Crystal structure of human carbonic anhydrase‐related protein VIII reveals the basis for catalytic silencing

Carbonic anhydrases (CAs; EC 4.2.1.1) catalyze the reversible hydration of carbon dioxide to bicarbonate and a proton.1 They are ubiquitous in prokaryotes and eukaryotes, and are encoded by four evolutionarily-unrelated gene classes (α, β, γ, δ). The human CA family belongs to the α class and contains 12 catalytic isozymes with different tissue distribution, subcellular localization and kinetic properties. These include the cytosolic (CA I, CA II, CA III, CA VII, CA XIII), membrane-bound (CA IV, CA IX, CA XII, CA XIV), mitochondrial (CA VA, CA VB) and secreted (CA VI) isoforms. They share an absolute requirement for a catalytic Zn2+ ion in the active site, coordinated by an hydroxide ion and three invariant histidine residues H94CAII, H96CAII and H119CAII (human CA II numbering) that are in turn hydrogen bonded to conserved partners Q92CAII, N244CAII and E117CAII, respectively. CAs participate in physiological processes such as respiration, metabolite biosynthesis and pH regulation, and are interesting pharmaceutical targets.2 Sulphonamide-based CA inhibitors are established diuretics and anti-glaucoma drugs, and may have further therapeutic potentials in anti-obesity and anti-cancer treatment.3 The human CA family includes a subclass of three noncatalytic isoforms (CA VIII, CA X, CA XI), also known as CA-related proteins (CA-RPs), based on sequence homology with the catalytic isozymes. CA-RPs lack one or more of the essential Zn2+-coordinating histidines and are devoid of CO2 hydration activity.4 Todate, the biological functions of CA-RPs remain undefined. The first identified CA-RP, CA VIII, replaces the Zn2+-coordinating H94CAII and its hydrogen-bonding partner Q92CAII with R116 and E114, respectively (human CA VIII numbering).5 CA VIII is highly expressed in the cerebellum,6 and a mouse gene deletion causes a motor coordination defect.7 Relevant to this, CA VIII has been identified as a binding partner for the inositol 1,4,5 triphosphate (IP3) receptor type 1 which is abundant in the cerebellum.8 To provide insights into the biological properties of CA-RPs, we have determined the 1.6 Å crystal structure of human CA VIII (hCA VIII). This work represents the first structural characterization of a CA-RP and offers a structural basis for its catalytic silencing effect. A DNA fragment encoding full-length hCA VIII (aa 1–290; GenBank entry 22027500) was subcloned into the pNIC28-Bsa4 vector which incorporates a TEV-cleavable His6-tag. The plasmid was transformed into BL21(DE3)pRARE, cultured in 1 L Terrific Broth at 37°C, and induced with 0.1 mM IPTG overnight at 18°C. Cells were homogenized in lysis buffer (50 mM KH2PO4 pH 8.0, 500 mM NaCl) and insoluble material was removed by centrifugation. The supernatant was purified by affinity (Ni-sepharose) and size exclusion (Superdex S75) chromatography. Purified protein was treated with His-tagged TEV protease (1:100 v/v) overnight at 4°C, and passed over 1 mL Ni-sepharose resin. Protein was concentrated to 10.5 mg/ml and stored in 10 mM HEPES pH 7.5, 100 mM NaCl, 5% (w/v) glycerol at −80°C. Crystals were grown at 4°C by vapor diffusion in sitting drops mixing 100 nL protein (10.5 mg/mL) and 200 nL well solution containing 10% (w/v) PEG 6000, 0.3 M NH4Cl pH 6.3 and 10% (v/v) ethylene glycol. Crystals were cryo-protected using 20% (v/v) ethylene glycol and flash-cooled in liquid nitrogen. Diffraction data to 1.6 Å resolution were collected at the SLS beamline X10SA. Data were indexed and integrated using MOSFLM,9 and scaled and merged using SCALA from the CCP4 program suite.10 Human CA-related protein VIII crystallized in the P212121 space group with one molecule in the asymmetric unit. The structure was solved by molecular replacement using the program PHASER11 and the hCA XIII structure (PDB code 3DA2) as search model. Automated model building was performed with ARP/wARP,12 followed by iterative cycles of restrained refinement and model building using COOT13 and REFMAC5.14 The final model consists of aa 24–290, whereas aa 1–23 were not observed in the 2Fo − Fc electron density map. During refinement a Fo − Fc difference peak was observed at the bottom of the active site cavity. A chloride ion was modeled into the peak, and refined at full occupancy with B factor comparable with the neighboring protein atoms. Modeling a Zn2+ ion into the peak resulted in negative Fo − Fc density. Atomic coordinates and structure factors have been deposited in the PDB under the accession code 2W2J. Data collection and refinement statistics are summarized in Table I. Figures were prepared using PYMOL (www.pymol.org) and ICM Pro (www.molsoft.com). The final model of hCA VIII [Fig. 1(A)] comprises part of an N-terminal Glu-rich region (E loop; aa 24–36) that has no counterparts in other CAs, as well as the central core domain (aa 37–290). The first 23 residues were not observed in the electron density map and are presumably disordered. The core domain of hCA VIII adopts the classical architecture of the mammalian CA enzymes, namely a 10-stranded central β-sheet (β11, β4, β5, β7, β9, β10, β13, β12, β15, β2) surrounded by several short α-helices (α1–α3) and β-strands (β1, β3, β6, β8, β14). The overall structure resembles closely the cytosolic isozymes CA II and CA XIII (RMSD 1.3 Å, 41% sequence identity). CA II is a well-characterized isozyme with the highest catalytic turnover.15 CA XIII has been identified recently16 and we have determined its structure (PDB code 3DA2; Pilka ES, Picaud SS, Oppermann U, Yue WW, unpublished). Comparison between the hCA VIII and hCA XIII structures reveals significant differences in two loop regions [Fig. 1(B)]. In hCA VIII, the unique E loop protrudes into the exterior and packs against the α3-β15 loop which incorporates a five-residue insertion compared with other isozymes. Together the two loops contribute to an extensive electronegative surface in hCA VIII, which contrasts with the more neutral surface in other isozymes (data not shown). Crystal structure of hCA VIII. (A) Ribbon diagram showing the secondary structure elements: α-helices (red) and β-strands (yellow). (B) Cα-superposition of hCA VIII (blue) and hCA XIII structures (cyan, PDB code 3DA2). The two loop regions that are unique in hCA VIII are shown in thick ribbons. The catalytic isozymes contain a conical cavity in the core domain with a 15-Å wide entrance, which tapers into the active site at the bottom of the cavity. This cavity provides a crucial transit route for the substrates, products and the Zn2+ ion. Compared with other CA structures, the hCA VIII cavity has a much narrower opening because of steric constriction by two bulky residues, R116 and I224, at the cavity entrance [Figs. 1(B) and 2(A)]. R116 replaces the essential Zn2+-coordinating H94CAII and I224 replaces T200CAII conserved in most isozymes [Fig. 2(B)]. Unlike H94CAII side-chain, which points into the active site for Zn2+ coordination, the R116 guanidino side-chain of hCA VIII is directed away from the cavity by polar interactions with D85 and E114 [Fig. 2(A)], which correspond to Asn (N244CAII) and Gln (Q92CAII), respectively, in most isozymes. Remarkably, mutations of R116, E114 and I224 to the corresponding residues in the catalytic isozymes rendered CA VIII enzymatically active.17, 18 Structural comparison of hCA VIII with the catalytic isozymes. (A) A close-up view showing the entrance of the hCA VIII cavity. hCA VIII is represented in blue ribbon, overlayed with a surface representation. (B) Active site of the CO2-bound hCA II (left, PDB code 3D92) and hCA VIII (right) structures. The hydrophobic CO2 pocket is shown in pink surface. Spheres represent the Zn2+ ion (black), water (red) and Cl− ion (green). In the right panel, the CO2 substrate from the hCA II structure (shown in dots) is superimposed onto the hCA VIII structure to illustrate the potential steric clash caused by I165 and I143 (green surface). (C) Schematic representation of the Zn2+ coordination in hCA II (left), and Cl− coordination in hCA VIII (right). Bond lengths are indicated in Å. In the catalytic isozymes, the active site constitutes a bipolar binding surface: a hydrophilic face (T199CAII, T200CAII, H96CAII, H94CAII, H119CAII) which contains the Zn2+ binding site and a hydrophobic face (V121CAII, V143CAII, L198CAII, V207CAII, W209CAII) which harbors the CO2 substrate pocket [Fig. 2(B), left]. Inspection of the hCA VIII structure reveals significant differences in the equivalent ‘active site’ [Fig. 2(B), right]. First, the hydrophobic pocket of CA VIII is more spatially constricted due to the substitution of two Val residues (V121CAII, V143CAII) to Ile (I143, I165). This reduces the pocket size and presumably precludes the binding of CO2 in the active site. Consistent with this, a V143CAII-to-Ile mutation diminished CA II activity by eightfold.19 Second, a difference Fourier peak was observed at the bottom of the active site cavity in hCA VIII, close to the position of Zn2+ in the catalytic isozymes, but is incompatible with the electronic configuration of Zn2+ (see “Materials and methods”). Taking into account (1) the inclusion of NaCl during purification and (2) the absence of Zn2+ in the crystallization solution, we have modeled and refined a chloride ion into the density [Fig. 2(B), right]. The Cl− ion is coordinated by the Nδ1 atom from H118 (bond length 3.02 Å), Nε2 atom from H141 (3.01 Å) and a water molecule W15 (3.15 Å). The observed bond lengths are significantly different from the typical Zn2+ coordination [Fig. 2(C)], but are consistent with the observed noncovalent Cl–N bond distances (2.9–3.2 Å) from our survey of PDB structures (resolution <2.0 Å, modeled with Cl−; data not shown). Selected examples from the PDB with observed Cl−-histidine interactions are listed in Table S1 of the Supporting Information. Our structural data revealed that CA VIII evolves to adopt the classic CA fold but substitutes key residues in the active site cavity that preclude the entry and binding of the catalytic Zn2+ and substrate CO2, hence silencing its activity. Nevertheless, the high sequence conservation between human and mouse CA VIII homologs (98% identity) lends support to the hypothesis that over the course of evolution the absence of CA activity may be associated with the gain of a new, yet unidentified cellular function for CA VIII. One possibility is the modulation of biological functions via protein–protein interactions. The type 1 IP3 receptor, identified as a CA VIII-binding protein, contains an electropositive IP3 binding site.20 It is possible that the electronegative surface in CA VIII, unique among CAs, may form a charge-complementary binding site for the receptor, thereby regulating IP3-dependent Ca2+ release.7 Consistent with this, a protein-interaction function has been reported for the CA-like domain in receptor protein tyrosine phosphatase β which serves as the contactin binding site.21 The Structural Genomics Consortium is a registered charity (Number 1097737) funded by the Canadian Institutes for Health Research, the Canadian Foundation for Innovation, Genome Canada through the Ontario Genomics Institute, GlaxoSmithKline, Karolinska Institutet, the Knut and Alice Wallenberg Foundation, the Ontario Innovation Trust, the Ontario Ministry for Research and Innovation, Merck and Co., Inc., the Novartis Research Foundation, the Swedish Agency for Innovation Systems, the Swedish Foundation for Strategic Research and the Wellcome Trust. Additional Supporting Information may be found in the online version of this article. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.