Pterin-sulfa conjugates as dihydropteroate synthase inhibitors and antibacterial agents

Abstract

The sulfonamide class of antibiotics has been in continuous use for over 70 years. They are thought to act by directly inhibiting dihydropteroate synthase (DHPS), and also acting as prodrugs that sequester pterin pools by forming dead end pterin-sulfonamide conjugates. In this study, eight pterin-sulfonamide conjugates were synthesized using a novel synthetic strategy and their biochemical and microbiological properties were investigated. The conjugates were shown to competitively inhibit DHPS, and inhibition was enhanced by the presence of pyrophosphate that is crucial to catalysis and is known to promote an ordering of the DHPS active site. The co-crystal structure of Yersinia pestis DHPS bound to one of the more potent conjugates revealed a mode of binding that is similar to that of the enzymatic product analog pteroic acid. The antimicrobial activities of the pterin-sulfonamide conjugates were measured against Escherichia coli in the presence and absence of folate precursors and dependent metabolites. These results show that the conjugates have appreciable antibacterial activity and act by an on target, anti-folate pathway mechanism rather than as simple dead end products.

Graphical Abstract

Sulfonamides were the some of the first successful antimicrobial agents and have been used continuously since the 1940s to treat a wide variety of Gram-positive, Gram-negative and protozoal infections. ^1–3 Sulfonamides are usually prescribed in combination with the dihydrofolate reductase inhibitor trimethoprim to maximize therapeutic response and reduce the development of resistance.^{4, 5} Today, this drug combination is relied upon as a valuable therapeutic option in two important areas: (1) as an inexpensive oral treatment for community acquired bacterial infections resistant to frontline agents such as penicillins and fluoroquinolones; and (2) as a prophylaxis treatment for the prevention of Pneumocystis jirovecii pneumonia in immune-compromised cancer and HIV patients.^{6, 7} Over time, sulfonamide use has been compromised by the emergence of resistance and the poor tolerance of patients to long-term treatment. ^{6, 8}

Sulfonamides target the enzyme dihydropteroate synthase (DHPS) and mimic the substrate p-aminobenzoic acid (pABA).^{1, 9} We previously reported detailed structural, biochemical and computational studies on the catalytic mechanism of DHPS.¹⁰ DHPS catalyzes the conjugation of 6-hydroxymethyl-7,8-dihydropterin pyrophosphate (DHPPP) with pABA to form 7,8-dihydropteroate, an intermediate in the folate biosynthesis pathway (Figure 1A).¹⁰ In addition to inhibiting DHPS, sulfonamides can also act as alternative substrates by forming dead end pterin-sulfonamide products, which may also inhibit enzymes later in the pathway (Figure 1B).^10–12 In this role, the contribution to the overall antimicrobial potency of this class of antibiotics is poorly defined. For the purpose of continuing our studies on the inhibition of DHPS and its therapeutic potential, we sought to chemically synthesize a panel of pterin-sulfonamide conjugates.

Figure 1.

1A: Reaction catalyzed by dihydropteroate synthase; 1B: General structure of pterinsulfonamide conjugates

Repeating the previously published alkylation based route¹³ to these conjugates, we obtained only incorrect isomeric products, challenging the previously ascribed pharmacological properties.^{14, 15} Consequently, a new synthetic route to rapidly and accurately generate pterin-sulfonamide conjugates was developed. The new conjugates were tested for inhibition of DHPS and co-crystallized with the enzyme to characterize their mode of binding. Microbiological profiling using E. coli indicated that pterin-sulfonamides are not only good inhibitors of DHPS but also have substantial antimicrobial activity that can be antagonized by folate precursors.

Our initial strategy to synthesize the conjugates was based on the method detailed by Piper and Montgomery.¹³ Briefly, commercially available (2,4-diaminopteridin-6-yl)methanol 1 was treated with triphenyl phosphine and bromine to afford the 6-bromomethyl pterin analog 2,¹³ which was expected to couple with the appropriate sulfonamide, followed by 4-deamination, to give target compounds in three steps (Scheme 1). However, when compound 2 reacted with sulfamethoxazole 3 in the presence of N,N-diisopropylethylamine (DIEA), only isomeric sulfonamide linked compound 4 was isolated as the main product, instead of the desired aniline linked pterin-sulfamethoxazole 15. Attempts to modify the synthesis by protection of the sulfonamide or alter the alkylating conditions failed.^16–18

Scheme 1.

Initial synthetic strategy.

An alternative synthetic route that provided the necessary flexibility to conjugate a large variety of sulfonamides to pterin was developed around the key acyl protected formyl-pterin intermediate 13 shown in Scheme 2. Treatment of commercially available 2-amino-5-(chloromethyl)-3-cyanopyrazine 1-oxide 5 with phosphorus trichloride at room temperature in tetrahydrofuran solution resulted in smooth deoxygenation to give 6,¹⁹ which was treated with pyridine with stirring overnight at room temperature to afford pyridinium salt 7.²⁰ Salt 7 reacted with N,N-dimethyl-4-nitrosoaniline in the presence of potassium carbonate to give nitrone 8, which was then treated with 6N HCl to give aldehyde 9. Quantitative conversion of 9 to its dimethyl acetal in the presence of DOWEX 50WEX8-100 ion-exchange resin, followed by condensation with guanidine carbonate in the presence of sodium methoxide, afforded 2,4-diamino pterin 10. Treatment of 10 with hot 5% sodium hydroxide gave 6-formylpterindine dimethyl acetal 11, which was treated with acetic anhydride at 100 °C to give 2-acetyl protected 12. The 6-formylpterine dimethyl acetal 12 was hydrolyzed with the treatment of formic acid to afford 6-formylpterin 13.²¹ Coupling of 6- formylpterin 13 with different sulfonamides or dapsone (4-[(4-aminobenzene)sulfonyl]aniline) was carried out smoothly in acetic acid to give the corresponding Schiff base compound, which was subsequently reduced with borane tert-butylamine complex to give 14 a–h.²² Deprotection of 14 ah with 0.1 N NaOH afforded target compounds 15-22.²³

Scheme 2.

Synthetic method for production of pterin-sulfonamide conjugates.

Compounds 15–22 were all tested for inhibition of B. anthracis DHPS using a fluorescence polarization (FP) based competition binding assay, performed in the absence or presence of 2 mM sodium pyrophosphate (Table 1 and Supplementary Material). ²⁵ Pyrophosphate (PP_i) is the biochemical byproduct of DHPS catalysis, and its inclusion in the FP assay allowed us to confirm the site-specific binding of pterin-sulfonamide conjugates. More specifically, we have previously observed that PP_i stabilizes two flexible loops in the DHPS active site that form the pABA-sulfonamide binding site.¹⁰

Table 1.

Inhibitory activities of pterin-sulfa conjugates.

Cmpd	R	BaDHPS in vitro FP Binding Inhibition IC50 (µM)		Minimum Inhibitory Concentration E. coli K12, in µMb,c
		(−) PPia	(+) 2 mM PPia	M9 media	M9 plus 5 µg/mL PABA	M9 plus 20 µg/mL Methionine
15		20.6±3.5	3.4±0.1	10.9 (0.8)c	467 (263)c	29 (1.5)c
16		22.5±1.0	8.1± 0.7	28.3 (1.5)	>412 (249)	47 (5.9)
17		19.3±0.6	9.4± 0.7	17.1 (1.6)	>470 (666)	78 (6.2)
18		>62.5	2.5± 0.2	29.5 (2.7)	>471 (669)	98 (6.3)
19		26.8±3.0	1.9± 0.1	17.2 (1.6)	>412 (537)	86 (3.4)
20		15.2±0.6	11.5±0.8	34.3 (3.8)	>412 (>644)	172 (10.1)
21		17.8±0.8	11.3±1.1	197 (25)	236 (>934)	236 (50)
22		18.6±0.8	17.1±1.2	136 (22)	>514 (>934)	>200 (58)
Ampicillin		ND	ND	1.58	2	2.1
Chloramphenicol		ND	ND	8.5	9.7	9.7
Trimethroprim		ND	ND	1.7	1.4	0.7

^aPresent as sodium pyrophosphate. Average of 3 experiments, ±SEM.

^bPerformed according to ²⁴.

^cCorresponding values for parent sulfonamides shown in parentheses.

As a group, the pterin-sulfonamide conjugated compounds are more efficient than SMX (FP IC₅₀ >62.5 µM) at displacing the probe from the active site in the FP competition assay. These compounds show an even more efficient displacement of probe when PP_i is added to the assay. Among all analogs, pterin-sulfadimethoxine (pterin-SDM) conjugate 19 and pterin-sulfamethoxazole (pterin-SMX) conjugate 15 have the highest potency against BaDHPS, with IC₅₀ values of 1.9 µM and 3.4 µM, respectively, in the presence of PP_i. Conjugation of pterin with other sulfa analogs resulted in compounds with good inhibitory activity with the weakest inhibition observed for the sulfanilamide conjugate 22.

To establish the mode of binding of the pterin-sulfonamide conjugates in the DHPS active site, we soaked compound 16 into apo crystals of Yersinia pestis DHPS (YpDHPS) and determined the structure of the complex to 2.1 Å resolution (PDB ID: 5JQ9). As anticipated, 16 occupies both the pterin and pABA binding pocket of DHPS (Figure 2A). The mode of binding is very similar to those of pteroic acid²⁶ and the sulfathiazole-pterin adduct that was generated catalytically in crystallo ¹⁰ (Figures 2B and 2C). Residues D96, N115, D185, K221 and R255 interact with the pterin moiety, and F190, K221 and S222 interact with the sulfonamide/pABA groups (YpDHPS residue numbering). We have also determined the structure of the DHPS near transition state structure,¹⁰ and its superposition onto the 16 complex shows how closely the compound mimics the binding modes of pterin and pABA. Note that weak electron density adjacent to 16 (Figure 2A) corresponds to the DHPS anion-binding pocket, which is occupied by sulfate and pyrophosphate in the DHP-STZ and transition state structures, respectively (Figures 2C and 2D). Occupancy of the anion-binding pocket by pyrophosphate and the resulting stabilization of loop2 (Figure 2D) would explain our observation that pyrophosphate augments the binding of the conjugate compounds.

Figure 2.

A. Yersinia pestis DHPS (YpDHPS) with compound 16 at the active site. Also shown is the omit map (mFo-DFc) showing the electron density of 16 contoured at 2σ (PDB ID: 5JQ9). B. YpDHPS with the product analog pteroic acid bound at the active site (PDB ID: 3TYU). C. Bacillus anthracis DHPS (BaDHPS) with the pterin-sulfathiazole (DHP-STZ) catalytic product bound at the active site (PDB ID: 3TYE). D. Superimposition of the YpDHPS 16 complex and the YpDHPS near transition state complex (PDB ID: 3TYZ). In all figures, the DHPS cartoon and carbon atoms are grey, the oxygen, nitrogen and sulfur atoms are red, blue and yellow respectively, and the compound carbons are cyan. In (D), the transition state structure is shown as pink cartoon and carbon atoms, the ordered loop2 is shown in green, and the essential Mg²⁺ ion is shown.

The antibacterial activities of these conjugate compounds and their corresponding parent sulfonamides were tested against Escherichia coli K12 (Table 1) in M9 minimal media ²⁷ suitable for testing sulfonamide activity. Under these conditions, most conjugates showed appreciable antibacterial activity, with the exception of the sulfanilamide and dapsone hybrids (21, 22) mirroring the weaker direct activity of the corresponding parent sulfonamides.²⁴ The sulfamethoxazole hybrid (15) showed the most potency with a MIC of 10.9 µM. Overall the minimal inhibitory concentration activities for the conjugates were usually around 10 times weaker than for the corresponding parent sulfonamides (compound 15 10.9 µM, vs sulfamethoxazole 0.8 µM).

We next explored whether the observed antimicrobial activities are a result of on target biochemical inhibition. MIC activities were examined in the presence of the added folate precursor pABA or the dependent metabolite methionine (Table 1). Addition of pABA completely ablated MIC activity for all conjugates²⁸, and a similar but less dramatic effect was seen when growth media was supplemented with methionine. These data mirrored the MIC shifts we observed with the corresponding sulfonamides, supporting the hypothesis that the pterin conjugates function as direct antibacterial compounds through inhibition of the folate pathway,

In conclusion, we have demonstrated that pterin-sulfonamide conjugates have antibacterial action through biochemical inhibition of DHPS and antagonism of the folate pathway. It is important to note that the conjugates synthesized in this study are in the oxidized form of pterin and not the reduced form dihydropterin that is formed by the natural biochemical action of DHPS on DHPPP and sulfonamides, which likely accounts for some of the reduction in pharmacological activity. However, our structures reveal that the oxidation state of the pterin ring has little effect on how it binds into the pterin pocket (compare Figures 2A and 2B with 2C and 2D).

The lower MIC activity of the conjugates compared to their corresponding sulfonamides may be attributable to differences in penetration and efflux of the pterin sulfonamide conjugates vs sulfonamides. Further studies are ongoing to study the susceptibility to efflux, capacity to inhibit enzymes later in the pathway and in cell metabolism of the conjugates. The pharmacological advantages that pterin-sulfonamide conjugates may have over sulfonamides are also worthy of further study. The conjugates may have a lower propensity for sulfonamide mediated anaphylaxis, as the oxidatively reactive aniline group in the sulfonamides is hindered by the pterin conjugation.²⁹ Another important observation from this study is the enhanced binding to DHPS by these conjugates that presumably results from the introduction of a pterin anchor. The pterin anchor also has implications for the development of resistance to these conjugates. As previously noted,³⁰ the pterin pocket is highly conserved and unlikely to tolerate mutations that might decrease binding affinity. This is not true of the isolated sulfonamides because the pABA-binding site can tolerate such mutations.¹⁰ Thus, the activity of the conjugates against defined sulfonamide resistant strains should be determined.