Crystal structure of carbon–nitrogen hydrolase from Helicobacter pylori G27

This article is part of a special issue celebrating early career researchers in structural science.

Carbon–nitrogen hydrolases (CNHs) are members of the diverse nitrilase superfamily of enzymes that facilitate cellular adaptation to environmental stress by metabolizing nitrogen, detoxifying xenobiotics and catabolizing environmentally derived metabolites. Helicobacter pylori CNH (HpCNH) may contribute to metabolic flexibility under acid stress, detoxification of reactive nitrogen species or nutrient scavenging in the nutrient-limited gastric environment. Here, we report the 2.1 Å resolution crystal structure of a CNH from H. pylori strain G27 (PDB entry 6mg6). HpCNH adopts the characteristic nitrilase-superfamily αββα-sandwich core and contains the conserved catalytic cysteine typical of enzymatically active CNHs. The overall structure and active site of HpCNH are most similar to those of carbamoylputrescine amido­hydrolase from the plant Medicago truncatula. Despite structural variations in loop regions, including near the active site, HpCNH retains the key residues required to bind putrescine and the prototypical N-carbamoylputrescine amidase active site.

1. Introduction

Helicobacter pylori is a Gram-negative, microaerophilic bacterium that colonizes the human gastric mucosa and infects more than half of the global population (Warren & Marshall, 1983; Malfertheiner et al., 2023; Moss et al., 2023). Persistent H. pylori infection is the primary cause of chronic gastritis, peptic ulcer disease and gastric adenocarcinoma, and the World Health Organization classifies H. pylori as a group 1 carcinogen (Malfertheiner et al., 2023; Moss et al., 2023; Cover & Blaser, 2009). H. pylori persists in the gastric mucosa by modulating environmental conditions and host responses using mechanisms including hydrolysis of urea to buffer acidic pH, modulation of gene expression in response to acid stress and alteration of gastric epithelial biology (Benoit et al., 2020; Kuhns et al., 2016; Warren & Marshall, 1983; Malfertheiner et al., 2023; Moss et al., 2023).

Structure–function studies of H. pylori proteins, including carbon–nitrogen hydrolases (CNHs), at the Seattle Structural Genomics Center for Infectious Disease (SSGCID) are under way to generate insights into the structural basis of H. pylori survival mechanisms. This is because CNHs facilitate cellular adaptation to environmental stress by metabolizing nitrogen, detoxifying xenobiotics and catabolizing environmentally derived metabolites (Bork & Koonin, 1994). CNHs hydrolyze nonpeptide C—N bonds in substrates such as amides, nitriles, carbamates and urea, and are widely distributed across bacteria, archaea and plants (Bork & Koonin, 1994). Many bacterial CNHs, including H. pylori CNH (HpCNH), remain poorly characterized, and structural characterization is essential to infer biochemical function, substrate specificity and potential physiological roles.

HpCNH is annotated as an N-carbamoylputrescine amidase, placing it within the putrescine biosynthetic process from arginine, a pathway critical for polyamine metabolism. Polyamines such as putrescine play key roles in cell proliferation, biofilm formation and acid resistance in many bacterial species, making this enzyme a potential player in the adaptation of H. pylori to gastric acidity and stress (Nair et al., 2025; Sagar et al., 2021). We present the recombinant production and crystal structure of HpCNH as part of efforts to clarify its functions.

2. Materials and methods

2.1. Macromolecule production

HpCNH was cloned, expressed and purified using standard protocols established at the SSGCID (Stacy et al., 2011; Serbzhinskiy et al., 2015; Rodríguez-Hernández et al., 2023). Briefly, the full-length gene for carbon–nitrogen hydrolase from H. pylori G27 (UniProt B5Z8D8) encoding amino acids 1–292 was PCR-amplified from genomic DNA using the primers shown in Table 1. The gene was cloned into the expression vector BG1861 encoding an N-terminal His-tag to generate plasmid DNA, which was transformed into chemically competent Escherichia coli BL21(DE3)R3 Rosetta cells. The plasmid containing His-HpCNH was tested for expression, and 2 l of culture was grown using auto-induction medium (Studier, 2005) in a LEX Bioreactor (Epiphyte Three) as described previously (Serbzhinskiy et al., 2015). Glycerol stocks of the expression clone may be requested at https://www.ssgcid.org/available-materials/expression-clones/.

Table 1
Macromolecule-production information

Source organism	Helicobacter pylori (strain G27)
DNA source	Nina Salama, FHCRC
Forward primer	5′-CTCACCACCACCACCACCATATGATTTATGCAGGCGTCCTCCA-3′
Reverse primer	5′-ATCCTATCTTACTCACTTAAATATAGCGTTTCAACAAATCGTTATA-3′
Expression vector	BG1861
Expression host	Escherichia coli BL21(DE3) Rosetta
Complete amino-acid sequence of the construct produced	MAHHHHHHMIYAGVLQHAYCGSRKKTIEHTANLLEQALKKHPKTNLVVLQELNPYSYFCQSENPKFFDLGEYFEEDKAFFSALAQKFQVVLIASLFEKRAKGLYHNSAVVFEKDGSIAGVYRKMHIPDDPGFYEKFYFTPGDLGFEPIITSVGKLGLMVCWDQWYPEAARIMALKGAEILIYPSAIGFLEEDSNEEKKRQQNAWETIQRGHAIANGLPLIATNRVGVELDPSGAIKGGITFFGSSFVVGALGEFLAKASDKEEILYAEIDLERTEEVRRMWPFLRDRRIDFYNDLLKRYI

HpCNH was purified using the previously described SSGCID two-step protocol of an immobilized metal (Ni²⁺) affinity chromatography (IMAC) step followed by size-exclusion chromatography (SEC) on an ÄKTApurifier 10 (GE Healthcare) using automated IMAC and SEC programs (Serbzhinskiy et al., 2015). Briefly, thawed bacterial pellets (∼25 g) were lysed by sonication in 200 ml lysis buffer [20 mM HEPES pH 7.0, 300 mM NaCl, 5%(v/v) glycerol, 0.5%(w/v) CHAPS, 30 mM imidazole, 21 mM MgCl₂, 1 mM TCEP and five protease-inhibitor cocktail tablets (cOmplete Mini, EDTA-free, Roche, Basel, Switzerland)]. The solution was incubated with 20 µl Benzonase nuclease (EMD Chemicals, Gibbstown, New Jersey, USA) for 40 min at room temperature under gentle agitation. The lysate was centrifuged with a Sorvall RC5 at 10 000 rev min⁻¹ for 60 min at 4°C in an F14S Rotor (Thermo Fisher, Waltham, Massachusetts, USA). The clarified lysate was filtered through a 0.45 µm cellulose acetate filter (Corning Life Sciences, Lowell, Massachusetts, USA). The IMAC step used a HisTrap FF 5 ml column (GE Bio­sciences, Piscataway, New Jersey, USA) equilibrated in binding buffer [20 mM HEPES pH 7.0, 300 mM NaCl, 5%(v/v) glycerol, 30 mM imidazole] and eluted over a 7 column-volume linear gradient with elution buffer [20 mM HEPES pH 7.0, 300 mM NaCl, 5%(v/v) glycerol, 500 mM imidazole]. IMAC peak fractions were pooled and concentrated to 5 ml using an Amicon Ultra centrifugal filter (Millipore, Billerica, Massachusetts, USA). The SEC step used a Superdex 75 26/60 column (GE Bio­sciences) equilibrated in SEC buffer [20 mM HEPES pH 7.0, 300 mM NaCl, 5%(v/v) glycerol, 1 mM TCEP] attached to an ÄKTAprime plus FPLC system (GE Bio­sciences). 100 ml SEC buffer was run over the column, and peak fractions were collected at 1.5 ml min⁻¹ for an additional 180 ml. Peak fractions were collected and assessed by SDS–PAGE on a 4–20% protein gel (Invitrogen) and visualized by Coomassie staining with InstantBlue colloidal stain (Expedeon, San Diego, California, USA). HpCNH eluted from SEC as a single monodisperse peak of ∼90 kDa (based on molecular-weight standards), suggesting the formation of the assigned dimer by the Protein Interfaces, Surfaces and Assemblies (PISA) service at the European Bioinformatics Institute (https://www.ebi.ac.uk/pdbe/prot_int/pistart.html). SEC peak fractions were pooled, concentrated to ∼20 mg ml⁻¹, flash-frozen in liquid nitrogen and stored at −80°C. Recombinant HpCNH protein is available for request online at https://www.ssgcid.org/available-materials/ssgcid-proteins/.

2.2. Crystallization

HpCNH crystallized at 290 K using sitting-drop vapor diffusion directly from Microlytic MCSG1 screen condition G8 (Table 2). Briefly, equal volumes of 19.9 mg ml⁻¹ HpCNH and 25%(w/v) PEG 3350, 200 mM ammonium sulfate, 100 mM Tris–HCl pH 8.5 were mixed in sitting-drop screens (Table 2). Before data collection, the crystals were harvested and cryoprotected with 15%(v/v) ethylene glycol (Table 2).

Table 2
Crystallization

Method	Vapor diffusion, sitting drop
Plate type	96-well plates, SwissSci MRC2 tray
Temperature (K)	290
Protein concentration (mg ml^−1)	19.9
Buffer composition of protein solution	20 mM HEPES pH 7.0, 300 mM NaCl, 5%(v/v) glycerol, 1 mM TCEP
Composition of reservoir solution	25%(w/v) PEG 3350, 200 mM ammonium sulfate, 100 mM Tris–HCl pH 8.5
Volume and ratio of drop	0.4 µl:0.4 µl
Volume of reservoir (µl)	80
Composition of cryoprotectant solution	21.25%(w/v) PEG 3350, 170 mM ammonium sulfate, 85 mM Tris–HCl pH 8.5, 15% ethylene glycol

2.3. Data collection and processing

Data were collected at 100 K on Advanced Light Source (ALS) beamline 5.0.1 at Lawrence Berkeley National Laboratory (Table 3). Data were integrated with XDS and reduced with XSCALE (Kabsch, 2010). Raw X-ray diffraction images are stored at the Integrated Resource for Reproducibility in Macromolecular Crystallography at https://www.proteindiffraction.org.

Table 3
Data collection and processing

Values in parentheses are for the outer shell.

Diffraction source	ALS beamline 5.0.1
Wavelength (Å)	0.97741
Temperature (K)	100
Detector	Dectris PILATUS3 6M
Space group	P21212
a, b, c (Å)	137.58, 91.38, 95.04
α, β, γ (°)	90, 90, 90
Resolution range (Å)	47.58–2.10 (2.15–2.10)
Completeness (%)	99.400 (100.000)
Multiplicity	6.543 (6.670)
〈I/σ(I)〉	25.59 (2.0)
Rr.i.m.	0.042 (0.589)
Overall B factor from Wilson plot (Å^2)	50.3

2.4. Structure solution and refinement

The structure of HpCNH was determined by molecular replacement with Phaser (McCoy et al., 2007) from the CCP4 suite of programs (Collaborative Computational Project, Number 4, 1994; Krissinel et al., 2004; Winn et al., 2011; Agirre et al., 2023) using PDB entry 5h8i (Sekula et al., 2016) as the search model. The structure was refined using iterative cycles of Phenix (Adams et al., 2011) followed by manual rebuilding of the structure using Coot (Emsley & Cowtan, 2004; Emsley et al., 2010). The structure quality was checked using MolProbity (Williams et al., 2018). Data-reduction and refinement statistics are shown in Table 4. The coordinates and structure factors have been deposited in the Worldwide PDB (wwPDB) as entry 6mg6.

Table 4
Structure refinement

Values in parentheses are for the outer shell.

Resolution range (Å)	47.58–2.10 (2.15–2.10)
Completeness (%)	99.4 (99.0)
No. of reflections, working set	70175 (4496)
No. of reflections, test set	2044 (161)
Final Rcryst	0.168 (0.219)
Final Rfree	0.208 (0.251)
No. of non-H atoms
Protein	9123
Ion	5
Water	393
Total	9521
R.m.s. deviations
Bond lengths (Å)	0.007
Angles (°)	0.845
Average B factors (Å^2)
Protein	56.3
Ion	83.1
Water	50.2
Ramachandran plot
Most favored (%)	98.3
Allowed (%)	2.6
Outliers (%)	0.09

3. Results and discussion

HpCNH crystallized in the orthorhombic space group P2₁2₁2, and the crystal structure was refined at 2.10 Å resolution with an R_work of 0.168 and an R_free of 0.208 (Tables 3 and 4). The final high-quality model has well defined electron density (Fig. 1a). The asymmetric unit contains four copies of the polypeptide (chains A–D), forming a dimer of two tightly packed dimers (Fig. 1b). Each chain forms a prototypical four-layered αββα nitrilase structure, with three β-sheets (Fig. 1c). The 294 amino acids of each chain fold with secondary-structure topology of 27.2% β-strand, 27.2% α-helix, 7.8% 3₁₀-helix and 37.8% other (mostly flexible loop regions). Each HpCNH chain contains three sheets (two six-stranded mixed sheets and one three-stranded antiparallel sheet), five β-hairpins, five β-bulges, 15 strands, 12 helices, 29 helix–helix interactions, 15 β-turns and four γ-turns. Further topological details from PDBsum are given in Supplementary Scheme S1. Pairwise PDBeFold superpositions reveal that the four copies are very similar, with an r.m.s.d. of 0.15–0.40 Å over all main-chain atoms. The four polypeptide chains exhibit minimal conformational heterogeneity at the termini and surface-exposed loops (Fig. 1d).

Figure 1 Overall structure of HpCNH. (a) The refined model fits well into the 1.2σ 2Fo − Fc electron-density map (blue mesh). (b) HpCNH has four protomers in the asymmetric unit, and they are comprised of two tightly packed prototypical CNH dimers, indicated by dimer AB and dimer CD. (c) A cartoon, colored in a rainbow from red at the C-terminus to blue at the N-terminus, of a representative HpCNH protomer reveals the canonical nitrilase fold. The predicted active site is indicated with a molecule of putrescine in a space-filling model based on the structure of MtrCPA (PDB entry 5h8i). (d) A ribbon diagram of superposed protomers, colored in a rainbow from blue at the N-terminus to red at the C-terminus, reveals nearly identical structures.

HpCNH is predicted to be a biological dimer based on PISA analysis (Krissinel, 2015) and agreement with SEC data (Supplementary Fig. S1). There are two HpCNH dimers in the asymmetric unit. Each has a large, buried surface area (Fig. 1b, Table 5). There are 50 interface amino acids per monomer, eight salt bridges, 39 hydrogen bonds and 375 nonbonded contacts at the dimer interface. Both dimers are consistent with biologically relevant dimers observed in other nitrilases (Bork & Koonin, 1994; Nair et al., 2025; Sagar et al., 2021; Sekula et al., 2016).

Table 5
The HpCNH structure contains two prototypical CNH dimers per asymmetric unit of the complex

	HpCNH (chains A:B)	HpCNH (chains C:D)
Multimeric state	Homodimer	Homodimer
Total solvent-accessible surface area of the complex (Å^2)	20679	20621
Buried surface area (Å^2)	5584	4746
Dissociation area (Å^2)	2111	2373
Dissociation energy (ΔGdiss) (kcal mol^−1)	31.83	30.17
Dissociation entropy (TΔSdiss) (kcal mol^−1)	13.85	13.67
No. of interfaces	1	1

ENDScript (Gouet et al., 2003; Robert & Gouet, 2014) analysis revealed the most similar crystal structure to HpCNH to be that of the molecular-replacement search model, M. truncatula carbamoylputrescine amidohydrolase (MtrCPA) in complex with N-(dihydroxymethyl)putrescine (PDB entry 5h8i), followed by the C158S mutant of the same protein, PDB entry 5h8k (Sekula et al., 2016). MtrCPA shares ∼41% sequence identity with HpCNH, as well as an r.m.s.d. of ∼1.4 Å of C_α atoms per protomer. A BlastP search against the PDB reveals it to be the only protein within the PDB that shares over 34% sequence identity with HpCNH. MtrCPA catalyzes the final step of putrescine biosynthesis: the hydrolysis of carbamoylputrescine to putrescine. The closest structure after MtrCPA is that of mouse nitrilase: PDB entry 2w1v (Barglow et al., 2008). The subsequent closest structures are of complexes of the C146A mutant of the amidase from Pyrococcus horikoshii: PDB entries 7ovg and 6ypa (Makumire et al., 2022). The structure of β-alanine synthase from Drosophila melanogaster, PDB entry 2vhh (Lundgren et al., 2008), is more similar to HpCNH than the wild-type amidase from P. horikoshii, PDB entry 1j31 (Sakai et al., 2004). The remaining similar structures are PDB entries 3ivz, 8pt4, 1ems, 6ftq, 2ggk, 1fo6, 2ggl, 1erz, 8hpc, 1uf4, 7elf, 4hg3, 3hkx, 4izt, 4izw, 4hgd, 4h5u and 4f4h (Raczynska et al., 2011; Cederfelt et al., 2023; Pace et al., 2000; Maurer et al., 2018; Chiu et al., 2006; Wang et al., 2001; Jin et al., 2021; Liu et al., 2013; Nel et al., 2011). None of these similar structures are of H. pylori proteins. ENDScript also reveals close alignment of core secondary-structure elements, while highlighting variations in surface loops and peripheral regions (Fig. 2). Conserved residues are shown in red, except for the catalytic cysteine, because the structures include mutant proteins where the cysteine has been changed to various amino acids or is covalently bound to a ligand (Fig. 2). An ENDScript-generated sausage plot reveals that HpCNH conserves the αββα core of other nitrilases (Fig. 3a). The least conserved regions identified by ENDScript alignment are in the loop regions (Fig. 2).

Figure 2 ENDScript analysis reveals the nearest structural neighbors of HpCNH. Identical and conserved residues are highlighted in red and yellow, respectively. The different secondary-structure elements shown are α-helices (α), 310-helices (η), β-strands (β) and β-turns (TT) (Gouet et al., 1999, 2003).

Figure 3 Comparison of HpCNH with its closest structural neighbors. (a) Sausage plot generated by ENDScript. The ribbon (sausage) shows relative secondary-structural conservation compared with other nitrilase structures. The ribbon thickness reflects the level of secondary-structure similarity, with thinner ribbons indicating more conserved regions and thicker ribbons suggesting lower conservation, as determined by r.m.s.d. alignment of structures shown in Fig. 2. The ribbon is colored based on sequence conservation, with red indicating identical residues. Also included is putrescine from PDB entry 5h8i, shown as yellow sticks. (b) HpCNH forms a typical nitrilase homodimer, with interaction between monomers involved in the catalytic cavity. One protomer is depicted as a surface plot, while the other is shown as a cyan cartoon. The protomer represented in the surface plot in (b) is in the same orientation as the protomer in (a). The surface plot is colored by sequence conservation, with red indicating identical residues. (c) Ribbon diagram of superposed tetramers of HpCNH (PDB entry 6mg6, gray) and MtrCPA (PDB entry 5h8i, cyan) reveals a similar overall quaternary structure.

The sausage plot reveals that the core strands exhibit the highest levels of both structural and sequence conservation (Fig. 3a). PDBeFold analysis (https://www.ebi.ac.uk/msd-srv/ssm) with the default threshold cutoff of 70% identified similar results as ENDScript (Supplementary Table S1). Notably, the closest structures to HpCNH were of carbamoylputrescine amidohydrolase (CPA) from the model legume plant M. truncatula (Sekula et al., 2016). Interestingly, the HpCNH tetramer and the MtrCPA tetramer align with an r.m.s.d. of ∼1.5 Å for all main-chain tetramer C_α atoms, revealing a conserved quaternary structure (Fig. 3c). This is comparable to the r.m.s.d. of ∼1.4 Å for all main-chain protomer C_α atoms. Thus, we can conclude that HpCNH and MtrCPA have similar tertiary and quaternary structures.

Although no biologically relevant compounds were soaked or co-crystallized with the HpCNH structure, comparison of the putative active site with that of the ternary complex of MtrCPA with (dihydroxymethyl)putrescine (PDB entry 5h8i) reveals an active site containing the catalytic cysteine and several other residues required for amidase activity (Fig. 4a). Residues that are identically conserved from MtrCPA are Cys158, Pro128, Trp159, Glu48, Ala183, Tyr54 and Lys121, which in HpCNH correspond to Cys152, Pro119, Trp153, Glu43, Ala177, Tyr49 and Lys115, respectively (Fig. 4b). Similar residues in the active site also help to retain charge and interactions. For example, HpCNH contains Phe124, whereas MtrCPA contains Tyr130. A second example is Glu187 in MtrCPA, which is from a loop that is missing in HpCNH. Instead, Asp121 from another loop in HpCNH is positioned similarly and may interact with ligands in the binding cavity in its place. Additionally, Trp277 in MtrCPA, which interacts with ligands across the dimer interface, aligns with Trp271 in HpCNH, as does the helix containing it (Figs. 4a and 4b).

Figure 4 Comparison of HpCNH with MtrCPA reveals conserved putrescine-binding residues. (a) The putrescine (green sticks) binding-cavity residues of both the HpCNH (gray sticks) and MtrCPA (blue sticks) structures are conserved. Also shown are residues from across the dimer interface of HpCNH (cyan sticks) and MtrCPA (black sticks). (b) Structure-based sequence alignment of HpCNH (PDB entry 6mg6) and MtrCPA (PDB entry 5h8i) structures. The secondary-structure elements are as follows: α-helices are shown as large coils, 310-helices are shown as small coils labeled η, β-strands are shown as arrows labeled β and β-turns are labeled TT. The identical residues are shown in white font on a red background, conserved residues are shown in red font and conserved regions are shown in blue boxes. (b) was generated using ESPript 3.0 (Gouet et al., 1999, 2003).

The structure of HpCNH reveals conserved residues that are required to bind putrescine comparably to MtrCPA. Overall, HpCNH has conserved secondary-structure elements typical of CNHs, with variable loops in proximity to the active site (Fig. 3a). HpCNH has the necessary active-site residues required to function as an N-carbamoylputrescine amidase. Future studies will include determination of the N-carbamoylputrescine amidohydrolase activity of HpCNH using established methods (Piotrowski et al., 2003). These methods will also be used to determine whether HpCNH is inhibited by known MtrCPA inhibitors. Once activity and inhibition are confirmed, co-crystallization with putrescine and selected inhibitors will be conducted.

4. Conclusion

The structure of HpCNH retains the conserved primary- and secondary-structure elements typical of CNHs, with variable loops in proximity to the active site. While HpCNH retains the catalytic site residues required to function as an N-carbamoylputrescine amidase, structure–function activity studies are required to determine the N-carbamoylputrescine amidase activity and establish HpCNH as a target for structure-based drug discovery.