Synthesis and evaluation of functionalized isoindigos as antiproliferative agents

Abstract

A series of functionalized isoindigos structurally related to meisoindigo (1-methylisoindigo), a therapeutic agent used for the treatment of a form of leukemia, were synthesized and evaluated for antiproliferative activities on a panel of human cancer cells. Two promising compounds (1-phenpropylisoindigo and 1-(p-methoxy-phenethyl)-isoindigo) that were more potent than meisoindigo and comparable to 6-bromoindirubin-3′-oxime on leukemic K562 and liver HuH7 cells were identified. Structure–activity relationships showed the importance of keeping one of the lactam NH in an unsubstituted state. Substitution of the other lactam NH with aryl or arylalkyl side chains retained or improved activity in most instances. An intact exocyclic double bond was also essential, possibly to maintain planarity and rigidity of the isoindigo scaffold. None of the compounds were found to inhibit CDK2 in an in vitro assay, in spite of reports linking the antiproliferative activities of meisoindigo and other isoindigos to CDK2 inhibition. Hence, these functionalized isoindigos disrupted cell growth and proliferation by other mechanistic pathways that did not involve CDK2 inhibition.

Introduction

The indigoids are a group of isomeric bisindoles that have recently emerged as a promising scaffold for anticancer activity. Interest in the indigoids was kindled when indirubin was identified as the active ingredient of a traditional Chinese recipe (Danggui Longhui Wan) that was used for the treatment of chronic myelogenous leukemia (CML).1, 2 The anti-tumor activity of indirubin and some of its analogs (notably indirubin-3′-oximes) were subsequently attributed to the inhibition of cyclin-dependant kinases (CDK).³ Co-crystallized structures of CDK2 and various indirubin analogs showed that these compounds interacted with the ATP binding site via hydrogen bonds to key amino acids in the hinge region.3, 4 Unfortunately, indirubin was not considered a good candidate for drug development because its poor water solubility would adversely affect oral bioavailability. Moreover, there were reports of gastrointestinal side effects associated with its use.² Hence, more drug-like substitutes of indirubin were sought and some promising candidates identified were the functionalized indirubin-3′-oximes,5, 6, 7 meisoindigo,2, 8, 9, 10, 11, 12 and Natura™ (1-β-d-triacetylxylopyranosylisoindigo)¹³ (Fig. 1). Meisoindigo and Natura™ are derivatives of isoindigo (3,3′-bisindole) which is a regioisomer of indirubin (3,2′-bisindole). Unlike the indirubin scaffold which is widely investigated as an anti-cancer lead, isoindigos have attracted less attention.

Meisoindigo (1-methylisoindigo) is used as an indirubin substitute for the treatment of CML in China.¹² It has a multi-targeting profile that included inhibition of DNA biosynthesis and microtubule assembly in tumor cells,¹⁰ induction of leukemic cell differentiation and c-myb gene downregulation,¹¹ and inhibition of angiogenesis.¹² Natura™ is an example of an isoindigo that is derivatized with a sugar moiety. Its anti-cancer agent activity was attributed to inhibition of several cyclin dependent kinases (CDK 2, 4, and 6).¹³ The presence of the protected sugar moiety was proposed to enhance the interaction with the ATP binding site of the kinase, since isoindigo was essentially inactive as a CDK inhibitor.¹⁴ When several 1-glycosylisoindigos were evaluated as CDK inhibitors, it was found that only compounds with O-benzyl groups on the sugar hydroxyl (OH) residues inhibited CDK while most entities with O-acetyl or unprotected OH groups on the sugar moiety were inactive.¹⁴ Notwithstanding their CDK inhibitory activities, the O-benzyl 1-glycosylisoindigos had only weak antiproliferative activities,¹⁵ possibly due to their poor aqueous solubilities which would affect cellular penetration.¹⁴ As these compounds had molecular weights that exceeded 500 Da and lipophilicities (evaluated by C log P) that were greater than 5, their poor permeabilities are anticipated from property profiling rules like the Lipinski’s rule of five.¹⁶ The same would be true for Natura™ which was also poorly soluble in water.¹³ Thus, designing pharmacologically active isoindigo derivatives with a desirable solubility–lipophilicity balance that would translate to an acceptable level of cellular permeability poses a formidable challenge. In this report, we describe our efforts to address this problem by synthesizing a series of isoindigos that have functionalized arylalkyl or alkyl substituents on the ring nitrogen. To maintain a pharmaceutically friendly profile, efforts were made to restrict the size and lipophilicity of the synthesized compounds. The compounds were evaluated for effects on the growth of several human cancer cell lines. As CDK2 inhibition has been widely associated with the growth inhibitory properties of isoindigos,13, 14, 15 we carried out in silico docking to assess their potential as CDK2 inhibitors and followed this up with experimental determination of CDK2 inhibitory activity.