Inhibition studies of bacterial α-carbonic anhydrases with phenols

Abstract The α-class carbonic anhydrases (CAs, EC 4.2.1.1) from the bacterial pathogens Neisseria gonorrhoeae (NgCAα) and Vibrio cholerae (VchCAα) were investigated for their inhibition by a panel of phenols and phenolic acids. Mono-, di- and tri-substituted phenols incorporating additional hydroxyl/hydroxymethyl, amino, acetamido, carboxyl, halogeno and carboxyethenyl moieties were included in the study. The best NgCAα inhibitrs were phenol, 3-aminophenol, 4-hydroxy-benzylalcohol, 3-amino-4-chlorophenol and paracetamol, with KI values of 0.6–1.7 µM. The most effective VchCAα inhibitrs were phenol, 3-amino-4-chlorophenol and 4-hydroxy-benzyl-alcohol, with KI values of 0.7–1.2 µM. Small changes in the phenol scaffold led to drastic effects on the bacterial CA inhibitory activity. This class of underinvestigated bacterial CA inhibitors may thus lead to effective compounds for fighting drug resistant bacteria.


Introduction
Phenol (PhOH) was first investigated for its interaction with the metalloenzyme cabonic anhydrase (CA, EC 4.2.1.1) by Koenig's group in 1980 1 , and few years later Lindskog's group 2 demonstrated that this compound is one of the few competitive inhibitors (with CO 2 as substrate) of these enzymes, more precisely of the human (h) isoform hCA II. However, only in 1994 Christianson's group 3 elegantly demonstrated by using X-ray crystallography that phenol shows a new inhibition mechanism against this enzyme: it is anchored to the zinc-coordinated water molecule through hydrogen bonds involving the phenolic OH group, without displacing the zinc-bound water as other classes of inhbitors known at that time 3 . Subsequently, a large number of simple, synthetic and natural product phenols and polyphenols, incorporating various ring systems and possessing a range of substitution patterns were investigated for their inhibitory effects against all mammalian isoforms (CA I-XV) as well as for their interactions with pathogenic bacterial or fungal such enzymes 4-10 . For example dodoneine (Figure 1), a dihydropyranone phenolic compound isolated from the African mistletoe Agelanthus dodoneifolius 11 was demonstrated to possess significant CA inhibitory effects and to induce vasorelaxation through interference with calcium channels blockade and CA inhibition in the vascular smooth muscle cells 11 .
Such results prompted an intense research in the synthesis of phenolic derivatives incorporating diverse scaffolds (e.g. sugars, steroids, Mannich bases, etc.) 12, 13 , which have been tested for the inhibition of all hCAs as well as of enzymes belonging to nona-CA classes, such as b-, c-, d-, g-CAs from bacteria, fungi, algae, diatoms and protozoans [14][15][16] . Recent X-tay crystallographic of aspirin (hydrolyzed to salicylic acid) or caffeic acid bound to hCA II 17 also allowed a better understanding of the binding mode of such compound to the enzyme ( Figure 2) and to rationalise their inhibition mechanism with useful hints for the drug design of novel classes of CA inhibitors (CAIs) 18,19 . It may be observed that salicylic acid binds with its carboxylic acid moiety to the zinc-coordinated water molecule through a network of hydrogen bonds, with two water molecules being observed coordinated toi the metal ion 17a . Caffeic acid anchors with its catechol moiety to the zinc-coordinated water and to the deep water from the CA active site 17b - Figure 2.
Such data prompted us to investigate the interactions of a series of simple phenols and some of their derivatives with bacterial CAs which have recently been proposed 20,21 as novel drug targets for fighting the emergence of drug resistant bacteria, which no longer respond to clinically used antibiotics 22 . We included in the study NgCAa from Neisseria gonorrhoeae and VchCAa from Vibrio cholerae.

Enzymology and CA activity and inhibition measurements
An Applied Photophysics stopped-flow instrument was used to assay the CA-catalysed CO 2 hydration activity 23 . Phenol red (0.2 mM) was used as a pH indicator, working at the absorbance maximum of 557 nm, with 10 mM HEPES (pH 7.4) as a buffer, and in the presence of 10 mM NaClO 4 to maintain constant ionic strength, in order to follow the initial rates of the CA-catalysed CO 2 hydration reaction for a period of 10-100 s. The CO 2 concentrations ranged from 1.7 to 17 mM for the determination of the kinetic parameters and inhibition constants. For each inhibitor, at least six traces of the initial 5-10% of the reaction were used to determine the initial velocity. The uncatalyzed rates were determined in the same manner and subtracted from the total observed rates. Stock solutions of inhibitors (10-20 mM) were prepared in distilled-deionized water, and dilutions up to 10 nM were done thereafter with the assay buffer. Inhibitor and enzyme solutions were preincubated together for 15 min prior to the assay, in order to allow for the formation of the E-I complex. The inhibition constants were obtained by non-linear least-squares methods using Prism 3 and the Cheng-Prusoff equation, as reported previously 21,22 , and represent the mean from at least three different determinations. The NgCAa concentration in the assay system was 7.1 nM whereas the VchCAa was 10.3 nM. The used enzymes were recombinant proteins obtained in-house, as described earlier 22,24 .

Results and discussion
Bacterial CAs were investigated in the last decade for their inhibition with the several types of classical CAIs, among which the sulphonamides and their isosteres, the metal complexing anions and more recently also the coumarins [18][19][20][21][22] . However, no detaled inhibition data with many other classes of inhibitors, including phenols, are available so far in the literature for many of these enzymes, among which NgCAa and VchCAa belong 22,24 . The first of this enzyme was recently validated 21 as a potential drug target for developing antibiotics able to reverse or at least to alleviate the extensive drug resistance phenomenon that has emerged for this N. gonorrhoeae 25 . Although phenols usually act as weaker CAIs compared to other classes such as the sulphonamides or the sulfamates, we decided to investigate here the susceptibility of these enzymes to inhibition by a series of simple phenols and phenolic acids, of types 1-22 (Table 1) investigated earlier for their interaction with the human CA isoforms 4,6 .
The following structure activity relationship (SAR) can be evidenced from the inhibition data presented in Table 1: i. For NgCAa, the compounds 1-22 investigated here showed inhibitory activity, with K I -s in the range of 0.6-76.0 mM. The most effective, submicromolar inhibitors were 1, 6, 12, 13, and paracetamol 8, with K I values of 0.6-1.7 mM. These compounds are the simple phenol, its 4-hydroxymethyl derivative (12) as well as 3-amino-phenol or 4-acetamido moiety in derivatives (6, 8 and 13). The presence of a second/third Hbond donating moiety such as NH 2 or OH lead to slight loss of inhibitory activity K I values in the range of 1.7-6.0 mM (Table 1)    The active site zinc ion is shown as a grey sphere, and the water molecules as red spheres. Amino acid residues coordinating the metal ion or involved in inhibitor binding are also evidenced.
13.8-81.6 mM, Table 21. Again, as for the enzyme discussed above, small structural changes in the molecule of the inhibitor lead to significant sifferences of the inhibitory action (for example compare the three difluorinated phenols 15, 16 and 17, with one isomer, 16 being an effective inhibitor -K I of 4.9 mMand the remaining two acting as much weaker inhibitors, with K I -s of 36.6 and 57.3 mM, respectively). iii. The inhibition pattern of the two bacterial enzymes was generally rather distinct, although some compounds, such as 1, 12, 13 and 16 showed effective inhibition for both of them. However, in most cases, the NgCAa was more susceptible to phenol containing analogs than VchCAa (compare 2, 8, 19 or 21 for their K I s). These data points suggest that presumably it may be possible to obtain both phenols that show a wide action against various bacterial CAs, but also selective bacterial CA inhibitors belonging to the phenol and phenolic acid classes of derivatives. iv. There were relevant differences of inhibitory activity of the investigated phenols towards the human over the bacterial enzymes ( Table 1). Few of the investigated derivatives (e.g. 18,21,22) showed effective hCA I inhibitory activity while moderately inhibiting the bacterial enzymes. Most derivatives were on the other hand poor or ineffective hCA I inhibitors (e.g. 2,3,5,9,12,14,[15][16][17]. hCA II was effectively inhibited by 4, 9, 18 and 20-22, with few compounds being low activity or inactive as inhibitors (5, 7, 12, 14-17) and most of  them being moderale, micromolar inhibitors. However, even on this small panel of tested phenols, some compounds showed selectivity form inhibiting the bacterial over the human isoforms, such as for example 12, and to a lower extent 13. In fact 4-hydroxy-benzylalcohol has a selectivity index for inhibiting NgCAa over hCA I of 114.8 and over hCA II of 158.8. For VchCAa, the selectivity ratios are of 57.4 (bacterial enzyme over hCA I) and of 79.4 (bacterial enzyme over hCA II). It should be mentioned that benzylalcohol, a compound structurally similar to 12, was crystallised in adduct with hCA II, and it binds to the enzyme in a similar manner to phenol, by anchoring with the OH group to the zinc-coordinated water molecule 26 .

Conclusions
We report a study of bacterial CA inhibition with simple phenols and few phenolic acids. Two a-class enzymes from bacterial pathogens which developed drug resistance to classical antibiotics, i.e. NgCAa from Neisseria gonorrhoeae and VchCAa from Vibrio cholerae were included in the study. The panel of 22 phenols and phenolic acids inhibited both enzymes with K I values in the range of 0.6-76.0 mM for NgCAa, and of 0.6-76.0 mM for VchCAa. The best NgCAa inhibitors were phenol, 3-aminophenol, 4-hydroxybenzylalcohol, 3-amino-4-chlorophenol and paracetamol, with K I s of 0.6-1.7 mM. The most effective VchCAa inhibitors were phenol, 3-amino-4-chlorophenol and 4-hydroxy-benzyl-alcohol, with K I -s of 0.7-1.2 mM. Small changes in the phenol scaffold led to drastic effects on the bacterial CA inhibitory activity. This class of underinvestigated bacterial CA inhibitors may thus lead to effective compounds for fighting drug resistant bacteria.

Disclosure statement
CT Supuran is Editor-in-Chief of the Journal of Enzyme Inhibition and Medicinal Chemistry. He was not involved in the assessment, peer review, or decision-making process of this paper. The authors have no relevant affiliations of financial involvement with any organisation or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Funding
The research program was also partially funded by a Purdue Institute for Drug Discovery Programmatic Grant (to D.P.F.) and NIH/NIAID 1R01AI148523 (to D.P.F.).