Crystallization and biochemical studies of the NYN domain of human KHNYN

The NYN domain of human KHNYN was crystallized with a single-stranded RNA and the crystal diffracted to 1.72 Å resolution. The RNase activity of the NYN domain was also demonstrated using different RNAs, together with the binding between the NYN domain of KHNYN and the zinc-finger domain of ZAP.

Abstract

KHNYN is composed of an N-terminal KH-like RNA-binding domain and a C-terminal PIN/NYN endoribonuclease domain. It forms a complex with zinc-finger antiviral protein (ZAP), leading to the degradation of viral or cellular RNAs depending on the ZAP isoform. Here, the production, crystallization and biochemical analysis of the NYN domain (residues 477–636) of human KHNYN are presented. The NYN domain was crystallized with a heptameric single-stranded RNA from the AU-rich elements of the 3′-UTR of interferon lambda 3. The crystal belonged to space group P4₁32, with unit-cell parameters a = b = c = 111.3 Å, and diffacted to 1.72 Å resolution. The RNase activity of the NYN domain was demonstrated using different single-stranded RNAs, together with the binding between the NYN domain of KHNYN and the zinc-finger domain of ZAP.

1. Introduction

Zinc-finger antiviral protein (ZAP) is commonly known as a viral RNA-binding protein that inhibits viral replication (Yu et al., 2021 ▸; Gao et al., 2002 ▸; Bick et al., 2003 ▸; Zhu et al., 2011 ▸; Ficarelli et al., 2021 ▸). ZAP has been demonstrated to specifically bind regions of HIV-1 RNA with high CpG dinucleotide content, effectively marking them for degradation (Takata et al., 2017 ▸).

Two distinct human ZAP isoforms, ZAP-L and ZAP-S, have been discovered (Kerns et al., 2008 ▸), and both isoforms share four CCCH-type zinc-finger motifs in the N-terminal RNA-binding domain, one CCCH-type zinc-finger motif in the central domain and two WWE domains (Ficarelli et al., 2021 ▸). The longer isoform, ZAP-L, contains an additional catalytically inactive C-terminal poly(ADP-ribose) polymerase (PARP)-like domain (Fig. 1 ▸ a; Ficarelli et al., 2019 ▸). ZAP-L plays a traditional antiviral role (Sauter & Kirchhoff, 2021 ▸), while ZAP-S is involved in negative feedback regulation of interferon responses by binding to the AU-rich elements (AREs) in the 3′-UTR of certain host interferon messenger RNAs, including interferons beta and lambda (Ficarelli et al., 2019 ▸; Schwerk et al., 2019 ▸). The specificity of these two isoforms is determined by their different intra­cellular localization: ZAP-S in the cytosol targets host interferon mRNA, while ZAP-L, which is found in endolysosomal membranes due to its C-terminal prenylation motif (CVIS), primarily targets viral RNA (Schwerk et al., 2019 ▸).

Figure 1.

Domain organizations of human KHNYN, ZAP-L, ZAP-S and ZC3H12A and sequence alignment of KHNYN, ZC3H12A, ZC3H12A and ZC3H12C. (a) Schematic representation of the domain structures of KHNYN, ZAP-L, ZAP-S and ZC3H12A. (b) Sequence alignment of human KHNYN, human ZC3H12A, mouse ZC3H12A, human ZC3H12B and mouse ZC3H12C using Clustal Omega (Madeira et al., 2022 ▸). Residues coordinating Mg²⁺ ions are marked with an asterisk.

Because neither isoform of ZAP has nuclease activity, they require another protein with nuclease activity to degrade viral or interferon RNAs. The protein KHNYN was identified as an interaction partner of ZAP using a yeast two-hybrid screen (Ficarelli et al., 2019 ▸). KHNYN contains an N-terminal KH-like domain and a C-terminal NYN endoribonuclease and CUBAN domain (Fig. 1 ▸ a). The NYN domain, which belongs to the PIN nuclease domain superfamily, has endonuclease activity. Several proteins with a potentially active NYN domain, such as ZC3H12A and MARF1, have been characterized (Ficarelli et al., 2019 ▸). ZC3H12A contains both a zinc-finger RNA-binding domain and an NYN nuclease domain (Fig. 1 ▸ a). The NYN domains of ZC3H12A, ZC3H12B and ZC3H12C share a negatively charged active site with four aspartic acid residues that coordinate one Mg²⁺ ion (Jolma et al., 2020 ▸; Yokogawa et al., 2016 ▸; Xu et al., 2012 ▸; Garg et al., 2021 ▸). They share significant sequence identity (56–57%) with the NYN domain of human KHNYN (Fig. 1 ▸ b).

KHNYN is an essential cofactor of ZAP with nuclease activity (Ficarelli et al., 2019 ▸), and the ZAP–KHNYN complex functions similarly to ZC3H12A, but with the zinc-finger RNA-binding domain and the NYN endonuclease domain in two different proteins (Jolma et al., 2020 ▸; Garg et al., 2021 ▸; Yokogawa et al., 2016 ▸; Xu et al., 2012 ▸; Ficarelli et al., 2021 ▸). Here, we report the production, crystallization and data collection of the NYN domain (residues 477–636) of human KHNYN in complex with a heptameric single-stranded RNA (UAUUUAU) from the 3′-UTR of interferon lambda 3 (IFNL3). We also demonstrate the endonuclease activity of the NYN domain using different single-stranded RNAs and the binding of the zinc-finger domain of ZAP to the NYN domain of KHNYN using the BLItz system.

2. Materials and methods

2.1. Macromolecule production

2.1.1. Cloning

The full-length human KHNYN gene was purchased from IDT. DNA for the NYN domain (residues 477–636) was synthesized by polymerase chain reaction (PCR) using the forward primer 5′-TACTTCCAATCCAATGCAGATTTGCGTCACATTGTTATTGAC-3′ and the reverse primer 5′-TTATCCACTTCCAATGTTATTAAGCAGGCTTTTTCAGGAACTCAT-3′. The PCR product was cloned via ligase-independent cloning (LIC; Bonsor et al., 2006 ▸) into pLIC-Tr3a-MHA vector, which contains an N-terminal His₆ tag (His₆), maltose-binding protein (MBP) and a Tobacco etch virus (TEV) protease cleavage site (Table 1 ▸).

Table 1. Macromolecule-production information.

Source organism	Homo sapiens
DNA source	Synthetic
Forward primer	TACTTCCAATCCAATGCAGATTTGCGTCACATTGTTATTGAC
Reverse primer	TTATCCACTTCCAATGTTATTAAGCAGGCTTTTTCAGGAACTCAT
Expression vector	pLIC-Tr3a-MHA
Expression host	Escherichia coli BL21(DE3)
Expression details	Selenomethionine instead of methionine
Tag	6×His + MBP + TEV cleavage site: MKSSHHHHHHGSSMKIEEGKLVIWINGDKGYNGLAEVGKKFEKDTGIKVTVEHPDKLEEKFPQVAATGDGPDIIFWAHDRFGGYAQSGLLAEITPDKAFQDKLYPFTWDAVRYNGKLIAYPIAVEALSLIYNKDLLPNPPKTWEEIPALDKELKAKGKSALMFNLQEPYFTWPLIAADGGYAFKYENGKYDIKDVGVDNAGAKAGLTFLVDLIKNKHMNADTDYSIAEAAFNKGETAMTINGPWAWSNIDTSKVNYGVTVLPTFKGQPSKPFVGVLSAGINAASPNKELAKEFLENYLLTDEGLEAVNKDKPLGAVALKSYEEELAKDPRIAATMENAQKGEIMPNIPQMSAFWYAVRTAVINAASGRQTVDEALKDAQTNSSSNNNNNNNNNNLGIEENLYFQ
Complete amino-acid sequence of the construct produced	MKSSHHHHHHGSSMKIEEGKLVIWINGDKGYNGLAEVGKKFEKDTGIKVTVEHPDKLEEKFPQVAATGDGPDIIFWAHDRFGGYAQSGLLAEITPDKAFQDKLYPFTWDAVRYNGKLIAYPIAVEALSLIYNKDLLPNPPKTWEEIPALDKELKAKGKSALMFNLQEPYFTWPLIAADGGYAFKYENGKYDIKDVGVDNAGAKAGLTFLVDLIKNKHMNADTDYSIAEAAFNKGETAMTINGPWAWSNIDTSKVNYGVTVLPTFKGQPSKPFVGVLSAGINAASPNKELAKEFLENYLLTDEGLEAVNKDKPLGAVALKSYEEELAKDPRIAATMENAQKGEIMPNIPQMSAFWYAVRTAVINAASGRQTVDEALKDAQTNSSSNNNNNNNNNNLGIEENLYFQSNADLRHIVIDGSNVAMVHGLQHYFSSRGIAIAVQYFWDRGHRDITVFVPQWAFSKDAKVRESHFLQKLYSLSLLSLTPSRVMDGKRISSYDDRFMVKLAEETDGIIVSNDQFRDLAEESEKWMAIIRERLLPFTFVGNLFMVPDDPLGRNGPTLDEFLKKPA

2.1.2. Site-directed mutagenesis

To obtain a structure of the complex between RNA and KHNYN, the RNase activity of KHNYN had to be eliminated. The NYN domains of KHNYN and ZC3H12A/B/C share several conserved residues (Fig. 1 ▸ b). Upon analyzing the structure of the complex of ZC3H12B with RNA (PDB entry 6sjd; Jolma et al., 2020 ▸), Arg237 is positioned close to the RNA, suggesting a possible role in catalytic activity. Therefore, site-directed mutagenesis was carried out to replace the corresponding Arg526 residue of KHNYN (477–636) with an alanine using the forward primer 5′-CCGTCTTCGTACCACAATGGGCGTTCTCAAAAGATGCTAAGG-3′ and the reverse primer 5′-CCTTAGCATCTTTTGAGAACGCCCATTGTGGTACGAAGACGG-3′.

2.1.3. Expression and purification

The KHNYN (477–636) R526A construct was transformed into Escherichia coli BL21(DE3) cells. The transformed cells were grown in LB medium with antibiotics at 37°C overnight. After growth, the cells were harvested at 1050g for 15 min using a centrifuge. To obtain selenomethionine-labeled protein, the cell pellet was grown in modified M9 medium using 1× M9 salt medium, 0.4% glucose, 0.0001% thiamine, 2 mM MgCl₂, 0.2 mM CaCl₂. When the cells reached an OD₆₀₀ of 0.6–0.7, the temperature was decreased to 20°C and 100 mg each of lysine, threonine and phenylalanine and 50 mg each of leucine, isoleucine, valine and selenomethionine were added and incubated for 1 h. Isopropyl β-d-1-thiogalactopyranoside (IPTG) was added to the medium to a final concentration of 0.4 mM and growth was continued overnight. The cells were harvested by centrifugation at 1050g for 15 min. The cell pellet was resuspended in buffer A (20 mM Tris–HCl pH 7.5, 250 mM NaCl, 1 mM MgCl₂, 20 mM imidazole, 5% glycerol) and then lysed by sonication. The cell lysate was centrifuged at 19 000g for 1 h to separate the soluble fraction. The soluble fraction was loaded onto an Ni Sepharose 6 Fast Flow column (Cytiva) equilibrated with 5 column volumes (CV) of buffer B (20 mM Tris–HCl pH 7.5, 250 mM NaCl, 1 mM MgCl₂, 5% glycerol), washed with 20 CV buffer A, and His₆-MBP-TEV-KHNYN (477–636) R526A was then eluted with 2.5 CV buffer consisting of 50, 100, 200, 400 and 800 mM imidazole in buffer B. The purified protein was dialyzed twice with buffer B at 4°C and treated with His₆-TEV protease. After cleavage with TEV protease, a second nickel-affinity chromatography step using an Ni Sepharose 6 Fast Flow column (Cytiva) was performed. The protein was then subjected to a Superdex 75 prep-grade column (Cytiva) using buffer C (20 mM Tris–HCl pH 7.5, 250 mM NaCl, 1 mM MgCl₂, 1 mM DTT, 5% glycerol). The purified selenomethionine-labeled KHNYN (477–636) R526A protein was concentrated to 4.0 mg ml⁻¹ and then mixed with heptameric RNA (5′-UAUUUAU-3′) in a 1:10 molar ratio of protein:RNA.

2.1.4. RNase activity assay

The RNase activity assay was initiated by mixing 0.41 mg ml⁻¹ wild-type KHNYN (477–636) or its R526A mutant with 2.5 mg ml⁻¹ ARE17 (5′-GAUUUUAUUUAUAAAUU-3′), RNA21 (5′-AACCUAAUAAUUAUCAAAAUG-3′) or RNA23 (5′-GAUGGUGCUUCAAGCUAGUACGC-3′) in a buffer consisting of 20 mM Tris–HCl pH 7.5, 250 mM NaCl, 1 mM MgCl₂, 1 mM DTT, 5% glycerol. The reaction mixture was incubated at 37°C for 1, 3, 9 and 27 h and the reaction mixture was then subjected to TBE (Tris–borate–EDTA) acrylamide gel electrophoresis. The gel was visualized using ethidium bromide stain.

2.1.5. Binding assay of KHNYN and ZAP

The binding affinity between the KHNYN (477–636) R526A mutant and ZAP (1–230) was measured using the BLItz system employing biolayer interferometry. In the initial step, the Ni–NTA probe was loaded with ZAP (1–230) in 20 mM Tris–HCl pH 7.5, 500 mM NaCl, 50 µM zinc acetate, 1 mM DTT, 5% glycerol. The wash step was performed with 100 mM NaHPO₄, 150 mM NaCl, 0.02% Tween 20, 1 mg ml⁻¹ BSA, 1 mM MgCl₂, 50 µM zinc acetate (BLItz assay buffer). Subsequently, the probe was transferred to 26.5 nM KHNYN (477–636) R526A in 20 mM Tris–HCl pH 7.5, 250 mM NaCl, 1 mM MgCl₂, 1 mM DTT, 5% glycerol during the association step and then transferred back to the BLItz assay buffer for dissociation. In experiments involving RNA, the Ni–NTA probe was initially loaded with ZAP (1–230) mixed with a 1.2-fold molar excess of ARE17. The subsequent wash, association and dissociation steps were performed in the same way as the experiments conducted without RNA.

2.2. Crystallization

Crystallization was performed using the sitting-drop vapor-diffusion method at 20°C. 0.5 µl protein solution consisting of 4 mg ml⁻¹ selenomethionine-labeled KHNYN (477–636) R526A and a tenfold molar excess of heptameric RNA (5′-UAUUUAU-3′) was mixed with 0.5 µl reservoir solution consisting of 90 mM Tris–HCl pH 8.5, 4.77%(w/v) PEG 4000, 360 mM CaCl₂, 4.8%(v/v) 1-propanol (Table 2 ▸). Crystals were harvested from the drops and cryoprotected in well solution containing 30%(v/v) glycerol before flash-cooling in liquid nitrogen.

Table 2. Crystallization.

Method	Vapor diffusion
Plate type	Sitting drop
Temperature (K)	293
Protein concentration (mg ml−1)	4
Buffer composition of protein solution	20 mM Tris–HCl pH 7.5, 250 mM NaCl, 1 mM MgCl2, 1 mM DTT, 5% glycerol
Composition of reservoir solution	90 mM Tris–HCl pH 8.5, 4.77%(w/v) PEG 4000, 360 mM CaCl2, 4.8%(v/v) 1-propanol
Volume of drop (µl)	1.0
Volume of reservoir (µl)	500

2.3. Data collection and processing

Diffraction data were obtained on beamline 5C at Pohang Accelerator Laboratory (PAL) using X-rays with a wavelength of 0.97495 Å. Diffraction data were processed with HKL-2000 (Otwinowski & Minor, 1997 ▸). The crystal belonged to the cubic space group P4₁32, with unit-cell parameters a = b = c = 111.3 Å (Table 3 ▸). The Matthews coefficient (Matthews, 1968 ▸; Kantardjieff & Rupp, 2003 ▸; Weichenberger & Rupp, 2014 ▸) suggested that there is one molecule in the asymmetric unit, with a solvent content of 57.7%.

Table 3. Data collection and processing.

Values in parentheses are for the outer shell.

Diffraction source	Beamline 5C, PAL
Wavelength (Å)	0.97495
Temperature (K)	77
Detector	EIGER 9M
Crystal-to-detector distance (mm)	150
Rotation range per image (°)	1
Total rotation range (°)	90
Exposure time per image (s)	1
Space group	P4132
a, b, c (Å)	111.3, 111.3, 111.3
α, β, γ (°)	90, 90, 90
Mosaicity (°)	0.8
Resolution range (Å)	50–1.72 (1.75–1.72)
Total No. of reflections	472842
No. of unique reflections	25659
Completeness (%)	100 (100)
Multiplicity	18.43
〈I/σ(I)〉	39.3 (1.35)†
R meas	0.154 (1.686)
CC1/2	0.997 (0.815)
Overall B factor from Wilson plot (Å2)	24.4

^†〈I/σ(I)〉 falls below 2.0 in the 1.82–1.78 Å resolution shell; the data resolution is cut based on CC_1/2, which is 0.815 in the highest resolution shell (1.75–1.72 Å).

3. Results and discussion

3.1. Production of the NYN domain of KHNYN

We tried to express the full-length human KHNYN protein; however, sufficient amounts of protein could not be produced, likely due to its instability. Therefore, we cloned the nuclease domain of KHNYN (residues 477–636) as an MBP-fusion protein. To obtain the structure of the KHNYN (477–636)–RNA complex, we mutated Arg526 to alanine after examination of the homologous complex structure of ZC3H12B (56% sequence identity to KHNYN) with RNA (PDB entry 6sjd). The KHNYN (477–636) R526A mutant showed significantly reduced the RNase activity, and structural studies of the KHNYN (477–636)–RNA complex could be conducted. However, the KHNYN (477–636) R526A mutant has residual RNase activity and a double mutant will be generated to completely abolish the RNase activity.

3.2. RNase activity assay of KHNYN (477–636)

The RNase activity of both wild-type KHNYN (477–636) and its R526A mutant was evaluated using three different RNAs (Fig. 2 ▸). ARE17, derived from the AU-rich elements (AREs) within the 3′-UTR of interferon lambda 3 (IFNL3), contains an AU-rich element (AUUUA) in the centre. RNA21 and RNA23 were used as negative controls. ARE17 was digested almost completely by wild-type KHNYN (477–636) protein after 27 h, while approximately 50% of ARE17 was digested by the R526A mutant. Surprisingly, RNA21, which lacks an AU-rich element or a CpG motif, was also digested by KHNYN (477–636). Conversely, RNA23 was not digested even after 27 h, despite containing a CpG motif near the 3′ end. From these results, we found that the NYN domain alone has RNase activity and that the R526A mutation significantly diminishes its activity. While KHNYN (477–636) seems to exhibit substrate specificity by digesting ARE17 and RNA21 but not RNA23, the sequence motif controlling this specificity has yet to be identified.

Figure 2.

RNase activity assay of wild-type KHNYN (477–636) and the R526A mutant. (a) The sequences of the RNAs used are shown. (b) The RNAs were digested with wild-type (WT) KHNYN (477–636) or the R526A mutant for 0, 1, 3, 9 and 27 h and loaded onto the PAGE gel. Lane M contains a DNA marker, with the smallest size corresponding to 50 base pairs. (c) The intensity of the bands was quantified and plotted against time.

3.3. Determination of the binding affinity between KHNYN (477–636) and ZAP (1–230)

The NYN domain of KHNYN is known to bind the zinc-finger domain of ZAP (Ficarelli et al., 2019 ▸). To quantify this interaction, we employed the BLItz system to measure the binding affinity between ZAP (1–230) and KHNYN (477–636) (Fig. 3 ▸). In the initial experiment, ZAP (1–230) with a His₆ tag was loaded onto the Ni–NTA probe, washed with buffer and subsequently subjected to interaction with KHNYN (477–636) R526A. The association and dissociation rates were measured, and K _d was calculated to be 12.5 nM. In the next experiment, ZAP (1–230) mixed with ARE17 RNA was loaded onto the probe and the binding of KHNYN (477–636) R526A was measured. The calculated K _d was 9.83 nM, suggesting that the binding of ZAP (1–230) and KHNYN (477–636) becomes stronger in the presence of substrate RNA.

Figure 3.

BLItz binding assay between ZAP (1–230) and KHNYN (477–636) R526A mutant in the absence (pink) and presence (purple) of 17-mer RNA (ARE17). The K _d and other parameters calculated by the BLItz software are shown.

3.4. Crystallization and data collection

Crystals of the KHNYN (477–636)–7-mer RNA complex grew at 20°C in 90 mM Tris–HCl pH 8.5, 4.77%(w/v) PEG 4000, 360 mM CaCl₂, 4.8%(v/v) 1-propanol (Fig. 4 ▸). Diffraction data were collected at PAL at a wavelength of 0.97495 Å to obtain the anomalous signal from the selenomethionine incorporated in the protein. The data were indexed in the cubic space group P4₁32, with unit-cell parameters a = b = c = 111.3 Å (Table 3 ▸).

Figure 4.

Crystals of the KHNYN (477–636) R526A–7-mer RNA (5′-UAUUUAU-3′) complex.

4. Conclusions

The ZAP–KHNYN system plays a crucial role in recognizing and degrading viral or cellular RNA, with the specificity of this process being dependent on the isoform of ZAP that interacts with KHNYN. Notably, ZAP lacks RNase activity, making KHNYN an indispensable component of the ZAP–KHNYN system. We have produced and crystallized a KHNYN (477–636)–7-mer RNA complex and obtained X-ray diffraction data to a resolution of 1.72 Å. This structural information will provide valuable insights into the mechanisms underlying the recognition and degradation of substrate RNA by KHNYN.

Funding Statement

This study was supported by a grant from the 2022 Research Fund of the University of Seoul to J. Choe.