Expression, Purification and Characterization of a GII.4 Norovirus Protease from Minerva Virus

Abstract

Background:

Noroviruses are the leading cause of acute gastroenteritis worldwide. Norovirus proteases, which are responsible for cleavage of the viral polyprotein, have become an attractive drug target to treat norovirus infections. Genogroup II (GII) noroviruses are responsible for a majority of outbreaks; however, limited data exists regarding GII norovirus proteases.

Methods:

We report here successful expression, purification, characterization, and inhibition of the Minerva virus protease (MVpro), a genogroup II genotype 4 (GII.4) norovirus protease. We observed MVpro as both a monomer and dimer in solution through size-exclusion chromatography. In addition, MVpro cleaves the synthetic substrate mimicking the MVpro NS2/NS3 cleavage site more efficiently than other norovirus proteases such as the Norwalk virus protease (GI.1) and the MD145 protease (GII.4).

Results and Conclusion:

Compound A, a potent inhibitor of MVpro, is a good starting point for the design of inhibitors to target GII.4 noroviruses. Furthermore, the results presented here will allow for future characterization of MVpro inhibitors as they are synthesized.

1. INTRODUCTION

Noroviruses, which belong to the Caliciviridae family, are the leading cause of gastroenteritis worldwide. Each year in the United States, noroviruses are responsible for greater than 20 million cases of acute gastroenteritis, leading to an estimated 800 deaths and 71,000 hospitalizations [1]. While most cases resolve within a week, immunocompromised patients, children, and the elderly have an elevated risk of long-term and potentially fatal infections [2–4]. Noroviruses are divided into seven genogroups (GI-GVII); GI is subdivided into genotypes 1-7 and GII is subdivided into genotypes 1-15 [5]. Genogroups GI, GII, and GIV are infectious in humans [6], with GII and GI predominantly associated with outbreaks [7, 8]. GII.4 viruses are responsible for the majority of human outbreaks, causing an estimated 60-70% of such cases [7, 9].

The positive-sense viral RNA genome is composed of three open reading frames (ORFs). ORF1 encodes a polyprotein that is processed by a 3C-like protease (3CLpro) into functional proteins including the helicase, protease and polymerase, ORF2 encodes the capsid protein, and ORF3 encodes a small basic protein. While no anti-norovirus therapy has yet been approved for human use, the 3CLpro, a cysteine protease, has emerged as an attractive drug target due to its essential role in viral maturation. Significant progress has been made targeting norovirus proteases: inhibitors of the 3CLpro have been reported with IC₅₀ values in the low nanomolar range [10–20]. However, much of this progress has been made with GI norovirus proteases, such as the Norwalk virus protease (GI.1) [12, 19, 21, 22], Chiba virus protease (GI.4) [23, 24], or Southampton virus protease (GI.2) [25, 26] serving as the target. To date, the MD145 remains the only GII.4 norovirus protease reported in the literature [5].

We report here for the first time the expression, purification, and characterization of a novel GII.4 norovirus protease – the Minerva virus protease (MVpro). MVpro was expressed using Escherichia coli and purified via 6x-His affinity and size-exclusion chromatography. Pure MVpro was characterized using a fluorescence resonance energy transfer (FRET) protease assay. The successful purification and characterization of MVpro increases our knowledge of GII.4 noroviruses and represents a new target to guide the synthesis of future anti-norovirus therapies.

2. MATERIALS AND METHODS

2.1. Cloning and Small-Scale Expression

The cDNA encoding the 19kD NS6 protease was obtained in a pET28a vector (Invitrogen) (Genbank accession no: EF684915, amino acids 1009-1188, which corresponds to the 2006b variant of the Minerva virus. Amplification by PCR used the following primers: 5’-GAATAAGAAGACATAGGTGCCCCACCAAGCATC-3’ (forward); 5’-GATACGCTCGAGTTATTCAAGTGTAGCTTCC-3’ (reverse). The PCR product was then ligated into a pSUMO vector (LifeSensors) containing a T7 promoter, an N-terminal His6-SUMO tag, and BbsI and Xho2 restriction sites. The resulting clone was transformed into Escherichia coli BL21 Codon Plus (DE3) cells for protein expression.

Small-scale (5 mL) cultures were prepared to optimize conditions for protein overexpression. Transformed E. coli BL21 Codon Plus (DE3) cells containing the MVpro insert were grown in 5 mL LB medium in the presence of streptomycin. Protein expression was induced by the addition of isopropyl-β-D-thiogalactoside (IPTG). Three variables were tested to optimize protein overexpression: 1) OD₆₀₀ before induction, 2) concentration of IPTG, and 3) temperature. Cell cultures were induced at either an OD₆₀₀ of 0.5 or 1.0 with the addition of either 0.1 mM, 0.4 mM, 0.6 mM, or 1.0 mM IPTG. After IPTG induction, protein expression was conducted at either 37°C for three hours or 15°C overnight. Cells were lysed and evaluated by SDS-PAGE (15% w/v polyacrylamide) for protein expression levels and protein solubility.

2.2. Large-Scale Protein Expression and Purification

Large-scale protein expression was performed using a 2-liter culture. The cultures were grown to an OD₆₀₀ of 1.0 at 37 °C in LB medium. Protein expression was induced by addition of 0.1 mM IPTG and was carried out at 37 °C for 3 hours. The cells were harvested by centrifugation and lysed by French Press. The soluble fraction was purified using a Ni²⁺ affinity column (HisTrap™ HP, GE). To separate the His6-SUMO tag from the MVpro, proteolytic cleavage of the His6-SUMO tag with 1x yeast SUMO Protease 1 (ULP-1) (LifeSensors) was performed overnight on the eluted fractions according to lab optimized protocol. MVpro was then separated from the cleaved His6-SUMO tag by a second round of Ni²⁺ affinity chromatography. MVpro was further purified by size-exclusion chromatography (Superdex 200, GE). The pure MVpro was concentrated to 20 mg/mL in a buffer containing 20 mM Tris-HCl (pH 8.0), 150 mM NaCl, 5 mM β-mercaptoethanol, and 5% glycerol. Eluted fractions for each step of purification were analyzed by SDS-PAGE (15% w/v polyacrylamide) and Coomassie Brilliant Blue staining.

2.3. Kinetic Enzyme Assay

Purified MVpro was tested for enzymatic activity with a FRET-based enzyme assay. The MVpro was diluted in reaction buffer (50 mM HEPES, Ph 8.0, 120 mM NaCl, 0.4 mM EDTA, 20% glycerol, and 4 mM DTT) to a final concentration of 128 nM. Each reaction was initiated by addition of FRET substrate (HiLyte Fluor 488 – DFELQGPK(QXL520)-NH2, purchased from Anaspec, Inc.). The FRET substrate is based on the P5-P2’ residues of the Minerva virus NS2/NS3 cleavage site. A C-terminal lysine was added in order to attach the QXL520 fluorescence quencher to the substrate. To determine kinetic parameters, the FRET substrate was serially diluted to final concentrations of 100 μM to 49 nM and added to the reaction. The final reaction volume was 100 μL. The fluorescence emitted by substrate cleavage was monitored by a microplate reader (Spectra-Max M5, Molecular Devices, Sunnyvale, CA, USA) at 488 nm excitation wavelength with an emission wavelength of 520 nm. Readings were taken every minute for 20 minutes, and the reactions were performed at 37°C. In order to convert relative fluorescence units (RFU) into μM, a standard curve was created by measuring fluorescence of free HiLyte Fluor 488, which was serially diluted from 250 nM to 3.9 nM. All data were plotted and analyzed with GraphPad Prism v. 6.07.

2.4. Enzyme Inhibition Assay with Compound A

Compound A was tested for anti-proteolytic activity against MVpro. Compound A, a white solid, is a peptidomimetic protease inhibitor with a molecular mass of 581.4 Da. The compound A structure is presented later (Fig. 5), and its organic synthesis is described in the Supplementary Information section (Fig. S3).

Fig. (5).

Inhibition of MVpro by compound A. Structure of and inhibition data for compound A. The data is represented as the mean ± one standard deviation. The experiment was performed in duplicate.

Purified MVpro was diluted to a final concentration of 128 nM in reaction buffer (50 mM HEPES, pH 8.0, 120 mM NaCl, 0.4 mM EDTA, 20% glycerol, and 4 mM DTT). Compound A was serially diluted in DMSO and was incubated with diluted enzyme for 90 min at 37°C to ensure complete inactivation. Each reaction was initiated by the addition of 15 μM FRET substrate. The final reaction volume was 100 μL. The fluorescence monitoring and analysis are the same as above. The Morrison equation for tight-binding inhibitors was used to calculate the K_i.

3. RESULTS AND DISCUSSION

3.1. Comparison of GI and GII Norovirus Proteases

The genome of noroviruses exists in three open reading frames (ORFs). ORF1, which contains non-structural proteins is cleaved in five locations by the viral polyprotein [5] (Fig. 1A). The Minerva virus protease (MVpro) recognizes Q399/G400 (NS2/NS3), Q762/G763 (NS3/NS4), E961/G962 (NS4/NS5), E1099/A1100 (NS5/NS6), and E1280/G1281 (NS6/NS7) as cleavage junctions (Fig. 1A). Comparison of the MVpro cleavage junctions to the Norwalk virus protease (NoVpro) cleavage junctions showed that the P2-P2’ residues are nearly universally conserved, differing only at the P2 residue of the NS4/NS5 junction (Fig. 1B). Norwalk virus, the most widely studied norovirus, is a GI.1 norovirus and serves as the prototype norovirus. Extending past the P2/P2’ residues display more sequence variability between the GI.1 norovirus (NoVpro) and the GII.4 norovirus (MVpro). Overall, the cleavage junctions are well conserved.

Fig. (1).

Genome organization and alignment to GI noroviruses. A) Norovirus genomes contain three open reading frames. Norovirus proteases are responsible for cleavage of the ORF1 polyprotein, which contains the viral non-structural proteins: N-terminal protein (N term,NS1-2), NTPase (NS3), p20 (NS4), VPg (NS5), viral 3C-like protease (Pro,NS6), RNA-dependent RNA polymerase (Pol, NS7). ORF2 encodes the major capsid protein (VP1) and ORF3 encodes the minor structural protein (VP2). B) Minerva virus protease (GII.4) cleavage site junctions aligned with Norwalk virus protease (GI.1) cleavage site junctions. C) Sequence alignment of Minerva virus protease (MVpro), Chiba virus protease (CVpro), Norwalk virus protease (NoVpro), and Southampton virus protease (SVpro). MVpro is a GII.4 norovirus whereas CVpro, NoVpro, and SVpro are GI.4, GI.1, and GI.2 noroviruses, respectively.

Additionally, the sequence of MVpro was aligned to the Norwalk virus protease (NoVpro) [22], the Southampton virus protease (SVpro) [26], and the Chiba virus protease (CVpro) [27]. NoVpro, SVpro, and CVpro represent GI.1, GI.2, and GI.4 noroviruses, respectively [28]. The alignment in Fig. 1C shows that the GI proteases are highly conserved (at least 90% sequence identity). A higher degree of variation exists between genogroups as MVpro is 68.51%, 67.05%, and 66.85% identical to SVpro, CVpro, and NoVpro, respectively. Highly conserved areas include the catalytic triad, consisting of His30, Glu54, and Cys139, the S1 substrate pocket (residues 134-138, and 157), the S2 pocket (residues 30, 109, 112113, and 158-159), the S4 pocket (residues 107, 118, 161, 166, 168), as well as residues which have been reported to contribute important hydrogen bonds to the substrate backbone (residues 110, 158, and 160) (Fig. 1C).

Due to the high sequence similarity, the structure of MVpro is expected to closely mimic the structures of the aforementioned GI norovirus proteases. Additionally, the conserved substrates and conserved substrate binding pockets suggest that the shape of the MVpro active site should also closely resemble the shape of the NoVpro active site. Therefore, potent inhibitors targeting NoVpro [10–20] may also be applicable to GII norovirus proteases, such as the MVpro.

3.2. Overexpression and Purification of Minerva Virus Protease

A 2-liter culture of E. coli was used to overexpress MVpro. Prior to the 2-liter culture, overexpression conditions were optimized as described in section 2.1. The soluble fraction of the cell lysate was purified according to the scheme shown in Fig. 2A. The expressed His6-SUMO-MVpro fusion protein appears at a molecular weight of ~40 kDa on an SDS-PAGE gel (Fig. 2B). After the sample was applied to the first HisTrap™ column, a gradient of imidazole was used to elute the His6-SUMO-MVpro fusion protein. The fusion protein started to elute at about 174 mM imidazole and reached completion at approximately 310 mM imidazole (Fig. 2C). Before the sample was applied a second time to the HisTrap™ column, ULP-1 was used to cleave the His6-SUMO tag from MVpro (Fig. 2D). Once separated from the His6-SUMO tag, nearly pure MVpro was eluted in the flow-through and wash of the second HisTrap™ column (Fig. 2D). The expected MW of MVpro is 19.3 kDa; MVpro displays an apparent MW of ~20 kDa on an SDS-PAGE gel (Fig. 2D).

Fig. (2).

Protein purification scheme. A) Schematic diagram of the purification process for Minerva virus protease. B) Uninduced and induced cells as well as cell lysate fractions and the 1st Ni2+ affinity column load, flow-through, and wash. C) 1st Ni2+ affinity column eluted fractions containing the His6-SUMO-MVpro fusion protein. D) Steps 2 and 3 of scheme A): after cleavage with ULP-1, MVpro appears at a MW of ~20 kDa. MVpro is eluted in the flow-through of the 2nd Ni2+ affinity column. SE = size exclusion column load. Panels B)-D) are 15% SDS-PAGE gels stained with coomassie blue. (The color version of the figure is available in the electronic copy of the article).

To ensure that pure MVpro was obtained, the sample was further purified with size-exclusion chromatography. Fig. (3A) shows a strong peak starting to elute at about 78 mL. Interestingly, instead of exhibiting a symmetrical peak, the chromatogram seems to show the presence of two protein species: the first species is evidenced by the first peak at 85 mL. A shoulder at 92.5 mL suggests a second protein species (Fig. 3). The first peak at 85 mL corresponds to an MW of about 36 kDa, or dimer MVpro. The shoulder at 92.5 mL corresponds to an MW of about 20 kDa, or monomer MVpro. Based on these data, it appears that MVpro, in solution, exists in both monomer and dimer states as reported previously [22]. The relative strength of the signal suggests that the dimer may be slightly favored (Fig. 3A). However, further testing must be carried out to confirm the monomeric and dimeric states.

Fig. (3).

Size exclusion chromatography. A) chromatogram showing the eluted peak of the size-exclusion column. The asymmetrical peak appears to contain two species, dimer (85 mL) and monomer (92.5 mL), of MVpro. B) 15% SDS-PAGE gels stained with Coomassie Brilliant Blue showing the eluted fractions of pure MVpro. Two bands are apparent: ~20 kDa, and ~ 19 kDa which suggests auto-proteolysis of the C-terminus.

Somewhat unexpectedly, the fractions containing MVpro eluted from the size-exclusion column appeared to show two distinct bands: one dark band at ~20 kDa and a lighter band just below at ~19 kDa (Fig. 3B). The presence of any nonspecific impurities is unlikely considering the rigorous purification. Inspection of the MVpro sequence revealed a Q/G, E/G, and an E/A cleavage junction at the C-terminus. Mutating the catalytic cysteine to an alanine (C139A) prevented the two bands from appearing on an SDS-PAGE gel suggesting that the faint band at ~19 kDa may be the result of autoproteolysis at the C-terminal end of the protein (Fig. S1).

3.3. Characterization of Enzymatic Activity

Following successful purification, MVpro was subjected to a FRET-based assay to test for enzymatic activity. A protocol described by Kyeong-Ok Chang’s group was used to design the assay to detect MVpro activity [5]. The substrate used for our assay was HiLyte Fluor 488 - DFELQGPK(QXL520) - NH2. This substrate represents the P5-P2’ residues of the MVpro NS2/NS3 cleavage site. MVpro efficiently cleaves the HiLyte Fluor 488 - DFELQGPK(QXL520) - NH2 substrate as evidenced by the kinetic data, in which the initial velocity is plotted as a function of substrate concentration (Fig. 4).

Fig. (4).

Enzyme kinetics of MVpro. Cleavage velocity of NS2/NS3 substrate (HiLyte Fluor 488 - DFELQGPK(QXL520) - NH2) plotted as a function of substrate concentration. MVpro exhibits classic MichaelisMenten kinetics. Data is represented as mean velocity ± standard error of the mean (SEM).

Compared to other norovirus proteases, MVpro appears to be more efficient than both NoVpro (GI.1) and MD145pro (GII.4) [5]. The enzymatic data for these proteases are shown in Table 1. MVpro is nearly twice as efficient as NoVpro and slightly more efficient than MD145pro: the specificity constants (k_cat/K_M) for MVpro, NoVpro, and MD145 are 2.024 x 10³ (M⁻¹min⁻¹), 1.111 x 10³ (M⁻¹min⁻¹), and 1.873 x 10³ (M⁻¹min⁻¹), respectively (Table 1). Although the specificity constants of the three proteases are similar, the turnover (k_cat) and Michaelis constant (K_M) are drastically different. MVpro cleaves the NS2/NS3 substrate at a rate 15-fold slower when compared to MD145pro and 18-fold slower when compared to NoVpro. However, MVpro displays a 23-fold increase in affinity for the NS2/NS3 substrate compared to MD145pro and a 39-fold increase compared to NoVpro (Table 1).

Table 1.

Kinetic parameters for MVpro, NoVpro, and MD145.

Protease	Substrate	kcat (min−1)	Km (μM)	kcat/Km (M−1min−1)
MVpro* (GII.4)	HiLyte Fluor 488 - DFELQGPK(QXL520) - NH2	0.0142 ± 0.00055	7.001 ± 1.075	2.024 x 103
NoVpro# (GI.1)	Edans-DFHLQGP-dabcyl	0.2626 ± 0.02562	272.3 ± 58.19	1.111 x 103
MD145# (GII.4)	Edans-DFHLQGP-dabcyl	0.2201 ± 0.02443	159.8 ±	1.873 x 103

^*MVpro data is represented as the mean ± standard error. Experiments were performed in triplicate.

^#Data extracted from Chang et al. 2012[5].

Differences in experimental set up may account for the variation seen in the enzymatic data (Table 1). First, the substrate presented here used HiLyte Fluor 488/QXL520 as the fluorogenic pair for the NS2/NS3 substrate compared to the edans/dabcyl pair used by Chang et al. [5]. It is likely that the different fluorescent pairs would have an effect on substrate binding. Second, the assay presented here is likely more sensitive in the nanomolar range. The MVpro NS2/NS3 substrate was serially diluted from concentrations ranging from 100 μM to 98 nM, compared to the NoVpro substrate which was serially diluted from ~640 μM to ~40 μM [5].

3.4. Inhibition of the Minerva Virus Protease by Compound A

Based on the high affinity of MVpro for the NS2/NS3 substrate, a peptidomimetic inhibitor resembling the substrate was synthesized. Compound A is a tetrapeptide inhibitor containing a hydrophobic cap and an aldehyde warhead (Fig. 5). Compound A potently inhibits MVpro, exhibiting an IC₅₀ of 112 ± 25 nM and a K_i of 49.7 ± 2.3 nM (Fig. 5); comparable to the inhibitory activity of compound 16 against Norwalk virus with an IC₅₀ of 0.1 μM described by Galasiti Kankanamalage et al. [18].

Previous studies have indicated that inhibitors containing an aldehyde warhead may interact covalently with the catalytic Cys139 of norovirus proteases [11, 12, 17–19] (Fig. S4). Compound A was therefore expected to inhibit MVpro covalently. Time course assays performed with compound A show that full inactivation of MVpro by compound A is achieved after 90 minutes incubation at 37°C (Fig. S2). A zero minute incubation resulted in an IC₅₀ value of ~700 nM, compared to the 112 nM activity after 90 minutes incubation (Fig. S1). This suggests that compound A may inhibit MVpro covalently. It is possible that the IC₅₀ at zero minutes incubation is representative of a non-covalent collision complex, whereas the 90-minute IC₅₀ may represent full, covalent inhibition. To date, limited data showing inhibition of GII.4 norovirus proteases exists [5, 10, 17]. Compound A, therefore, represents an advanced starting point for the design of against GII norovirus protease inhibitors.

CONCLUSION

We report here for the first time successful expression, purification, characterization, and inhibition of the Minerva virus protease (MVpro), a GII.4 norovirus protease. Pure MVpro, which appears to exist as both monomer and dimer in solution, cleaves the NS2/NS3 synthetic substrate more efficiently than other norovirus proteases such as the Norwalk virus protease (GI.1) and the MD145 protease (GII.4). Inhibition studies involving GII.4 norovirus proteases remain underrepresented in the literature, despite GII.4 noroviruses being responsible for a majority of outbreaks. Compound A, a potent inhibitor of MVpro, may be a good starting point for the design of inhibitors to target GII norovirus proteases. Furthermore, the results presented here will allow for future characterization of MVpro inhibitors as they are synthesized.