Activation of α-, β-, γ- δ-, ζ- and η- class of carbonic anhydrases with amines and amino acids: a review.

Eight genetically distinct carbonic anhydrase (EC 4.2.1.1) enzyme families (α-, β-, γ- δ-, ζ-, η-, θ- and ι-CAs) were described to date. On the other hand, 16 mammalian α-CA isoforms are known to be involved in many diseases such as glaucoma, edema, epilepsy, obesity, hypoxic tumors, neuropathic pain, arthritis, neurodegeneration, etc. Although CA inhibitors were investigated for the management of a variety of such disorders, the activators just started to be investigated in detail for their in vivo effects. This review summarizes the activation profiles of α-, β, γ-, δ-, ζ- and η- CAs from various organisms (animals, fungi, protozoan, bacteria and archaea) with the most investigated classes of activators, the amines and the amino acids.

Introduction

Carbonic anhydrases (CAs; EC 4.2.1.1) are metalloproteins present virtually in all living organisms. CA enzymatic activity was first observed in the early 1930s, when experiments performed with hemolyzed blood samples have demonstrated that the rate of carbon dioxide release from the hemolyzed blood was higher than expected, indicating that blood could contain a catalyst for the dehydration of bicarbonate, which allows the formation of CO2 [1]. This catalyst, named carbonic anhydrase, was thereafter extracted from erythrocytes in 1933 by Meldrum and Roughton [2]. Upon the discovery in 1940 that zinc ions are an intrinsic cofactor of the protein, CA became the first recognized metalloenzyme. This enzyme efficiently catalyzes the reversible hydration of carbon dioxide (CO2) to yield bicarbonate (HCO3-) and protons (H+) [2,3].

It has been known since the 1940s that CA is ubiquitous in plants [4], where it performs an essential role in CO2 fixation [5]. CAs, under the form of many enzyme families and isoforms, are virtually found in all living organisms, from the unicellular ones to higher vertebrates including humans. Their structure is encoded by eight evolutionary unrelated gene families, leading thus to the α-, β-, γ-, δ-, ζ-, η-, Θ-, and ι-CA classes [6-13]:

α-CAs are Zn2+ metalloproteins expressed in animals, vertebrates, prokaryotes, fungi, algae, protozoa and plants [9].

β-CAs are Zn2+ metalloproteins present in bacteria, plants, fungi, chloroplasts of mon-/dicotyledons [6].

γ-CAs are Zn2+ or Fe, Co metalloproteins present in some plants, fungi, bacteria and archarea [6].

δ-CAs are Co metalloproteins present in marine diatoms [7,10].

ζ-CAs are Cd or Zn metalloproteins identified only in some marine diatoms [11].

η-CA are Zn metalloproteins identified in Plasmodium spp. [12].

Θ-CA are Zn metalloproteins identified in Marine diatoms [11].

ι-CAs were only recently reported to be present in diatoms and bacteria and seem to be Mn(II) proteins [13].

CA inhibitors (CAIs) targeting mammalian CAs, are in clinical use as diuretics, antiglaucoma, antiepileptic or antiobesity agents for decades [3,6,14-18]. These diverse applications are due to the fact that at least 15 different α-CA isoforms are present in humans, being involved in critical physiological and pathological processes [14-18].

In the current review, we focused our attention on recent activation studies on α-, β-, γ-, δ-, ζ-, and η-CA classes which were explored with at least two classes of modulators of activity, amines and amino acids. The catalytic mechanism of these enzymes is in fact well understood [3]. A metal hydroxide species present in the active site of these enzymes as the fourth ligand (Figure 1(A,B)) acts as a strong nucleophile (at physiologic pH) converting the CO2 to bicarbonate, bis-coordinated to Zn(II), in a trigonal bipyramidal geometry (Figure 1(C)). This adduct is not very stable and reaction with a water molecule leads to liberation of bicarbonate in solution and generation of an acidic form of the enzyme incorporating a M2+(OH2) species at the metal center, which is catalytically ineffective for the hydration of CO2 (Figure 1(D)). In order to generate the nucleophilic, M 2+(OH_) species, a proton transfer reaction occurs, which is rate determining for the catalytic cycle in many of these quite rapid enzymes. CA enzymes typically use a metal ion (Zn2+ in α-, β- and γ-CAs, Fe2+/Co2+/Zn2+) which favors in the reduction pKa of H2O from 14 to 7 [6-8]. Human CAs use a Zn2+ ion to decrease the pKa of H2O bound with Zn2+ ion which also binds to histidine residues (His94, His96 and His119). For many α-CAs this step is assisted by a proton shuttle residue, which is His64 in most mammalian isoforms. Possessing a flexible orientation, inwards (the in conformation) or outwards (the out conformation) the zinc ion center, the imidazole moiety of this histidine, with a pKa of 6.0–7.5 is an appropriate proton shuttling residue and crucially important for the entire catalytic cycle. The process can be also assisted by endogenous molecules, which bind within the enzyme active site (as proven by X-ray crystallography and other techniques) which have been termed CA activators (CAAs) [19]. They facilitate the proton transfer reactions between the metal ion center and the external medium. It was understood that CA activators act by speeding up the deprotonation of zinc-bound water (the rate-determining step, Equation (2) in the catalytic mechanism) [19-21], with the generation of the active form of the enzyme [22] (see equations below):

Figure 1.

Catalytic mechanism of α-CAs [3]. A. The zinc hydroxide form of the enzyme. B. The bucleophilic attack on CO2 bound in the hydrophobic pocket. C. Bicarbonate bound to the active site metal ion. D. Acidic form of the enzyme. B in the last step of the cycle is a buffer molecule or the imidazole moiety of a His64 residue from the enzyme active site, acting as proton shuttle.

In the presence of an activator ‘A’, Equation (2) becomes (3):

enzyme - activator complexes

CAAs may have pharmacologic applications, activation of the mammalian enzymes was shown to enhance cognition and memory in experimental animals [23], likewise its inhibition has an opposite effect [24].

In order to better understand the catalytic mechanism of CAs belonging to the β-, δ- γ-, δ- ζ-, η-CA and Θ-CA classes, it is of crucial importance to see if these enzymes act, similar to the α-CAs, which can be activated by compounds that shuttle protons between the active site and the environment. The activation of CAs from pathogenic bacteria may be relevant for understanding the factors governing virulence and colonization of the host, because pH in the tissues surrounding the pathogens likely plays a key role in such processes and many compounds that are CAAs (biogenic amines and amino acid derivatives) are abundant in such tissues. In this review, we have carefully analyzed the activation potential of different natural, non natural, aromatic/heterocyclic amino acids and amines (compounds 1–19) across 6 different families of CAs that were investigated based on the existing literature (Chart 1) [19-24]. These compounds have functional groups similar to their endogenous proton shuttlers, and can participate in proton transfer processes during the catalytic cycle. This study is relevant as no X-ray crystal structures of enzyme activator complexes have been reported so far for β- γ-, δ-, ζ-, η-CA and Θ-CAs.

Chart 1.

Amino acids 1–11 and amines 12–19.

Activation of α-CAs with amino acids and amines

Activation of the twelve catalytically active human (h) or murine (m) CA isoforms, hCA I, hCA II, hCA III, hCA IV, hCA VA, hCA VB, hCA VII, hCA IX, hCA XII, mCA XIII, hCA XIV and mCA XV with amino acids and amines (1–19) has been investigated by stopped flow CO2 hydrase assay method and are shown in Table 1 [25-29]. This bioassay is in excellent agreement with results from native mass spectrometry [30]. The following structure-activity relationship (SAR) can be summarized from data presented in Table 1 based on the activation profile of these derivatives.

Table 1.

In vitro hCA I [25], hCA II [25], hCA III [26], hCA IV [26], hCA VA [27], hCA VB [27], hCA VII [28], hCA IX [29], hCA XII [29], mCA XIII [25], hCA XIV [28] and mCA XV [30] activation data with amines and amino acids (1–19) by a stopped-flow CO2 hydrase assay.

		KA (µM)a
No	Compound	hCA I	hCA II	hCA III	hCA IV	hCA VA	hCA VB	hCA VII	hCA IX	hCA XII	mCA XIII	hCA XIV	mCA XV
1	L-His	0.03	10.9	35.9	7.30	1.34	0.97	0.92	9.71	37.5	0.13	0.90	32.1
2	D-His	0.09	43	1.13	12.3	0.12	4.38	0.71	12.5	24.7	0.090	2.37	14.1
3	L-Phe	0.07	0.013	34.7	36.3	9.81	10.45	10.93	16.3	1.38	1.02	0.24	33.4
4	D-Phe	86	0.035	15.4	49.3	4.63	0.072	9.74	9.30	0.37	0.051	7.21	9.5
5	L-DOPA	3.1	11.4	13.5	15.3	0.036	0.063	58.3	51.3	1.67	43	12.1	6.5
6	D-DOPA	4.9	7.8	28.7	34.7	4.59	3.71	34.7	54.7	0.89	0.73	36.8	4.0
7	L-Trp	44	27	20.5	37.1	1.13	0.89	57.5	37.5	26.0	16	16.5	13.5
8	D-Trp	41	12	19.0	39.6	1.24	1.35	39.6	43.6	28.1	0.81	18.0	8.7
9	L-Tyr	0.02	0.011	34.1	25.1	2.45	0.044	20.3	25.3	25.8	–	21.8	8.9
10	D-Tyr	0.04	0.013	–	–	–	–	–	–	–	–	–	–
11	4-H2N-L-Phe	0.24	0.15	43.2	0.079	2.76	2.17	18.7	48.7	1.09	–	2.90	16.3
12	Histamine	2.1	125	36.9	25.3	0.010	3.52	37.5	35.1	27.9	4.6	0.010	18.5
13	Dopamine	13.5	9.2	33.2	30.9	0.13	7.85	0.89	0.92	0.67	27	14.6	7.1
14	Serotonin	45	50	0.78	3.14	6.33	0.11	0.93	33.1	0.30	0.51	6.5	7.5
15	2-Pyridyl-methylamine	26	34	1.03	5.19	23.56	0.24	43.7	1.07	41.5	3.8	21.7	11.6
16	2–(2-Aminoethyl)pyridine	13	15	1.10	7.13	7.62	0.094	27.8	0.013	0.69	46	6.9	11.9
17	1–(2-Aminoethyl)-piperazine	7.4	2.3	0.32	24.9	6.04	0.91	32.5	0.009	48.3	54	18.3	10.4
18	4–(2-Aminoethyl)morpholine	0.14	0.19	0.091	1.30	0.089	1.15	64.3	0.43	0.24	0.013	5.4	9.3
19	L-Adrenaline	0.09	96	36.4	45.0	–	–	–	60	0.87	–	36.1	6.9

Mean from 3 different determinations (errors in the range of 5–10% of the reported values.

Compounds 1–19 generally activated, these CA isoforms in a very different manner based on their structures. Nanomolar potencies were observed for several isozymes. For example, hCA I was activated by compounds 1 (L-His), 2 (D-His), 3 (L-His), 9 (L-Tyr), 10 (D-Tyr), and 19 (L-adrenaline) with KAs ranging from 20 to 90 nM. The best activation profile was observed against one of the most abundant cytosolic isoform hCA II with KAs ranging from 125 µM to 11 nM. Specifically, compounds 3 (L-His), 9 (L-Tyr), and 10 (D-Tyr) showed good activation potency with KAs of 13, 11 and 13 nM, respectively. Other cytosolic isoforms hCA III and hCA VII were weakly activated, in general, by these series of amines and amino acids 1–19. The remaining cytosolic isoform mCA XIII was moderately activated by most of the compounds with KAs ranging from 0.24 to 48.3 µM. Among the mitochondrial isoforms hCA VB was slightly better activated than hCA VA by these amines and amino acids. Interestingly, compound 5 (L-DOPA) showed nanomolar potency against both isozymes, hCA VA and VB, with KAs of 36 and 63 nM, respectively. Only one compound 11 (4-H2N-L-Phe) had nanomolar activity against membrane-bound isoform hCA IV with a KA of 79 nM. On the other hand, another transmebrane-bound tumor overexpressed isoform hCA IX was moderately activated by most of the tested compounds, except the compound 16 and 17 which showed one of the best activation profile from the Table 1 with KAs of 13 and 9 nM, respectively. The CA activating effects of amines and amino acids 1–19 on the remaining membrane-bound isoforms hCA XII, hCA XIV and mCA XV were moderate to weak and most of the results were very close the each others (Table 1).

Activation of β-CAs with amino acids and amines

In literature, there are many β-CAs which were investigated in details, among which Cab (from Methanobacterium thermoautotrophicum), scCA (from Saccharomyces cerevisiae), CgCA (from Candida glabrata), MgCA (from Malassezia globosa), VchCAβ (from Vibrio cholerae, mtCA 3 (from Mycobacterium tuberculosis), BsuCA1 (from Brucella suis), FtuCA (from Francisella tularensis), LdcCA (from Leishmania donovani chagasi), and EhiCA (from Entameba histolytica). Their activation by by amines and amino acid derivatives was investigated in the last several years [31-38] – Table 2. In general, good to moderate activation effects were obtained against all β-CAs, except FtuCA, by using amino acid and amine derivatives 1–19. Among the β-CAs, the best activation profile was observed against LdcCA, for which two compounds, 16 and 17, showed nanomolar potency, with KAs of 12 and 9 nM, respectively. Interestingly, these two compounds have (2-aminoethyl) groups in their structures. Other interesting results were obtained against scCA for which the activation profile was better with amines (KAs: 0.95–21.3 µM) than with amino acids (KAs: 82–91 µM). Furthermore, VchCAβ and BsuCA1 was also activated efficiently, with KAs of 0.18–20.3 and 0.70–43.1 µM, by amino acids and amines, respectively. Specifically, VchCAβ was activated slightly more effectively by amines (KAs: 0.18–12.8 µM) than by amino acid derivatives (KAs: 0.94–20.3 µM). For BsuCA1 activities of most compounds are close to each other, except the compounds 2, 8, and 17 with KAs of 12.3, 13.7 and 43.1 µM, respectively, which are the least effective CAAs. In the case of FtuCA, most of the amines and amino acid derivatives (compounds 5, 9–14, 16, 18 and 19) investigated so far showed weak activation effects, with activation constants >100 µM. The remaining activators were also moderately active against FtuCA, with KAs ranging between 30.5 to 78.3 µM. Other β-CAs (Cab, CgCA, MgCA, mtCA 3 and EhiCA) were activated in a different manner, as seen from Table 2, with most of the activation constants in a limited range of values.

Table 2.

In vitro β-CA (Cab [31], scCA [31–34], CgCA [34], MgCA [32], VchCAβ [35], mtCA 3 [36], BsuCA1 [37], FtuCA [37], LdcCA [38, 39], and EhiCA [38]) activation data with amines and amino acids (1–19).

		KA (µM)a
No	Compound	Cab	scCA	CgCA	MgCA	VchCAβ	mtCA 3	BsuCA1	FtuCA	LdcCA	EhiCA
1	L-His	69	82	37	29.3	20.3	18.2	1.76	40.7	8.21	78.7
2	D-His	57	85	21.2	18.1	18.0	32.5	12.3	78.3	4.13	9.83
3	L-Phe	70	86	24.1	34.1	15.4	30.6	1.16	69.1	9.16	16.5
4	D-Phe	10.3	86	15.7	10.7	5.12	44.1	1.21	75.0	3.95	10.1
5	L-DOPA	11.4	90	23.3	8.31	8.36	30.0	2.07	>100	1.64	16.6
6	D-DOPA	15.6	89	15.1	13.7	6.27	9.74	2.34	44.8	5.47	4.05
7	L-Trp	16.9	91	22.8	10.1	4.18	8.98	1.25	34.1	4.02	5.24
8	D-Trp	41	90	12.1	12.5	5.89	43.7	13.7	30.5	6.18	4.95
9	L-Tyr	10.5	85	9.5	15.7	6.15	28.9	1.38	>100	8.05	4.52
10	D-Tyr	19.2	84	7.1	25.1	0.94	17.6	0.95	>100	1.27	1.07
11	4-H2N-L-Phe	89	21.3	31.6	13.4	7.21	40.5	1.18	>100	15.9	8.12
12	Histamine	76	20.4	27.4	10.9	9.50	34.2	3.71	>100	0.74	7.38
13	Dopamine	51	13.1	27.6	9.43	1.24	12.1	1.54	>100	0.81	30.8
14	Serotonin	62	15.0	16.7	14.2	1.37	10.3	4.26	>100	0.62	4.94
15	2-Pyridyl-methylamine	18.7	16.2	15.0	6.12	0.18	43.3	1.62	46.3	0.23	>100
16	2–(2-Aminoethyl)pyridine	40	11.2	16.3	7.30	1.00	45.9	5.20	>100	0.012	>100
17	1–(2-Aminoethyl)-piperazine	13.8	9.3	14.9	0.81	0.24	50.3	43.1	51.8	0.009	43.8
18	4–(2-Aminoethyl)morpholine	18.5	10.2	10.1	5.82	12.8	52.0	9.56	>100	0.94	>100
19	L-Adrenaline	11.5	0.95	10.8	0.72	8.73	52.2	0.70	>100	4.89	25.6

Mean from 3 different determinations (errors in the range of 5–10% of the reported values, data not shown).

Activation of γ-, δ-, ζ-, and η-CAs with amino acids and amines

Activation studies were also performed recently against γ-CAs, such as Zn-Cam and Co-Cam (from the Archaeon Methanosarcina thermpophila), BpsγCA (from the pathogenic bacterium Burkhalderia pseudomallei), PhaCA (from the cyanobacterium Pseudoalteromonas haloplanktis), and CpsCA (from another cyanobacteriu, Colwellia psychrerythraea), as well as δ-CAs, such as TweCAδ (from the diatom Thalassiosira weissflogii)], ζ-CA, such as ZnTweCAζ (from the same diatom, Thalassiosira weissflogii)], and η-CAs, such as PfaCA (from Plasmodium falciparum) [31, 40–44]. Among them, an interesting activation profile was observed for some of the γ- class CAs, such as BpsγCA. Most of the tested compounds showed nanomolar potency against this enzyme. Specifically, BpsγCA was efficiently activated by compounds 2, 5, 8, 11, 13, and 16–19 with activation constants ranging between 9 to 86 nM. Interestingly, the ζ- class CA, ZnTweCAζ was activated slightly more efficiently by amines (KAs of 92 nM to 10.1 µM) than by amino acids (KAs of 0.62 to 15.4 µM), which is just the opposite in the case of the η- class CA PfaCA, for which KAs ranging from 0.12 to 8.55 µM were obtained for amino acid derivatives and between 0.71 and 9.97 µM for amines (Table 3). A wide range of activities of the various activators for the remaining CAs was observed, such as for γ- class of CAs, Co-Cam and PhaCA, which were moderately activated by amino acid derivatives and amines with KAs of 0.72–135 µM (Table 3). Other γ-CAs, such as Zn-Cam and CpsCA were less prone to be activated, as compared to other γ- CAs investigated so far, with activation constants ranging between 4.79 to >100 µM. The unique δ- class CA investigated in details at this moment, TweCAδ, was efficiently activated by most of the amino acid derivatives and amines 1–19, with KAs ranging between 51 nM and 18.9 µM.

Table 3.

In vitro γ-CA (Zn-Cam [31], Co-Cam [31], BpsγCA [40], PhaCA [41], and CpsCA [41]), δ-CA (TweCAδ) [42], ζ-CA (ZnTweCAζ) [43], and η-CA (PfaCA) [44] activation data with amines and amino acids (1–19).

		KA (µM)a
No	Compound	Zn-Cam	Co-Cam	BpsγCA	PhaCA	CpsCA	TweCAδ	ZnTweCAζ	PfaCA
1	L-His	68	135	24.7	12.6	47.5	0.75	0.81	1.06
2	D-His	46	73	0.086	9.41	35.9	4.90	7.16	2.19
3	L-Phe	68	70	1.73	15.8	>100	2.15	15.4	0.43
4	D-Phe	42	24	0.13	3.19	15.4	1.96	9.63	0.75
5	L-DOPA	39	38	0.072	1.08	4.79	2.11	3.19	0.12
6	D-DOPA	37	41	0.98	0.72	11.2	6.24	2.87	0.39
7	L-Trp	38	47	0.43	7.12	21.3	0.93	8.54	5.21
8	D-Trp	33	68	0.052	13.9	36.8	0.69	1.79	8.47
9	L-Tyr	24	53	0.20	1.02	19.5	1.52	0.98	1.02
10	D-Tyr	–	–	32.8	7.35	18.4	0.051	0.62	8.55
11	4-H2N-L-Phe	72	22	0.009	3.27	17.2	18.9	7.90	1.00
12	Histamine	63	9.2	0.12	6.48	20.6	1.34	1.27	9.86
13	Dopamine	54	18.4	0.014	8.70	32.1	0.51	10.1	9.97
14	Serotonin	38	0.97	0.10	9.05	34.8	0.90	3.06	7.18
15	2-Pyridyl-methylamine	11.4	8.7	2.36	2.39	21.5	5.28	0.88	3.69
16	2–(2-Aminoethyl)pyridine	24	18.5	0.034	18.7	38.2	8.16	0.85	6.75
17	1–(2-Aminoethyl)-piperazine	10.1	16.1	0.018	15.1	33.0	4.37	0.12	0.71
18	4–(2-Aminoethyl)morpholine	45	38	0.015	10.1	34.2	7.39	0.15	5.33
19	L-Adrenaline	39	8.9	0.019	17.5	79.8	2.43	0.092	2.40

Mean from 3 different determinations (errors in the range of 5–10% of the reported values, data not shown).

Conclusions and future perspective

To our knowledge, this is the first article that summarizes the activation profile of all classes of CAs (the α-, β-, γ-, δ-, ζ-, and η-CA) with a small library of amines and amino acid derivatives. This panel of investigated amino acids and amines showed considerable activating properties, with a well defined structure–activity relationship, but without net differences between the various CA families. Even if the available activators are not isoform-selective (for the many α-CAs of human or other origins), as already mentioned above, in the last period, their possible use as pharmacological agents for memory therapy or for artificial tissues engineering started to be explored [23,24], with very promising results being obtained. There is however a stringent need for having more effective, isoform-selective and possibly non-autacoid or amino acid derived compounds, which may possess a rather complicated polypharmacology [3]. Furthemore, the investigations of the activating effects of non-human CAs are still in their infancy, with very few in vitro studies being available on the non-α-CA activators. Indeed, only in the few several years the first activation studies of β-, γ-, δ-, ζ-, and η-CAs from various organisms have been reported, which allowed the identification of compounds active in the nanomolar to micromolar range. However, no drug design studies of CAAs targeting these enzymes were performed so far, which is one of the future objectives of research in this area. In addition, almost nothing is known regarding the in vivo effects of CAAs in organisms other than the vertebrates (human and rodents). As briefly mentioned, many pathogenic bacteria, fungi or protozoans live in various niches which are potentially rich in endogenous activators of the amine and amino acid type. A deep understanding of the role that these modulators of activity may play in the interaction between the host and the pathogen, may lead to relevant biomedical discoveries, but this is an entire new field to be explored in the future.