In previous reports, a series of analogues were prepared and their in vitro anti-proliferative activities in HL-60 leukemia cells were established and compared with those of JS-K.^4,7,8 The ability to inhibit cancer cell proliferation by this class of compounds was sensitive to structure. Bulkier substituents on the piperazine ring,⁸ or removal of the carbamate functionality altogether in moving to a piperidine ring,⁴ or an acyclic secondary amine bearing the arylated diazeniumdiolate resulted in diminution of inhibitory activity.^7,8 Furthermore, the size of the carbamate group was also an important determinant of the cytotoxicity; an alkyl group longer than ethyl decreased the inhibitory activity.^4,8 Finally, two nitro groups on the aryl ring were essential for nitric oxide release or any observable anti-proliferative activity in the in vitro cell viability assay.^4,8 In a recent study, it was found that JS-K and its homopiperazine analogue were identical in activity against a number of cancer cell lines; these compounds did not display any cytotoxicity against a normal renal epithelial cell line at concentrations where they inhibited the proliferation of a panel of renal cancer cell lines.⁸ In order to explore the stereoelectronic accommodation at N-4 of the (homo)piperazine ring of such compounds, a number of variably N-substituted heterocyclic O²-(2,4-dinitrophenyl) diazeniumdiolates were prepared and studied.

Using a reported method, compound 1a was prepared and deprotection of 1a afforded the amine hydrochloride, 2a.⁴ A similar procedure was used to prepare the homopiperazines, 1b and 2b (Scheme 1).⁴

Compounds 2a and 2b were then independently treated with a range of corresponding electrophiles to form the desired compounds 3–15 (Table 1).^4,7,8 Compound 16 was prepared using a reported method from N-carboethoxy(homopiperazine) in two steps (Fig. 1).⁴ The quaternary ammonium salt 17 was obtained by methylation of 1c¹² with excess methyl iodide in dichloromethane (Scheme 2).

JS-K was designed to be activated by glutathione (GSH) to release potentially cytotoxic nitric oxide intracellularly (Scheme 3).¹ This reaction was found to be catalyzed by glutathione S-transferase (GST), a class of enzymes frequently over-expressed in certain cancers.^1,8

First, glutathione-activated nitric oxide yields from these compounds were determined using a chemiluminescence assay. All compounds were found to release nearly quantitative amounts of NO on reaction with GSH (Table 2).

Next, JS-K and its analogues were tested for their in vitro anti-proliferative activities against human leukemia cell lines HL-60 and U937 (Table 2). Compounds 3–9 were found to display nearly identical anti-proliferative activities (Table 2) but changing the ethyl carbamate to a diethyl phosphamate group resulted in a lowering of inhibitory activity (Table 2, compound 10). All the homopiperazine analogues prepared in this study were found to have activity comparable with that of JS-K in both HL-60 and U937 leukemia cell lines (Table 2, compounds 11–16). The quaternary ammonium salt 17 that had improved aqueous solubility was lower in potency than JS-K (Table 2).

Finally, in order to test the role of cell permeability and NO in cytotoxicity of these compounds, we measured the levels of intracellular nitric oxide formed upon treating cells with selected compounds and compared these with JS-K.^13,14 The intracellular NO was estimated using the nitric oxide-sensitive fluorophore, 4-amino-5-methylamino-2′,7′-difluorofluorescein diacetate (DAF-FM diacetate); briefly, human leukemia HL-60 cells were pre-loaded with DAF-FM diacetate, followed by treatment with DMSO solutions of various JS-K analogues; fluorescence measurements after 40 min provided estimates of levels of intracellular NO (Fig. 2).

NMR