RNA 5′-Triphosphatase Activity of the Hepatitis E Virus Helicase Domain^▿

Hepatitis E virus (HEV) is the sole member of the genus Hepevirus in the family Hepeviridae. HEV is a nonenveloped virus with a single-stranded positive-sense RNA genome of ∼7.2 kb, consisting of a 5′ noncoding region (5′-NCR) of 27 to 35 nucleotides (nt) followed by three open reading frames (ORFs), ORF1, ORF2, and ORF3, and a 3′ NCR of 65 to 74 nt ending with a poly(A) tail of variable length. ORF1 is known to encode the viral nonstructural polyprotein of ∼186 kDa (1). Based on protein sequence homology, the ORF1 polyprotein is proposed to contain domains indicative of methyltransferase, papain-like cysteine protease, RNA helicase, RNA-dependent RNA polymerase, and two domains (X, Y) of unknown function (17). The HEV genome has a 5′ m7G cap (15). ORF2 and ORF3 proteins of HEV are encoded by a 2.2-kb subgenomic RNA, which is also capped (8).

HEV belongs to the “alpha-like” superfamily of positive-sense single-stranded RNA viruses. These viruses multiply in the cytoplasm and are equipped with their own capping machinery to carry out mRNA capping reactions. As per the current knowledge, GTP is first methylated at the N7 position of guanine, and the m7GMP moiety of the m7GTP is then transferred to RNA through a two-step mechanism involving a covalent (enzyme-m7GMP) intermediate. The presence of four universal motifs within the methyltransferase domain is the hallmark of this superfamily (23). Enzymatic activities involved in capping have been demonstrated for the members of this group, namely, Semliki Forest virus (SFV) (2), bamboo mosaic virus (BaMV) (12, 18), tobacco mosaic virus (TMV) (22), and brome mosaic virus (BMV) (3). The conventional cellular mRNA capping involves the following steps: (i) removal of γ-phosphate from the 5′ end of the primary transcripts (pppRNA to ppRNA) by RNA triphosphatase, (ii) reaction of guanylyltransferase with GTP to form a covalent complex containing GMP, (iii) transfer of GMP onto 5′-ppRNA to form G(5′)ppp(5′)RNA, and (iv) addition of the methyl group by guanine-7N-methyltransferase to form the cap-0 structure m7GpppRNA (24).

Methyltransferase activity was demonstrated for the N-terminal 979-amino-acid-long recombinant protein of HEV ORF1. The protein catalyzed transfer of the methyl group from S-adenosylmethionine to GTP to yield m7GTP and also formed a covalent complex with m7GMP, indicating associated guanylyltransferase activity (20). To transfer m7GMP to the 5′ end of nascent RNA, it is essential to remove 5′-γ-phosphate from the RNA by RNA triphosphatase. So far, biochemical evidence is lacking for the protein acting as RNA 5′-triphosphatase in HEV.

RNA triphosphatases specifically cleave 5′-γ-phosphate of the nascent mRNAs, without attacking the β-phosphoryl group. The documented RNA 5′-triphosphatases can be grouped into three classes according to the catalytic mechanisms they employ. (i) The metal-independent cysteine phosphatases have a characteristic HCC(X)₅R(S/T) motif in which the cysteine residues act as a catalytic nucleophile within the catalytic pathway. Their enzymatic activity is independent of divalent cations. These enzymes are found in plants and metazoa (7, 25). (ii) The metal-dependent phosphohydrolases possess two essential glutamate-containing motifs and require divalent cations for their activity. They also possess the ability to hydrolyze nucleoside triphosphates (NTPs). These enzymes are found in DNA viruses, fungi, and protozoa (9, 11, 21). (iii) The RNA triphosphatases from RNA viruses are helicases or helicase-like proteins in which the active site of the RNA triphosphatase is either shared or overlaps with the helicase/NTPase catalytic site. These include enzymes identified from RNA viruses such as reovirus (6), alphaviruses (27), flaviviruses (5, 28), coronaviruses (13, 14), and BaMV (10, 19).

Previously, we demonstrated that the helicase domain of HEV expressed in Escherichia coli (amino acids 960 to 1204 from ORF1) (HEV Hel) exhibits NTPase and 5′-to-3′ RNA duplex unwinding activities (16). Helicase-associated RNA 5′-triphosphatase (RTPase) activity in the members of alpha-like virus superfamily led us to check this activity in HEV Hel.

Two RNA substrates, a nonspecific sequence (5′-GGGA₂₄-3′) and 5′-NCR (27 bases) of the HEV genome (genotype 1, isolate PM2000; collected by the National Institute of Virology, Pune, India, during an outbreak in Lonawala, Western India, in 2000 [GenBank accession no. DQ459342]) with a 5′-end label were generated by in vitro transcription (Riboprobe in vitro transcription system; Promega) in the presence of 50 μCi of either [γ-³²P]GTP or [α-³²P]GTP (each 3,200 Ci/mmol; BRIT, India) using T7 RNA polymerase. The double-stranded DNA templates used for transcription reactions were prepared by annealing the following pairs of oligonucleotides: A, nonspecific RNA, 5′-AAAATAATACGACTCACTATAGGG(A)₂₄-3′ (T7 promoter sequence underlined) and 5′-(T)₂₄CCCTATAGTGAGTCGTATTATTTT-3′; and B, 5′-NCR, 5′-AAAATAATACGACTCACTATAGCAGACCACATATGTGGTCGATGCCAT-3′ and 5′-ATGGCATCGACCACATATGTGGTCTGCTATAGTGAGTCGTATTATTTT-3′. Synthesized RNAs were separated from unincorporated radioactive nucleotides by two rounds of gel filtration through Sephadex G50 columns. The RTPase assay was carried out with a 20-μl volume containing 50 mM HEPES (pH 7.0), 2 mM MgCl₂, 0.05 mg of bovine serum albumin (BSA) per ml, 2 mM dithiothreitol (DTT), 100 pmol of HEV Hel protein, and 10 nmol of 5′-γ-³²P- or 5′-α-³²P-labeled RNA. Following incubation at 37°C for 30 min, products were analyzed by thin-layer chromatography (TLC) and exposed to X-ray film. HEV Hel released labeled the 5′-phosphate group from 5′-γ-³²P-labeled RNA (Fig. 1A). Released radiolabel comigrated with the orthophosphate (Pi) produced by [γ-³²P]GTP hydrolysis, indicating that it cleaved the β-γ-phosphate bond of the 5-terminal nucleotide of the 27-nt RNA substrate. This result suggested that HEV Hel has RNA 5′-triphosphatase (RTPase) activity. To determine the specificity of RTPase, 5′-[γ-³²P]RNA and 5′-[α-³²P]RNA were used as substrates. Each 5′-radiolabeled RNA was incubated with the HEV Hel and shrimp alkaline phosphatase (SAP) separately. As expected, SAP could remove the ³²P moiety from both RNA substrates, but HEV Hel showed no release of label from 5′-[α-³²P]RNA, indicating that HEV Hel hydrolyzes only the γ-β-triphosphate bond and, unlike SAP, does not have a general phosphohydrolase activity that would also remove the β- and α-phosphate groups (Fig. 1B). Release of the 5′-phosphate group from the nonspecific RNA oligonucleotide and that from the HEV 5′-NCR oligonucleotide were equally efficient. This confirmed that HEV Hel 5′-RTPase does not have any sequence specificity. Hence, for all further experiments, RNA with nonspecific sequence was used as a substrate. To exclude the possibility that contaminating E. coli RNase activities degraded the RNA substrate and thereby released [γ-³²P]GTP that was then hydrolyzed by the GTPase activity of HEV Hel, the products were also analyzed by polyacrylamide gel electrophoresis (PAGE). There was no RNase contamination, as both RNA substrates remained intact during the course of the reaction and no products smaller than substrate RNA were observed by PAGE (Fig. 1C).

Further, to rule out the possibility of contamination of bacterial phosphatases in the HEV Hel, the enzyme was treated at different temperatures prior to incubation with 5′-[γ-³²P]RNA. HEV Hel showed complete inhibition after incubation at 55°C (Fig. 1D). Since bacterial phosphatases are characteristically thermoresistant at 65°C (26), the results affirm the existence of RTPase activity in HEV Hel.

As reported previously, mutations (underlined) in the Walker A motif (AGVPGSGKS to AGVPGSGAS for Hel mutant [mut] I) and Walker B motif (VIDEAP to VIAAAP for Hel mut II) of HEV Hel result in a significant loss of ATPase activity and the complete loss of unwinding ability of the mutants (16). To find out the effects of mutations on RTPase activity, mutant enzymes were incubated with 5′-[γ-³²P]RNA substrate. For the quantitation of released Pi from RNA, the corresponding spots from TLCs were excised and counted with a Packard Tri-Carb 2810TR liquid scintillation counter. For each reaction, a sum of the counts of 5′-[γ-³²P]RNA plus Pi spots was taken as total counts, and the fraction of radioactivity released was determined. Hel mut I and Hel mut II showed significantly less activity than wild-type HEV Hel (Fig. 2A). Overall, effects of mutations on the ATPase and RTPase activities were similar, suggesting that the NTPase and RTPase activities shared a common reaction mechanism. To further support this, HEV Hel was incubated with 5′-[γ-³²P]RNA in the presence of different concentrations of ATP, ADP, AMP, and two nonhydrolyzable ATP analogs, adenosine-5′-O-thiotriphosphate (ATP-γ-S) and adenylyl-imidodiphosphate (AMP-PNP), to see the extent of competition. ATP showed very effective inhibition of RTPase even at a 0.1 mM concentration, indicating that ATP and RNA shared the binding site (Fig. 2B and C). ADP and AMP did not affect the RTPase activity even at a 5 mM concentration. This result was expected; as ATP is the authentic substrate, it would have higher binding affinity than that of ADP or AMP. This confirmed that the 5′-terminal γ-phosphate group in the substrate is the major determinant for the competition. AMP-PNP and ATP-γ-S were comparatively less effective competitors than ATP, probably due to weak binding.

It was previously demonstrated by us that NTPase and helicase activities of HEV Hel are Mg²⁺ dependent (16). To see the effect of Mg²⁺ on RTPase activity of HEV Hel, incubations were done in the presence of different MgCl₂ concentrations (0 to 5.0 mM). Enzyme could release phosphate even in the absence of Mg²⁺, and the released phosphate levels increased further in the presence of Mg²⁺. The maximum level of enhancement in the enzymatic activity in the presence of Mg²⁺ was not more than 2-fold (Fig. 3A and B). This clearly indicated that there is no absolute requirement for Mg²⁺ for the RTPase activity of HEV Hel. This was further confirmed by adding different concentrations of EDTA (0.5 to 5.0 mM) in the activity assay. There was no effect of EDTA on the enzymatic activity (Fig. 4A and B). BaMV (19), SFV, Sindbis virus (27), and reovirus (6) have helicases which harbor Mg²⁺-dependent RTPase activities. Nsp13-associated RTPase activities in human coronavirus 229E and severe acute respiratory syndrome (SARS) coronavirus in the presence of Mg²⁺ were demonstrated (13, 14). However, there are contradictory reports on the Mg²⁺ requirements of NS3 protein of Dengue and West Nile viruses with respect to RTPase activity (4, 5, 28).

In summary, HEV Hel has enzymatic functions as NTPase, helicase, and RNA 5′-triphosphatase. The ability of the enzyme to bind and hydrolyze NTPs, deoxynucleoside triphosphates (dNTPs) and 5′-RNA-triphosphate suggests that the binding site of HEV Hel is determined primarily by the triphosphorylated nucleotides. Since helicase-associated RNA 5′-triphosphatase activity involved in 5′ cap formation has been demonstrated for many RNA virus families (6, 19, 27, 28) and HEV is not known to have any other RNA 5′-triphosphatase domain, it seems reasonable to suggest that HEV employs the helicase to mediate the first step of 5′ cap synthesis.