Sulfa and trimethoprim-like drugs – antimetabolites acting as carbonic anhydrase, dihydropteroate synthase and dihydrofolate reductase inhibitors

Abstract

Recent advances in microbial genomics, synthetic organic chemistry and X-ray crystallography provided opportunities to identify novel antibacterial targets for the development of new classes of antibiotics and to design more potent antimicrobial compounds derived from existing antibiotics in clinical use for decades. The antimetabolites, sulfa drugs and trimethoprim (TMP)-like agents, are inhibitors of three families of enzymes. One family belongs to the carbonic anhydrases, which catalyze a simple but physiologically relevant reaction in all life kingdoms, carbon dioxide hydration to bicarbonate and protons. The other two enzyme families are involved in the synthesis of tetrahydrofolate (THF), i.e. dihydropteroate synthase (DHPS) and dihydrofolate reductase. The antibacterial agents belonging to the THF and DHPS inhibitors were developed decades ago and present significant bacterial resistance problems. However, the molecular mechanisms of drug resistance both to sulfa drugs and TMP-like inhibitors were understood in detail only recently, when several X-ray crystal structures of such enzymes in complex with their inhibitors were reported. Here, we revue the state of the art in the field of antibacterials based on inhibitors of these three enzyme families.

• Antibacterial
• carbonic anhydrase
• dihydrofolate reductase
• dihydropteroate synthase
• sulfa drug
• trimethoprim

Introduction

Carbon biological cycle

Living organisms play an important role in the movement of carbon between land, ocean and atmosphere through the processes of photosynthesis and respiration1. Carbon dioxide is the final product of the respiration in all domains of life, from microorganisms to mammals2,3. Virtually all life on earth depends on the production of sugars from sunlight and carbon dioxide (photosynthesis) and the metabolic breakdown (respiration) of those sugars to produce the energy needed for movement, growth and reproduction. Carbon dioxide (CO₂) and its hydration product bicarbonate () are substrates and products of various metabolic reactions in the cells of different living organisms4,5. The interconversion of CO₂ and is spontaneously balanced to maintain the equilibrium between dissolved inorganic carbon dioxide, carbonic acid and carbonate of which is physiologically the most important one as it is a substrate for several carboxylating enzymes involved in biosynthetic pathways such as biosynthesis of fatty acids, amino acids and nucleotides6,7. The uncatalyzed hydration–dehydration of CO₂- is slow at physiological pH and thus, in biological systems, the reaction is accelerated by an enzymatic catalyst, called carbonic anhydrase (CA). CAs (EC 4.2.1.1) catalyze a simple but physiologically relevant reaction in all life kingdoms, carbon dioxide hydration to bicarbonate and protons8,9:

These enzymes are thus involved in many physiologic processes in microorganisms, such as photosynthesis, respiration, CO₂ transport, as well as metabolism of xenobiotics (e.g. cyanate in Escherichia coli)10–13.

Folic acid synthesis pathway

Prokaryotes and unicellular eukaryotes such as yeasts or Plasmodium are able to synthesize folate de novo via the folic acid synthesis pathway (Figure 1). The steps of this pathway include: (i) condensation of dihydropteridine pyrophosphate (DHPP) 1 with p-amino benzoic acid (pABA) 2 by dihydropteroate synthase (DHPS, EC 2.5.1.15) to form dihydropteroate (DHP) 3; (ii) addition of glutamate to DHP by dihydrofolate synthase (DHFS, EC 6.3.2.12) to form dihydrofolate (DHF) 4 and (iii) reduction of DHF to form tetrahydrofolate (THF) 5, catalyzed by dihydrofolate reductase (DHFR, EC 1.5.1.3)14.

Figure 1. Biosynthesis of THF 5 from dihydropteridine pyrophosphate 1 (see text for details).

DHPS is thus an essential enzyme responsible for the de novo synthesis of the folate molecule. THF produced by this pathway is required for one carbon transfer reactions in the biosynthesis of a range of biomolecules, such as nucleotides and amino acids15. DHFR is the enzyme responsible for the maintenance of folate pools in their physiologic, reduced states. DHFR, in fact, is needed for the intracellular conversion of synthetic folic acid into the THF forms that can participate in the folate/homocysteine metabolism. Reduction of DHFR enzymatic activity diminishes the THF pool inside the cells, affecting the level of folate coenzymes and thus purine and pyrimidine biosynthesis16. This may as well influence homocysteine levels and methylation processes, since methyl-THF is needed for the re-methylation of homocysteine to form methionine, thereby ensuring the provision of S-adenosylmethionine (SAM) necessary for most biological methylation processes17.

In contrast to prokaryotes and lower eukaryotes, higher eukaryotic cells cannot synthesize folates de novo, being totally dependent on exogenous folic acid. Thus, the folic acid synthesis pathway represents a convenient target for obtaining drugs for the treatment of infectious diseases, caused by bacteria and protozoan parasites. Indeed, at least two steps in this pathway are targeted by antibacterial drugs in clinical use for decades: DHPS is the target of the sulfa drugs, whereas DHFR is targeted by trimethoprim (TMP) (and its congeners)5–7. These enzymes have been validated as antimicrobial therapeutic targets by chemical and genetic means at the beginning of research in the field of antibacterials18–20.

Enzymatic targets for sulfa- and TMP-like drugs

Carbonic anhydrases

CAs (EC 4.2.1.1) catalyze a simple but physiologically relevant reaction in all life kingdoms, carbon dioxide hydration to bicarbonate and protons8. Five genetically distinct CA families are known to date, the α-, β-, γ-, δ- and ζ-CAs, and all of them are metalloenzymes, using Zn(II), Cd(II) or Fe(II) at their active sites8,12,21. Bacteria encode for enzymes belonging to the α-, β- and γ-classes. Figure 2 shows the phylogenetic relation between the α-, β- and γ-classes from different species. The phylogenetic analysis also shows three different clusters denoting the existence of these three evolutionary classes of CAs, as mentioned above (we did not include δ- and ζ-CAs in our phylogenetic analysis as these enzymes are present only in diatoms)10. CAs thus represent an excellent example of Darwinian convergent evolution of catalytic function. Intriguingly, the phylogenetic analysis produced two clusters, one of which includes the α- and γ-CAs (Figure 2) suggesting that these classes of enzymes might have originated from an ancestral gene different from the gene, which led to the β-CAs.

Figure 2. Phylogenetic tree of the α, β and γ CAs from selected eukaryotic and prokaryotic species. The tree was constructed using the program PhyML 3.0. Branch support values are reported at branch points. Organisms, accession numbers and cryptonyms of the sequences used for the phylogenetic analysis have been indicated in Table 1.

Table 1. CA class, organisms, accession numbers and cryptonyms of the sequences used in the phylogenetic analysis.

CA class	Organism	Accession number	Cryptonym
alpha	Helicobacter pylori J99	NP_223829.1	HpylCA_alpha
	Homo sapiens, isoform II	AAH11949.1	hCAII_alpha
	Homo sapiens, isoform I	NP_001158302.1	hCAI_alpha
	Sulfurihydrogenibium yellowstonense YO3AOP1	ACD66216.1	SspCA_alpha
	Streptococcus salivarius PS4	EIC81445.1	SsalCA_alpha
	Neisseria gonorrhoeae	CAA72038.1	NgonCA_alpha
beta	Schizosaccharomyces pombe	CAA21790	SpoCA_beta
	Brucella suis 1330	NP_699962.1	BsuCA_beta
	Burkholderia thailandensis Bt4	ZP_02386321	BthCA_beta
	Coccomyxa sp.	AAC33484.1	CspCA_beta
	Chlamydomonas reinhardtii	XP_001699151.1	CreCA_beta
	Acinetobacter baumannii	YP_002326524	AbaCA_beta
	Porphyromonas gingivalis ATCC 33277	YP_001929649.1	PgiCA_beta
	Myroides injenensis	ZP_10784819	MinCA_beta
	Zea mays	NP_001147846.1	ZmaCA_beta
	Vigna radiata	AAD27876	VraCA_beta
	Flaveria bidentis, isoform I	AAA86939.2	FbiCA_beta
	Arabidopsis thaliana Arabidopsis thaliana	AAA50156	AthCA_beta
	Helicobacter pylori Helicobacter pylori	BAF34127.1	HpyCA_beta
	Legionella pneumophila 2300/99	YP_003619232	LpnCA_beta
	Escherichia coli Escherichia coli	ACI70660	EcoCa_beta
	Methanobacterium thermoautotrophicum Escherichia coli	GI:13786688	Cab_beta
	Saccharomyces cerevisiae	GAA26059	SceCA_beta
	Dekkera bruxellensis AWRI1499	EIF49256	DbrCA_beta
gamma	Pseudomonas sp. PAMC 25886	ZP_10427314.1	PseCA_gamma
	Burkholderia gladioli BSR3	YP_004359911.1	BglCA_gamma
	Methanosarcina thermophila TM-1	ACQ57353.1	CAM_gamma
	Chlamydomonas reinhardtiiChlamydomonas reinhardtii	XP_001703237.1	CreCA_gamma
	Arabidopsis thaliana	NP_564091.1	AthCA_gamma

Most of these enzymes contain zinc ion (Zn²⁺) in their active site, coordinated by three histidine residues and a water molecule/hydroxide ion (in the α- and γ-classes) or by two Cys and one His residues (in the β-CA classes), with the fourth ligand being a water molecule/hydroxide ion8,12,22,23. Some of the catalytically active α-CAs can also catalyze the hydrolysis of esters, for example, 4-nitrophenyl acetate (4-NpA) and other hydrolytic reactions. Recently, CA enzymes were investigated in detail in pathogenic (as well as nonpathogenic) bacteria such as Brucella spp., Mycobacterium tuberculosis, Streptococcus spp., Helicobacter pylori, Salmonella enterica and Sulfurihydrogenibium spp.in the search for antibiotics with a novel mechanism of action, since it has been demonstrated in many of these organisms. CAs are essential for their life cycle8,21,24. For some of them, X-ray crystal structures of the encoded CAs were also determined, and in vitro and in vivo inhibition studies with various classes of inhibitors, such as anions, sulfonamides and sulfamates have been reported8,9,24–28. Although the bacterial CA inhibition studies are in their infancy at this moment, cloning of more bacterial genomes may lead to the discovery of genes and proteins, which may have interesting applications both in the biomedical and biotechnological fields8,9,21,24–26.

DHP synthase

DHPS activity was identified in E. coli by Brown29 and later in crude cell extracts of several organisms by Shiota et al.30,31. Humans lack an equivalent to DHPS, which is an advantage from the drug design viewpoint, since no toxicity to the host should emerge by using compounds targeting the bacterial/protozoan enzyme. DHPSs are inhibited by sulfonamide, which function as competitive inhibitor of one of the DHPS substrates, pABA 2. The DHPS encoded by the folP gene in Bacillus anthracis is also inhibited by sulfonamides32. The protein was shown to be a homodimer of two 30 kDa subunits. The chromosomal gene that codes for DHPS in Streptococcus pneumoniae was cloned, sequenced, and shown to code for a protein of 34 kDa33. A similar sequence was also identified in a Bacillus subtilis folic acid biosynthetic operon34. Two other genes (sulI and sulII) that code for plasmid-borne sulfonamide-resistant DHPSs have also been sequenced35,36. DHPS possesses a classic (β/α) TIM barrel structure, and the pABA binding site is at the edge of the barrel, being comprised of loop regions where the mutations that confer sulfonamide resistance are found37. In contrast, the other substrate of DHPS dihydropterin pyrophosphate (DHPP) 1, binds in a deep, structured pocket within the DHPS β-barrel. This pocket is highly conserved among various bacterial species37.

Dihydrofolate reductase

DHFR is the enzyme responsible for the nicotinamide adenine dinucleotide phosphate (NADPH)-dependent reduction of 5,6-DHF to 5,6,7,8-THF, an essential cofactor used in the biosynthetic pathways of purines, thymidylate, methionine, glycine, pantothenic acid and N-formyl-methionyl tRNA38,39. DHFR is ubiquitously expressed in all organisms, being found in all dividing cells of prokaryotes and eukaryotes. The mammalian enzymes are all highly similar in sequence, while each bacterial form is distinct. The DHFR sequence in humans is 30% similar to that of E. coli and 70% similar to other mammalian DHFRs40. Human DHFR is a small enzyme with a primary structure of 186 amino acid residues and a molecular weight of about 20 kDa40. Structural studies show that DHFR is a monomeric molecule with many secondary structural elements. The protein is divided into two subdomains, the adenosine-binding subdomain and the loop subdomain41. The adenosine-binding subdomain is the larger of the two and binds the adenosine moiety of NADPH. The loop subdomain contains three loops. Between the two subdomains is the active site, where folate and NADPH bind. The size of the active site is regulated by the movements of the two subdomains41–43. At the transcriptional level, DHFR is characterized by a TATA-less promoter that is controlled by numerous transcription factors, including Sp1 and E2F, which are important for their regulation throughout the cell cycle16,44. In addition to key roles in folate metabolism, DHFR can reduce 7,8-dihydrobiopterin (BH2) to 5,6,7,8-tetrahydrobiopterin (BH4)45, the natural cofactor required for the hydroxylation of aromatic amino acids, such as phenylalanine, tyrosine and tryptophan in higher animals46. Also, BH4 is an essential cofactor for nitric oxide synthesis involved in oxidative stress as a regulator of nitric oxide synthase or as a direct radical scavenger47. In bacteria, there are DHFR encoded within mobile elements (e.g. plasmids), which are rapidly spread within a bacterial community48. They are grouped into two main families, A and B48,49. The A gene (dfrA) family encodes proteins of 152–189 amino acids with identity levels of 20–90% and some structural and sequence similarities to the chromosomal enzyme. At least 20 different dfrA sequences have been reported so far50,51. The B gene (dfrB) family, by contrast, encodes a unique group of enzymes referred to as DfrB, which are completely unrelated to the chromosomal DHFRs or to other DHFRs in terms of sequence and structure. The dfrB genes encode similar and much shorter proteins (78 residues) with identity levels of 75% or above, which are extremely resistant to TMP52–54.

Clinically used inhibitors

Sulfa drug inhibitors

The primary sulfonamides were the first antimicrobial drugs, discovered in 1935 by Domagk, and they paved the way for the antibiotic revolution in medicine55. The first sulfonamide showing effective antibacterial activity, Prontosil 6, was in fact a prodrug, with the real antibacterial agent being sulfanilamide 7, a compound isosteric/isostructural with pABA 2. Sulfanilamide is generated by in vivo reduction of prontosil55,56. In fact the aniline moiety is common in both compounds 2 and 7, whereas the carboxyl moiety of pABA is mimicked by the primary sulfamoyl moiety of sulfanilamide. In the following years after the discovery of sulfanilamide as a bacteriostatic agent56, a range of analogs have entered into clinical use (constituting the so-called sulfa drug class of antibacterials), and many of these compounds are still widely used (mainly in combination with DHFR inhibitors)57.

The clinically used sulfa drugs (including prontosil which is no longer used) are shown in Figure 3, and they include Sulfamethoxazole 8, Sulfisomidine 9, Sulfacetamide 10 and Sulfathiazole 11 among others57,58. All these derivatives are competitive inhibitors (with pABA) of DHPS, the first enzyme in the biosynthesis of THF. However, significant resistance developed over the years to these drugs, which drastically limit their use in the clinic (see later in the text). In fact an intermediate obtained by reaction of the sulfa drug with DHP led to a covalent product, which has been characterized by means of X-ray crystallography (Figure 4)57.

Figure 3. Sulfa drugs in clinical use (7–11) and prontosil 6, the lead molecule generating this class of pharmacological agents. These compounds are structurally similar to pABA 2 and compete with this compound for the biosynthesis of DHP 3⁵⁸. Many other analogs are known⁵⁸.

Figure 4. Covalent adduct of DHPP with the sulfa drug sulfathiazole, of type 14, as determined by X-ray crystallography (A) and the detailed interactions in which 14 participates when bound to the enzyme (B)⁵⁸.

DHFR inhibitors

DHFR inhibitors such as TMP 12, are in clinical use since the 1960s54, whereas newer analogs (e.g., brodimoprim, tetroxoprim, iclaprim) were constantly discovered and developed57 (Figure 5). Such compounds compete with the pteridine moiety of DHF for binding to the enzyme (Figure 1) and thus impair the generation of THF, a compound crucial in many biosynthetic processes (as shown in the “Introduction” section). Sulfa drugs and DHFR inhibitors are commonly used in combination as antibacterial drugs, just to avoid the drug-resistance problems typical of these pharmacologic agents57,58.

Figure 5. Chemical structure of the DHFR inhibitor trimethoprim 12. Its congeners (brodimoprim, tetroxoprim, etc.) have slightly different moieties replacing the 4-methoxy group of 1257. The new DHFR inhibitor iclaprim 13 is also shown57.

CA inhibitors with antibacterial activity

One of the best studied bacterial α-CA is the one from the gastric pathogen provoking ulcer and gastric cancer, H. pylori, hpαCA59,60. The genome project of H. pylori identified, in fact, two different classes of CAs, with different subcellular localization: a periplasmic α-class CA (hpαCA) and a cytoplasmic β-class CA (hpβCA)59,60. These two CAs were shown to be catalytically efficient with almost identical activity to that of the human isoform hCA I, for the CO₂ hydration reaction, and highly inhibited by many sulfonamides/sulfamates, including acetazolamide, ethoxzolamide, topiramate and sulpiride, all clinically used drugs (Figure 6)59,60. Furthermore, certain CAIs, such as acetazolamide and methazolamide, were shown to inhibit the bacterial growth in cell cultures59,60. Since the efficacy of H. pylori eradication therapies currently employed has been decreasing due to drug resistance and side effects of the commonly used drugs, the dual inhibition of α- and/or β-CAs of H. pylori could be applied as an alternative therapy in patients with Helicobacter pylori infection or for the preve of gastroduodenal diseases caused by this widespread pathogen59,60.

Figure 6. Inhibition data of hpCA (α isoform), bsCA 1, bsCA 2 and VchCA with sulfonamides/sulfamates in clinical use12,13,59,61.

As mentioned in the introduction, the β-CA class is most widespread in bacteria8,10,21. The proof-of-concept study that such an enzyme may be a drug target has been published by Nishimori et al.59 who cloned and purified the H. pylori enzyme (hpβCA), showing that it is highly susceptible to be inhibited by sulfonamides and sulfamates, of the same type as those mentioned above for hpαCA (Figure 6).

Recently, Capasso’s group61,62 reported the purification and characterization of another interesting enzyme, the α-CA from V. cholerae (denominated VchCA), a Gram-negative bacterium, which is the causative agent of cholera. This bacterium colonizes the upper small intestine where sodium bicarbonate is present at a high concentration. Sodium bicarbonate is an inducer of virulence gene expression. V. cholerae utilizes the CA system to accumulate bicarbonate into its cells, suggesting a pivotal role of this metalloenzymes in the microbial virulence. VchCA showed a low esterase activity with 4-nitrophenyl acetate as substrate, and a high activity for the hydration of CO₂ to bicarbonate61,62. The inhibition study with sulfonamides and sulfamates led to the detection of a large number of low nanomolar inhibitors, among which are methazolamide, acetazolamide, ethoxzolamide, dorzolamide, brinzolamide, benzolamide and indisulam (K_I values in the range 0.69–8.1 nM) (Figure 6). As bicarbonate is a virulence factor of this bacterium and since ethoxzolamide was shown to inhibit the in vivo virulence, it was proposed that VchCA may be a target for antibiotic development, exploiting a mechanism of action rarely considered until now61,62. Some glycosylsulfanilamide derivatives (in addition to the sulfonamides/sulfamates shown in Figure 6) were also shown to have potent antibacterial effects, in vivo against Brucella suis, inhibiting the growth of the pathogen in cultures63. These compounds act as effective, nanomolar inhibitors of the two β-CAs present in this pathogen, bsCA 1 and bsCA 212,13.

Molecular mechanisms of drug resistance

The molecular mechanisms of drug resistance both to sulfa drugs58 and TMP-like DHFR inhibitors52,57 are understood in detail due to the fact that several X-ray crystal structures of DHPS and/or DHFR (both wild-type and mutated proteins) are available, in complex with substrates and/or inhibitors.

Resistance to sulfa drugs

Crystallographic work with Bacillus anthracis (BaDHPS) and Yersinia pestis (YpDHPS) DHPS, complexed with DHPP 1, pABA 2 and the pABA analog p-hydroxybenzoic acid, afforded a deep understanding of the catalytic, inhibition and resistance mechanisms to sulfa drugs of this enzyme58. First, the reaction between 1 and 2 occurs not via an S_N-2 like but via an S_N-1 reaction mechanism with formation of a DHP carbocation (at C-9) which then attacks the weakly nucleophilic amino group of pABA (or its analog p-hydroxybenzoic acid), with formation of DHP 3 (the adduct of 3 and BaDHPS has been isolated and characterized at atomic level). Furthermore, when BaDHPS was soaked in DHPP 1 and sulfathiazole 11, the covalent compound 14 (Figure 4) was observed in the X-ray crystal structure, bound within the active site of the enzyme and stabilized by the so-called loop 1–loop 2 substructure of the protein58. Figure 4 shows the detailed interactions between the covalent adduct 14 and the enzyme and also the amino acid residues involved in the sulfa drug resistance.

Indeed, the same crystallographic study from White’s group58 showed that in the YpDHPS – sulfa drug complexes, the sulfa drug perfectly fits the pABA binding pocket, with the SO₂ oxygens matching the COOH group of pABA, and the phenylene moieties of the two type of compounds being superposable and engaging the same hydrophobic pocket in the loop 1 – loop 2 substructure of the enzyme58. The resistance mutations in DHPS all cluster around this substructure, and comprise residues Phe33, Thr67 and Pro69. These amino acids form key elements for the pABA binding site. The sulfa drugs have been observed positioned (with their heterocyclic ring, which is attached to the SO₂NH moiety) outside the DHPS substrate envelope, being located in such a way that mutations at Phe33 and pro69 can impede the binding of the drug58. Thus, the presence of the heterocyclic ring (or the acetamido moiety in 10) in the sulfa drugs (which has no counterpart in the substrate molecule, pABA) leads to the generation of mutations within the DHPS active site and the consequent drug-resistance problems.

Resistance to TMP-like DHFR inhibitors

TMP is a synthetic drug that was introduced for clinical use in Western Europe in the early 1960s54. Although its structural similarity to DHF makes it a competitive inhibitor of the ubiquitous chromosomal DHFR in bacteria, fungi and protozoa, mammalian DHFRs are resistant to TMP54. Therefore, the specificity and selectivity of this antifolate drug have led to its widespread use in the treatment of human infections. However, bacteria have developed resistance mechanisms including the acquisition of plasmid-derived versions of DHFR, leading to high-level resistance or the point mutations of the chromosomal DHFR, which have been demonstrated to confer intermediate resistance52.

Just a single amino acid substitution, Phe98 to Tyr98, in DHFR was shown to be the molecular origin of TMP 12 resistance (in Staphylococcus aureus52). This active site amino acid substitution was found in all S. aureus TMP-resistant clinical isolates tested so far. Ternary complexes of the chromosomal S. aureus DHFR (SaDHFR) with methotrexate and TMP, in the presence of NADPH as well as that of mutant Phe98Tyr DHFR SaDHFR (F98Y) ternary folate-NADPH complex have been obtained and their structure determined by high-resolution X-ray crystallography52. Critical evidence concerning the resistance mechanism has also been provided by NMR spectral analyses of 15N-labelled TMP in the ternary complexes of both wild-type and mutant enzyme. These studies show that the mutation results in loss of a hydrogen bond between the 4-amino group of TMP and the carbonyl oxygen of Leu5. This mechanism of resistance is predominant in both transferable plasmid-encoded and non-transferable chromosomally encoded resistance52. Knowledge of the resistance mechanism at a molecular level could help in the design of antibacterials against multi-resistant Staphylococcus aureus (MRSA), one of today’s most serious problems in clinical infectology, which is, in fact, not yet resolved.

Resistance to CAIs

Resistance to the sulfonamide or sulfamate CAIs has not been reported up until now.

Clinical impact of resistance to sulfa drugs and DHFR inhibitors

The wide use (and misuse) of antibacterial treatment led to the selection for bacterial strains with physiologically or genetically enhanced capacity to survive high doses of antibacterials of several pharmacological classes including the DHFR inhibitors and the sulfa drugs52,58,64,65. The biochemical mechanisms by which these processes occur have been outlined earlier.

Co-trimoxazole, a combination of TMP 12 and sulfamethoxazole 8 (in the ratio of 1:5) is still widely clinically used, although significant resistance problems to both drugs emerged64,65. The synergistic effects of the two drugs are due to the fact they inhibit successive steps in folate synthesis (Figure 1).

Presently, sulfa drugs/DHFR inhibitors are primarily used for the treatment of uncomplicated urinary tract infections (cystitis) (mainly TMP-sulfa-methoxazole), this regimen being preferred to the quinolones (even though approximately 20% of E. coli are resistant to co-trimoxazole)66. Skin infections are sometimes treated with sulfacetamide or a sulfacetamide-silver(I) salt67. With its greater efficacy against a limited number of bacteria, co-trimoxazole remains indicated for some infections; for example, it is used as prophylaxis in patients at risk for Pneumocystis jirovecii pneumonia and as therapy in Whipple's disease64,65,68. Gram-positive bacteria are generally moderately susceptible to this combination therapy68. However, these drugs were not combined, as far as we know, with CAIs which specifically and potently inhibit bacterial CAs, as those discussed here. It would be of interest to investigate whether such a combination of three different enzyme inhibitors has a more potent antibacterial effect compared to the separate compounds used alone. We hypothesize that a synergistic effect between these different classes of pharmacological agents may have a beneficial and strong antibacterial effect against some pathogens.

Conclusions

Since the introduction of the sulfa drugs into clinical use in the 1940s, large numbers of antibiotics have been developed, possessing a wide range of mechanisms of action. However, their extensive use and misuse raised a serious public health problem due to the emergence of multi-antibiotic resistant bacterial pathogens worldwide. There is thus a stringent need to develop new antibiotics to keep pace with the bacterial resistance. Recent advances in microbial genomics, synthetic organic chemistry and X-ray crystallography provided opportunities to identify novel antibacterial targets for the development of new classes of antibiotics and to design more potent antimicrobial compounds derived from existing antibiotics in clinical use for decades, such as the sulfa drugs or the TMP-like DHFR inhibitors. Although resistance to these pharmacological agents constitutes a serious problem, sulfa drugs and DHFR inhibitors still have a firm place in the antibacterials’ armamentarium and may lead to new agents in which the resistance problem may be overcome by structured-based drug design. Furthermore, the recent cloning and characterization of many CAs in pathogenic bacteria, and the proof-of-concept studies showing that these targets are druggable and can lead to the inhibition of growth of these bacteria, offer interesting new alternatives that were not yet exploited clinically. Indeed, low nanomolar CAIs targeting the α- and/or β-CAs from many pathogenic bacteria are available, and further studies are necessary to see whether they may be synergistic with other antibacterials, such as the sulfa drugs or the DHFR inhibitors. Undoubtedly, these are fascinating research fields that can lead to interesting developments in the antibacterial drug research.

Declaration of interest

The authors declare no conflict of interest.

This research was financed by a grant of the “Accordo di Programma CNR-MSE” and by two EU projects (Metoxia and Dynano).

Acknowledgements

The authors are grateful to Dr Atilla Akdemir, Bezliaalem University, Istanbul, Turkey, for generating Figure 4 for us.