Functional and Structural Characterization of Acquired 16S rRNA Methyltransferase NpmB1 Conferring Pan-Aminoglycoside Resistance

ABSTRACT

Posttranslational methylation of the A site of 16S rRNA at position A1408 leads to pan-aminoglycoside resistance encompassing both 4,5- and 4,6-disubstituted 2-deoxystreptamine (DOS) aminoglycosides. To date, NpmA is the only acquired enzyme with such a function. Here, we present the function and structure of NpmB1, whose sequence was identified in Escherichia coli genomes registered from the United Kingdom. NpmB1 possesses 40% amino acid identity with NpmA1 and confers resistance to all clinically relevant aminoglycosides, including 4,5-DOS agents. Phylogenetic analysis of NpmB1 and NpmB2, its single-amino-acid variant, revealed that the encoding gene was likely acquired by E. coli from a soil bacterium. The structure of NpmB1 suggests that it requires a structural change of the β6/7 linker in order to bind to 16S rRNA. These findings establish NpmB1 and NpmB2 as the second group of acquired pan-aminoglycoside resistance 16S rRNA methyltransferases.

INTRODUCTION

Aminoglycosides, since their initial discoveries in the 1940s, continue to serve as an important class of antibiotics, with clinical use in Gram-negative, Gram-positive, and mycobacterial infections. Pathogenic bacteria have since acquired resistance to aminoglycosides through several mechanisms, including production of aminoglycoside-modifying enzymes (AMEs), upregulation of efflux pumps, and reduced affinity to the target nucleotide in the 16S rRNA, either through nucleotide substitution or posttranslational methylation (1).

Of these mechanisms, posttranslational methylation of the 16S rRNA by acquired 16S rRNA methyltransferases confers high-level resistance to multiple aminoglycosides. This mechanism is intrinsic to many Streptomyces and Micromonospora species that produce aminoglycosides. However, acquired 16S rRNA methyltransferases have been reported from pathogenic Gram-negative bacteria (e.g., Enterobacterales, Pseudomonas aeruginosa, and Acinetobacter baumannii) over the last 2 decades (1–6). These acquired enzymes are categorized into two groups based on whether they methylate G1405 or A1408. Methylation of G1405 confers resistance to 4,6-disubstituted 2-deoxystreptamine (DOS) aminoglycosides (e.g., kanamycin, gentamicin, tobramycin, and amikacin) but not 4,5-disubstituted 2-DOS (e.g., neomycin and paromomycin). A total of nine acquired G1405 16S rRNA methyltransferases have been identified and reported, which include ArmA (7) and RmtA (8) through RmtH (9–16). On the other hand, NpmA is the only known acquired A1408 16S rRNA methyltransferase, initially reported in 2007 and identified in Escherichia coli (NpmA1 [17]) and Clostridioides difficile (NpmA2 [18]). C. difficile has been suggested as the source of NpmA in the clinical environment (18). Unlike G1405 16S rRNA methyltransferases, NpmA confers resistance to both 4,5- and 4,6-disubstituted 2-DOS aminoglycosides, and, thus, can be considered a pan-aminoglycoside resistance enzyme.

In the process of screening for NpmA-like enzymes in the NCBI sequence databases, we identified two highly related, putative A1408 16S rRNA methyltransferase genes in several E. coli genome sequences deposited from the United Kingdom. We therefore proceeded with functional and structural characterization of this new enzyme. Here, we describe the resistance to aminoglycosides conferred by this enzyme, its molecular mechanism, and its unique structural features compared with NpmA1 (19–21) in addition to Kmr (22), KamB (20), and CacKam (23), which are the intrinsic A1408 16S rRNA methyltransferases of aminoglycoside-producing soil bacteria.

RESULTS AND DISCUSSION

Identification of new A1408 16S rRNA methyltransferase genes.

The reference sequence of NpmA1 (WP_032492089.1) was queried against the NCBI database using the blastp (24) function to identify amino acid sequences with similarities to pathogenic bacteria. We found two sequences having 40% sequence identities with NpmA1 and described as class I S-adenosylmethionine (SAM)-dependent methyltransferase. We designated them NpmB1 and NpmB2 (Fig. 1). NpmB1 is assigned to GenBank accession numbers EGD5494512.1 and MJZ53392.1, and these entries were identified in the E. coli strains 437948 and 392824, respectively. NpmB2 is also found in two individual entries deposited under GenBank accession numbers EAB8931753.1 and EAB9582428.1, and these entries were identified in E. coli strains 316274 and 365344, respectively. All E. coli strains carrying npmB1 or npmB2 were isolated in the United Kingdom. NpmB1 and NpmB2 consist of 217 amino acids, and only one amino acid substitution is identified in their sequences at position 21 (arginine for NpmB1 and cysteine for NpmB2). The residue at position 20 of NpmA1 is isoleucine, which is equivalent to position 21 of NpmB1 and NpmB2, and located away from the active site for methyltransferase activity and the binding sites for SAM or 16S rRNA. We assumed that the residues at position 21 of NpmB1 and NpmB2 are not critical for their function.

FIG 1.

Alignments of amino acid sequences of A1408 16S rRNA methyltransferases. Identical and similar amino acids among all enzymes are red and dark red letters, respectively. The black dots indicated above the sequence are critical residues for the NpmA1 function. Position 21 of NpmB1 and NpmB2 is indicated as an asterisk.

Previous studies indicated that functional residues of NpmA1 are classified into three groups related to binding to SAM or 16S rRNA and two tryptophan residues, which stack the flipped-out adenine base at position 1408 of 16S rRNA and directed it to the active site. In particular, residues D30 and D55 for SAM binding, R208 for RNA binding, and W107 and W197 for stacking of A1408 are critical in conferring resistance to aminoglycosides, and single substitutions of these residues have been shown to abolish resistance (19, 21). Sequence alignment of NpmB1, NpmB2, and NpmA1 showed that they share the SAM-binding GxGxG motif and functional residues of NpmA1 described above (Fig. 1). We therefore hypothesized that NpmB methylates A1408 of 16S rRNA and confers resistance to aminoglycoside in a manner similar to that of NpmA1.

NpmB likely originated in Acidobacteria.

On the phylogenetic tree, the amino acid sequences of NpmB1 and NpmB2 belonged to a cluster consisting of amino acid sequences associated with Acidobacteria, especially an unclassified Blastocatellia species (accession number HCA58615) (Fig. 2). Both upstream and downstream regions of npmB1 and npmB2 were also sequences related to Acidobacteria species (Fig. 3). The contig including the npmB1 gene (AAVRUE010000189) was identical to that of npmB2, except for the one nucleotide difference in npmB1. The gene cluster associated with Acidobacteria species was interrupted by an IS91 family pseudogene on the contig that included npmB2 (AAAGDO010000044). It is therefore likely that the npmB genes originated from Acidobacteria species and were integrated into the E. coli genome in the soil that is inhabited by the former species.

FIG 2.

Phylogenetic tree of NpmB-related proteins. The phylogenetic tree is constructed from amino acid sequences of NpmB1 and NpmB2, closely related proteins (HCA58615 and QQS40840), NpmA1, and 30 randomly selected putative proteins. NpmB1 and NpmB2 (blue) belong to a cluster consisting of Acidobacteria-associated proteins (red).

FIG 3.

Genetic structure surrounding npmB2. Open reading frames associated with Acidobacteria species are indicated in red, and those associated with Enterobacterales are in blue. Upstream and downstream regions of npmB2 resembled sequences from Acidobacteria species.

NpmB1 confers resistance to aminoglycosides through A1408 methylation of 16S rRNA.

To examine whether NpmB1 as the representative NpmB enzyme actually confers resistance to aminoglycosides, we generated an NpmB1-producing E. coli JM109(DE3) laboratory strain by transforming it with the pUC19–promoter_npmB1 plasmid, which includes the intrinsic promoter sequence found in the genome sequence of the source strain. We tested susceptibilities of this NpmB1-producing strain to 4,6-disubstituted 2-DOS aminoglycosides amikacin, arbekacin, tobramycin, and gentamicin, a 4,5-disubstituted 2-DOS aminoglycoside neomycin, and a 4-monosubstituted aminoglycoside apramycin. The MICs of these aminoglycosides are shown in Table 1. The MICs of aminoglycosides against the NpmB1-producing strain were at least 16-fold higher than that of the control strain harboring the pUC19 plasmid. In addition, we tested susceptibility to non-A-site binder streptomycin by the disk-diffusion method, and both NpmB1-producing and control strains were susceptible. These results indicated that NpmB1 confers resistance to both 4,5- and 4,6-disubstituted 2-DOS aminoglycosides as well as 4-monosubstituted aminoglycosides, as expected, and that the levels of resistance to aminoglycosides conferred by NpmB1 are similar to those conferred by NpmA1 (17). We further performed the primer extension assay and confirmed the methylation position of 16S rRNA in the NpmB1-producing strain. The reverse transcriptase reaction using 16S rRNA isolated from the NpmB1-producing strain was aborted at position C1409 of 16S rRNA (G1409 of synthesized cDNA), while the reaction was normally terminated when using 16S rRNA isolated from the control strain (see Fig. S1a and b in the supplemental material). In addition, we examined the level of N1 methylated adenosine in the NpmB1-producing and control strains by an enzyme-linked immunosorbent assay (ELISA) with monoclonal antibody specific for 1-methyladenosine. This result indicated that the amount of 1-methyladenosine included in the NpmB1-producing strain is larger than that included in the control strain (Fig. S1c). Thus, we conclude that NpmB1 is functionally equivalent to NpmA1 and methylates the N1 position of A1408 of 16S rRNA in vivo to confer pan-resistance to aminoglycosides.

TABLE 1.

MICs for aminoglycosides against laboratory-generated strains with and without NpmB1 production^a

Strain	MICs (μg/ml)
	ABK	AMK	TOB	GEN	NEO	APR
E. coli JM109(DE3) (p_promoter_npmB1)	16	64	128	32	128	>256
E. coli JM109(DE3) (pUC19)	1	2	0.5	0.5	2	4

^aABK, arbekacin; AMK, amikacin; TOB, tobramycin; GEN, gentamicin; NEO, neomycin; APR, apramycin.

Crystal structure of NpmB1.

To obtain structural insights, we determined the crystal structure of NpmB1 at 1.5-Å resolution and refined it to final R and R_free factors of 16.6% and 19.3%, respectively (Table S1). The NpmB1 crystal belongs to the orthorhombic space group P2₁2₁2₁ and contains one NpmB1 molecule in a crystallographic asymmetric unit. We could not obtain the clear electron density corresponding to the residues N182 to P192; thus, these residues were omitted from the final structure (Fig. 4a). The crystal structure of NpmB1 showed that it forms a class I methyltransferase fold consisting of a central seven-stranded β-sheet surrounded by eight α-helices, four 3₁₀-helices, and a two-stranded β-sheet (Fig. 4a). The sequence identities and overall root mean square deviation (RMSD) values between NpmB1 and the other A1408 16S rRNA methyltransferases are summarized in Table 2. These results show that the structures of the β5/6 linker and the β6/7 linker are not superimposable among the A1408 16S rRNA methyltransferases. We discuss below the characteristic features of the cofactor binding and the 16S rRNA binding of NpmB1 based on these structure comparisons of A1408 16S rRNA methyltransferases.

FIG 4.

Crystal structure of NpmB1. (a) Overall structure of NpmB1. Missing residues due to the poor electron density map are connected by a black dashed line, and the position of the terminal residues of missing regions are labeled and indicated as black dots. (b) Comparisons of the cofactor-binding site of A1408 16S rRNA methyltransferase. Stereo view of the cofactor-binding site. The SAH molecule is shown as a cyan stick model. Residues consisting of the cofactor-binding site are colored in blue (NpmB1), magenta (NpmA1; PDB entry 3P2E) and orange (KamB; PDB entry 3MQ2). (c) Recognition of A1408 of 16S rRNA by NpmB1 and NpmA1. The structures of 16S rRNA bound to NpmA1 is shown as a white stick representation. NpmB1 and NpmA1 are colored blue and magenta, respectively.

TABLE 2.

Sequence identities and overall RMSD values for A1408 16S rRNA methyltransferases against the NpmB1 structure

Enzyme	Identity (%)	Alignment length (aa)	RMSD	Alignment length (aa)
NpmA1	40	174	1.24	178
Kmr	43	200	1.19	176
KamB	37	145	1.58	189
CacKam	37	158	1.47	178

Cofactor SAM binds to NpmB1 in a manner similar to that of the other A1408 16S rRNA methyltransferases.

The structure of NpmA1 complex with SAM or S-adenosylhomocysteine (SAH) showed that the cofactor-binding site consists of residues D30, G32, N38, D55, P56, A87, E88, L104, T109, and S195 (Fig. 4b) (19). Importantly, the side chains of residues D30, D55, E88, and T109 contribute to the recognition of SAM via the hydrogen bond formations. In NpmB1, D31, D56, and E89 are equivalent residues to D30, D55, and E88 of NpmA1, respectively, and are located at analogous positions enabling formation of hydrogen bonds (Fig. 1 and 4b). Among the A1408 16S rRNA methyltransferases, the conserved residue corresponding to position T109 of NpmA1 is serine (Fig. 1), and the structures of NpmA1 and KamB complexed with SAM or SAH showed that both serine and threonine residues contribute to the hydrogen bond formation with SAM through the hydroxyl group in their side chain structures (Fig. 4b) (19, 20). On the whole, while some amino acid substitutions are observed in the cofactor-binding site, there are no residues that appear to interfere with the binding of SAM in the NpmB1 structure.

Comparison of the target recognition sites among A1408 16S rRNA methyl-transferases.

Residues W107 and W197 of NpmA1 are critical for recognition of the target A1408 base (19). The crystal structure of the NpmA1 complex with 30S ribosome revealed that residues W107 and W197 stabilize the target A1408 base via stacking interactions (Fig. 4c) (21). In NpmB1, residues W108 and W197 correspond to residues W107 and W197 of NpmA1, and these residues are located at the β4/5 linker and the β6/7 linker, respectively. Comparison of the NpmB1 structure with the structure of the NpmA1 complex with 30S ribosome showed that the position of residue W197 of NpmB1 does not coincide with W197 of NpmA1 (Fig. 4c), while residue W108 of NpmB1 is located at almost an analogous position to residue W107 of NpmA1 and will stack to the target A1408 base by changing the rotamer of W108 (Fig. 4c). The structural and mutational studies of Kmr indicated that its variants containing a single substitution of residue W194 retained the enzymatic activity, suggesting that residue F144 would come into close proximity to residue W194 and complement the function of residue W194 (22). However, we could not identify a residue having an aromatic ring, such as tryptophan, phenylalanine, or tyrosine, that would stack the target A1408 base in the active site of NpmB1.

NpmB1 requires a structural change of the β6/7 linker to recognize A1408 of 16S rRNA.

To deduce whether W197 of NpmB1 participates in the stacking interaction, we generated plasmids producing NpmB1 variant NpmB1_W108A or NpmB1_W197A, introduced them to E. coli JM109(DE3), and tested the susceptibilities to tobramycin. The MICs of E. coli JM109(DE3) producing NpmB1_W108A and NpmB1_W197A were 0.5 μg/ml and 1 μg/ml, respectively, values that were significantly lower than that for the NpmB1-producing strain and comparable to those of the control strain (Table 1). These results suggested that both residues W108 and W197 of NpmB1 are critical for the resistance to aminoglycosides in a manner similar to that of NpmA1. In addition, A1408 16S rRNA methyltransferases possess the positively charged molecular surfaces interacting with 16S rRNA (Fig. 5). Based on the structure of the NpmA1 complex with 30S ribosome, the β2/3 linker and the β6/7 linker of NpmA1 are the primary binding site for rRNA of the 30S ribosome (21). Comparisons of electrostatic potential molecular surfaces among A1408 16S rRNA methyltransferases showed that the rRNA-binding surface of NpmA1 is larger than that of the other enzyme, and the widespread positively charged region is observed at the β2/3 linker through the β1/2 linker to the β6/7 linker (Fig. 5b). The positively charged molecular surfaces of NpmB1, Kmr, KamB, and CacKam are narrower than that of NpmA1 and mainly consist of the β2/3 linker and the β1/2 linker (Fig. 5a and c to e). Taken together, we suggest that a structural change of the β6/7 linker of NpmB1 occurs where residue W197 stacks the target A1408 base, and basic amino acids such as arginine or lysine will be assembled to the rRNA-binding surface.

FIG 5.

Comparisons of the 16S rRNA-binding site of A1408 16S rRNA methyltransferase. Electrostatic surface potential representations of A1408 16S rRNA methyltransferases. (a) NpmB1. (b) NpmA1 (PDB entry 4OX9). (c) Kmr (PDB entry 4RWZ). (d) KamB (PDB entry 3MQ2). (e) CacKam (PDB entry 5D1N). (f) Scale for the electrostatic surface potential.

Conclusions.

The presented study describes NpmB as the second group of acquired pan-aminoglycoside 16S rRNA methyltransferases after NpmA. The npmB genes likely originated from Acidobacteria species and were integrated into the E. coli genome in the soil. Like NpmA1, NpmB1 methylates the N1 position of A1408 of 16S rRNA in vivo and confers resistance to aminoglycosides, including 4,5-disubstituted 2-DOS, 4,6-disubstituted 2-DOS, and 4-monosubstituted aminoglycosides, but it appears to require a structural change of the β6/7 linker in order to recognize the target A1408 base and bind to 16S rRNA. Although the prevalence of NpmB-producing bacteria remains unknown, these findings highlight the potential of the future spread of this pan-aminoglycoside resistance mechanism into pathogenic Gram-negative bacteria.

MATERIALS AND METHODS

Cloning of npmB1.

The npmB1 sequence was obtained from GenBank accession number AAVRUE-010000189.1 (locus tag DTY44_23675). The nucleotides 1370 to 1295 of AAVRUE010000189.1 was expected to be an intrinsic promoter for npmB1 based on promoter prediction performed with a G4PromFinder algorithm (25). The promoter_npmB1 gene corresponding to the nucleotides 226 to 1410 of AAVRUE010000189.1 was synthesized by Eurofins Genomics (Tokyo, Japan) and ligated to the plasmid vector pUC19 at the BamHI restriction site. Plasmids with genes encoding NpmB1 variants NpmB1_W108A and NpmB1_W197A were generated by the standard site-directed mutagenesis method using the pUC19–promoter_npmB1 plasmid vector as a template. All inserted sequences in the recombinant plasmids were confirmed by Sanger sequencing entrusted to Eurofins Genomics (Tokyo, Japan).

Phylogenetic analysis of npmB genes.

A total of 100 amino acid sequences with similarities with NpmB1 and NpmB2 were identified by blastp using nonredundant protein sequence (nr) database and downloaded on 15 April 2021. Multiple alignment of the amino acid sequences was carried out by MUSCLE (26). The phylogenetic tree was generated by FastTree (http://www.microbesonline.org/fasttree/) using the Le-Gascuel 2008 model and visualized by MEGA7 (https://www.megasoftware.net/).

Susceptibility testing.

E. coli JM109(DE3) harboring the recombinant plasmids of pUC19 or pUC19–promoter_npmB1 variants were grown on lysogenic broth agar plates containing ampicillin 50 μg/ml at 37°C overnight. MICs were determined in triplicated by standard broth microdilution methods with Mueller-Hinton broth by following the guidelines established by the Clinical and Laboratory Standards Institute (CLSI) (27).

Primer extension.

Lysogenic broth with 50 μg/ml ampicillin was inoculated with E. coli JM109(DE3) harboring the recombinant plasmids of pUC19 or pUC19–promoter_npmB1 and grown to an optical density at 600 nm (OD₆₀₀) of 0.8 to 1.0 at 37°C. The cells were harvested by centrifugation at 1,500 × g for 5 min, and total RNA was extracted by using PureLink RNA minikit (Thermo Fisher Scientific, Waltham, MA). Approximately 300 μg of purified total RNA was hybridized with 10 pmol primer (5′-[6-carboxyfluorescein]-TGAATCACAAAGTGGTAAGCG-3′) complementary to nucleotides 1480 to 1460 of 16S rRNA. Hybridization reactions were performed at 80°C for 2 min followed by slow cooling at the rate of 0.1°C/s to 58°C and then kept at 58°C for 1 h followed by slow cooling at the rate of 0.1°C/s to 42°C. Extension reactions were performed with 1 U avian myeloblastosis virus reverse transcriptase (Promega, Madison, WI) per reaction at 42°C for 30 min. The cDNA transcripts were analyzed on an 15% denatured polyacrylamide gel containing 8 M urea, and bands were visualized on a fluorescent image analyzer FLA-3000G (Fujifilm, Japan).

Measurement of 1-methyladenosine.

Total RNA isolated from the NpmB1-producing and control strains were prepared as described in the primer extension assay. To remove any RNA secondary structure, RNA was heated at 95°C for 5 min and rapidly chilled on ice. Five micrograms of RNA was digested to nucleosides by incubation with 5 U of nuclease P1 (New England Biolabs, Ipswich, MA) and 1 U of bacterial alkaline phosphatase derived from E. coli strain C75 (TaKaRa, Japan) in the nuclease P1 buffer (50 mM sodium acetate, pH 5.5) for 1 h at 37°C. The amount of 1-methyladenosine was determined using a CircuLex m¹A (N1-methyladenosine) competitive ELISA kit (Medical & Biological Laboratories, Co. Ltd., Japan) according to the manufacturer’s instructions.

Purification of NpmB1.

The gene encoding NpmB1 was cloned into the protein expression plasmid vector pET-32a between the restriction enzyme sites of NdeI and BamHI. Lysogenic broth with 50 μg/ml ampicillin and 34 μg/ml chloramphenicol was inoculated with E. coli Rosetta2(DE3) harboring the pET-32a–NpmB1 recombinant plasmid and grown to an OD₆₀₀ of 1.0 at 30°C. A final concentration of 0.1 mM isopropyl-β-d-thiogalactopyranoside (IPTG) was added, and the culture was incubated for an additional 2 h. The cells were harvested by centrifugation at 6,000 × g for 10 min, and the pellet was resuspended in 20 mM Tris-HCl, pH 7.0, buffer, sonicated, and centrifuged at 7,000 × g for 30 min at 4°C to remove cell debris. The supernatants were further centrifuged at 48,000 × g, and the supernatant was filtered through a 0.45-μm Minisart syringe filter (Sartorius, Gottingen, Germany). The crude extract was loaded onto a HiTrap SP HP (Cytiva, Marlborough, MA) previously equilibrated with 20 mM Tris-HCl, pH 7.0, buffer containing 150 mM NaCl. NpmB1 was then eluted with a NaCl gradient at a flow rate of 5 ml/min. To remove the endogenous nucleic acids bound to NpmB1, fractions containing NpmB1 were pooled, diluted with 20 mM Tris-HCl, pH 7.0, buffer containing 1 M NaCl, and stored at 4°C for 30 min. The NpmB1 solution was then concentrated using a Vivaspin turbo 15 centrifugal concentrator (molecular weight cutoff [MWCO], 10,000; Sartorius) and washed twice with 20 mM Tris-HCl, pH 7.0, buffer containing 1 M NaCl. Further purification was performed with HiTrap Blue HP column (Cytiva) chromatography. The NpmB1 solution was diluted to an NaCl concentration of 150 mM and loaded onto the HiTrap Blue HP column, and NpmB1 was eluted with a linear gradient of 150 mM to 2 M NaCl in 20 mM Tris-HCl, pH 7.0, buffer. The fractions containing NpmB1 were collected and concentrated using a Vivaspin turbo 15 centrifugal concentrator (MWCO, 10,000; Sartorius), and the buffer was exchanged to 20 mM Tris-HCl, pH 7.0, buffer containing 1 M NaCl and 20% glycerol by several rounds of dilution and concentration. The purified NpmB1 solution was finally concentrated to approximately 13 mg/ml and stored in 20-μl aliquots at −80°C until use in crystallization experiments.

Crystallization.

Initial screening of NpmB1 was performed with the commercial crystal screening kits Index, Crystal screen and Crystal screen 2 (Hampton Research, Aliso Viejo, CA) using the hanging-drop vapor diffusion method by mixing 1.0 μl of the NpmB1 solution (12 mg/ml in 20 mM Tris-HCl, pH 7.0, buffer containing 500 mM NaCl and 10% glycerol) and 1.0 μl of the reservoir solution at 20°C. The cluster crystals of NpmB1 appeared in several crystallization conditions containing polyethylene glycol 3350 (PEG3350), PEG4000, PEG8000, or PEG monomethyl ester 2000 as a precipitant within 1 week. Crystals of NpmB1 suitable for the X-ray crystallography were obtained by mixing 1.0 μl of the NpmB1 solution (5 mg/ml in 20 mM Tris-HCl buffer containing 500 mM NaCl and 20% glycerol) and 1.0 μl of the reservoir solution containing 30% PEG3350, 100 mM HEPES, pH 7.0, and 5% ethylene glycol at 4°C.

Data collection, structure determination, and refinement.

Prior to the X-ray experiments, the NpmB1 crystals were transferred into a cryoprotectant solution composed of reservoir solution containing 20% ethylene glycol and then flash frozen in liquid nitrogen. Synchrotron experiments were carried out using Photon Factory BL-17A (High Energy Accelerator Research Organization, Tsukuba, Japan). Diffraction data sets were collected at −173°C using an EIGER X16M detector (Dectris Ltd., Switzerland), and data sets were processed and scaled using XDS (28). The initial phase of the NpmB1 structure was determined by the molecular replacement method using Molrep (29) from the CCP4 program suite (30), with the coordinate (PDB entry 4RWZ) serving as the search model. Manual model rebuilding was performed with COOT (31). Structure refinement was performed with phenix.refine from the PHENIX package (32) using the atomic displacement parameters using the translation, liberation, and screw (TLS) method, and TLS groups were determined by using phenix.find_tls_groups. The stereochemical quality of the final structure was evaluated by MolProbity (33). All molecular graphics were prepared using PyMOL v 2.4.0. (Schrödinger, LLC., New York, NY, USA). Electrostatic potentials were calculated with APBS (34) by using a PyMOL plug-in APBS tools.