|Home | About | Journals | Submit | Contact Us | Français|
Replacement of the central isophthalamide core in a synthetic HCl receptor, with a 2,6-dicarboxamidopyridine, leads to a more preorganized molecular structure that exhibits higher chloride affinity and membrane transport flux.
There is currently strong interest in the design of synthetic membrane transporters for anions.1 The goal of our collaborative research programme is to develop HCl co-transport systems. These compounds are designed to act as functional mimics of the prodigiosin family of natural products (e.g., 1) which have been shown to promote the co-transport of H+/Cl− across bilayer membranes.2 These natural products exhibit a range of potentially useful biological activities, including immunosuppression, induction of tumor cell apoptosis, and toxicity against bacteria, protozoa, fungi and malaria parasite.3 Recently, we demonstrated that compound 2 is a weak chloride receptor at neutral pH but when protonated shows a significantly enhanced chloride binding affinity.4 Transport studies showed that 2 functions as an HCl co-transporter in vesicles, and that transport was accelerated by the presence of a pH gradient. More recently, J.T. Davis, Gale, Quesada and co-workers reported that an isophthalamide-derived carrier, with hydroxyl groups in the 4- and 6-positions, functions as a very efficient chloride transporter across vesicle membranes.5 This functionalised system is more highly preorganised than a simple isophthalamide due to the presence of intramolecular hydrogen bonds. Furthermore, the internal hydrogen bonds decrease the molecular polarity. We decided to test the generality of this preorganisation concept and evaluate a series of putative HCl co-transporters 3 and 4 that contain either an isophthalamide or 2,6-dicarboxamide core and a pendant methylimidazole ring. As with 1 and 2, compounds 3 and 4 contain two hydrogen bond donor groups and a basic site. However, the pyridyl analogue 3 was expected to possess a higher degree of preorganisation than the isophthalamide 4, due to the well-known propensity of pyridine-2,6-dicarboxamides to adopt a syn-syn bis(amide) conformation that creates a convergent hydrogen bonding pocket.6 We also prepared and evaluated the negative control compound 5 which lacks the basic imidazole ring.
Compound 3 was synthesized by conversion of 6-(methoxycarbonyl)pyridine-2-carboxylic acid7 to the acid chloride, followed by addition of (1-methyl-1H-imidazol-2-yl)methanamine8, subsequent saponification of the methyl ester, conversion to the acid chloride and finally addition of 4-butylaniline to afford compound 3 (47% yield). Compound 4 was obtained by addition of one equivalent of 4-butylaniline to N1,N3-bis(2-mercaptothiazolides)-isophthalamide9 followed by the addition of (1-methyl-1H-imidazol-2-yl)methanamine8 to afford 4 (yield 18%).
The anion binding affinities of receptors 3 and 4 were studied by standard 1H NMR titration techniques with tetrabutylammonium chloride in DMSO-d6, and the stability constants determined using EQNMR computer program.10 Chloride binding constants were found to be < 10 M−1 for both compounds at 298K (the titration data was fitted to a 1:1 binding model but the formation of a weak 1:2 receptor:anion complex cannot be ruled out). In other words, receptors 3 and 4 have a weak affinity for Cl− under neutral conditions. The binding constants in DMSO-d6 were also determined in the presence of one equivalent of HPF6 at 298K. Compound 4 also interacts weakly with chloride in the presence of HPF6 (<10 M−1), while compound 3 exhibits an enhanced binding affinity for chloride (59 M−1) under these conditions. Such an enhancement was previously observed with compound 2.¶ The influence of the imidazole in the structure of 3 was illustrated by comparison to control structure 5. This compound was obtained in one step by addition of 4-butylaniline on pyridine-2,6-dicarbonyl dichloride. Binding studies showed that compound 5 has a weak affinity for chloride (< 10 M−1) in both the absence and presence of HPF6, demonstrating the ability of the imidazole ring to enhance chloride affinity under acidic conditions.
Crystals of the HCl complex of 3 were obtained by slow evaporation of a solution of 3 in a mixture of toluene/dichloromethane/methanol/HClaq(2M)/isopropanol.‡ The X-ray crystal structure is shown in Figure 1 and reveals that in the solid state 3•HCl forms a ‘2+2 dimer’. The chloride is bound by the two amide NH groups (N•••Cl 3.11(3) Å, N•••Cl 3.48(3) Å). These nitrogen atoms also form hydrogen bonds with the nitrogen of the pyridine ring (N•••Npy 2.770(3) Å, N•••Npy 2.727(3) Å). The protonated imidazole does not directly hydrogen bond with the bound chloride, but rather is involved in an intermolecular hydrogen bond with the amide carbonyl of the adjacent molecule (N•••O 2.67(7) Å). Hydrogen bonding to the amide carbonyl oxygens can enhance chloride transport in two ways. First, it increases chloride affinity by increasing the acidity of the amide NH residues.11 Second, the lipophilicity of the molecule is raised because the amide carbonyl oxygens cannot readily interact with water.
Initially, compounds 3, 4, and 5 were tested for their abilities to transport Cl− across unilamellar vesicles (200 nm diameter) composed of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) or POPC/cholesterol (7/3 molar ratio) under neutral and pH gradient conditions. The neutral vesicles contained NaCl (500 mM), pH 7.0 and were immersed in NaNO3 (500 mM), pH 7.0; whereas, the pH gradient vesicles contained NaCl (500 mM), citric acid (5 mM), pH 4 and were immersed in NaNO3 (500 mM), sodium phosphate (5 mM), pH 7.0. A chloride selective electrode was used to detect chloride release from the vesicles.4 Under neutral conditions, only compound 3 was an effective chloride transporter, with efflux rate slightly inhibited by the presence of cholesterol (Fig 2). Structurally, 3 and 4 only vary by a single nitrogen atom but their transport activities are quite different, lending support to the concept of enhanced transport by a receptor (3) that has a more preorganised and convergent binding pocket. In the presence of a pH gradient, the rates of chloride efflux were enhanced (Fig 3); furthermore, compounds 3 and 4 (but not 5) are able to deacidify the acidic interior of the vesicles (Fig 4).§ These results are consistent with a transport mechanism that involves simultaneous co-transport of HCl by the carrier. With the data in hand we cannot state if the kinetically active transport complex (under pH gradient conditions) is a 1:1 ion pair of 3.HCl, or the “2+2’ dimer shown in Fig 1. The fact that compound 5 shows little or no H+ or Cl− transport activity indicates that the pyridine nitrogen (in compounds 3 and 5) is not protonated under the transport conditions.
To gain additional mechanistic insight, the transport mediated by 3 under neutral conditions was examined in more detail. A transporter concentration study (Fig 5) showed that chloride efflux increases linearly with increasing concentration of 3, supporting a mobile carrier mechanism (i.e., the carrier diffuses across the bilayer membrane).1(a),12 The observation that transport rate is diminished by the presence of cholesterol (Fig 2), is also consistent with a carrier diffusion process, assuming no change in carrier partitioning.13 Finally, we find that the chloride efflux promoted by 3 under neutral conditions is strongly dependent on the identity of the anion in the vesicle external solution. Specifically, substitution of NaNO3 in the external solution for Na2SO4 leads to almost complete loss of chloride efflux. This is evidence for an anion exchange mechanism that only allows chloride efflux if there is a corresponding influx of external anion (i.e., A−/Cl− antiport). Thus, there appears to be a change in the transport mechansim with assay conditions, from H+/Cl− co-transport under pH gradient conditions to A−/Cl− antiport under neutral conditions, a feature exhibited by prodigiosin.2c Further work is needed to elucidate these finer mechanistic details.
The results of this study contribute to an emerging picture in the design of membrane transport carriers.5 Subtle changes in the conformation of a transport carrier can produce large changes in binding affinity, carrier lipophilicity, and as a consequence, enhancements in transport flux.
PAG thanks the EPSRC for a post-doctoral fellowship (J.G.) and the EPSRC together with Professor Mike Hursthouse for access to the crystallographic facilities at the University of Southampton. BDS acknowledges funding support from the NIH (USA).
†Electronic Supplementary Information (ESI) available: [Synthesis details, NMR titration curves and crystallographic data]. See http://dx.doi.org/10.1039/b000000x/
¶ With 2 in acetonitrile-d3, chloride association constants increased from 60 M−1 to 397 M−1 in the presence of HPF6. When these studies were repeated in DMSO-d6, it was found that the affinity for chloride was enhanced from <10 M−1 to 40 M−1. Solubility problems prevented titration studies with compounds 3 and 4 in acetonitrile-d3.
‡ Crystal data for the HCl complex of 3: C29H34ClN5O2, Mr = 520.06, T = 12(2) K, triclinic, space group P-1, a = 7.3811(3), b = 12.1320(4), c = 17.4317(7) Å, α = 76.000(2), β = 85.986(2), γ = 85.254(2)°, V = 1507.39(10) Å3, ρcalc = 1.146 g cm −3, μ = 0.159 mm−1, Z = 2, reflections collected: 20076, independent reflections: 5309 (Rint = 0.0515), final R indices [I > 2σI]: R1 = 0.0680, wR2 = 0.1570, R indices (all data): R1 = 0.0862. wR2 = 0.1678.
§ The acidity inside the vesicles was measured using the encapsulated fluorescent pH indicator Lysosensor Blue™ (Molecular Probes Inc), whose emission intensity decreases as pH increases. The data in Figure 4 shows that the background proton efflux rate is quite substantial due to the large pH gradient. This feature has been observed previously and is discussed in reference 4.