Structure-activity relationships of bumetanide derivatives: correlation between diuretic activity in dogs and inhibition of the human NKCC2A transporter

Abstract

Background and purpose: The N-K-Cl cotransporters (NKCCs) mediate the coupled, electroneutral movement of Na⁺ , K⁺ and Cl^- ions across cell membranes. There are two isoforms of this cation co-transporter, NKCC1 and NKCC2. NKCC2 is expressed primarily in the kidney and is the target of diuretics such as bumetanide. Bumetanide was discovered by screening ∼5000 3-amino-5-sulfamoylbenzoic acid derivatives, long before NKCC2 was identified in the kidney. Therefore, structure-activity studies on effects of bumetanide derivatives on NKCC2 are not available.

Experimental approach: In this study, the effect of a series of diuretically active bumetanide derivatives was investigated on human NKCC2 variant A (hNKCC2A) expressed in Xenopus laevis oocytes.

Key results: Bumetanide blocked hNKCC2A transport with an IC₅₀ of 4 μM. There was good correlation between the diuretic potency of bumetanide and its derivatives in dogs and their inhibition of hNKCC2A (r² = 0.817; P < 0.01). Replacement of the carboxylic group of bumetanide by a non-ionic residue, for example, an anilinomethyl group, decreased inhibition of hNKCC2A, indicating that an acidic group was required for transporter inhibition. Exchange of the phenoxy group of bumetanide for a 4-chloroanilino group or the sulfamoyl group by a methylsulfonyl group resulted in compounds with higher potency to inhibit hNKCC2A than bumetanide.

Conclusions and implications: The X. laevis oocyte expression system used in these experiments allowed analysis of the structural requirements that determine relative potency of loop diuretics on human NKCC2 splice variants, and may lead to the discovery of novel high-ceiling diuretics.

Figure 1

Structures of bumetanide derivatives and their code numbers and, for comparison, bumetanide (PF-1593). For all compounds, lipophilicity (logP) is indicated. Furthermore, the acidic dissociation constant, pKa, of the carboxylic group is shown for bumetanide. Bumetanide derivatives had similar pKa values, except PF-2178 (0.7) and BUM13 (7.0).

Figure 2

Effect of bumetanide on hNKCC2A-mediated ⁸⁶Rb⁺ uptake in Xenopus oocytes. (A) A representative experiment demonstrating the inhibitory effect of 0.03–100 μM bumetanide on ⁸⁶Rb⁺ uptake (in cpm, counted for 10 min) in hNKCC2A-expressing oocytes and on batch-matched uninjected oocytes, n = 5–10 oocytes per condition and error bars are SDs. (B) Dose–inhibition curve of bumetanide on hNKCC2A-mediated ⁸⁶Rb⁺ uptake (corrected for endogenous NKCC contribution in uninjected oocytes) normalized to control (0 μM bumetanide) and averaged across six experiments, with the IC₅₀ calculated from each individual experiment prior to averaging.

Figure 3

Inhibitory effect of bumetanide derivatives on hNKCC2A activity in the Xenopus oocyte assay. ⁸⁶Rb⁺ uptake was measured in the absence or presence of different drug concentrations. Data are shown as a representative experiment (out of at least three identical experiments) with means ± SD from 5 to 15 oocytes per condition in each experiment. For PF-1659 and BUM13, the highest concentration of the compound (100 μM) was only tested once due to lack of inhibitory effect.

Figure 4

Correlation between log IC₅₀ (or eIC₅₀) for inhibition of hNKCC2A in the Xenopus oocyte assay and diuretic potency in the dog assay for the 10 compounds used in this study (see Figure 1 and Table 2013).