DNA gel electrophoresis was run at 50 V for 2 h using a Consort EV243 power pack, to determine the relative amounts of decatenated DNA bands obtained in each compound lane.
Positive (water), as well as negative controls (ellipticine) were incorporated in order to validate the results of each run [33]. The resulting gels were viewed under UV light using a DNR Bio-Imaging System and photographed using GelCapture software.

NCI-60 Anticancer Screening
The experimental methodology involves initial growth of the tumour cell lines in RPMI 1640 medium containing 5% foetal bovine serum and 2 mM L-glutamine. For a typical screening experiment, cells are inoculated into 96-well microtiter plates in 100 µL of medium at plating densities ranging from 5000-40,000 cells/well, depending on the doubling time of individual cell lines. After cell inoculation, the microtiter plates are incubated at 37 • C, 5% CO 2 , 95% air and 100% relative humidity for 24 h prior to the addition of candidate compounds.
After 24 h, two plates of each cell line are fixed in situ with trichloroacetic acid (TCA), to represent a measurement of the cell population for each cell line at the time of drug addition (time zero (Tz)). Candidate compounds are dissolved in DMSO at 400-fold the desired final maximum test concentration and stored frozen prior to use. The single dose screen is carried out at a concentration of 10 µM.
Following drug addition, the plates are incubated for an additional 48 h at 37 • C, 5% CO 2 , 95% air and 100% relative humidity. For adherent cells, the assay is terminated by the addition of cold TCA. Cells are fixed in situ by the gentle addition of 50 µL of cold 50% (w/v) TCA and incubated for 60 min at 4 • C. Sulforhodamine B (SRB) solution (100 µL) at 0.4% (w/v) in 1% acetic acid is added to each well, and plates are incubated for 10 min at room temperature. Absorbance is read on an automated plate reader at a wavelength of 515 nm, and using the absorbance measurements of time zero (Tz), control growth (C) and test growth in the presence of the drug at a concentration of 10 µM, the percentage of growth is calculated [34].

Chemical Data
The data for Compounds 17-21 are available in the Supplementary Information.
6-(1H-Indol-1-yl)hexanenitrile (22). To a solution of indole (2.51 g, 21.4 mmol) in dry DMF (60 mL) at 0 • C was added sodium hydride (1.31 g, 32.75 mmol) in a portion-wise manner. The resultant mixture was allowed to stir at room temperature for 30 min after which time 6-bromohexanitrile (4.25 mL, 1.328 g/mL, 32 mmol) was added with care. The reaction mixture was then allowed to warm to room temperature and stirred overnight. The reaction mixture was subsequently and carefully poured into ice-cold water, and this resulting mixture was extracted with ethyl acetate (6× 50 mL). Combined organic layers were then washed with water (5× 50 mL) and brine (3× 50 mL) before being dried over anhydrous magnesium sulphate and concentrated under reduced pressure to yield a brown oil, which was subject to flash column chromatography (65:35, hexane/ethyl acetate) to yield a viscous yellow oil, which was used without further purification 2-(1-(5-Cyanopentyl)-1H-indol-3-yl)-2-oxoacetic acid (23). To a solution of 6-(1H-indol-1-yl)hexanenitrile 6 (3.49 g, 16.4 mmol) in diethyl ether (100 mL) at 0 • C was added oxalyl chloride (2.20 mL, 1.48 g/mL, 24.64 mmol) dropwise over a 20 min. The resultant deep orange slurry was allowed to stir at room temperature for 30 min. After this time saturated, aqueous sodium bicarbonate solution was added very carefully (15 mL). The resulting biphasic mixture was left to stir overnight. Following this, the reaction mixture was vacuum filtered, and the product cake was washed with ice-cold diethyl ether (10 mL) and ice-cold water (5 mL) before being allowed to dry to yield the oxoacetic acid as a pink solid (4.03 g, 14.2 mmol, 86%), m.p. 133-135 • C: ν max /cm −1 (KBr) 3428, 2952, 2241, 1752, 1628  3-(1-Acetyl-1H-indol-3-yl)-4-(3,4,5-trimethoxyphenyl)furan-2,5-dione (28). Potassium 2-(1H-indol-3-yl)-2-oxoacetate 2 (703 mg, 3.09 mmol) was added to a round-bottomed flask along with 3,4,5-trimethoxyphenylacetic acid (700 mg, 3.09 mmol). Acetic anhydride was added (25 mL) and the reaction mixture was heated for 24 h at 80 • C. Following this, the acetic anhydride was removed under reduced pressure, and to the crude residue was added saturated aqueous sodium bicarbonate solution. This mixture was extracted with ethyl acetate (5× 50 mL). The ethyl acetate was then washed with saturated aqueous sodium bicarbonate solution (4× 50 mL), water (4× 50 mL) and brine (2× 100 mL) before being dried over anhydrous magnesium sulphate and concentrated under reduced pressure to yield a brown residue. were placed in a round-bottomed flask. Acetic anhydride (55 mL) was added, and the resulting mixture was heated for 24 h at 80 • C. Following this, the excess acetic anhydride was removed under reduced pressure, and to the solid residue was added saturated aqueous sodium bicarbonate solution. The crude organic material was extracted with ethyl acetate (4× 50 mL). The combined ethyl acetate washings were then washed with saturated aqueous sodium bicarbonate solution (3× 100 mL), water (3× 100 mL) and brine (2× 100 mL) before being dried over anhydrous magnesium sulphate and concentrated under reduced pressure to yield a sticky red solid. Purification by flash chromatography Acetic anhydride (15 mL) was added, and the mixture was heated at 80 • C for 24 h. Following this, the excess acetic anhydride was removed under reduced pressure. Saturated aqueous sodium bicarbonate solution was added to the residue to remove any residual acetic anhydride. Crude organic material was then extracted with ethyl acetate (4× 50 mL). The combined ethyl acetate extracts were then washed with saturated aqueous sodium bicarbonate solution (3× 50 mL), water (3× 50 mL) and brine (3× 50 mL) before being dried over anhydrous magnesium sulphate and concentrated under  Synthesis of aryl hydroxymaleimides: General Procedure (i): To a solution of the maleic anhydride (1 eq.) in acetonitrile (or DMF) was added hydroxylamine hydrochloride (5 eq.) and triethylamine (5 eq.). The mixture was heated for 24 h at 80 • C. Following this, 1 M aqueous HCl was added to quench the reaction. Organic components were extracted with ethyl acetate, and the combined extracts were washed with saturated aqueous sodium bicarbonate solution, water, brine, dried over anhydrous magnesium sulphate and concentrated under reduced pressure to yield the corresponding hydroxymaleimide.

Conclusions
We report the synthesis and evaluation of a novel template for the investigation of the inhibition of cancer cell growth. The synthetic route allows for late stage differentiation to explore the effect of substitution on the headgroup between indole and trimethoxyphenyl subgroups. While disappointing in terms of hydroxymaleimide raw potency, the results of the cell growth inhibition NCI-60 cell screen illustrate the possibility of a similar mode of bioactivity across the four series of 3,4-diaryl heterocyclic derivatives. Appreciable selectivity was recorded for the three parent aminopyrazoles 40, 46 and 51. Upon extension of the 5-aminopyrazole core towards bicyclic systems such as triazinedione 41, the overall mean antiproliferative potency was diminished. However, it was highly notable that conversion of aminopyrazole 46 to pyrimidine 48 resulted in an extraordinary increase in both selectivity and potency against HOP-92 lung cancer cells.
In contrast to their bicyclic analogues, monosubstitution of the 5-aminopyrazole core with thiourea and acetyl functionalities resulted in an increase in overall mean growth inhibition in each case, except for acetylated derivative 49. Also of particular note is the substituent directed progression seen from 33-45, 50 and finally 55, the most potent compound of the entire series. Taking the core hydroxymaleimide 33 and converting to a 5-aminopyrazole thiourea 55 yields significant growth inhibition, and the difference in 45 and 50 confirms which orientation is most potent. It is evident that conversion of the headgroup to novel substituted heterocycles can imbue potency and some selectivity in cancer cell lines. This finding is of particular relevance to Compounds 9 and 10 seen earlier, albeit that the substituent orientation that delivers potency against VEGF-R2 kinase is reversed.
Despite the overall mean growth inhibition evident in all three series of aminopyrazole derivatives, the lack of cytotoxicity points to a cytostatic (or cell cycle) mechanism of action. Taking this into account, further SAR study using other monosubstituents is imperative towards the future realization of this class of compounds as chemotherapeutic agents. Similarly, further investigative work will aim to identify the molecular targets associated with the cancer cell lines against which considerable selectivity was observed, namely HOP-92 lung cancer, SNB-75 CNS and UO-31, CAKI-1 renal cancer cell lines. As a result of the topological assays, it is clear topo II inhibition is not a relevant mechanism of action, and to further this, screening of protein kinase and tubulin binding are proposed for the active novel 3,4-diaryl-5-aminopyrazole derivatives.