Vibrio Phage KVP40 Encodes a Functional NAD⁺ Salvage Pathway

ABSTRACT

The genome of T4-type Vibrio bacteriophage KVP40 has five genes predicted to encode proteins of pyridine nucleotide metabolism, of which two, nadV and natV, would suffice for an NAD⁺ salvage pathway. NadV is an apparent nicotinamide phosphoribosyltransferase (NAmPRTase), and NatV is an apparent bifunctional nicotinamide mononucleotide adenylyltransferase (NMNATase) and nicotinamide-adenine dinucleotide pyrophosphatase (Nudix hydrolase). Genes encoding the predicted salvage pathway were cloned and expressed in Escherichia coli, the proteins were purified, and their enzymatic properties were examined. KVP40 NadV NAmPRTase is active in vitro, and a clone complements a Salmonella mutant defective in both the bacterial de novo and salvage pathways. Similar to other NAmPRTases, the KVP40 enzyme displayed ATPase activity indicative of energy coupling in the reaction mechanism. The NatV NMNATase activity was measured in a coupled reaction system demonstrating NAD⁺ biosynthesis from nicotinamide, phosphoribosyl pyrophosphate, and ATP. The NatV Nudix hydrolase domain was also shown to be active, with preferred substrates of ADP-ribose, NAD⁺, and NADH. Expression analysis using reverse transcription-quantitative PCR (qRT-PCR) and enzyme assays of infected Vibrio parahaemolyticus cells demonstrated nadV and natV transcription during the early and delayed-early periods of infection when other KVP40 genes of nucleotide precursor metabolism are expressed. The distribution and phylogeny of NadV and NatV proteins among several large double-stranded DNA (dsDNA) myophages, and also those from some very large siphophages, suggest broad relevance of pyridine nucleotide scavenging in virus-infected cells. NAD⁺ biosynthesis presents another important metabolic resource control point by large, rapidly replicating dsDNA bacteriophages.

IMPORTANCE T4-type bacteriophages enhance DNA precursor synthesis through reductive reactions that use NADH/NADPH as the electron donor and NAD⁺ for ADP-ribosylation of proteins involved in transcribing and translating the phage genome. We show here that phage KVP40 encodes a functional pyridine nucleotide scavenging pathway that is expressed during the metabolic period of the infection cycle. The pathway is conserved in other large, dsDNA phages in which the two genes, nadV and natV, share an evolutionary history in their respective phage-host group.

INTRODUCTION

Many bacteria encode both de novo and salvage pathways for pyridine nucleotide (NAD⁺) synthesis (1, 2). From the paradigm pathways studied in Escherichia coli and Salmonella enterica serovar Typhimurium (S. Typhimurium), nad genes typically encode enzymes of the de novo pathway that convert aspartate to quinolinic acid to NAD⁺, whereas the pyridine nucleotide salvage pathway is encoded by several pnc and pnu genes. The latter enzymes scavenge nicotinic acid (NA), nicotinamide (NAm), or the derivative mononucleotides (NAMN/NMN) for conversion into NAD⁺ (2, 3). These pathways are detailed at KEGG (http://www.genome.jp/kegg-bin/show_pathway?map00760).

Certain bacteria lack the capacity for de novo synthesis of the pyridine ring and encode only an abbreviated form of the pyridine scavenging pathway. This was first described for members of the Pasteurellaceae that have the capacity to assimilate NAm or NMN to yield NAD⁺ (4–6). Other bacteria of this group lack this ability, such as Haemophilus influenzae, other Haemophilus species, and Actinobacillus pleuropneumoniae; these bacteria must be provided an amidated pyridine-ribose intermediate, such as NMN or NAm-ribose (5).

In 2001, Martin et al. (7) described a single gene, nadV, present on a plasmid from Haemophilus ducreyi 27722 that encodes a nicotinamide phosphoribosyltransferase (NAmPRTase) and catalyzes the conversion of NAm to NMN. nadV was originally characterized on a plasmid, but homologues occur on the chromosome of several bacterial genera (7); interestingly, H. ducreyi nadV is on an integrating plasmid or phage-like mobile element (8). The distribution of these genes among microbes and the enzyme activities they encode have been reviewed (2).

Bacteriophage KVP40 is a lytic, double-stranded DNA (dsDNA) phage that infects Vibrio cholerae, Vibrio parahaemolyticus, and other Vibrio species (9). KVP40 is a myovirus that is distantly related to phage T4, although only 26% of all KVP40 genes are shared with T4 (10, 11). Indeed, more than 60% of KVP40-encoded proteins are hypothetical and of unknown function. Among the predicted 386 KVP40 proteins, five appear to be enzymes for pyridine nucleotide metabolism, which led to the prediction of a phage-encoded NAD scavenging pathway (10). Two of these enzymes, NadV (NAmPRTase) and NatV nicotinamide mononucleotide adenylyltransferase (NMNATase), would suffice for a two-reaction phage-encoded scavenging system (Fig. 1). We describe their purification, confirm their activity, and characterize their expression during KVP40 infection of V. parahaemolyticus.

FIG 1.

Phage-encoded pyridine nucleotide scavenging pathway. Two reactions, NadV nicotinamide phosphoribosyltransferase and NatV nicotinamide mononucleotide adenyltransferase, together yield NAD⁺ (1, 2, 10). The pyridine compounds are nicotinamide (NAm), nicotinamide mononucleotide (NMN), and nicotinamide dinucleotide (NAD⁺). PRPP, 5-phospho-d-ribosyl-1-pyrophosphate.

RESULTS

Prevalence of NAD scavenging pathways encoded by phage genomes.

When the nadV and natV genes were annotated in the KVP40 genome (10), a pyridine nucleotide scavenging pathway coded by a viral genome had not been described and few related genes had been identified. To update a previous database survey for these enzymes (2), GenBank was searched with BLASTP using the 497- and 341-amino-acid proteins of KVP40, NadV and NatV, respectively. NadV enzymes contain a common NAmPRTase domain (cd01569, pfam04095, PHA02594, COG1488) over their entire length; these enzymes are encoded by hundreds of different bacterial genomes in the database (e.g., Proteobacteria, Actinobacteria, Firmicutes, etc.). NatV enzymes contain two domains, an NMN nucleotidyl transferase (cd02165 and COG1056) and a Nudix hydrolase (cd02883, pfam00293, and COG1051) aligning with the ADP-ribose pyrophosphatases. The bifunctional NMNATase-Nudix enzymes (cd02168) (12) are encoded by many bacteria and phages, with >66% KVP40 NatV query coverage required to include the two domains; protein hits that did not include both domains were excluded from further analysis.

At least 169 different phages infecting hosts across the kingdom Bacteria encode an NadV-like NAmPRTase. Only 50 documented phages encode a bifunctional NatV-like enzyme, and all of these also contain a nadV gene in their genomes (Table 1). Hosts of the 50 phages that have both nadV and natV are found within three classes of Proteobacteria, i.e., Alphaproteobacteria (Caulobacter, Reugeria, and Sinorhizobium), Betaproteobacteria (Ralstonia), and Gammaproteobacteria (Aeromonas, Citrobacter, Cronobacter, Enterobacter, Escherichia, Klebsiella, Stenotrophomonas, and Vibrio). All phages with both genes have relatively large dsDNA genomes (from Ruegeria phage DSS3P8 of 146,135 bp to Cronobacter sakazakii phage GAP32 of 358,663 bp). Nearly all of the phages with the two-gene system are Myoviridae phages, with the exception of the Siphoviridae phage DSS3P8 and the large Siphoviridae Caulobacter crescentus phiCbK and phiCbK-like phages (13, 14). Clustal Omega alignments and neighbor-joining trees (Fig. 2; see Tables S1 and S2 in the supplemental material) show that both NadV and NatV protein sequences cluster with enzymes from phages that infect the same host (or a closely related one), and both enzymes show approximately the same phylogenetic relationships and distances. The exceptions are two enzymes from Stenotrophomonas maltophilia phage IME13 that cluster with enzymes from Aeromonas salmonicida phages (Aes012, Aes025, etc.). BLAST searches show that the capsid and DNA polymerase of IME13 (GenBank accession no. JX306041) also align with phages infecting Aeromonas strains. Either IME13 has a unique genome structure and host range, or a revision to the host taxonomy is required. A similar observation was seen with the two enzymes from Vibrio phage 32; most of the identified proteins encoded by this phage align most closely to Aeromonas phages clustering with Aeh1.

TABLE 1.

Bacteriophage genomes with both nadV and natV genes^a

Phage	Accession no.	Genome size (bp)	Hostb	Morphotype
CcrColossus	JX100810	279,967	Caulobacter crescentus	S
CcrKarma	JX100811	221,828	C. crescentus	S
CcrRogue	JX100814	223,720	C. crescentus	S
CcrSwift	JX100809	219,216	C. crescentus	S
CcrMagneto	JX100812	218,929	C. crescentus	S
phiCbK	JX163858	205,504	C. crescentus	S
DSS3P8	KT870145	146,135	Ruegeria pomeroyi	S
phiM7	KR052480	188,427	Sinorhizobium meliloti	M
phiM19	KR052481	188,047	S. meliloti	M
phiM12	KF381361	194,701	S. meliloti	M
phiN3	KR052482	206,713	S. meliloti	M
RSL1	AB366653	231,255	Ralstonia solanacearum	M
RSP15	LC121084	167,619	R. solanacearum	M
Aeh1	AY266303	233,234	Aeromonas hydrophila	M
Aes508	JN377894	160,646	A. salmonicida	M
phiAS4	HM452125	163,875	A. salmonicida	M
phiAS5	HM452126	225,268	A. salmonicida	M
PX29	GU396103	222,006	A. salmonicida	M
25	DQ529280	161,475	A. salmonicida	M
31	AY962392	172,963	A. salmonicida	M
44RR2.8t	AY375531	173,591	A. salmonicida	M
Aes012	JN377895	161,978	Aeromonas sp.	M
IME-CF2	KR869820	177,688	Citrobacter freundii	M
Margaery	KT381880	178,182	C. freundii	M
Miller	KM236237	178,171	C. freundii	M
vB_CfrM_CfP1	KX245890	180,219	Citrobacter sp.	M
S13	KC954775	182,145	Cronobacter sakazakii	M
vB_CsaM_GAP161	JN882287	178,193	C. sakazakii	M
vB_CsaM_GAP32	JN882285	358,663	C. sakazakii	M
phiEap-3	KT321315	175,814	Enterobacter aerogenes	M
Lw1	KC801932	176,227	Escherichia coli	M
RB16	HM134276	176,788	E. coli	M
RB43	AY967407	180,500	E. coli	M
K64-1	AB897757	346,602	Klebsiella pneumoniae	M
KP15	GU295964	174,436	K. pneumoniae	M
KP27	HQ918180	174,413	K. pneumoniae	M
Matisse	KT001918	176,081	K. pneumoniae	M
Miro	KT001919	176,055	K. pneumoniae	M
vB_KleM-RaK2	JQ513383	345,809	K. pneumoniae	M
IME13	JX306041	162,327	Stenotrophomonas maltophilia	M
vB_SmaS-DLP_6	KU682439	168,489	S. maltophilia	M
IME-SM1	KR560069	159,514	Stenotrophomonas sp.	M
KVP40	AY283928	244,834	Vibrio parahaemolyticus	M
phi-pp2	JN849462	246,421	V. parahaemolyticus	M
phi-Grn1	KT919972	248,605	V. alginolyticus	M
phi-ST2	KT919973	250,485	V. alginolyticus	M
ValKK3	KP671755	248,088	V. alginolyticus	M
nt-1	HQ317393	247,511	V. natriegens	M
VH7D	KC131129	246,964	V. harveyi	M
vB_VmeM-32	KU160494	199,912	V. metschnikovii	M

^aGenome size, host, and morphotype (M, Myoviridae; S, Siphoviridae) are from GenBank accession files. NadV and NatV protein sequences were identified in GenBank using BLASTP (Blosum45; word size, 6; expect threshold, 10) (45).

^bThe Caulobacter, Ruegeria, and Sinorhizobium species belong to the class Alphaproteobacteria, Ralstonia solanocearum belongs to the Betaproteobacteria, and the remaining species in the list belong to the Gammaproteobacteria.

FIG 2.

Phylogenetic trees of phage NadV and NatV proteins. Protein sequences were obtained from GenBank (Table 1) (45). NadV and NatV from 50 phages that encoded both proteins were aligned using Clustal Omega (46). SplitsTree4 was used to generate the unrooted phylogenetic trees using neighbor joining, 0.85 maximum sequence difference, and the Kimura protein distance method (47). NadV (A) and NatV (B) proteins in the trees are labeled by their respective phage names (some names are abbreviated for clarity), and clusters are labeled and color-shaded with closely related hosts. Phage IME13 (*) NadV and NatV cluster with the Aeromonas phage genes, although the phage is described as infecting Stenotrophomonas maltophilia (48). A similar observation was made with “Vibrio” phage 32; most all of its identified proteins align with the Aeromonas phage cluster represented by Aeh1. The scale bars indicate the number of substitutions per position.

KVP40 nadV complements bacterial pyridine nucleotide auxotrophy.

nadV and natV were cloned from KVP40 genomic DNA using proofreading (Pfu) DNA polymerase in PCR to obtain the respective genes, which were then inserted into pET101/D-TOPO (nadV) or pSMART HK80 (natV). Both configurations create C-terminal six-histidine (His6) tags on the proteins and provide inducible high-level expression (see Materials and Methods).

Salmonella Typhimurium LT2 TT13007 (nadB499::MudJ pncA278::Tn10cam) is a characterized mutant strain that is unable to synthesize NAD⁺ through either the de novo pathway or the recycling pathways due to mutations in nadB and pncA (15). No growth occurs for TT13007 in M9 minimal medium without the addition of an intermediate compound of the NAD⁺ pathway, such as quinolinic acid (QA). pZL405nadV⁺ was predicted to provide a nicotinamide phosphoribosyltransferase activity that can catalyze conversion of nicotinamide (NAm) to nicotinamide mononucleotide (NMN), an intermediate in the NAD salvage pathway that can restore NAD⁺ synthesis in strain TT13007. Strain TT13007/pZL405nadV⁺ grew on M9 minimal agar medium across the dilution series, while strain TT13007 failed to grow (Fig. 3A). In LB broth, both TT13007/pZL405nadV⁺ and TT13007 grew effectively, reaching an optical density at 600 nm (OD₆₀₀) of ∼1.3 to 1.4 (data not shown). In supplement-free M9 minimal medium, both OD₆₀₀ values were below 0.420 (Fig. 3B). TT13007/pZL405nadV⁺ was able to scavenge small amounts of NAm for effective growth in comparison to TT13007 lacking the plasmid. At an NAm concentration as low as 0.01 μM, the generation times and maximum cell densities were notably different: for TT13007, the generation time (g) was 391 min and the OD₆₀₀ maximum was 0.109; for TT13007/pZL405nadV⁺, the generation time was 125 min and the OD₆₀₀ maximum was 0.532. KVP40 nadV clearly complemented the pyridine nucleotide auxotrophy of the nadB pncA mutant. Induction of the strong expression promoter on pZL405nadV⁺ was not needed, suggesting transcription read-through into nadV on the pET vector.

FIG 3.

Complementation of a Salmonella Typhimurium NAD⁺ auxotroph with nadV. pZL405nadV⁺ transformed into S. enterica Typhimurium strain TT13007 (nadB499::MudJ pncA278::Tn10cam) complements the nutrient requirement on solid (A) or in liquid (B) medium. (A) Cells were serially diluted 10-fold and spotted across a plate with M9 medium plus 10 μM NAm. (B) Cells were grown at 37°C in M9 medium without or with NAm as indicated.

NAmPRTase activity of NadV-His6.

The identity and sequence of NadV-His6 and NatV-His6 enzymes purified with Ni-Sepharose fast performance liquid chromatography (FPLC) (Fig. 4) were confirmed by liquid chromatography-mass spectrometry (LC-MS) electrospray ionization of tryptic fragments (Table S3). Reaction constituents of NadV-His6 NAmPRTase assays (7) were analyzed using reverse-phase C₁₈ high-performance liquid chromatography (HPLC); ATP and ADP were included because this family of enzymes has been shown to have facultative ATPase activity (16) (Fig. 5). In an evaluation of pH effects, comparable rates at pH 6.8 and 7.4, with reduced activity at pH 5.8 and 8.9, were observed (data not shown). NMN synthesis from NAm was detected at 254 nm, with the highest specific activity measured of 2.89 μmol NMN min⁻¹ μg⁻¹ NadV-His6. With ATP added, production of ADP was also detected, although the configuration of the assay did not allow detecting whether the reaction rate is higher in the presence of ATP. The KVP40 NadV NAmPRTase is therefore active, possessing properties similar to those of other nicotinate/nicotinamide phosphoribosyltransferases, including the facultative ATPase activity (16).

FIG 4.

Purified KVP40 NadV-His6 and NatV-His6 enzymes. Proteins expressed in E. coli were purified on Ni-Sepharose, dialyzed, and analyzed on SDS-PAGE as described in Materials and Methods. (A) NadV-His6 (55.5 kDa). Lanes: 1, protein markers; 2, induced BL21-AI/pZL405nadV⁺ cells; 3, soluble induced cell extract; 4, Ni-Sepharose wash fraction; 5 and 6, elution fractions used in assays; 7, concentrated sample. (B) NatV-His6 (40 kDa). Lanes: 1, protein markers; 2, induced BL21-AI/pZL166natV⁺ cells; 3, soluble induced cell extract; 4, Ni-Sepharose wash fractions; 5, pooled elution fractions used in assays.

FIG 5.

HPLC assay of purified NadV-His6 NAmPRTase activity. Reactions were initiated by the addition of enzyme and mixtures resolved by HPLC as detailed in Materials and Methods. Panels show separation of reference substrate and product solutions (A and B), substrates in reaction buffer without enzyme (C), reaction with NAm, no ATP, and added enzyme (D), and reaction with NAm, ATP, and enzyme (E). Phosphoribosylpyrophosphate (PRPP), which is present in panels C and D, is not resolved under these conditions.

NMNATase activity of NatV-His6.

Two assays were used to confirm the activity of the NMNATase domain of purified KVP40 NatV-His6, namely, direct LC-MS quantification of reaction products and a coupled assay measuring NAD⁺ yield following its conversion to fluorescent NADH by alcohol dehydrogenase (ADH). By mass spectrometry, 12.5 pmol of NatV-His6 converted NMN to NAD⁺ at a maximum average specific activity of 350 μmol NAD⁺ s⁻¹ μg⁻¹ NatV-His6. In an assay of NMNATase coupled to ADH, the ADH indicator reaction was optimized at 52 μmol NADH s⁻¹ U⁻¹ ADH at 25°C. With 50 pmol NatV-His6 in the reaction (NMN+ATP≫NAD⁺≫NADH), a specific activity of 50 μmol s⁻¹ μg⁻¹ NatV-His6 (Fig. S1) was observed. Although the coupled assay showed lower specific activity (and may be rate limited by the ADH reaction), it provides a convenient activity assay of this domain and confirms that KVP40 NatV is an active NMNATase.

Nicotinamide to NAD(H) in a multicomponent reaction.

The two reactions of NadV and NatV were combined with ADH to measure the appearance of NADH as an assay for the reconstituted scavenging pathway. This three-part reaction mixture contained NAm, other reaction components but without NMN, and 50 pmol of NatV-His6, and then the reaction was started by adding various amounts of NadV-His6. The yield of NADH was dependent on the amount of added NadV-His6 enzyme (Fig. 6). Under these conditions, the maximum specific activity observed was 5.1 μmol NADH min⁻¹ pmol⁻¹ NadV-His6.

FIG 6.

NADH production in a three-component NadV-NatV-ADH coupled assay. Reaction mixtures contained various amounts of NadV-His6, 2.5 mM PRPP, 50 pmol NatV-His6, 2.5 mM ATP, 5 U ADH, and 1.5% ethanol, at 25°C. NADH produced was dependent on the amount of NadV-His6 added. The maximum rate measured was 2.7 μmol NADH s⁻¹ μg⁻¹ NadV-His6. Data are from two assays.

Nudix pyrophosphatase activity of NatV-His6.

The C-terminal half of NatV is predicted to be a Nudix hydrolase domain similar to the cyanobacterial Synechocystis sp. PCC 6803 slr0787 Nudix domain, a protein that is also bifunctional, possessing both NMNATase and Nudix hydrolase activity (17) (accession no. BAA10693). A monofunctional Nudix hydroloase, NudE.1, from phage T4 has been characterized (18), but no activity of the 50 bifunctional phage enzymes related to KVP40 NatV (Fig. 3) has been reported. Two assays were used to evaluate the substrate specificity of the NatV Nudix hydrolase: phosphate release and mass spectrometry.

Figure 7 summarizes the substrate specificity and relative activity of 50 pmol of NatV-His6 as measured by P_i released from a 1.5 mM concentration of the respective substrate. At 37°C, the preferred substrates are ADP-ribose, NAD⁺, and NADH (0.5 to 0.6 μmol AMP s⁻¹ μg⁻¹ NatV-His6), with little difference in activity between them. The stoichiometry of P_i released from all of these Nudix hydrolase reactions is 1:1, most yielding an AMP. In reactions where AMP could be directly measured using mass spectrometry, ADP-ribose was also the preferred substrate, again at 0.6 μmol AMP s⁻¹ μg⁻¹ NatV-His6, with NAD⁺, NADH, and flavin adenine dinucleotide (FAD) showing reactivity but at a lower rate of 0.15 to 0.2 μmol AMP s⁻¹ μg⁻¹ NatV-His6. Only AMP-yielding Nudix hydrolase substrates were used in the mass spectrometry assay.

FIG 7.

Substrate specificity of the KVP40 NatV Nudix pyrophosphatase. (A) Activity was measured in a phosphate release assay using 1.5 mM substrate and 50 pmol NatV-His6 present in reaction mixtures incubated at 37°C. Values shown are averages from two assays. Stoichiometry of phosphate per substrate is 1:1. (B) Mass spectrometry analysis of KVP40 NatV Nudix hydrolase substrate specificity. Specific activity was obtained with 1.5 mM substrate and 1 μg NatV, incubated at 37°C. Values are averages from duplicate assays (triplicate for ADP-ribose). The specific activity on ADP-ribose was 0.6 μmol AMP s⁻¹ μg⁻¹ NatV-His6.

Transcription of nadV and natV during phage infection.

Expression of nadV and natV during KVP40 infection of V. parahaemolyticus was measured using reverse transcription-quantitative PCR (qRT-PCR) of phage transcripts. A single-cycle growth curve, initiated with a multiplicity of infection (MOI) of 7, showed completion of phage development and cell lysis by 50 to 60 min at 32°C (Fig. S2). mRNA from infected cells was collected throughout this infection period and then assayed using primers for phage T4-like early (td), delayed-early (asiA, frd, and gene 55), and late (wac, gene 23, and gene 25) genes as reference transcripts, along with nadV and natV primer pairs. The overall transcription pattern of the KVP40 genome is not experimentally known, and it is especially likely to differ from the T4 middle period (10, 19, 20). Nonetheless, these genes provide a temporal context in which to evaluate expression of the pyridine nucleotide scavenging genes.

No phage transcripts were detected before infection or in the 1-min postinfection sample. Plots of mRNA copy number measurements throughout the infection are available in Fig. S3. Table 2 summarizes the time of appearance and maximum levels of transcripts, with nadV, natV, and td appearing the earliest (3-min sample), followed by asiA, frd, and gene 55 (5 min), and then wac, gene 23, and gene 25 (10 otor 15 min). The most abundant mRNA measured was for gene 23, which encodes the major capsid protein that is present in hundreds of copies in each mature phage. The relative expression levels and time of appearance for all nine gene mRNAs (Fig. 8) reveal that nadV and natV are expressed early in infection, at times when other genes for DNA precursor metabolism (td and frd) are expressed. Their expression precedes the appearance of mRNA for the transcriptional regulator (AsiA) and sigma factor (gene 55), which are typically involved in directing transcription to phage structural genes (gene 23, gene 25, and wac).

TABLE 2.

Summary of gene expression data from KVP40-infected V. parahaemolyticus^a

Transcription period	Gene	Time of appearance (min)	Peak (min)	No. of transcripts
Early	nadV	3–5	20	1.9 × 105
	natV	3–5	10	6.0 × 105
	td	3–5	8	2.7 × 105
Delayed early	asiA	5–8	15	1.3 × 105
	frd	5–8	30	3.7 × 105
	55	5–8	20	1.8 × 105
Late	wac	10–15	30	1.2 × 105
	23	10–15	30	1.4 × 106
	25	10–15	20	2.7 × 104

^aqRT-PCR data presented are derived from Fig. S8.

FIG 8.

Relative mRNA levels from KVP40 genes by qRT-PCR. Genes were grouped according to their expression periods as early, delayed early, or late. Relative expression is given as the percent maximum transcript level for each gene. (A) Expression patterns of KVP40 early genes nadV, natV, and td; (B) delayed-early genes asiA, frd, and 55; (C) late genes 25, 23, and wac. Values are averages obtained from triplicate assays.

NMNATase activity in phage-infected cells.

A final confirmation of pyridine nucleotide scavenging directed by the phage genome was obtained by measuring NMNATase activity in KVP40-infected cells using the NatV-ADH coupled reaction. Samples from infected V. parahaemolyticus cells were collected over the infection cycle and sonicated, and NatV activity was measured (see Materials and Methods). Figure 9 shows that NMNATase appears and then increases over 5 to 20 min during infection and is detectable through 40 min, when the infection cycle is well into the late stages. A lysate of E. coli phage RB43, a dsDNA phage that also harbors both nadV and natV (Table 1), was also used to infect an E. coli B strain at an MOI of 7; levels of NatV activity were also detected (albeit 25% of the level directed by KVP40) during infection that were maximum at 15 min (data not shown). These data show that active NatV expressed during phage infection increases NMNATase enzyme amounts, with the potential to boost NAD⁺/NADH levels during the metabolic period of the phage developmental cycle.

FIG 9.

NAD(H) synthesis in KVP40-infected cells. NMNATase activity yielding NAD⁺ in KVP40-infected cells was measured using the coupled NatV-ADH assay (see Materials and Methods) that detects NADH fluorescence. Data are average values from duplicate assays per infection sample (sets 1 and 2). 0, uninfected V. parahaemolyticus EB101; 5 to 40, minutes postinfection of cells at 32°C. The enzyme assay was performed at 25°C.

DISCUSSION

During infection, viruses directly capture host metabolites, supplement metabolic pathways, or redirect metabolic intermediates into new products that often distinguish viral macromolecules from those of its host. The paradigm bacteriophage T4 system illustrates this well, with the numerous metabolic enzymes it encodes, including those that lead to glucosyl-hydroxymethylcytosine in its DNA (21). More recently, the photosynthetic apparatus in cyanobacteria (e.g., Prochlorococcus and Synechococcus) has been shown to be augmented by the cyanomyovirus genes psbA and psbD, which encode the D1 and D2 proteins of photosystem II (reviewed in reference 22). These and many other examples emphasize the metabolic resource allocations that viral genome products control and manipulate for developmental success.

In this study, we addressed whether previously identified phage KVP40 genes nadV and natV encode active enzymes that are capable of scavenging nicotinamide into the pyridine nucleotide pool during infection of Vibrio parahaemolyticus. We asked the following questions. (i) By use of the annotated KVP40 genome (accession no. AY283928), could nadV and natV be cloned and expressed to show demonstrable enzyme activity, providing possible implications for the properties of related enzymes from other phages? And (ii) can the expression pattern during phage infection support the relevance of pyridine nucleotide scavenging enzymes to phage development and suggest possible roles?

Both genes were successfully cloned and expressed as active enzymes by using the annotated KVP40 GenBank file, from which few genes have received direct study (23–25). Although >60% of KVP40 genes are of unknown function, NadV and NatV proteins aligned sufficiently well with bacterial NAmPRTases and NMNATases, respectively, so that their function could be predicted. HPLC, direct mass spectrometry of reaction products, and a coupled fluorimetric assay of NAD(H) show that the purified enzymes are indeed active. Together, the two enzymes provide a scavenging pathway from nicotinamide to NAD⁺. Not surprisingly, similar two-reaction pathways of pyridine salvage have been described in bacteria (2), while this appears to be the first functional confirmation of the pathway encoded by a phage or viral genome.

Each of these two KVP40 pyridine nucleotide enzymes has interesting properties that warrant further enzymatic characterization. Like other nicotinate/nicotinamide phosphoribosyl transferases, the KVP40 NadV enzyme has a conserved histidine shown in other systems to participate in a high-energy phosphohistidine intermediate during ATP hydrolysis. Although not required for the synthesis of NMN from NAm and 5-phospho-d-ribosyl-1-pyrophosphate (PRPP), energy from the ATPase activity improves catalysis and lowers the K_m for NAm. Previously demonstrated only for human and bacterial (16, 26) enzymes, phage NAmPRTases such as that from KVP40 can provide phylogenetically distant and tractable enzymes for further study of energy coupling. In our preliminary analysis, KVP40 NadV does show ATPase activity, and interestingly, the pZL405nadV⁺ clone complements the Salmonella Typhimurium pyridine nucleotide auxotroph even in the absence of promoter induction and noticeable enzyme synthesis (Fig. 3 and 5). Possibly due to “leaky” expression without induction from the clone, even low enzyme levels combined with energy coupling appear to provide high affinity for nicotinamide. The complementation data also confirm that the vibriophage enzyme functions in an enterobacterium.

The two-domain NatV enzyme, having both Nudix hydrolase (pyrophosphatase) and NMNATase activities, is not as common from phage genomes as is NadV, although single-domain Nudix hydrolase phage genes are often observed. What appears to be the first characterized phage Nudix hydrolase, the nudE.1-encoded enzyme of phage T4 (18), is a single-domain enzyme half the length of NatV with preferred substrate specificity for FAD, AP₃A (adenosine 5′-triphospho 5′-adenosine, or diadenosine triphosphate), and ADP-ribose. This enzyme resembles NudE of the E. coli host, and similar genes are identified using default BLASTP parameters in several dsDNA bacteriophage and bacterial genomes. NatV, the bifunctional Nudix enzyme of KVP40, showed a preference for ADP-ribose, NAD⁺, NADH, and to a lesser extent FAD. Therefore, the Nudix hydrolase activity of NatV differs from that of the single-domain NudE-type enzymes in other phages and more closely resembles the bifunctional cyanobacterium Synechocystis sp. strain PCC 6803 enzyme Slr0787, which also shows a dominant ADP-ribose pyrophosphatase activity (12, 17). Many of the phages encoding the bifunctional NatV enzymes infect hosts occurring in aquatic environments (e.g., Aeromonas, Caulobacter, and Vibrio), which may have influenced the origin and evolutionary history of this form of the phage gene.

The demonstrated NMNATase activity of KVP40 NatV confirms that with NadV, the two enzymes catalyze the formation of NAD⁺ from nicotinamide, whether measured independently or in the three-enzyme coupled assay. Interestingly, the two activities of NatV—the biosynthetic NMNATase and the Nudix “metabolic scrubbing” pyrophosphatase acting on ADP-ribose—involve NAD⁺. Covalent enzyme and protein modification via ADP-ribosylation utilizes NAD⁺ as the substrate, which is well documented in the T4-related phage transcription cycle (reviewed in references 19 and 20). However, the characterized ADP-ribosylation enzymes of T4 (Alt, Mod, etc.) controlling T4 transcription are not obviously encoded by the KVP40 genome; this source of ADP-ribose during phage infection may not be likely. Further, if NAD⁺ were hydrolyzed by the Nudix domain, then together with the NMNATase activity the bifunctional NatV would catalyze a futile cycle (27) and possibly modulate the optimum level of NAD⁺.

Transcription of nadV and natV coincides with the expression periods of well-characterized T4 phage early and delayed-early genes that are involved in DNA precursor synthesis (td and frd) and in redirecting RNA polymerase toward transcription of predominately phage genes (asiA and gene 55) (19, 20). High-level transcription of the abundant major capsid protein from gene 23 occurs after mRNA for the phage late sigma factor (gene 55) first appears, as expected from the T4 model and the presence of conserved T4-like late promoters upstream of these genes that are predicted to be recognized by the KVP20 gp55 late sigma factor (10, 19). This overall pattern of transcription (Table 2) shows that nadV and natV mRNAs are present early in the infection cycle during the metabolic phase when the pyridine nucleotides NAD⁺ and NADH would be most relevant for phage development. Compared to the scavenging enzymes in uninfected cells, the in vivo enzyme assays show that the scavenging enzymes are elevated during KVP40 infection, appearing shortly after the respective mRNAs are detected and rising into the late period.

T4-type large dsDNA genome phages studied to date invariably encode ribonucleoside diphosphate reductase, thioredoxin/glutaredoxin, and enzymes of the thymidine pathway (thymidylate synthase, dihydrofolate reductase, and thymidine kinase). These reductive and reductive methylation reactions are used extensively during phage metabolism for deoxyribonucleotide synthesis, providing precursors for phage DNA replication rates that can be four times greater or more than the rate of host DNA replication (28). NADH and NADPH are typical redox cofactors that contribute to electron transfers directly or that lead to these reactions. Similarly, although KVP40 encodes a T4-type, ATP-dependent DNA ligase (gene 30), host DNA ligases (lig) use NAD⁺ as the adenylate donor, and although probably not essential, the host NAD⁺-dependent DNA ligase may contribute to effective phage DNA replication. Overall, the phage pyridine nucleotide scavenging enzymes contribute yet another metabolic channel (electron carriers or adenylate donor) to enhance DNA synthesis of lytic phage genomes.

Although the Alt, ModA, and ModB ADP-ribosylation enzymes encoded by phage T4 are not recognized from the KVP40 genome, similar reactions, which use NAD⁺ as the ADP-ribose donor, may be carried out in the infected cell to modify protein activity. That KVP40 encodes a predicted CobB/Sir2 type NAD⁺-dependent deacetylase suggests yet another role for NAD⁺ in phage developmental cycles, where acetylation/deacetylation can modulate enzyme activity and gene expression during viral infection (29, 30).

Efficient DNA sequencing methods are yielding numerous complete phage genomes that allow robust approaches to comparative genomics and mechanistic studies of specific phage proteins and pathways (31, 32). Phage T4 provided a beautiful model for laying the foundations of molecular biology and for unraveling the biochemical processes directed by T4-type phages during lytic infections (reviewed in reference 33). However, other myoviruses that structurally resemble T4 and infect disparate hosts often have only ∼30 “core genes” that represent at most only about 12% of the genome; for Myoviridae with >300-kbp genomes, this portion is even smaller. Most of the predicted genes in these phage genomes encode proteins of unknown function or are unique. So, it is not surprising that entirely new biochemical processes not previously seen in the T-even phages are being observed. Thus, we encounter the aforementioned psbA and psbD (and hli) photosynthesis-related genes in the cyanomyovirus genomes that appear to be important for phage in the marine environmental niche/host (22, 34) and the novel massive whole-genome antisense RNA suppression of phage mRNA degradation by host RNase E by these same types of phages (35). Interestingly, whenever natV is observed in a phage genome, there is always a nadV in that genome as well (Table 1); the converse is not the case. When present, NadV and NatV proteins show the same phylogenetic relationships (Fig. 2) that generally align with their respective host, suggesting that these two genes in any phage group share an isolated (if not long) history together. Although there are three other KVP40 genes of pyridine nucleotide metabolism (pnuC, nadR, and cobB/sir2) yet to be studied, the insights presented in this work show that bacteriophages with large dsDNA genomes encode active pathways that contribute to oxidation-reduction reactions, covalent protein modification (ADP-ribosylation), or both.

MATERIALS AND METHODS

Bacteria, phage, and growth conditions.

The bacteria used were Vibrio parahaemolyticus strain EB101 (9), E. coli strains TOP10, E. cloni (Lucigen, Inc.), and BL21-AI (Invitrogen, CA), and Salmonella enterica serovar Typhimurium LT2 strain TT13007 (nadB499::MudJ pncA278::Tn10cam) (J. Roth, University of California, Davis, CA). Vibrio phage KVP40 (9) and its genome sequence (10) have been described previously. KVP40 was propagated by the confluent plate lysate method using YP medium (9). Phage were added to early-log-phase EB101 cells at an MOI of 0.01 in 3 ml of 0.4% YP top agar, applied to 1% YP agar plates, and incubated overnight at 30°C. Lysates (∼2 to 3 ml) of 10¹⁰ PFU ml⁻¹ were typically obtained from each plate.

Standard phage and phage DNA isolation procedures were used (36). pET101/D-TOPO (Invitrogen, CA) was used for cloning KVP40 nadV, and pSMART HK80 (Lucigen, Inc.) was used for cloning natV. Culture media and antibiotics were from Fisher. Other reagents were ammonia carbonate, HPLC-grade methanol, potassium phosphate, and magnesium chloride (from Fisher), and ADP, ATP, nicotinamide, NAD⁺, NMN, and PRPP (from Sigma).

Cloning of nadV and natV and complementation analysis.

nadV was amplified from phage KVP40 DNA using nadVcacc 5′ primer 5′-CAC CAT GCT AAA TCT TAA TCA AAA-3′ and nadVHis×6 3′ primer 5′-TCA GTG ATG GTG ATG GTG GTG CGC AGT TTG AAT CTT TTT CG -3′. PCR mixtures were 50 μl containing 0.8 mM deoxynucleoside triphosphate (dNTPs), 0.4 μM 5′ and 3′ primers, 70 ng KVP40 genomic DNA, Pfu polymerase reaction buffer (Stratagene, CA), and 2 μl Pfu DNA polymerase (kindly provided by S. Libby). PCR conditions were 2 min at 94°C, followed by 25 cycles for 1.5 min at 94°C, 1.5 min at 55°C, and 1.5 min at 72°C, and a final 10-min extension at 72°C. The 1.5-kbp blunt-ended DNA was used for cloning into pET101/D-TOPO in accordance with the protocol provided by the plasmid manufacturer (Invitrogen, Carlsbad, CA). Chemically competent TOP10 cells were transformed using the heat shock method and used to propagate cells in LB medium with ampicillin at 100 μg ml⁻¹. nadV expression plasmid pZL405nadV⁺ was obtained and sequenced (using T7 forward and T7 reverse primers), confirming the appropriate orientation, C-terminal His6 tag, and correct nadV sequence. Expression is under T7 promoter/lacO-IPTG control when the plasmid is transformed into E. coli strain BL21-AI.

Primers used to amplify natV for cloning into pSMART HK80 (Lucigen, Inc.) were natVrbs 5′ primer 5′-AGG AGA TTA ATA CAT ATG TCA CAC GCA ATC TTT ATC-3′ and natVHis×6 3′ primer 5′-TCA GTG ATG GTG ATG GTG GTG GAT GCC AGT GAA ATA CTG AA-3′. Underlined sequences indicated the ribosome binding site added to the 5′ primer. To obtain the natV expression plasmid pZL166natV⁺ in the pSMART vector, natV primers were first phosphorylated using T4 polynucleotide kinase (T4 PNK; New England BioLabs, MA) in a 40-μl reaction mixture with 1 μM natVrbs 5′ primer, 1 μM natVHis×6 3′ primer, 1× T4 PNK buffer, 0.02 mM ATP, and 3 μl T4 PNK, followed by a 30-min incubation on ice. Then, 20 μl of the T4 PNK reaction mixture was transferred to a 100-μl PCR mixture under conditions as described above for nadV. A 1.3-μg PCR product was then mixed with 1× CloneSMART vector premix and 0.2 U of CloneSmart T4 DNA ligase in a total volume of 10 μl. The reaction mixture was incubated at room temperature for 30 min, and then 1 μl of the ligation reaction mixture was transformed into 20 μl of electrocompetent E. cloni cells or 3 μl of the ligation reaction mixture was transformed into 50 μl of chemically competent TOP10 cells. For electroporation of electrocompetent E. cloni cells, 0.2-cm Bio-Rad cuvettes and a GenePulser electroporation system (Bio-Rad, CA) set at 25 μF, 200-Ω resistance, and 2.5 kV were used. The electroporated E. cloni cells (Lucigen, Inc.) were incubated in 980 μl of terrific broth (ThermoFisher) at 37°C, 250 rpm, for 1 h. Cells were then plated on LB agar with kanamycin (50 μg ml⁻¹). For transformation of chemically competent TOP10 cells, cloning mixtures were incubated with TOP10 cells for 30 min on ice and then heat shocked at 42°C for 30 s, followed by incubation at 37°C and 250 rpm in 900 μl SOC broth (ThermoFisher) for 30 min, and then plated on LB agar with ampicillin (100 μg ml⁻¹).

For complementation analysis, Salmonella strain TT13007 (nadB499::MudJ pncA278::Tn10cam) was grown in supplemented LB medium (10 mM quinolinic acid [QA]; 50 μg ml⁻¹ chloramphenicol) and transformed with pZL405nadV⁺. The confirmed transformants were grown in LB medium plus QA, ampicillin, and chloramphenicol or in M9 medium supplemented with nicotinamide (NAm; 10⁻⁷ to 10⁻³ M) and antibiotics. For liquid complementation assays, overnight cell cultures were prepared in LB medium plus QA and antibiotics as needed and in M9 medium plus 3 mM NAm or QA (for strain TT13007). Cell pellets from overnight cultures (37°C; OD₆₀₀ = 1.3 to 2.6) were washed 3 times with 1 ml LB or M9 medium and then suspended in LB or M9 medium, respectively, for inoculation in growth medium. Growth curves were obtained using a Bioscreen C (Growth Curves, NJ) to measure growth (OD₆₀₀) over 36 h. Cultures were started at an OD₆₀₀ of ∼0.100 to 0.180 and incubated at 37°C with vigorous shaking, and the OD₆₀₀ was measured every 10 min. Samples and controls were replicated in nine Bioscreen wells (200 μl) in two different experiments. Growth on solid medium was done on M9 medium agar plates spread with 50 μl of 10 μM NAm. Serial dilution transfers were performed across the plate.

NadV and NatV overproduction and purification.

For NadV, pZL405nadV⁺ from TOP10 cells was transformed into strain BL21-AI and the transformant pool was grown overnight at 37°C in LB broth with ampicillin (100 μg ml⁻¹) and tetracycline (12.5 μg ml⁻¹). The overnight culture was transferred to fresh medium to an OD₆₀₀ of 0.1. NadV synthesis was induced by adding 1 mM IPTG and 0.02% l-arabinose after the OD₆₀₀ reached ∼0.4 to 0.6, with incubation continued for another 3 h. For NatV, BL21-AI cells harboring pZL166natV⁺ were grown overnight in LB broth containing kanamycin (50 μg ml⁻¹), tetracycline (12.5 μg ml⁻¹), and d-(+)-glucose (0.02% final concentration). Cells were diluted to an OD₆₀₀ of ∼0.1 into 16 ml of the same, fresh medium. When cultures reached an OD₆₀₀ of ∼0.5, 8 ml of each culture was taken as an uninduced control. By adding 1 mM IPTG and 0.2% l-arabinose, the remaining culture was induced, and then it was incubated for another 4 h. Cell pellets were collected by centrifugation at 5,000 rpm for 10 min at 4°C and stored at −20°C until disrupted.

The pellet from 8 ml of induced cell culture was suspended in 1 ml buffer A (50 mM Na₂HPO₄, 300 mM NaCl, pH 8.0) and then disrupted 3 times in a French press at 1,290 lb/in², followed by centrifugation at 12,000 rpm for 30 min. The supernatant (about 1.5 ml) was filtered through a 0.22-μm filter and then loaded onto a 1-ml FPLC Ni²⁺-Sepharose column (HisTrap HP; GE Healthcare, NJ) previously charged with 0.1 M NiCl₂ in buffer A using a Bio-Rad BioLogic DuoFlow chromatography system. Bound protein was washed with 5 ml buffer A (flow rate, 1 ml min⁻¹) and then subjected to gradient elution with buffer A plus 100 mM imidazole and buffer B (50 mM Na₂HPO₄, 300 mM NaCl, 500 mM imidazole, pH 8.0) as the eluent. Fractions were collected, analyzed by 12% SDS-PAGE, and stored in 20% glycerol at −20°C. Purified proteins were dialyzed against 800 ml 1 mM Tris-HCl (pH 7.5), 1 mM MgCl₂, 1 mM EDTA, 1 mM dithiothreitol (DTT), and 10% glycerol, using a Slide-A-Lyzer 10,000-molecular-weight-cutoff dialysis cassette (Pierce, Rockford, IL). Two-hundred-microliter aliquots in 20% glycerol were stored at −80°C. Protein concentration was determined by the Bradford protein assay (Bio-Rad, CA).

NadV enzyme assays.

NadV nicotinamide phosphoribosyltransferase (NAmPRTase) assays were performed in 160 mM phosphate, pH 6.8, 2 mM NAm, 5 mM PRPP, and 16 mM MgCl₂, with or without 1 mM ATP. Different pH conditions were examined with phosphate buffer at pH 5.8, 6.8, 7.4, and 8.9. Purified NadV-His6 was used at 2.45 nM. Reaction mixtures were incubated at 37°C. At designated time points, samples were removed and quickly frozen on dry ice-ethanol to stop the reactions. Samples were stored at −20°C and thawed immediately before HPLC analysis. NMN and NAm standards of 0.1 mM to 1.0 mM were separated and quantified on an HPLC system (Waters Millipore) (Turbochrom Navigator software version 4.1; PerkinElmer Corp.), with a Millipore Nova-PAK C₁₈ column of 4-μm particle size (3.9 by 75 mm). The mobile phase was composed of eluent A (0.1 M phosphate buffer, pH 6.0, 6 mM tetrabutylammonium bromide) and eluent B (100% methanol). The gradient profile was as follows: 7% B for 5 min, increase to 30% B in 1 min, and then back to 7% B after 4 min. Initial conditions were restored in 8 min. The column was washed with eluent B for 3 min after each run, followed by a 3-min wash with eluent A. The amount of NMN produced in each reaction was calculated from the standard curve of A₂₅₄ traces.

MS analysis was performed on NadV reaction constituents carried out in a mixture of 80 mM ammonium carbonate, pH 7.5, 8 mM MgCl₂, 5 mM PRPP, 2 mM NAm, and 1 mM ATP. Reaction mixtures were incubated at 37°C with shaking at 250 rpm for various times and analyzed using a Phenomenex Synergi column (4-μm particle size, 2 by 100 mm). Volatile mobile phases used were eluent A (99% 5 mM ammonium carbonate, pH 7.5, 1% methanol) and eluent B (10% 5 mM ammonium carbonate, pH 7.5, 90% methanol). Gradient settings were the same as those using phosphate buffer and the Nova-PAK C₁₈ column. NMN and ATP standard profiles in the carbonate buffer system were established at 0.1 mM to 2 mM. Reactions were analyzed by tandem LC-MS (NCSU Mass Spectrometry Facility) using a Phenomenex Synergi column and ammonium carbonate.

NatV NMNATase assays.

NatV-His6 NMNATase reactions were performed in buffer containing 50 mM Tri-HCl, pH 7.64, 12 mM MgCl₂, and 0.02% bovine serum albumin (BSA) (37, 38) with NMN and rATP at 2.5 mM each in a volume of 50 μl. The reaction was started by adding 0.5 μg (12.5 pmol) of NatV-His6, followed by incubation at 25°C for 0.5 or 1 min. Reactions were terminated by adding an equal volume of chilled 100% methanol (MeOH) and instantly transferring the mixture to a dry ice-ethanol bath.

Mass spectrometry of NatV reaction products used a Thermo LTQ linear ion trap mass spectrometer and electrospray ionization, with a Surveyor autosampler HPLC using Thermo Hypersil and MH⁺ detection or Thermo Hypercarb [M-H]⁻ detection for phosphorylated compounds (performed at the NC State University Proteomics Facility). A standard curve of NAD⁺ (0.001 to 0.5 mM) was generated to determine product yield.

A coupled enzyme reaction using alcohol dehydrogenase (ADH) was used to measure NAD produced by NatV-His6, similar to the method described previously (37, 39). ADH (Worthington, NJ, USA) was prepared in 0.1 M phosphate buffer, pH 7.5, and filtered through a 0.2-μm filter. Using the NatV-His6 reaction conditions described above plus 5 U of ADH and 1.5% ethanol, 50 pmol of purified NatV-His6 was added to start the reaction. Fluorescence of NADH was measured at 340-nm excitation and 460-nm emission. Reactions were read in a Polarstar Galaxy (BMG Labtech, Offenburg, Germany) plate reader with a positioning delay of 0.2 s, 20 flashes per cycle, gain of 80, and cycle time of 20 s in fluorescence intensity mode. The total reaction volume was 200 μl in a 96-well Corning microtiter plate with an empty well between each reaction well. The incubation temperature was 25°C, with no substantial difference in activity between 25 and 37°C. A standard curve of NADH (1 to 100 μM) fluorescence was generated to determine product yield.

Coupled NadV–NatV-ADH assay.

The reconstituted scavenging pathway used the same NatV-His6 reaction buffer described above, with NAm and PRPP added at 2.5 mM, 50 pmol of purified NadV-His6 and NatV-His6, 5 U ADH, and 1.5% ethanol (NMN is omitted). NADH fluorescence was measured as described above.

NatV Nudix hydrolase assays.

For the colorimetric assay (with phosphate released from AMP, a Nudix reaction product, by the addition of alkaline phosphatase), a standard curve of phosphate was generated with Na₂HPO₄ at concentrations of 0.01 to 1 mM. Phosphate solutions (300 μl) of various concentrations were mixed with 700 μl of P_i reagent (1 part 10% ascorbic acid and 6 parts of 0.42% ammonium molybdate·4H₂O in 1 N H₂SO₄), followed by 15 min of incubation at 37°C for color development (40) and subsequent measurement of A₆₀₀. A standard Nudix hydrolase-alkaline phosphatase assay contained 50 mM Tri-HCl, pH 8.0, 5 mM MgCl₂, 1.5 mM substrate, 5 U of bovine alkaline phosphatase (Sigma-Aldrich, MO), and 2 μg (50 pmol) of purified NatV-His6. Substrates (all from Sigma-Aldrich, MO) were ADP-ribose (ADP ribose), rATP, dATP, rGTP, dGTP, NAD⁺, NADH, NADP⁺, NADPH, FAD, and AP₃A (adenosine 5′-triphospho 5′-adenosine, or diadenosine triphosphate), all at 1.5 mM (12, 41). Tubes contained 300 μl of reaction mixture incubated at 37°C for 1 min, quenched with 700 μl of P_i reagent, followed by 15 min of incubation at 37°C for color development. When deoxynucleoside triphosphates were used as the substrates, the reaction was terminated with 30 μl EDTA (10 mM final concentration) and 50 μl of 20% activated charcoal (Sigma-Aldrich) to absorb residual nucleoside triphosphates, followed by 2 min of gentle shaking (42, 43). The supernatant was collected in a new 1.5-ml microcentrifuge tube, and then 700 μl of P_i reagent was added, followed by 15 min of incubation at 37°C for color development. Average values were obtained from duplicate samples.

Mass spectrometry assays of Nudix activity used 50-μl reaction mixtures that contained buffer (50 mM Tris-HCl, pH 8.0, 5 mM MgCl₂), 1.5 mM substrate, and 1.1 μg (27 pmol) of NatV-His6 (0.56 μg μl⁻¹) to start the reaction. Mixtures were incubated at 37°C for 1 min and terminated by the addition of 50 μl of cold 100% methanol (50% final concentration), followed by rapid transfer to a dry ice-ethanol bath. Samples were stored on dry ice at −80°C until analyzed by mass spectrometry. An injection volume of 5 μl was used. A standard curve of AMP (0.5 μM to 0.5 mM concentration in 50% MeOH) was generated.

mRNA measurements from KVP40-infected cells.

V. parahaemolyticus EB101 cultures grown at 32°C to an OD₆₀₀ of ∼0.5 were infected with KVP40 at an MOI of 7 in a final volume of 12 ml YP medium. One milliliter was immediately collected in a 2-ml tube containing 0.4 ml of an RNAlater-like solution (Ambion-ThermoFisher, MA; 10 M [NH₄]₂SO₄, 25 mM Na₃-citrate, 10 mM EDTA, pH 8.0, in diethyl pyrocarbonate-treated H₂O). Samples were collected through a time course of 1, 3, 5, 8, 10, 15, 20, 30, and 40 min postinfection. Samples were incubated with the RNA preservation solution for 5 min on ice, centrifuged at 5,000 × g, and washed with 0.4 ml of Tris-EDTA (TE) buffer. Pellets were immediately used to extract RNA by suspending them in 1 ml TE buffer containing 0.4 mg lysozyme by using an RNeasy minikit (Qiagen, Valencia, CA). RNA was eluted with 30 μl of nuclease-free water. Samples were then treated with 8 U of DNase RQ1 (Promega, Madison, WI) at 37°C for 20 min in a volume of 80 μl, followed by 10 min of enzyme inactivation at 65°C. Elimination of DNA was confirmed by agarose gel analysis and PCR using 200 ng of RNA and phage-specific primers for KVP40 genes 23 and frd. Concentrations of RNA (A₂₆₀) were determined with a NanoDrop ND-1000 UV-Vis spectrophotometer (ThermoFisher Scientific, MA) in triplicate readings of 2-μl diluted samples. Five hundred nanograms of RNA was reverse transcribed using 100 U of Moloney murine leukemia virus (M-MLV) reverse transcriptase (Promega, WI), 0.25 μg of random hexamer (New England BioLabs, MA, USA), and 20 U of RNasin (Promega, WI) in a final volume of 25 μl. The reaction mixture was incubated at 37°C for 1 h, followed by 10 min of enzyme inactivation at 80°C.

Based on the 500 ng of RNA in each reverse transcription reaction mixture, 0.4 ng was transferred from the cDNA product to qRT-PCRs. Nine phage-specific primer sets and V. parahaemolyticus 16S rRNA primers were used. Primer pairs were designed to amplify 100- to 150-bp products using software from Integrated DNA Technologies; the primers are listed in Table S4 in the supplemental material. Using a reaction volume of 10 μl, real-time PCR was performed in an iCycler (Bio-Rad, Hercules, CA) using SYBR green (Absolute Blue; Thermo Scientific, MA). Forward and reverse primers were added at a final concentration of 2.5 μM. The amplification conditions were as follows: 15 min of enzyme activation at 95°C and 40 cycles of 95°C for 30 s, 51°C for 30 s, and 72°C for 1 min, followed by a melt curve of 95°C for 1 min, 55°C for 1 min, and then an increase in the set point temperature after cycle 2 by 0.5°C every 10 s. Gene nadV, 23, frd, and 55 primers were used to amplify KVP40 genomic DNA for average C_T values as calibrators in the calculation of mRNA copy number. Reactions were performed in triplicate with each primer set.

C_T values were acquired at a threshold of 1/3 maximum relative fluorescent units. An average C_T value was obtained from triplicate samples. The copy number of mRNA at each infection time was calculated using the 2ΔΔ^CT method as described previously (44). Outlier data were detected and excluded using the Grubbs' test (GraphPad).

NMNATase assay of infected cells.

V. parahaemolyticus EB101 cells were grown, infected with KVP40 (MOI of 7), and sampled as described above for qRT-PCR. Infected cell pellets from 1-ml infection samples were washed in cold TE buffer and stored at −20°C until assayed. Pellets were suspended in 100 μl of lysis buffer (50 mM Tris-HCl, pH 7.52, 1 mM EDTA, 1 mM DTT) and disrupted by sonication (20 s, three times) on ice. The supernatant was obtained by centrifugation at 10,000 rpm for 5 min at 4°C. NMNATase activity was measured in a reaction buffer containing 50 mM Tri-HCl, pH 7.64, 2.5 mM NMN, 2.5 mM ATP, 12 mM MgCl₂, 1.5% ethanol, 5 U ADH, and 0.02% BSA. Reactions were initiated by adding 20 μl of cell extract, performed at 25°C in a total reaction volume of 200 μl. The fluorescence of NADH was measured as described above. Total protein at each time point was determined by the Bradford assay.