Novel Therapeutic Applications for Inhibitors and Activators

REVIEWS

Carbonic anhydrases: novel therapeutic applications for inhibitors and activators

Claudiu T. Supuran Abstract | Carbonic anhydrases (CAs), a group of ubiquitously expressed metalloenzymes, are involved in numerous physiological and pathological processes, including gluconeogenesis, lipogenesis, ureagenesis, tumorigenicity and the growth and virulence of various pathogens. In addition to the established role of CA inhibitors (CAIs) as diuretics and antiglaucoma drugs, it has recently emerged that CAIs could have potential as novel anti-obesity, anticancer and anti-infective drugs. Furthermore, recent studies suggest that CA activation may provide a novel therapy for Alzheimer’s disease. This article discusses the biological rationale for the novel uses of inhibitors or activators of CA activity in multiple diseases, and highlights progress in the development of specific modulators of the relevant CA isoforms, some of which are now being evaluated in clinical trials.

Carbonic anhydrases (CAs; also known as carbonate Many of the CA isozymes involved in these processes Glaucoma A chronic, degenerative eye dehydratases EC 4.2.1.1) are ubiquitous metalloenzymes are important therapeutic targets with the potential to be disease, characterized by high present in prokaryotes and eukaryotes that are encoded inhibited to treat a range of disorders including oedema, intraocular pressure that by four evolutionarily unrelated gene families. These are glaucoma, obesity, cancer, epilepsy and osteoporosis. causes irreversible damage to the α‑CAs (present in vertebrates, bacteria, algae and Two main classes of CA inhibitors (CAIs) are known: the optic nerve head, resulting in the progressive loss of cytoplasm of green plants); the β‑CAs (predominantly the metal-complexing anions and the unsubstituted visual function and eventually in bacteria, algae and chloroplasts of monodicotyle‑ sulphonamides and their bioisosteres — for example, blindness. dons and dicotyledons); the γ‑CAs (mainly in archaea sulphamates and sulphamides compounds 1–25 (FIG. 1; and some bacteria); and the δ‑CAs (present in some TABLE 1). These inhibitors bind to the Zn2+ ion of the marine diatoms)1–7. In mammals, 16 α‑CA isozymes or enzyme either by substituting the non-protein zinc CA‑related proteins have been described (TABLE 1), with ligand to generate a tetrahedral adduct or by addition different catalytic activity, subcellular localization and to the metal coordination sphere to generate a trigonal- tissue distribution8–18. There are five cytosolic forms (CA I, bipyramidal species1–7 (FIG. 2). At least 25 clinically used CA II, CA III, CA VII and CA XIII), five membrane- drugs have been reported to possess significant CA bound isozymes (CA IV, CA IX, CA XII, CA XIV and inhibitory properties (discussed below), in addition to CA XV), two mitochondrial forms (CA VA and CA VB), many other derivatives belonging to the sulphonamide, and a secreted CA isozyme (CA VI)19–25. sulphamate or sulphamide families1–3,5,7,8,19–25. CAs catalyse a simple physiological reaction (BOX 1): Furthermore, the potential use of CAIs to fight infec‑

the conversion of CO2 to the bicarbonate ion and pro‑ tions caused by protozoa, fungi and bacteria has recently Laboratorio di Chimica tons. The active site of most CAs contains a zinc ion emerged as a new research direction. CAs belonging Bioinorganica, Università 2+ degli Studi di Firenze, (Zn ), which is essential for catalysis. The CA reaction to various families were cloned and characterized in Rm 188, Via della is involved in many physiological and pathological pro many such organisms (such as Plasmodium falciparum,

Lastruccia 3, I‑50019 Sesto cesses, including respiration and transport of CO2 and Helicobacter pylori, Mycobacterium tuberculosis, Fiorentino (Firenze), Italy. bicarbonate between metabolizing tissues and lungs; Candida albicans and Cryptococcus neoformans)26–34 and e‑mail: pH and CO homeostasis; electrolyte secretion in vari‑ have been shown to be crucial for the virulence, growth [email protected] 2 doi:10.1038/nrd2467 ous tissues and organs; biosynthetic reactions (such as or acclimatization of the parasite. In addition, impor‑ Published online gluconeogenesis, lipogenesis and ureagenesis); bone tant advances have been achieved in the understanding 2 January 2008 resorption; calcification; and tumorigenicity8–18. of CA activation by several classes of activators35–41.

REVIEWS

Table 1 | Inhibition data with selected sulphonamides/sulphamates/sulphamides 1–25 against isozymes I–XIV*

KI Isozyme (h = human, m = mouse) (nm) hCA I‡ hCA II‡ hCA III‡ hCA IV‡ hCA VA‡ hCA VB‡ hCA VI‡ hCA VII‡ hCA IX§ hCA XII§ mCA XIII‡ hCA XIV‡ 1 250 12 2 × 105 74 63 54 11 2.5 25 5.7 17 41 2 50 14 7 × 105 6,200 65 62 10 2.1 27 3.4 19 43 3 25 8 1 × 106 93 25 19 43 0.8 34 22 50 2.5 4 374 9 6.3 × 105 95 81 91 134 6 43 56 1,450 1,540 5 1,200 38 6.8 × 105 15,000 630 21 79 26 50 50 23 345 6 50,000 9 7.7 × 105 8,500 42 33 10 3.5 52 3.5 18 27 7 45,000 3 1.1 × 105 3,950 50 30 0.9 2.8 37 3.0 10 24 8 31 15 10,400 65 79 23 47 122 24 3.4 11 106 9 250 10 7.8 × 105 4,900 63 30 45 0.9 58|| 3.8 47 1,460 10 56 35 2.2 × 106 8,590 20 6,033 89 117 5.1 11,000 430 5,250 11 12,000 40 10,600 6.5 × 105 174 18 0.8 3,630 46 3.9 295 110 12 3,450 21 7.0 × 105 24 765 720 653 23 34 12 1,050 755 13 37 10 6.5 × 105 NT NT NT NT NT 30 7.5 NT NT 14 50,000 21 7.4 × 104 880 794 93 94 2,170 16 18 98 689 15 54,000 43 7.8 × 104 1,340 912 88 572 3,900 27 13 425 107 16 18,540 5,950 1.0 × 106 7,920 10,060 7,210 935 10 103 633 12,100 773 17 1,300 45 1.3 × 106 650 134 76 145 18 24 5 76 33 18 4,000 21 3.1 × 105 60 88 70 65 15 14 7 21 13 19a 328 290 7.9 × 105 427 4,225 603 3,655 5,010 367 355 3,885 4,105 20 35,000 1,260 NT NT NT NT NT NT NT NT NT NT 21 54,000 2,000 6.1 × 105 216 750 312 1,714 2.1 320 5.4 15 5,432 22 348 138 1.1 × 104 196 917 9 1,347 2.8 23 4.5 15 4,130 23 51,900 2,520 2.3 × 105 213 890 274 1,606 0.23 36 10 13 4,950 24 62 65 3.2 × 106 564 499 322 245 513 420 261 550 52 25 4,930 6,980 3.4 × 106 303 700 NT NT NT 25.8 21.2 2,570 250 *The isoforms CA VIII, X and XI are devoid of catalytic activity and probably do not bind sulphonamides as they do not contain Zn2+ ions. ‡Full-length enzyme. §Catalytic domain. ||The data against the full-length enzyme is of 1,590 nM. NT, not tested, data not available.

Such compounds might lead to pharmacological agents compounds 19a–e, as well as derivatives 20–25 are still that have the potential to treat Alzheimer’s disease, ageing widely clinically used2,3 (FIG. 1). However, some of these and other conditions involving memory deficits40. enzyme inhibitors could also be used for the systemic treatment of glaucoma (see below), and more recently, Carbonic anhydrase inhibitors newer derivatives have been discovered that have the CAIs include the classical inhibitors acetazolamide (com‑ potential as topical antiglaucoma agents, as well as anti‑ pound 1), methazolamide (compound 2), ethoxzolamide tumour, anti-obesity or anti-infective drugs1–3,5,7,9–26. (compound 3), sulthiame (compound 4) and dichloro The inhibitory effects of some of these clinically phenamide (compound 5). CAIs also include more used drugs against the mammalian isoforms CA I–XIV recent drugs/investigational agents such as dorzolamide — of human or mouse origin — are shown in TABLE 1. (compound 6), brinzolamide (compound 7), indisulam As specific isozymes are responsible for different bio‑ (compound 8), topiramate (compound 9), zonisamide logical responses, the diverse inhibition profiles of the (compound 10), sulpiride (compound 11), COUMATE various isozymes may explain the different actual and (compound 12), EMATE (compound 13), celecoxib potential clinical applications of the CAIs, which range (compound 14), valdecoxib (compound 15) and saccharin from diuretics and antiglaucoma agents, to anticancer, (compound 16) (FIG. 1; TABLE 1). Derivatives 17 and 18 are anti-obesity and anti-epileptic drugs. However, a crucial investigational agents for targeting the tumour-associated problem in CAI design is related to the high number of isoform CA IX (see later in the text). Many of these isoforms, their diffuse localization in many tissues and compounds were initially developed years ago during organs (TABLE 2), and the lack of isozyme selectivity of the search for diuretics, among which the thiazides, the presently available inhibitors. It can be observed that

REVIEWS

Box 1 | Mechanism of action of carbonic anhydrases there are sulphonamide- and sulphamate-avid isoforms, such as CA II, VI, VII, IX, XII and XIII, which generally Carbonic anhydrases (CAs) catalyse the following reaction: show low nanomolar affinity for most of these inhibitors. Other isozymes, however, such as CA I, IV, VA, VB and – + CO2 + H2O HCO3 + H (1) XIV show less propensity to be inhibited by these com‑ pounds, with inhibition constants in the nanomolar to The metal ion (which is a Zn2+ ion in all α‑CAs investigated up to now) is essential for catalysis1–8. X‑ray crystallographic data show that the ion is situated at the bottom of a micromolar range. This leaves CA III as the only isoform 15 Å deep active-site cleft (shown as a pink sphere in the figureNature below; Reviews left | Drug panel), Disc andove ry that is not susceptible to inhibition by sulphonamides or is coordinated by three histidine residues (His94, His96 and His119; shown in green in sulphamates. the figure) and a water molecule/hydroxide ion1–8. The histidine cluster involved Few of the derivatives 1–16 (FIG. 1; TABLE 1) show in the proton-shuttling processes between the active site and the environment, selectivity for a specific isoform: the classical inhibi‑ comprising residues His64, His4, His3, His17, His15 and His10, is also evidenced. tors, such as compounds 1–5, and the topically acting Amino-acid residues 92 and 131 involved in the binding of many sulphonamide/ antiglaucoma sulphonamides 6 and 7 together with sulphamate inhibitors are shown in yellow2,3. indisulam 8, are promiscuous CAIs, with strong affini‑ The zinc-bound water is also engaged in hydrogen-bond interactions with the ties for isoforms II, VA, VB, VI, VII, IX, XII, XIII and hydroxyl moiety of Thr199, which in turn is bridged to the carboxylate moiety of 7–18 XIV . Topiramate (compound 9) is a subnanomolar Glu106; these interactions enhance the nucleophilicity of the zinc-bound water CA VII inhibitor but it also effectively inhibits isoforms molecule, and orient the substrate (CO2) in a favourable location for nucleophilic 7–18 attack1–8 (figure, right panel). The active form of the enzyme is the basic one, with II, VB and XII . Zonisamide (compound 10) shows 2+ a good affinity for CA IX but appreciably inhibits also hydroxide bound to Zn (a). This strong nucleophile attacks the CO2 molecule that is bound in a hydrophobic pocket in its neighbourhood (the elusive substrate-binding CA II and VA, while having lower affinities for the other site comprises residues Val121, Val143 and Leu198 in the case of the human isozyme isoforms. Sulpiride (compound 11) is a potent CA VI CA II) (b), leading to the formation of bicarbonate coordinated to Zn2+ (c). The and CA XII inhibitor and shows lower affinity for other bicarbonate ion is then displaced by a water molecule and liberated into solution, isoforms. Valdecoxib (compound 15) is a strong CA XII leading to the acid form of the enzyme, with water coordinated to Zn2+ (d), which is inhibitor7–18, whereas saccharin (compound 16) is a CA catalytically inactive1–8. VII-specific inhibitor (KI of 10 nM against this isoform and much higher for the other CAs)19. His10 O - - OH OH O Progress in the design of CA‑selective and isozyme- + specific CAIs has recently been made. Owing to the His15 2+ CO2 Zn Zn2+ His119 His119 extracellular location of some CA isozymes, such as His17 His94 His94 His3 His96 His96 CA IV, IX, XII and XIV (TABLE 2), it is possible to design a b membrane-impermeant CAIs, which would therefore specifically inhibit membrane-associated CAs without interacting with the cytosolic or mitochondrial isoforms. + His4 – BH B This possibility has been explored through the design His96 of positively charged sulphonamides that generally His64 OH2 O H incorporate pyridinium moieties, of which compound + H2O 20–22 2+ O Zn - 18 is a representative . The inhibitors obtained in this His94 His119 O His119 - + way showed nanomolar affinities for CA II as well as His94 – HCO3 Zn2 His96 His119 CA IV and CA IX, and, more importantly, they were His94 His96 unable to cross the plasma membranes in vivo20–22. This d c Gln92 new class of potent, positively charged CAIs, was able Phe131 to discriminate between the membrane-bound and the cytosolic isozymes, selectively inhibiting only CA IV, in To regenerate the basic form (a), a proton transfer reaction from the active site to the two model systems20–22. Another approach for the design environment takes place, which may be assisted either by active-site residues (such as of isoform-selective CAIs exploited the presence of an His64, the proton shuttle in isozymes I, II, IV, VI, VII, IX and XII–XIVNature amongReviews others) | Drug Discor byov ery Ala65 amino-acid residue, which is present only in the buffers present in the medium. The process may be schematically represented by 23 reactions 2 and 3: ubiquitous CA II mammalian isoform . Compared with topiramate, its sulphamide analogue is a 210-times less H2O potent inhibitor of isozyme CA II, but effectively inhibits E–Zn2+–OH– E–Zn2+–HCO – E–Zn2+–OH + HCO – (2) 3 2 3 isozymes CA VA, VB, VII, XIII and XIV (KIs in the range of 21–35 nM). The weak binding of the sulphamide E–Zn2+–OH2 E–Zn2+–HO– + H+ (3) analogue of topiramate to CA II was shown to be due to Nature Reviews | Drug Discovery a clash between one methyl group of the inhibitor with the Ala65 amino-acid residue, which might therefore be The rate-limiting step in catalysis is reaction 3, that is, the proton transfer that exploited for the design of compounds with lower affinity regenerates the zinc-hydroxide species of the enzyme1–8. In the catalytically very active Nature Reviews | Drug Discovery 23 isozymes, such as CA II, IV, VI, VII, IX, XII, XIII and XIV, the process is assisted by a histidine for this isoform . residue placed at the entrance of the active site (His64), or by a cluster of histidines A further approach for selectively inhibiting the (figure, left panel), which protrudes from the rim of the active site to the surface of the tumour-associated isoforms CA IX (and XII) present enzyme, thus assuring efficient proton-transfer pathways1–8. This may explain why CA II is in hypoxic tumour tissues envisaged bioreductive 8 –1 –1 24,25 one of the most active enzymes known (with a kcat/Km of 1.5 x 10 M s ), approaching the prodrugs that are activated by hypoxia . The chosen limit of diffusion-controlled processes1–8. strategy was to use the disulphide bond as a bioreducible

REVIEWS

HC3 N N N N N N SO2NH2 S CH3CONH SO2NH2 S CH3CON SO NH EtO S SO NH S 2 2 2 2 O O 1 2 3 4

SO NH 2 2 NHEt NHEt

SO2NH2 SO2NH2 N Cl SO2NH2 Me S S MeO(CH2)3 S S Cl OO OO 5 6 7

O NH 2 OMe O O S O O NH O O SO2NH2 S O N Cl HN H N N O O SO2NH2 O O SO2NH2 8 9 10 11

O O O S NH2 O O S NH2

CH3 N O O O N O H3C SO2NH2 SO2NH2 F O N 12 13 F F 14 15

SO NH SO2NH2 S 2 2

HN N O H

- NH ClO4 S COOH + OO N

HO O O 16 17 18

H H H R3 N R6 Et N Cl Me N Cl Me

N HN N 2 SO NH R S SO2NH2 SO2NH2 2 2 OO O O 19 20 21 a R2 = R3 = H, R6 = Cl, Hydrochlorothiazide 2 3 6 b R = R = H, R =CF3 , Hydroflumethiazide 2 3 6 c R =H, R = PhCH2, R = CF3, Bendroflumethiazide 2 3 6 d R = H, R = CHCl2, R = Cl, Trichloromethiazide 2 3 6 e R = Me, R = CH2SCH2CF3, R = Cl, Polythiazide

Cl Cl Me H NH H N H N N Cl O O SO2NH2 N SO2NH2 O OH O

HOOC SO2NH2 HOOC SO2NH2

22 23 24 25

REVIEWS

function. The reducing conditions present in hypoxic a O R tumours, in combination with the presence of the redox - S HN protein thioredoxin 1, mediates the reduction of the O Tetrahedral adduct disulphide bond with the formation of thiols24,25. The Zn 2 + (sulphonamide) His119 His94 reduced compounds (thiols) are less bulky and show His96 excellent CA inhibitory activity (in the low nanomolar 2+ 2+ E–Zn –OH2 + I E–Zn –I + H2O range) compared with the corresponding sterically hin‑ (substitution reaction) dered disulphides, which have difficulty entering the 24,25 limited space of the enzyme active site . b S C - OH CAIs as diuretics N Trigonal-bipyramidal adduct (thiocyanate) CAs are highly abundant in the kidney, and the isoforms Zn 2+ His119 present in this organ play a crucial function in at least His94 His96 three physiological processes: the acid–base homeostasis 2+ 2+ E–Zn –OH2 + I E–Zn –OH2(I) balance (by secreting and excreting protons, due to the (addition reaction) CO2 hydration reaction catalysed by these enzymes); the + bicarbonate reabsorption process; and the NH4 out‑ Figure 2 | Mechanisms of inhibition of carbonic 42,43 put . Acetazolamide (compound 1) was the first non- anhydrase. a | UnsubstitutedNatur sulphonamidese Reviews | Drug and Disc theirovery mercurial diuretic to be used clinically in 1956 (Ref. 2). bioisosteres bind to the Zn2+ ion of the enzyme by It represents the prototype of a class of pharmacological substituting the non-protein zinc ligand to generate agents with relatively limited therapeutic use, but which a tetrahedral adduct. b | Anionic inhibitors add to the metal coordination sphere, generating trigonal- played a major role in the development of fundamental bipyramidal species. renal physiology and pharmacology, and in the design of many of the current widely used diuretic agents, such as the thiazide and the high-ceiling (loop) diuretic. Following the administration of a CAI, such as acetazolamide, the urine of this secretion to be sodium bicarbonate1,2,44–46. CAs volume increases and becomes alkaline2,42. Increased were identified in the anterior uvea of the eye and were bicarbonate is eliminated into the urine, together with shown to be responsible for the bicarbonate secretion. Na+ and K+ as accompanying cations, whereas the CAIs represent the most physiological treatment of amount of chloride excreted is diminished. This sequence glaucoma, as by inhibiting the ciliary-process enzyme of events is due to the inhibition of CA in the proximal — the sulphonamide susceptible isozyme CA II (TABLE 1) tubule, which leads to the inhibition of H+ secretion — the rate of bicarbonate and aqueous humour secretion by this segment of the nephron. Inhibition of cytosolic is reduced, resulting in a 25–30% decrease in IOP44–51. (CA II) and membrane-bound (CA IV, XII and CA XIV) Indeed, systemic acetazolamide (compound 1), meth‑ enzymes seems to be involved in the diuretic effects of azolamide (compound 2), ethoxzolamide (compound 3) the sulphonamides2,43. The net effect of these processes or dichlorophenamide (compound 4) are extensively is the transport of sodium bicarbonate from the tubular used to treat this disease2,45. The best-studied drug is lumen to the interstitial space, followed by movement of acetazolamide, which is frequently administered long- the isotonically obligated water, and augmented diuresis. term owing to its efficient reduction of IOP, minimal Acetazolamide, methazolamide (compound 2), ethox‑ toxicity and ideal pharmacokinetic properties. However, zolamide (compound 3) and dichlorophenamide (com‑ as CAs are ubiquitously expressed in vertebrates, the pound 5) are used to treat oedema due to congestive systemic administration of sulphonamides leads to heart failure and for drug-induced oedema42,43. Many nonspecific CA inhibition and is associated with unde‑ of the other diuretics, such as the benzothiadiazines sired side effects, including numbness and tingling of (compounds 19a–e; for example, chlorothiazide and extremities; metallic taste; depression; fatigue; malaise; hydrochlorothiazide); quinethazone (compound 20); weight loss; decreased libido; gastrointestinal irritation; metolazone (compound 21); chlorthalidone (compound 22); metabolic acidosis; renal calculi; and transient myo‑ indapamide (compound 23); furosemide (compound 24); pia. The development of water-soluble sulphonamide and bumetanide (compound 25) act as CA inhibi‑ CAIs to be used as eye drops began in the 1990s, and tors with varying efficiencies2,3,42 (FIG. 1). This is to be by 1995 the first such pharmacological agent, dorzola‑ expected, as all of them have unsubstituted primary mide (compound 6), was launched45 by Merck under sulphonamide moieties acting as effective zinc-binding the tradename Trusopt. A second structurally similar High-ceiling (loop) diuretic moieties19. compound, brinzolamide (compound 7), has also been Diuretics are drugs that approved (by Alcon under the tradename Azopt) for the increase the rate of urine 45 excretion. A high-ceiling (loop) CAIs as drugs for eye disorders topical treatment of glaucoma . diuretic inhibits the kidney’s Glaucoma is a chronic, degenerative eye disease, char‑ Dorzolamide and brinzolamide are potent water-soluble ability to reabsorb sodium in acterised by high intraocular pressure (IOP) that causes CAIs that are sufficiently liposoluble to penetrate the the ascending loop. This leads irreversible damage to the optic nerve head, resulting cornea, and may be administered topically as the hydro‑ to an increased loss of sodium and water in the urine, as water in the progressive loss of visual function and eventually chloride salt (at a pH of 5.5) or as the free base, respec‑ 44–46 45 normally follows sodium back blindness . Studies on the chemistry and dynamics tively . The two drugs are effective in reducing IOP and into the extracellular fluid. of aqueous humour have identified the main constituent show fewer side effects as compared with systemically

REVIEWS

Table 2 | Kinetic parameters for CO2 hydration reaction catalysed by the 16 vertebrate α-CA isozymes* –1 –1 –1 Isozyme Kcat (s ) Km (mM) Kcat/Km (M s ) KI (nM) Subcellular Tissue/organ Refs localization localization hCA I 2.0 × 105 4.0 5.0 × 107 250 Cytosol Erythrocytes, GI tract 1–3 hCA II 1.4 × 106 9.3 1.5 × 108 12 Cytosol Erythrocytes, eye, 1–3 GI tract, bone osteoclasts, kidney, lung, testis, brain hCA III 1.0 × 104 33.3 3.0 × 105 2.1 × 105 Cytosol Skeletal muscle, 7,8 adipocytes hCA IV 1.1 × 106 21.5 5.1 × 107 74 Membrane- Kidney, lung, 6 bound pancreas, brain capillaries, colon, heart muscle hCA VA 2.9 × 105 10.0 2.9 × 107 63 Mitochondria Liver 9 hCA VB 9.5 × 105 9.7 9.8 × 107 54 Mitochondria Heart and skeletal 10 muscle, pancreas, kidney, spinal cord, GI tract hCA VI 3.4 × 105 6.9 4.9 × 107 11 Secreted Salivary and 11 (saliva, milk) mammary glands hCA VII 9.5 × 105 11.4 8.3 × 107 2.5 Cytosol CNS 12 hCA VIII ND ND ND ND Cytosol CNS 13 hCA IX 3.8 × 105 6.9 5.5 × 107 25 Transmembrane Tumours, GI mucosa 14,18 hCA X ND ND ND ND Cytosol CNS 13 hCA XI ND ND ND ND Cytosol CNS 13 hCA XII 4.2 × 105 12.0 3.5 × 107 5.7 Transmembrane Renal, intestinal, 15 reproductive epi- thelia, eye, tumours hCA XIII 1.5 × 105 13.8 1.1 × 107 16 Cytosol Kidney, brain, 16 lung, gut, reproductive tract hCA XIV 3.1 × 105 7.9 3.9 × 107 41 Transmembrane Kidney, brain, liver 17 mCA XV 4.7 × 105 14.2 3.3 × 107 72 Membrane- Kidney Unpublished bound observations *At 20°C and pH 7.5, their inhibition data with acetazolamide 1 (5-acetamido-1,3,4-thiadiazole-2-sulphonamide), and their subcellular localization. α-CA, α-carbonic anhydrase; h, human; GI, gastrointestinal; m, mouse; ND, not determined.

applied drugs. The observed side effects include sting‑ The use of CAIs in the treatment of macular oedema is ing, burning or reddening of the eye, blurred vision and based on the observation that systemically administered pruritus, which are probably due to the acidic pH of acetazolamide (sodium salt) is effective in the treatment the dorzolamide eyedrops solution. Also, a bitter taste of this condition52. Similar efficiency has also recently is experienced with both systemic as well as topical been reported for topical administration of dorzolamide CAIs, which is probably due to drug-laden lachrymal and brinzolamide45. It is generally assumed that the dis‑ fluid draining into the oropharynx and inhibiting CAs appearance of oedema and improvement of visual func‑ present in the saliva (CA VI) and taste buds (CA II and tion are independent of the hypotensive activity of the CA VI) with the consequent accumulation of bicarbo‑ sulphonamide, due to direct effects on the circulation in nate. Therefore, novel, topically effective CAIs as possible the retina. Acetazolamide, dorzolamide or brinzolamide antiglaucoma agents are being investigated46. One recent probably act as local vasodilators, improving blood flow approach involved attaching water-solubilizing moieties in this organ and consequently clearing metabolic waste and ring-system derivatives to aromatic and heterocyclic products and drusen. Vision following such treatment (in sulphonamides, which produced compounds that were early phases of the disease) is markedly improved45,52. 53 Macular oedema two to three times more effective than dorzolamide in Gao et al. recently reported that the slow cytosolic Swelling of the retina due to lowering IOP in rabbits46–51. These compounds were isozyme CA I mediates haemorrhagic retinal and the leakage of fluid from blood water-soluble (as hydrochlorides, triflates or trifluoro‑ cerebral vascular permeability through activation of vessels within the central macula region of the retina, acetates), potent human CA II inhibitors, penetrated prekallikrein and generation of the highly active serine causing blurred vision and the cornea and significantly lowered IOP in both normo protease factor XIIa. These phenomena contribute to loss of visual function. tensive and glaucomatous rabbits46–51. the pathogenesis of proliferative diabetic retinopathy

REVIEWS

Mitochondrion Cytosol overexpressed in the eyes of patients with glaucoma, and probably plays an important role in the elevated IOP 57 [Pyruvate]c characteristic of the disease . We have subsequently shown that CA XII is highly inhibited by all the antiglau‑ – [Pyruvate] – CO 15 HCO3 m 2 coma sulphonamides used clinically (TABLE 1), and this PC Oxaloacetate – CO Oxaloacetate is probably the membrane-bound isoform involved in 2 glaucoma that is targeted by these agents. Further work Acetyl-CoA is warranted to better understand the involvement of various CA isoforms in eye pathologies such as glaucoma,

[Citrate]m [Citrate]c retinopathy and macular degeneration.

CoA CAIs as potential anti-obesity drugs α Krebs cycle Acetyl-CoA Among the ‑CA isoforms found in animals, two CA isozymes, VA and VB, are present in mitochondria (TABLE 2). These isozymes are involved in several biosyn‑ ACC HCO – 3 thetic processes, such as ureagenesis, gluconeogenesis and lipogenesis, in vertebrates (for example, rodents) Malonyl-CoA and in invertebrates (for example, the locust)58. The pro‑

Pyruvate transporter – CO2 vision of enough of the substrate, bicarbonate, in several Tricarboxylate transporter biosynthetic processes involving pyruvate carboxylase Fatty acids (PC), acetyl-CoA carboxylase (ACC) and carbamoyl Figure 3 | Fatty-acid biosynthesis and the role of carbonic anhydrase isozymes. phosphate synthetases I and II, is assured mainly by the Mitochondrial pyruvate carboxylase (PC) is needed for the efflux of acetyl groups from catalytic reaction involving the mitochondrial isozymes the mitochondria to the cytosol where fatty-acid biosynthesisNatur takese Revie placews | Drug1,58. Pyruvate Discover y CA VA and VB — probably assisted by the high activity is carboxylated to oxaloacetate in the presence of bicarbonate under the catalytic cytosolic isozyme CA II58 (FIG. 3). influence of the mitochondrial isozymes (carbonic anhydrase (CA) VA and/or CA VB). Several studies have provided evidence that CAIs have The mitochondrial membrane is impermeable to acetyl-CoA, which reacts with potential as anti-obesity drugs, which might be due to oxaloacetate, leading to the formation of citrate, which is then translocated to the their effects on CA isozymes. Topiramate (compound 9) cytoplasm by means of the tricarboxylic acid transporter. As oxaloacetate is unable to is an anti-epileptic drug possessing potent anticonvulsant cross the mitochondrial membrane, its decarboxylation regenerates pyruvate, which effects due to a multifactorial mechanism of action: block‑ can then be transported into the mitochondria by means of the pyruvate transporter. The acetyl-CoA thus generated in the cytosol is in fact used for de novo lipogenesis, by ade of sodium channels and kainate/AMPA (α-amino-3- carboxylation in the presence of acetyl-CoA carboxylase (ACC) and bicarbonate, with hydroxy-5-methyl-4-isoxazole propionic acid) receptors, CO retention secondary to inhibition of the red blood formation of malonyl-CoA, the conversion between CO2 and bicarbonate being assisted 2 by CA II. Subsequent steps involving the sequential transfer of acetyl groups lead to cell and brain CA isozymes, as well as enhancement of longer-chain fatty acids. Therefore, CA isozymes are critical to the entire process of fatty- GABA (γ-aminobutyric acid)-ergic transmission59. A side acid biosynthesis: VA and/or VB within the mitochondria (to provide enough substrate to effect of this drug observed in obese patients was the PC), and CA II within the cytosol (for providing sufficient substrate to ACC). loss of body weight, although no pharmacological explanation of this phenomenon has been provided60. Furthermore, topiramate was shown to reduce energy and diabetic macular oedema, which represent leading and fat gain in lean (Fa/?) and obese (fa/fa) Zucker rats60. causes of vision loss, and for which there are no phar‑ It was recently demonstrated that topiramate is also a macological treatments currently available. Therefore, potent inhibitor of several CA isozymes, such as II, VA, as suggested by the authors, CA I inhibition might be a VB, VI, VII, XII and XIII (TABLE 1), and the X‑ray crystal therapeutic target for the treatment of these conditions53. structure of its complex with human CA II has been In fact, potent CA I inhibitors are currently available54. determined, revealing the molecular interactions that Some of the membrane-associated isoforms, such as explain the high affinity of this compound for the CA CA IV, IX and XII, have also been considered as possible active site59. As topiramate also acts as an efficient inhib‑ targets of the antiglaucoma sulphonamides55–57. The itor of the human mitochondrial isozymes CA VA and bovine CA IV isozyme was shown to be susceptible to VB, this inhibition of both mitochondrial and cytosolic inhibition by most sulphonamides and sulphamates, CA isozymes involved in lipogenesis may provide 55 58–60 with KIs in the low nanomolar range . However, the a new approach to control weight loss . corresponding human isoform, CA IV, shows different Zonisamide (compound 10) is another anti-epileptic behaviour, with some drugs, such as acetazolamide, drug used as adjunctive therapy for refractory partial ethoxzolamide and sulthiame, acting as effective seizures1,61,62. It has multiple mechanisms of action, and inhibitors, whereas other antiglaucoma agents, such as exhibits a broad spectrum of anticonvulsant activity. methazolamide, dichlorophenamide, dorzolamide and Similar to topiramate, recent clinical studies have dem‑ brinzolamide, act as weak inhibitors56 (TABLE 1), which onstrated additional potential for therapeutic use for suggests that it is improbable that CA IV plays a role neuropathic pain, bipolar disorder, migraine, obesity, in aqueous humor secretion, as it is so weakly inhibited eating disorders and Parkinson’s disease1. Zonisamide is by most antiglaucoma sulphonamides. However, CA an aliphatic sulphonamide, which also potently inhibits XII (but not CA IX) was recently shown to be highly cytosolic and mitochondrial CAs involved in lipogenesis61

REVIEWS

Hypoxia Normoxia be strongly associated with tumour propagation, malig‑ nant progression, and resistance to chemotherapy and HIFα HIFα 5,63–68 VHL radiotherapy . Hypoxia regulates the expression of several genes, including a CA isozyme, CA IX, through 5,63–71 Stabilization PHD O2 Hydroxylation the hypoxia inducible factor 1 (HIF1) cascade . The expression of CA IX is strongly upregulated by hypoxia and is downregulated by the wild-type von Hippel- HIFα HIFα OH VHL Interaction with VHL Lindau tumour suppressor protein (pVHL) (FIG. 4). HIFα CA IX expression is strongly increased in many types of tumours, such as gliomas/ependymomas64, mesothelio‑ 64 64 HIF mas , papillary/follicular carcinomas , carcinomas of the α HIFα OH Ubiquitylation bladder72, uterine cervix73,74, nasopharyngeal carcinoma75, 76 70,77,78 64 79 Ubiquitin head and neck , breast , oesophagus , lungs , brain64, vulva64, squamous/basal cell carcinomas64, and kidney80 tumours, among others. In some cancer cells, the Degradation VHL gene is mutated leading to the strong upregulation Cytoplasm of CA IX (up to 150-fold) as a consequence of constitutive HIF activation5,64,77. Nucleus CA IX belongs to the highly active human α‑CAs, Constitutive subunit HIFα HIFβ HIFβ its catalytic properties for the CO2 hydration reaction being comparable with those of the highly evolved Active 5,64 α transcription catalyst CA II . As for all ‑CAs, CA IX is suscep‑ factor tible to inhibition by anions and sulphonamides and sulphamates1,2,5,14,81–89, with the inhibitors coordinating HIFα HIFβ directly to the zinc ion within the active-site cavity and GLUT1/3 (anaerobic glycolysis) participating in various other favourable interactions HRE VEGF (angiogenesis) EPO1 (erythropoesis) with amino-acid residues that are situated in the hydro‑ CA IX (pH regulation) phobic and hydrophilic halves of the active site. Many low nanomolar CA IX inhibitors have been identified Figure 4 | Mechanism of hypoxia-induced gene expression mediated by the HIF in the past several years1,2,5,14 (TABLE 1). Among them, transcription factor. At normal oxygen levels (normoxia), prolyl‑4-hydroxylase (PHD) some sulphamates81,88,89 and sulphonamides82–86 were hydroxylates the P564 on hypoxia inducible factor-α (HIFαNa).tur Thee Revonvie Hippel-Lindauws | Drug Discove ry characterized by X‑ray crystallography and homology protein (VHL) binds hydroxylated HIFα and targets it for degradation by the ubiquitin– modelling. Such studies also evidenced compounds proteasome system. Under hypoxia, HIFα is not hydroxylated, because PHD is inactive that are membrane impermeable (and thus specifically in the absence of dioxygen. Non-hydroxylated HIFα is not recognized by the VHL inhibit CA IX in vivo)90–93 or act as dual CA IX–COX2 protein, it is stabilized and accumulates. After translocation to the nucleus, HIFα 82,93 94,95 dimerizes with the HIF constitutive subunit to form an active transcription factor. (cyclooxygenase 2) inhibitors . Both heterocyclic , β 96,97 98 The HIF transcription factor then binds the hypoxia response element (HRE) in target aromatic sulphonamides as well as aliphatic sulphon 99–101 genes and activates their transcription. Target genes include glucose transporters amides/sulphamates/sulphamides possessing low (GLUT1 and GLUT3) that participate in glucose metabolism, vascular endothelial nanomolar inhibitory activity against CA IX have been growth factor (VEGF) that triggers neoangiogenesis, erythropoietin (EPO1) involved detected so far. Some sulphonamides incorporating in erythropoiesis, carbonic anhydrase (CA) IX involved in pH regulation and various sugar moieties were also reported102,103, but up tumorigenesis, and additional genes with functions in cell survival, proliferation, until now the most useful CAIs for understanding the 64–69 metabolism and other processes . function of this protein in vivo were the fluorescent com‑ pounds of type 17 (see later in the text)84,94,104. As described previously, hypoxia, through the HIF (TABLE 1). Furthermore, zonisamide in conjunction cascade, leads to a strong overexpression of CA IX in with a reduced-calorie diet (deficit of 500 kcal per many tumours. The overall consequence of this is a pH day), resulted in an additional mean 5 kg (11 pound) imbalance, with most hypoxic tumours having acidic weight loss compared with diet alone in obese female pH values around 6, in contrast to normal tissue, which patients62. Thus, inhibition of mitochondrial isoforms has characteristic pH values around 7.4 (REFS 84,85,104). CA VA and VB, probably in conjunction with that of the Constitutive expression of human CA IX was recently

ubiquitous cytosolic isoform CA II, may represent targets shown to decrease extracellular pH (pHe) in Madin-Darby for novel anti-obesity drugs that reduce lipogenesis by canine kidney (MDCK) epithelial cells84. CA IX-selective inhibiting CA58. sulphonamide inhibitors (of type 17 and 18) reduced the medium acidity by inhibiting the catalytic activity of Anticancer CAIs the enzyme, and thus the generation of H+ ions, bind‑ A key feature of many tumours is hypoxia63–68. The ing specifically only to hypoxic cells expressing CA IX84. inadequate supply of oxygen is primarily a pathophysio Deletion of the CA active site was also shown to reduce logical consequence of structurally and functionally the acidity of the medium, but a sulphonamide inhibitor disturbed microcirculation and deteriorated oxygen did not bind to the active site of such mutant proteins. 5,63–68 diffusion processes . Tumour hypoxia appears to Therefore, tumour cells decrease their pHe both by nature reviews | drug discovery volume 7 | february 2008 | 175 © 2008 Nature Publishing Group

REVIEWS

production of lactic acid (due to the high glycolysis rates), + – CO2 + H2O ↔ H + HCO3 and by CO2 hydration catalysed by the tumour-associated 5,84,104 CA IX CA IX, possessing an extracellular catalytic domain + Cl– H CA XII (FIG. 5). Low pH has been associated with tumorigenic e e f g transformation, chromosomal rearrangements, extracell ular matrix breakdown, migration and invasion, induc‑ tion of the expression of cell growth factors and protease activation5,64,104. CA IX probably also plays a role in providing bicarbonate to be used as a substrate for cell H+ CA II growth, as it is established that bicarbonate is required in d + – CO2 + H2O ↔ H + HCO3 the synthesis of pyrimidine nucleotides5,64. K+ a Indisulam (compound 10), a sulphonamide deriva‑ Cancer cell H+ tive (originally called E7070) with powerful anticancer Na+ c activity, was recently shown to act as a nanomolar inhibi‑ pH ~7.2 tor of CA IX105–108 (TABLE 1). Its detailed mechanism of Lactate Na+ action is not clear, but it is known to be involved in the b Glycolysis pH ~6.8 perturbation of the cell cycle in the G1 and/or G2 phases, the downregulation of cyclins, the reduction of cyclin- + dependent kinase 2 (CDK2) activity, the inhibition of H retinoblastoma protein (pRb) phosphorylation and differential expression of molecules known to participate in cell adhesion, signalling and immune response, in Endothelial cells Glucose pH ~7.4 addition to its CA IX inhibitory properties. Indisulam showed in vivo efficacy against human tumour xenografts Figure 5 | Proteins and processes involved in pH regulation within the tumour cell. CO hydration to in nude mice, exhibiting a significant antitumour effect Nature Reviews2 | Drug Discovery and progressing to Phase I and II clinical trials for the bicarbonate and protons is catalysed in these cells by 105–108 the transmembrane isozymes possessing an extracellular treatment of solid tumours . active site, carbonic anhydrase (CA) IX and/or CA XII. Among the many derivatives reported so far, some Other involved proteins include monocarboxylate carrier of the most interesting potent CA IX inhibitors are the + + + + 84 (a); Na – H antiporter (b); ATP-dependent Na –K compounds investigated by Svastova et al. (structures antiporter (c); H+ channels (d); plasma-membrane 17 and 18). Derivative 17 is a fluorescent sulphonamide proton pump H+–ATPase (e); acquaporins (f); and anion that binds only to CA IX under hypoxic conditions exchangers (AE1-AE3 isoforms) (g)5,64–69. in vivo84,85,104. This compound may therefore be used as a fluorescent probe in hypoxic tumour imaging. Compound 18 belongs to a class of positively charged, membrane-impermeable compounds. Therefore, as such management of hypoxic tumours that do not respond compounds do not inhibit intracellular CAs, they may to classical chemotherapy or radiotherapy5,64,84,104. exhibit fewer side effects as compared with the presently Therefore, the use of the CAIs described above provide available compounds (such as acetazolamide), which possibilities of developing both diagnostic tools for the indiscriminately inhibit all CAs48–50. The X‑ray crystal non-invasive imaging of these tumours, as well as thera‑ structure of compound 18 in adduct with CA II (the peutic agents that probably perturb the extratumoral active site of which is similar to that of CA IX)6 has been acidification in which CA IX is involved5,64,84,104. Many reported recently. The positively charged pyridinium types of highly effective in vitro CA IX inhibitors have derivative 18 favourably binds within the enzyme active been developed and evaluated in vitro80–104. However, site, coordinating with the deprotonated sulphonamide although a large number of derivatives with enhanced moiety to the catalytically critical Zn2+ ion. It also par‑ affinity for the tumour-associated isozyme IX over the ticipates in many other favourable interactions with ubiquitously expressed CA I and II isozymes have been amino-acid residues present in the active-site cavity, discovered94, further studies are warranted in order to including stacking between the trimethylpyridinium understand the behaviour of such compounds in vivo, ring of the inhibitor 18 and the phenyl ring of Phe131, in cell cultures or animal models of the disease. an amino acid important for the binding of inhibitors to CAs86. A similar binding mechanism was subsequently CAIs as potential drugs for osteoporosis reported for the fluorescein derivative 17 (Ref. 18). Thus, The highly active CA II is abundant in the bone, and is such structures can be used for the rational drug design present only in osteoclasts at concentrations of the same of more selective and potent isozyme IX inhibitors. As order of magnitude as those present in the kidneys109. Its the X‑ray structure of CA IX is not yet available, most role there is to provide hydrogen ions, formed from the

studies have used the CA II structure for modelling and hydration of CO2, to an ATP-dependent proton pump, designing CA IX inhibitors. which uses them in the mobilization of calcium from In summary, biochemical, physiological and the bone. These activities are required for inorganic pharmacological data indicate that inhibition of the matrix dissolution that precedes the enzymatic removal tumour-associated CA isozyme IX may be useful in the of organic bone matrix.

REVIEWS

Table 3 | Catalytic activity of non-vertebrate CAs belonging to the α-, β- and γ-classes*

–1 –1 –1 Enzyme Class Activity level Kcat (s ) Km (mM) Kcat/Km (M s ) KI (nM) Subcellular Refs localization pfCA‡ α Very low 0.173 3.7 46.75 99 Cytoplasm 26 hpαCA§ α Low 2.5 × 105 16.6 1.5 × 107 21 Periplasm 27 hpβCA|| β Medium 7.1 × 105 14.7 4.8 × 107 40 Cytoplasm 28 cab¶ β Low 3.1 × 104 1.7 1.8 × 106 12,100 Unknown 110 cam# γ Low 7.1 × 104 1.8 3.9 × 106 63 Unknown 111 *Their inhibition by acetazolamide; most carbonic anhydrases (CAs) show low activity as esterases1,2. ‡Plasmodium falciparum CA enzyme; esterase activity with 4-nitrophenyl acetate as substrate at 25°C. §Helicobacter pylori α-class CA enzyme; at pH 8.9 and || ¶ 25°C (CO2 hydration reaction). Helicobacter pylori β-class CA enzyme; at pH 8.3 and 20°C (CO2 hydration reaction). β-CA from the # archaeon Methanobacterium thermoautotrophicum; at pH 7.5 and 20°C (CO2 hydration reaction). γ-CA from the archaeon Methanosarcina thermophila; at pH 7.1 and 25°C (CO2 hydration reaction).

To evaluate the physiological role of membrane-bound are distinct to that of the human host enzymes CA I CAs in osteoclasts, a novel membrane-impermeable and II. 4‑(3,4-Dichlorophenylureido-ethyl)-benzene CA inhibitor structurally related to compound 18 was sulphonamide was the most effective antimalarial used. Increased osteoclast number and bone resorption sulphonamide CAI against growth of P. falciparum

activity was observed in rat osteoclast cultures exposed in vitro, with a KI in the nanomolar range (80 nM) and

to a low concentration of inhibitor, whereas higher con‑ EC50 (ex vivo) in the low micromolar range (20 µM). centrations affected cell survival109. Inhibitor treatment Thus, sulphonamide CAIs targeting the protozoan also disturbed intracellular acidification in osteoclasts. enzyme may have the potential for the development of Membrane-bound isoenzymes CA IV and CA XIV are novel antimalarial drugs26. expressed in osteoclasts in vivo and in vitro. In addition, H. pylori α‑class CA (hpαCA) has been cloned and the inhibitor experiments provide novel evidence to sup‑ sequenced from patients with different gastric mucosal port the hypothesis that intracellular pH regulation in lesions, including gastritis, ulcer and cancer27,28. Some transport metabolons osteoclasts may involve and that a potent hpaCA inhibitors were detected (KIs of 12–84 nM), possible use of such inhibitors would be in the design of among which were acetazolamide, 4‑amino‑6-chloro‑ novel anti-osteoporosis therapies109. 1,3-benzenedisulphonamide and some newly designed compounds incorporating lipophilic tails. As hpaCA Non-vertebrate CAs and their inhibition is essential for the survival of the pathogen in acid, Sequencing of eukaryotic/prokaryotic genomes, in par‑ its inhibition might be used as a new pharmaco‑ ticular of pathogens causing widespread diseases (for logical tool in the management of drug-resistant example, malaria, tuberculosis, as well as other bacterial H. pylori infection27. DNA clones for the β‑class CA and fungal infections), has revealed that CAs are also of H. pylori (hpβCA) were isolated from independent present in these organisms1–4. However, although verte strains obtained from patients with various gastric brates possess only CAs belonging to the α‑class9–16, such mucosal lesions, including patients with gastritis, gas‑ ‘simpler’ organisms have enzymes belonging to several tric ulcer and gastric cancer28. hpβCA was also highly

CA families, for example the α‑ and β‑CAs, β‑ and inhibited (KIs in the range of 24–45 nM) by many sul‑ γ‑CAs, γ‑ and δ‑CAs, or even representatives from three phonamides/sulphamates, including the clinically used such gene families (α–γ-CAs)27–34,110–112. drugs acetazolamide, ethoxzolamide, topiramate and Recently, representatives of the α‑ or β‑CA class have sulpiride. The dual inhibition of α‑ and/or β‑class CAs been cloned and characterized in several pathogens, such of H. pylori could therefore be a useful alternative for as the protozoan P. falciparum26, the bacteria H. pylori27,28 the management of gastritis/gastric ulcers, as well as and M. tuberculosis29, and the fungi C. albicans30 and gastric cancer. This was also the first study showing that C. neoformans31,32. As it has been proved that these CAs a bacterial β‑CA can be a druggable target28. The crucial are critical for the growth or virulence of these patho‑ role played by these two CAs present in H. pylori in acid gens, their capacity to be inhibited has also been investi acclimatization of the pathogen within the stomach gated (TABLE 3). As many such organisms are highly is shown schematically in FIG. 6, and this also helps to pathogenic, and present different degrees of resistance to understand why inhibition of the two enzymes leads to the currently available drugs targeting them, inhibition the death of the bacteria and a possible eradication of of their CAs may constitute novel approaches to fighting H. pylori from the stomach27,28. these diseases26–34,110–112. The M. tuberculosis Rv3588c gene, which has been Transport metabolons Carbonic anhydrase isozymes P. falciparum is responsible for the majority of life- shown to be required for in vivo growth of the pathogen, that interact with anion threatening cases of human malaria. The global emer‑ was discovered to encode a β‑CA with activity that is exchanger proteins to form gence of drug-resistant malarial parasites necessitates highly dependent on pH, being active at pH 8.4 but not at transport metabolons that the identification and characterization of novel drug pH 7.5 (Ref. 29). This β‑CA was observed to be a dimeric regulate intracellular and/or targets. The CAs present in P. falciparum belongs to the protein with a blocked active site, being able to switch extracellular pH. α‑CA class (pfCA) and possess catalytic properties that between two states with an opened or closed active site.

REVIEWS

Although no inhibition studies were performed with this H+ Urea β‑CA, it is probable that potent inhibitors may be detected Porin and designed. This could lead to a completely new class Gastric lumen of antituberculosis drugs, a disease for which resistance to the presently available drugs is a worldwide problem29. The NCE103 gene of the yeast Saccharomyces cerevisiae + + + – NH4 ← NH3 + H ← H + HCO3 also encodes a CA34. The main physiological role of CA during growth of S. cerevisiae in glucose-ammonium α-CA Urea H O salts media is the provision of inorganic carbon for the 2 NH CO 3 2 Urea channel bicarbonate-dependent carboxylation reactions catalysed Periplasm by PC, ACC and CPSase (carbamoyl-phosphate syn‑ thetase)30–32. However, the most interesting findings in this field regard the signalling role of CAs that is important Cytoplasm for the virulence of fungal pathogens such as C. albicans Urease and C. neoformans, as identified by Muhlschlegel’s and 30–32 2NH + CO Urea Heitman’s groups . It has been demonstrated that 3 2 – physiological concentrations of CO /HCO induce fila‑ H2O 2 3 β-CA mentation in C. albicans by a direct stimulation of the FIG. 7 + + + – adenylyl cyclase activity, as shown schematically in . NH4 ← NH3 + H ← H + HCO3 Furthermore, it has been shown that CO /HCO – equili‑ 2 3 Figure 6 | A model for the role of urease and α‑ and bration by the β‑CAs present in the organism is essential β‑CA in the maintenance of periplasmic pH in for the pathogenesis of C. albicans in niches where the Helicobacter pylori. Under acidicNature conditions, Reviews | Drug urea Disc movesovery

available CO2 is limited. Muhlschlegel’s group also dem‑ into the cytoplasm through the urea channel. In the

onstrated that adenylyl cyclase from C. neoformans is cytoplasm, 2NH3 and CO2 are produced from urea owing to – the activity of urease. CO in the periplasm and cytoplasm sensitive to physiological concentrations of CO2/HCO3 . 2 Thus, the link between cyclic AMP signalling and CO / is hydrated by H. pylori α‑ and β‑carbonic anhydrases (CAs), 2 + HCO – sensing is conserved in fungi, with CO sensing respectively, resulting in production of H and bicarbonate. 3 2 The proton is then consumed by NH to form NH + in the being an important mediator of fungal pathogenesis. 3 4 periplasm and cytoplasm, respectively27,28. Although no inhibitors of these new β‑CAs have yet been tested, novel therapeutic agents targeting this pathway at several levels could act as new antifungals30–32. Several X‑ray crystallographic structures of adducts CA activators in drug design of the main human isoforms, CA I and II, with activators Although CAIs have been extensively studied, and were reported recently, in addition to that of histamine, exploited clinically for the prevention and treatment reported in 1997 (Refs 35–41). The X‑ray crystal struc‑ of several diseases, the field of CA activators (CAAs) tures of the human CA II–l‑His/d-His adducts and the is largely unexplored. However, in the past decade, by CA II–l‑Phe/d-Phe adducts showed the activators to be means of electronic spectroscopy, X‑ray crystallography anchored at the entrance of the active site, participating and kinetic measurements, activators have been shown in extended networks of hydrogen bonds and hydro‑ to bind within the CA active cavity, at a site distinct phobic interactions with specific amino-acid residues to that of the inhibitor or substrate-binding sites, to or water molecules present in the cavity, thus explaining facilitate the proton-transfer step of the catalytic cycle. their different potency and interaction patterns with vari‑ Recently, activation of some members of the α‑CA family ous isozymes35–39,41. Many drug design studies of CAAs (human CA I and CA II) was shown to constitute a have also been reported, and histamine or carnosine as possible therapeutic approach for the enhancement of lead molecules has been considered114–121. synaptic efficacy, which may represent a conceptually CAAs may lead to the design of pharmacologically new treatment for Alzheimer’s disease, ageing and other useful derivatives for the enhancement of synaptic effi‑ conditions in which the achievement of spatial learn‑ cacy, which may represent a conceptually new approach ing and memory therapy is necessary40,113. Sun and for the treatment of Alzheimer’s disease, ageing and Alkon reported that phenylalanine, an activator first other conditions in which spatial learning and memory investigated by our group to target isozymes I and II114, therapy need to be enhanced. when administered to experimental animals produces a relevant pharmacological enhancement of synaptic Conclusions efficacy, spatial learning and memory, proving that this In the field of CA research, the past few years have been class of unexplored enzyme modulators may be used for highly dynamic and productive. The presumed last mam‑ the management of conditions in which learning and malian α‑CA isoform (CA XV) has now been isolated, memory are impaired40. A multitude of physiologically characterized and purified; along with representatives relevant compounds such as biogenic amines (histamine, of the α‑, β‑, γ‑ and δ‑classes from various pathogenic serotonin, catecholamines), amino acids, oligopeptides organisms from across the phylogenetic tree. Many of or small proteins, among others, act as efficient CAAs for these new CAs are druggable targets, and therefore pro‑ many of the human CA isozymes35–41,115–121. vide potential for the development of pharmacological

REVIEWS

for any CA isoform with clinical relevance, although CO2 important advances have been reported in the design of compounds with high selectivity for CA VA, IX and XIII over CA II (CA II is the ubiquitous, catalytically highly Capsule effective isoform, and so its inhibition is often regarded as detrimental). Significant advances have been reported in the design Cytoplasm of fluorescently labelled and membrane-impermeable H2O – + CO2 HCO3 + H compounds inhibiting only the membrane-associated Can2 CA isoforms (such as CA IX) and not those present in the cytosol (such as CA I and II). The use of fluores‑ cent CAIs has indicated the important role of CA IX in Cac1 tumour acidification processes, and the possibility of ATP cAMP reversing this phenomenon by inhibiting the catalytic activity of the enzyme with a CA IX‑selective and potent inhibitor. Regulation Various ophthalmologic and anti-obesity applications of capsule of the CAIs have also been reported, together with novel biosynthesis derivatives that act as anticonvulsants. The reported Figure 7 | A model of the regulation of filamentation X‑ray crystal structures of many human CA II and I

and fungal capsule biosynthesis. CO2 is catalytically adducts with sulphonamides and sulphamates has pro‑ converted to bicarbonate inNa thetur epresence Reviews | ofDrug the Disc overy vided a better understanding of the molecular interac‑ β‑carbonic anhydrase isozyme Can2 from Candida tions between the inhibitor and the enzyme, which may albicans. This bicarbonate directly stimulates Candida lead to a rational drug design of inhibitors with reduced adenylate cyclase (Cac1) activity, which converts ATP side effects and selectivity for the target isoform. to cyclic AMP, which regulates filamentation and fungal Finally, important progress has been achieved in the capsule biosynthesis30,31. field of the CAAs, with a large number of X‑ray crystallo graphic structures of adducts of isozymes I and II with amino-acid activators reported. Interesting differences agents with a new mechanism of action. Indeed, resist‑ in the binding of enantiomeric activators such as l‑/ ance to currently available drugs constitutes a serious d-Phe and l‑/d-His to human CA II have been observed; medical problem for infections caused by P. falciparum, and the first X‑ray crystal structure of an adduct of H. pylori, M. tuberculosis and C. albicans, and CAIs tar‑ human CA I with an activator (l-His) has recently been geting enzymes from these pathogens may overcome this reported, showing great differences between the bind‑ owing to their novel mechanism of action. Over the past ing of activators within the active site of the two main few years, the inhibitory profile of most of the important isoforms, CA I and II. For the first time, isoforms other classes of CAIs — the sulphonamides, sulphamates and than CA I and II, such as CA IV, VA, VII, XIII and XIV sulphamides — against the catalytically active isozymes have been investigated for their interaction with a large have been investigated in detail. This is valuable in the number of amine and amino-acid activators. Owing to search for isozyme-selective compounds and for reducing positive effects on spatial learning and memory, CAAs side effects of the presently available drugs. However, few may provide a novel mechanism for treating disorders compounds have been identified that exhibit selectivity such as Alzheimer’s disease.

1. Scozzafava, A., Mastrolorenzo, A. & Supuran, C. T. 7. Pastorekova, S. et al. Carbonic anhydrases: current state 13. Nishimori, I. in Carbonic Anhydrase — Its Inhibitors Carbonic anhydrase inhibitors and activators and of the art, therapeutic applications and future prospects. and Activators (eds Supuran, C. T., Scozzafava, A. their use in therapy. Expert Opin. Ther. Pat. 16, J. Enzyme Inhib. Med. Chem. 19, 199–229 (2004). & Conway J.) 25–43 (CRC, Boca Raton, 2004). 1627–1664 (2006). 8. Nishimori, I. et al. Carbonic anhydrase inhibitors. 14. Vullo, D., et al. Carbonic anhydrase inhibitors. 2. Supuran, C. T., Scozzafava, A. & Conway, j. Cloning, characterization and inhibition studies of Inhibition of the tumor-associated isozyme IX with Carbonic Anhydrase — Its inhibitors and Activators the cytosolic isozyme III with sulfonamides. aromatic and heterocyclic sulfonamides. Bioorg. Med. 1–363 (CRC, Boca Raton, 2004). Bioorg. Med. Chem. 15, 7229–7236 (2007). Chem. Lett. 13, 1005–1009 (2003). 3. Supuran, C. T., Scozzafava, A. & Casini, A. Carbonic 9. Vullo, D. et al. Carbonic anhydrase inhibitors. First CA IX inhibition study reported showing the anhydrase inhibitors. Med. Res. Rev. 23, 146–189 Inhibition of mitochondrial isozyme V with aromatic enzyme to be a druggable target. (2003). and heterocyclic sulfonamides. J. Med. Chem. 47, 15. Vullo, D. et al. Carbonic anhydrase inhibitors. A comprehensive review on the development of CAIs. 1272–1279 (2004). Inhibition of the transmembrane isozyme XII with 4. Smith, K. S. & Ferry, J. G. Prokaryotic carbonic 10. Nishimori, I. et al. Carbonic anhydrase inhibitors. sulfonamides — a new target for the design of anhydrases. FEMS Microbiol. Rev. 24, 335–366 The mitochondrial isozyme VB as a new target for antitumor and antiglaucoma drugs? Bioorg. Med. (2000). sulfonamide and sulfamate inhibitors. J. Med. Chem. Chem. Lett. 15, 963–969 (2005). 5. Thiry, A. et al. Targeting tumor-associated carbonic 48, 7860–7866 (2005). First CA XII inhibition study reported. This isozyme anhydrase IX in cancer therapy. Trends Pharmacol. Sci. 11. Nishimori, I. et al. Carbonic anhydrase inhibitors. is involved in glaucoma and cancer. 27, 566–573 (2006). DNA cloning, characterization and inhibition studies 16. Lehtonen, J. et al. Characterization of CA XIII, An up-to-date review regarding the targeting of of the human secretory isoform VI, a new target for a novel member of the carbonic anhydrase the tumour-associated CA IX by small-molecule sulfonamide and sulfamate inhibitors. J. Med. Chem. isozyme family. J. Biol. Chem. 279, 2719–2727 inhibitors. 50, 381–388 (2007). (2004). 6. Stams, T. & Christianson, D. W. in The Carbonic 12. Vullo, D. et al. Carbonic anhydrase inhibitors. 17. Nishimori, I. et al. Carbonic anhydrase inhibitors. Anhydrases — New Horizons (eds Chegwidden, W. R., Inhibition of the human cytosolic isozyme VII with Inhibition of the transmembrane isozyme XIV Carter, N. D. & Edwards, Y. H.) 159–174 (Birkhauser, aromatic and heterocyclic sulfonamides. Bioorg. Med. with sulfonamides. Bioorg. Med. Chem. Lett. 15, Basel, 2000). Chem. Lett. 15, 971–976 (2005). 3828–3833 (2005).

REVIEWS

18. Alterio, V. et al. Carbonic anhydrase inhibitors: X‑ray 35. Supuran, C. T. & Scozzafava, A. in The Carbonic 54. Supuran, C. T. et al. in Carbonic Anhydrase – and molecular modeling study for the interaction of a Anhydrases — New Horizons (eds Chegwidden, W. R., Its Inhibitors and Activators (eds Supuran, C. T., fluorescent antitumor sulfonamide with isozyme II and Carter, N. D. & Edwards, Y. H.) 197–219 Scozzafava, A. & Conway, J.) 67–147 (CRC, Boca IX. J. Am. Chem. Soc. 128, 8329–8335 (2006). (Birkhauser, Basel, 2000). Raton, 2004). X-ray crystal structure of a fluorescent, 36. Ilies, M. et al. Carbonic anhydrase activators. in A comprehensive review of the development of antitumour sulphonamide in adduct with CA II and Carbonic Anhydrase — Its Inhibitors and Activators sulphonamdes as CAIs. homology modelling for the binding to CA IX. (eds Supuran, C. T., Scozzafava, A. & Conway, J.) 55. Maren, T. H. The development of topical carbonic Opens new prospectives for the design of 317–352 (CRC, Boca Raton, 2004). anhydrase inhibitors. J. Glaucoma 4, 49–62 (1995). isoform-specific CAIs. 37. Briganti, F. et al. Carbonic anhydrase activators: X‑ray 56. Innocenti, A. et al. Carbonic anhydrase inhibitors. 19. Köhler, K. et al. Saccharin inhibits carbonic crystallographic and spectroscopic investigations for Inhibition of the membrane-bound human and bovine anhydrases: possible explanation for its unpleasant the interaction of isozymes I and II with histamine. isozymes IV with sulfonamides. Bioorg. Med. Chem. metallic aftertaste. Angew. Chem. Int. Ed. Engl. 46, Biochemistry 36, 10384–10392 (1997). Lett. 15, 1149–1154 (2005). 7697–7699 (2007). First study characterizing CA activators by means 57. Liao, S. Y. et al. Expression of cell surface 20. Supuran, C. T., Ilies M. A. & Scozzafava, A. Carbonic of X-ray crsytallography. transmembrane carbonic anhydrase genes CA9 and anhydrase inhibitors. Part 29. Interaction of isozymes 38. Temperini, C. et al. Carbonic anhydrase activators. CA12 in the human eye: overexpression of CA12 I, II and IV with benzolamide-like derivatives. Activation of isozymes I, II, IV, VA, VII and XIV with (CA XII) in glaucoma. J. Med. Genet. 40, 257–261 Eur. J. Med. Chem. 33, 739–752 (1998). l‑ and d‑histidine and crystallographic analysis of (2003). 21. Supuran, C. T., Scozzafava, A., Ilies M. A. & their adducts with isoform II: engineering proton 58. Supuran, C. T. Carbonic anhydrase inhibitors in the Briganti, F. Carbonic anhydrase inhibitors. transfer processes within the active site of an enzyme. treatment and prophylaxis of obesity. Expert Opin. Synthesis of sulfonamides incorporating Chemistry 12, 7057–7066 (2006). Ther. Pat. 13, 1545–1550 (2003). 2,4,6‑trisubstituted‑pyridinium-ethylcarboxamido 39. Temperini, C. et al. Carbonic anhydrase activators. 59. Casini, A. et al. Carbonic anhydrase inhibitors: SAR moieties possessing membrane-impermeability and Activation of isoforms I, II, IV, VA, VII and XIV with and X‑ray crystallographic study for the interaction of in vivo selectivity for the membrane-bound (CA IV) l‑ and d‑phenylalanine and crystallographic analysis of sugar sulfamates/sulfamides with isozymes I, II and IV. versus the cytosolic (CA I and CA II) isozymes. their adducts with isozyme II: sterospecific recognition Bioorg. Med. Chem. Lett. 13, 841–845 (2003). J. Enz. Inhib. 15, 381–401 (2000). within the active site of an enzyme and its 60. Picard, F. et al. Topiramate reduces energy and fat 22. Scozzafava, A., Briganti, F., Ilies, M. A. & Supuran, C. T. consequences for the drug design. J. Med. Chem. 49, gains in lean (Fa/?) and obese (fa/fa) Zucker rats. Carbonic anhydrase inhibitors. Synthesis of 3019–3027 (2006). Obesity Res. 8, 656–663 (2000). membrane-impermeant low molecular weight 40. Sun, M.‑K. & Alkon, D. L. Carbonic anhydrase gating 61. De Simone, G. et al. Carbonic anhydrase inhibitors. sulfonamides possessing in vivo selectivity for the of attention: memory therapy and enhancement. Zonisamide is an effective inhibitor of the cytosolic membrane-bound versus the cytosolic isozymes. Trends Pharmacol. Sci. 23, 83–89 (2002). isozyme II and mitochondrial isozyme V: solution and J. Med. Chem. 43, 292–300 (2000). 41. Temperini, C. et al. Carbonic anhydrase activators: X‑ray crystallographic studies. Bioorg. Med. Chem. A report of membrane-impermeable CAIs, which l-adrenaline plugs the active site entrance of Lett. 15, 2315–2320 (2005). selectively inhibit extracellular and not cytosolic isozyme II, activating better isoforms I, IV, VA, VII, 62. Gadde, K. M. et al. Zonisamide for weight loss in isozymes. and XIV. Bioorg. Med. Chem. Lett. 17, 628–635 obese adults: a randomized controlled trial. JAMA 23. Winum, J. Y. et al. Carbonic anhydrase inhibitors: (2007). 289, 1820–1825 (2003). clash with Ala65 as a means for designing inhibitors 42. Splendiani, G. & Condo, S. Diuretic therapy in heart 63. Maxwell, P. H., et al. The tumour suppressor protein with low affinity for the ubiquitous isozyme II, failure. G. Ital. Nefrol. 23, S74–S76 (2006). VHL targets hypoxia-inducible factors for oxygen- exemplified by the crystal structure of the topiramate 43. Kyllonen, M. S. et al. Localization of carbonic dependent proteolysis. Nature 399, 271–275 sulfamide analogue. J. Med. Chem. 49, 7024–7031 anhydrase XII to the basolateral membrane of (1999). (2006). H+-secreting cells of mouse and rat kidney. 64. Pastorekova, S. & Pastorek, J. in Carbonic Anhydrase 24. Saczewski, F. et al. Carbonic anhydrase inhibitors. J. Histochem. Cytochem. 51, 1217–1224 (2003). — Its Inhibitors and Activators (eds Supuran, C. T., Inhibition of the cytosolic human isozymes I and II, 44. Mincione, F. et al. The development of topically Scozzafava, A. & Conway, J.) 255–281 (CRC, Boca and the transmembrane, tumor-associated isozymes acting carbonic anhydrase inhibitors as anti- Raton, 2004). IX and XII with substituted aromatic sulfonamides glaucoma agents. Curr. Top. Med. Chem. 7, 65. Semenza, G. L. Hypoxia and cancer. Cancer activatable in hypoxic tumors. Bioorg. Med. Chem. 849–854 (2007). Metastasis Rev. 26, 223–224 (2007). Lett. 16, 4846–4851 (2006). 45. Sugrue, M. F. Pharmacological and ocular hypotensive 66. Brahimi-Horn, M. C. & Pouyssegur, J. Oxygen, a 25. De Simone, G. et al. Carbonic anhydrase inhibitors: properties of topical carbonic anhydrase inhibitors. source of life and stress. FEBS Lett. 581, 3582–3591 hypoxia-activatable sulfonamides incorporating Prog. Retin. Eye Res. 19, 87–112 (2000). (2007). disulfide bonds that target the tumor-associated 46. Scozzafava, A. et al. Carbonic anhydrase inhibitors. 67. Jaakkola, P. et al. Targeting of HIF-α to the von Hippel-

isoform IX. J. Med. Chem. 49, 5544–5551 (2006). Part 74. Synthesis of water-soluble, topically effective, Lindau ubiquitylation complex by O2-regulated prolyl 26. Krungkrai, J. et al. Carbonic anhydrase inhibitors. intraocular pressure-lowering aromatic/heterocyclic hydroxylation. Science 292, 468–472 (2001). Inhibition of Plasmodium falciparum carbonic sulfonamides containing cationic or anionic moieties: 68. Hon, W. C. et al. Structural basis for the recognition anhydrase with aromatic sulfonamides: towards is the tail more important than the ring? J. Med. of hydroxyproline in HIF‑1α by pVHL. Nature 417, antimalarials with a novel mechanism of action? Chem. 42, 2641–2650 (1999). 975–978 (2002). Bioorg. Med. Chem. 13, 483–489 (2005). 47. Scozzafava, A. et al. Carbonic anhydrase inhibitors: 69. Ratcliffe, P. J. et al. Targeting tumors through the HIF 27. Nishimori, I. et al. Carbonic anhydrase inhibitors. synthesis of water-soluble, aminoacyl/dipeptidyl system. Nature Med. 6, 1315–1316 (2000). DNA cloning and inhibition studies of the α‑carbonic sulfonamides possessing long-lasting intraocular 70. Trastour, C. et al. HIF-1α and CA IX staining in invasive anhydrase from Helicobacter pylori: a new target for pressure-lowering properties via the topical route. breast carcinomas: prognosis and treatment outcome. developing sulfonamide and sulfamate gastric drugs. J. Med. Chem. 42, 3690–3700 (1999). Int. J. Cancer 120, 1451–1458 (2007). J. Med. Chem. 49, 2117–2126 (2006). 48. Ilies, M. et al. Carbonic anhydrase inhibitors: 71. Pouyssegur, J. et al. Hypoxia signalling in cancer and 28. Nishimori, I. et al. Carbonic anhydrase inhibitors. sulfonamides incorporating furan‑, thiophene- and approaches to enforce tumour regression. Nature The β‑carbonic anhydrase from Helicobacter pylori is pyrrole-carboxamido groups possess strong topical 441, 437–443 (2006). a new target for sulfonamide and sulfamate inhibitors. intraocular pressure lowering properties as aqueous 72. Ord, J. J. et al. An investigation into the prognostic Bioorg. Med. Chem. Lett. 17, 3585–3594 (2007). suspensions. Bioorg. Med. Chem. 8, 2145–2155 significance of necrosis and hypoxia in high grade and A proof-of-concept study showing that enzymes (2000). invasive bladder cancer. J. Urol. 178, 677–682 belonging to the β‑CA class are also druggable 49. Scozzafava, A. et al. Carbonic anhydrase inhibitors: (2007). targets. perfluoroalkyl/aryl-substituted derivatives of aromatic/ 73. Swietach, P. et al. Regulation of tumor pH and the role 29. Suarez Covarrubias, A. et al. Structure and function heterocyclic sulfonamides as topical intraocular of carbonic anhydrase 9. Cancer Metastasis Rev. 26, of carbonic anhydrases from Mycobacterium pressure-lowering agents with prolonged duration of 299–310 (2007). tuberculosis. J. Biol. Chem. 280, 18782–18789 action. J. Med. Chem. 43, 4542–4551 (2000). 74. Hutchison, G. J. et al. Hypoxia-inducible factor 1‑α (2005). 50. Scozzafava, A. et al. Carbonic anhydrase inhibitors. expression as an intrinsic marker of hypoxia: 30. Klengel, T. et al. Fungal adenylyl cyclase integrates A general approach for the preparation of water- correlation with tumor oxygen, pimonidazole

CO2 sensing with cAMP signaling and virulence. soluble sulfonamides incorporating polyamino- measurements, and outcome in locally advanced Curr. Biol. 15, 2021–2026 (2005). polycarboxylate tails and of their metal complexes carcinoma of the cervix. Clin. Cancer Res. 10, 31. Morgensen, E.G. et al. Cryptococcus neoformans possessing long-lasting, topical intraocular pressure- 8405–8412 (2004).

senses CO2 through the carbonic anhydrase Can2 and lowering properties. J. Med. Chem. 45, 1466–1476 75. Sung, F. L. et al. Genome-wide expression the adenylyl cyclase Cac1. Eukaryot. Cell 5, 103–111 (2002). analysis using microarray identified complex (2006). 51. Winum, J. Y. et al. Carbonic anhydrase inhibitors: signaling pathways modulated by hypoxia in d 32. Bahn, Y. S. et al. Carbonic anhydrase and CO2 N‑(p-sulfamoylphenyl)‑a‑ -glycopyranosylamines as nasopharyngeal carcinoma. Cancer Lett. 253, sensing during Cryptococcus neoformans growth, topically acting antiglaucoma agents in hypertensive 74–88 (2007). differentiation, and virulence. Curr. Biol. 15, rabbits. Bioorg. Med. Chem. Lett. 14, 225–229 76. Koukourakis, M. I. et al. Hypoxia-activated tumor 2013–2020 (2005). (2004). pathways of angiogenesis and pH regulation 33. Soto, A. R. et al. Identification and preliminary 52. Cox, S. N. et al. Treatment of chronic macular independent of anemia in head‑and‑neck cancer. characterization of two cDNAs encoding unique edema with acetazolamide. Arch. Ophthalmol. 106, Int. J. Radiat. Oncol. Biol. Phys. 59, 67–71 (2004). carbonic anhydrases from the marine alga Emiliania 1190–1195 (1988). 77. Potter, C. P. & Harris, A. L. Diagnostic, prognostic and huxleyi. Appl. Environ. Microbiol. 72, 5500–5511 53. Gao, B. B. et al. Extracellular carbonic anhydrase therapeutic implications of carbonic anhydrases in (2006). mediates hemorrhagic retinal and cerebral vascular cancer. Br. J. Cancer 89, 2–7 (2003). 34. Aguilera, J. et al. Carbonic anhydrase (Nce103p): permeability through prekallikrein activation. 78. Hussain, S. A. et al. Hypoxia-regulated carbonic an essential biosynthetic enzyme for growth of Nature Med. 13, 181–188 (2007). anhydrase IX expression is associated with poor Saccharomyces cerevisiae at atmospheric carbon An important study showing the involvement of CA I survival in patients with invasive breast cancer. dioxide pressure. Biochem. J. 391, 311–316 (2005). (an orphan target) in retinal and cerebral disease. Br. J. Cancer 96, 104–109 (2007).

REVIEWS

79. Swinson, D. E. et al. Carbonic anhydrase IX expression, 94. Garaj, V. et al. Carbonic anhydrase inhibitors: 109. Riihonen, R. et al. Membrane-bound carbonic a novel surrogate marker of tumor hypoxia, is synthesis and inhibition of cytosolic/tumor-associated anhydrases in osteoclasts. Bone 40, 1021–1031 associated with a poor prognosis in non‑small‑cell lung carbonic anhydrase isozymes I, II, and IX with (2007). cancer. J. Clin. Oncol. 21, 473–482 (2003). sulfonamides incorporating 1,2,4-triazine moieties. 110. Innocenti, A. et al. Carbonic anhydrase inhibitors. 80. Dorai, T. et al. Role of carbonic anhydrases in the Bioorg. Med. Chem. Lett. 14, 5427–5433 (2004). Inhibition of the β-class enzyme from the progression of renal cell carcinoma subtypes: 95. Innocenti, A. et al. Carbonic anhydrase inhibitors: the methanoarchaeon Methanobacterium proposal of a unified hypothesis. Cancer Invest. 24, first on-resin screening of a 4‑sulfamoylphenylthiourea thermoautotrophicum (Cab) with anions. Bioorg. Med. 754–779 (2006). library. J. Med. Chem. 47, 5224–5229 (2004). Chem. Lett. 14, 4563–4567 (2004). 81. Abbate, F. et al. Carbonic anhydrase inhibitors: 96. Cecchi, A. et al. Carbonic anhydrase inhibitors: 111. Zimmerman, S. A. et al. Carbonic anhydrase X ray crystallographic structure of the adduct of synthesis and inhibition of cytosolic/tumor- inhibitors. Inhibition of the prokariotic b and g-class human isozyme II with a EMATE, a dual inhibitor of associated carbonic anhydrase isozymes I, II, enzymes from Archaea with sulfonamides. Bioorg. carbonic anhydrases and steroid sulfatase. and IX with sulfonamides derived from Med. Chem. Lett. 14, 6001–6006 (2004). Bioorg. Med. Chem. Lett. 14, 231–234 (2004). 4‑isothiocyanato‑benzolamide. Bioorg. Med. Chem. First study showing that β- and γ-CA enzymes are 82. Di Fiore, A. et al. Carbonic anhydrase inhibitors: Lett. 14, 5775–5780 (2004). also inhibited by sulphonamides. valdecoxib binds to a different active site region of 97. Winum, J. Y. et al. Carbonic anhydrase inhibitors: 112. Zimmerman, S. A., Ferry, J. G. & Supuran C. T. the human isoform II as compared to the structurally synthesis and inhibition of cytosolic/membrane- Inhibition of the archaeal β‑Class (Cab) and γ‑Class related, cyclooxygenase II “selective” inhibitor associated carbonic anhydrase isozymes I, II, and IX (Cam) carbonic anhydrases. Curr. Top. Med. Chem. 7, celecoxib. Bioorg. Med. Chem. Lett. 16, 437–442 with sulfonamides incorporating hydrazino moieties. 901–908 (2007). (2006). J. Med. Chem. 48, 2121–2125 (2005). 113. Sun, M. K. & Alkon, D. L. Pharmacological 83. Abbate, F. et al. Carbonic anhydrase inhibitors: X ray 98. Winum, J. Y. et al. Carbonic anhydrase inhibitors: enhancement of synaptic efficacy, spatial learning, and crystallographic structure of the adduct of human synthesis and inhibition of cytosolic/tumor-associated memory through carbonic anhydrase activation in isozyme II with the antipsychotic drug sulpiride. carbonic anhydrase isozymes I, II, and IX with rats. J. Pharmacol. Exp. Ther. 297, 961–967 (2001). Bioorg. Med. Chem. Lett. 14, 337–341 (2004). bis-sulfamates. Bioorg. Med. Chem. Lett. 15, 114. Supuran, C. T. & Clare, B. W. Carbonic anhydrase 84. Svastova, E. et al. Hypoxia activates the capacity of 579–584 (2005). activators. Part 3. Structure–activity correlations for a tumor-associated carbonic anhydrase IX to acidify 99. Winum, J. Y. et al. Carbonic anhydrase inhibitors: series of isozyme II activators. J. Pharm. Sci. 83, extracellular pH. FEBS Lett. 577, 439–445 (2004). synthesis and inhibition of cytosolic/tumor-associated 768–773 (1994). A proof-of-concept study showing the involvement carbonic anhydrase isozymes I, II, IX, and XII with 115. Supuran, C. T. & Scozzafava, A. Carbonic anhydrase of CA IX in tumour acidification and its reversal by N‑hydroxysulfamides — a new zinc-binding function in activators: amino acyl/dipeptidyl histamine inhibition of the enzyme. the design of inhibitors. Bioorg. Med. Chem. Lett. 15, derivatives bind with high affinity to isozymes I, II and 85. Cecchi, A. et al. Carbonic anhydrase inhibitors. 2353–2358 (2005). IV and act as efficient activators. Bioorg. Med. Chem. Sulfonamides inhibit isozyme IX mediated acidification 100. Winum, J. Y. et al. Carbonic anhydrase inhibitors. 7, 2915–2923 (1999). of hypoxic tumors. Fluorescent sulfonamides design as Synthesis and inhibition of cytosolic/tumor- 116. Scozzafava, A. & Supuran, C. T. Carbonic anhydrase probes of membrane-bound carbonic anhydrase associated carbonic anhydrase isozymes I, II, and IX activators: high affinity isozymes I, II, and IV isozymes involvement in tumorigenesis. J. Med. Chem. with boron-containing sulfonamides, sulfamides, and activators, incorporating a beta‑alanyl‑histidine 48, 4834–4841 (2005). sulfamates: toward agents for boron neutron capture scaffold. J. Med. Chem. 45, 284–291 (2002). 86. Menchise, V. et al. Carbonic anhydrase inhibitors: therapy of hypoxic tumors. Bioorg. Med. Chem. Lett. 117. Ilies, M. et al. Carbonic anhydrase activators: design Stacking with Phe131 determines active site binding 15, 3302–3306 (2005). of high affinity isozymes I, II, and IV activators, region of inhibitors as exemplified by the X‑ray crystal 101. Thiry, A. et al. Indanesulfonamides as carbonic incorporating tri-/tetrasubstituted‑pyridinium‑azole structure of a membrane-impermeant antitumor anhydrase inhibitors. Toward structure-based design moieties. J. Med. Chem. 45, 504–510 (2002). sulfonamide complexed with isozyme II. J. Med. of selective inhibitors of the tumor-associated 118. Scozzafava, A. & Supuran, C. T. Carbonic anhydrase Chem. 48, 5721–5727 (2005). isozyme CA IX. J. Med. Chem. 49, 2743–2749 activators: human isozyme II is strongly activated by 87. Ilies, M. A., et al. Carbonic anhydrase inhibitors. (2006). oligopeptides incorporating the carboxyterminal Inhibition of tumor-associated isozyme IX by 102. Wilkinson, B. L. et al. A novel class of carbonic sequence of the bicarbonate anion exchanger AE1. halogenosulfanilamide and anhydrase inhibitors: glycoconjugate benzene Bioorg. Med. Chem. Lett. 12, 1177–1180 (2002). halogenophenylaminobenzolamide derivatives. sulfonamides prepared by “click-tailing”. J. Med. 119. Parkkila, S. et al. Carbonic anhydrase activators: J. Med. Chem. 46, 2187–2196 (2003). Chem. 49, 6539–6548 (2006). activation of isozyme XIII with amino acids and amines. 88. Winum, J. Y. et al. Carbonic anhydrase inhibitors. 99. Wilkinson, B. L. et al. Carbonic anhydrase inhibitors: Bioorg. Med. Chem. Lett. 16, 3955–3959 (2006). Inhibition of cytosolic isozymes I and II and inhibition of isozymes I, II, and IX with triazole-linked 120. Temperini, C. et al. Carbonic anhydrase activators: transmembrane, tumor-associated isozyme IX with O‑glycosides of benzene sulfonamides. J. Med. Chem. the first X‑ray crystallographic study of an adduct of sulfamates including EMATE also acting as steroid 50, 1651–1657 (2007). isoform I. Bioorg. Med. Chem. Lett. 16, 5152–5156 sulfatase inhibitors. J. Med. Chem. 46, 2197–2204 104. Dubois, L. et al. Imaging the hypoxia surrogate (2006). (2003). marker CA IX requires expression and catalytic activity First X-ray crystallographic report of an activator 89. Winum, J. Y. et al. Carbonic anhydrase inhibitors: for binding fluorescent sulfonamide inhibitors. of CA I, showing important differences in the inhibition of transmembrane, tumor-associated Radiother. Oncol. 83, 367–373 (2007). activation mechanisms of CA I and II. isozyme IX, and cytosolic isozymes I and II with 105. Abbate, F. et al. Carbonic anhydrase inhibitors: 121. Nishimori, I. et al. Carbonic anhydrase activators: aliphatic sulfamates. J. Med. Chem. 46, 5471–5477 E7070, a sulfonamide anticancer agent, potently the first activation study of the human secretory (2003). inhibits cytosolic isozymes I and II, and isoform VI with amino acids and amines. Bioorg. Med. 90. Pastorekova, S. et al. Carbonic anhydrase inhibitors: transmembrane, tumor-associated isozyme IX. Chem. Lett. 17, 5351–5357 (2007). the first selective, membrane-impermeant inhibitors Bioorg. Med. Chem. Lett. 14, 217–223 (2004). targeting the tumor-associated isozyme IX. Bioorg. 106. Owa, T. et al. Array-based structure and gene Acknowledgements Med. Chem. Lett. 14, 869–873 (2004). expression relationship study of antitumor Research from the author’s laboratory was financed in part 91. Vullo, D. et al. Carbonic anhydrase inhibitors: sulfonamides including N‑[2‑[(4-hydroxyphenyl) by two EU grants of the 6th framework programme (EUROXY inhibition of the tumor-associated isozyme IX with amino]‑3‑pyridinyl]‑4‑methoxy-benzenesulfonamide and DeZnIt projects). fluorine-containing sulfonamides. The first and N‑(3‑chloro‑7-indolyl)‑1,4-benzenedisulfonamide. subnanomolar CA IX inhibitor discovered. Bioorg. J. Med. Chem. 45, 4913–4922 (2002). Competing interests statement Med. Chem. Lett. 14, 2351–2356 (2004). 107. Owa, T. et al. Discovery of novel antitumor The authors declare competing financial interests: see web 92. Casey, J. R. et al. Carbonic anhydrase inhibitors. sulfonamides targeting G1 phase of the cell cycle. version for details. Design of selective, membrane-impermeant inhibitors J. Med. Chem. 42, 3789–3799 (1999). targeting the human tumor-associated isozyme A study reporting the discovery of indisulam, an IX. J. Med. Chem. 47, 2337–2347 (2004). anticancer sulphonamide in clinical development. DATABASES 93. Supuran, C. T. et al. COX‑2 selective inhibitors, 108. Talbot, D. C. et al. A randomized phase II UniProtKB: http://ca.expasy.org/sprot carbonic anhydrase inhibition and anticancer pharmacokinetic and pharmacodynamic study of CA I | CA II | CA III | CA IV | CA VA | CA VB | CA VI | CA VII | properties of sulfonamides belonging to this class of indisulam as second-line therapy in patients with CA IX | CA XII | CA XIII | CA XIV | CAXV pharmacological agents. Mini Rev. Med. Chem. 4, advanced non-small cell lung cancer. Clin. Cancer Res. All links are active in the online pdf 625–632 (2004). 13, 1816–1822 (2007).