A Novel Cyanophage with a Cyanobacterial Nonbleaching Protein A Gene in the Genome

Abstract

A cyanophage, PaV-LD, has been isolated from harmful filamentous cyanobacterium Planktothrix agardhii in Lake Donghu, a shallow freshwater lake in China. Here, we present the cyanophage's genomic organization and major structural proteins. The genome is a 95,299-bp-long, linear double-stranded DNA and contains 142 potential genes. BLAST searches revealed 29 proteins of known function in cyanophages, cyanobacteria, or bacteria. Thirteen major structural proteins ranging in size from 27 kDa to 172 kDa were identified by SDS-PAGE and mass-spectrometric analysis. The genome lacks major genes that are necessary to the tail structure, and the tailless PaV-LD has been confirmed by an electron microscopy comparison with other tail cyanophages and phages. Phylogenetic analysis of the major capsid proteins also reveals an independent branch of PaV-LD that is quite different from other known tail cyanophages and phages. Moreover, the unique genome carries a nonbleaching protein A (NblA) gene (open reading frame [ORF] 022L), which is present in all phycobilisome-containing organisms and mediates phycobilisome degradation. Western blot detection confirmed that 022L was expressed after PaV-LD infection in the host filamentous cyanobacterium. In addition, its appearance was companied by a significant decline of phycocyanobilin content and a color change of the cyanobacterial cells from blue-green to yellow-green. The biological function of PaV-LD nblA was further confirmed by expression in a model cyanobacterium via an integration platform, by spectroscopic analysis and electron microscopy observation. The data indicate that PaV-LD is an exceptional cyanophage of filamentous cyanobacteria, and this novel cyanophage will also provide us with a new vision of the cyanophage-host interactions.

INTRODUCTION

Cyanophages are one kind of planktonic viruses that infect cyanobacteria (blue-green algae). Cyanophages and phages have amazing amounts of genetic diversity and biological activity in water environments (33, 34, 40, 54). In either a freshwater or saltwater environment, cyanophages are ubiquitous and play an important role in water ecosystems (46, 52, 61, 66). Generally, the complete genome sequences of cyanophages can provide significant clues for better understanding of the biological properties, ecological effects, and coevolutionary relationships between cyanophages and their hosts (10, 17, 18, 21, 27, 29). Some cyanophage genomes have been sequenced (32, 35, 44, 49, 51, 60, 64), which has revealed the presence of cyanobacterial genes involved in central energy metabolism and their host's survival. For examples, some photosynthesis-related genes (psbA, hliP, and PSII) and stress-response genes (coding for chaperones and genes associated with bacterial motility and chemotaxis) have been described in cyanophages (4, 5, 31, 36, 47), most of which are transcribed together with essential cyanophage replication-related genes (6, 13, 30, 65). Moreover, a nonbleaching protein A (NblA) gene has been found from a lytic phage, Ma-LMM01, infecting Microcystis aeruginosa (50, 64), but its function has not been demonstrated.

Despite different hosts and habitats, more than 95% of the known cyanophages belong to the tailed phages (16, 68). The tailed phages are classified into three families: Myoviridae with long contractile tails, Siphoviridae with long noncontractile tails, and Podoviridae with short tails. Various members in each family are morphologically similar and share similar genome organizations, gene contents, and conserved structural protein genes, including capsid protein genes and tail-related genes (21, 23, 26, 37–39, 42, 58).

PaV-LD can infect and lyse the filamentous cyanobacterium Planktothrix agardhii (14). To uncover the biological properties and interaction with the host, we sequenced the PaV-LD genome, identified the major structural proteins, and especially analyzed the functional role of a host-like gene that encodes an NblA homologue, which mediates phycobilisome degradation.

MATERIALS AND METHODS

Cyanobacterial strains and culture conditions.

Host cyanobacterium Planktothrix agardhii was clonally grown in medium BG-11 at 25°C under a 14-h/10-h light/dark cycle of 30 to 40 μE m⁻² s⁻¹. Synechocystis sp. strain PCC 6803 cells were grown in BG-11 medium at 30°C under continuous illumination of 30 μE m⁻² s⁻¹.

DNA preparation and sequencing.

Cyanophage DNA was prepared for sequencing according to methods previously described (62). Briefly, the crude lysates were treated with chloroform and centrifuged to remove cell debris. The cyanophage particles were precipitated using NaCl (5.8% [wt/vol]) and polyethylene glycol 8000 (9.3% [wt/vol]). The precipitates were then purified using a sucrose density gradient. The purified cyanophage particles were broken using sodium dodecyl sulfate-proteinase K. Genomic DNA was then subjected to phenol-chloroform extraction and ammonium acetate-isopropanol precipitation. The PaV-LD genome was sequenced using the Genome Sequencer 20 (GS20) system. Briefly, after the quality of PaV-LD genome DNA had been assessed by agarose gel electrophoresis, 20-μg samples were broken into fragments of approximately 150 bp by nebulization. The whole genomic library was amplified using GS20 emPCR kits and sequenced with the 454 Life Science GS20 instrument according to the manufacturer's recommendations. The consensus sequence of the whole DNA sample was generated by assembly of de novo shotgun sequencing reads into contigs followed by subsequent ordering of these contigs into scaffolds. The average reading frame length was about 100 bp, with 20-fold genome coverage. The primer-walking technique was used to fill the gaps.

Genome annotation.

The open reading frames (ORFs) were predicted using GeneMarkS (3) and ORF Finder. Predicted protein comparisons with entries in the nonredundant GenBank databases were carried out using the programs BLASTp and PSI-BLAST (E value, ≤10⁻³). tRNA genes were identified by using tRNAscan-SE. Multiple-sequence alignments were performed using the Clustalx1.83 program and manual editing, and MEGA 4 was used to construct phylogenetic trees. The softwares TMHMM, SOSUI, and PSORT-B were used to predict the transmembrane proteins, and the SignalP 3.0 server was used to find possible signal peptide cleavage sites.

SDS-PAGE and mass spectrometry.

The purified PaV-LD virions were treated with SDS buffer, boiled for 5 min, and loaded onto a 15% polyacrylamide gel containing SDS. The proteins were electrophoresed at constant voltage. Then the gels were stained with Coomassie brilliant blue. The most intensely stained bands were excised, digested with trypsin, and analyzed by matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS). The resulting peptide mass fingerprints were searched against the translated ORFs of PaV-LD and Swiss-Prot data using the Mascot search program (Matrix Science, Ltd.).

Prokaryotic expression, His₆-tagged NblA purification and antibody preparation.

A fragment containing PaV-LD nblA was amplified using primers PaVF (5′-T GTT GGT ACCCTT ATA ACA GCG GAA-3′) and PaVR (5′-AGT TCT CGA GTT TGG GTT GTT TTC C-3′) (KpnI and XhoI sites underlined). The amplified fragment was restricted by KpnI and XhoI and cloned into prokaryotic vector pET-32a (Novagen) restricted with KpnI and XhoI previously. Then the constructed plasmid pET-NblA was transformed into Escherichia coli BL21(DE3), and the bacteria were induced for 5 h by adding 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside) at 37°C to produce a fusion protein. The fusion protein was purified using a His·Bind purification kit (Novagen) and mixed with an equal volume of Freund's complete adjuvant (Sigma) to immunize mice by intraperitoneal injection once a week. After 6 weeks, the anti-NblA serum was collected and tested by Western blotting with lysates from infected cells.

Western blot analysis.

For Western blot analysis, the extracted proteins from the infected cells at 0, 12, 24, 36, 48, and 60 h postinfection were separated using a Tricine-SDS-PAGE system as described previously (48) and transferred electrophoretically to 0.2-μm-pore polyvinylidene difluoride (PVDF) membrane using a semidry transfer system. The membranes were then incubated with anti-NblA mouse serum as the primary antibody and then with horseradish peroxidase-conjugated goat anti-mouse as the secondary antibody and immunostained according to a previous report (69).

Expression of PaV-LD NblA in Synechocystis sp. strain PCC 6803.

The DNA fragment amplified by the primers PaVF and PaVR was cloned into pMD18-T vector (Takara) to make plasmid pNblA. A fragment containing Omega-P_6803rbcL with spectinomycin resistance was excised from plasmid pHB2759 (22) with PstI and XbaI, and blunted with T4 DNA polymerase (Promega). The plasmid pNblA was cut with BamHI, blunted with T4 DNA polymerase, and ligated with the blunted Omega-P_6803rbcL fragment to make plasmid pOPNblA. The Omega-P_6803rbcL-NblA fragment in pOPNblA was excised with KpnI and PstI, blunted with T4 DNA polymerase and ligated with EcoRI-cut and blunted pKW1188 to make plasmid KpOPNblA. The wild-type Synechocystis sp. strain PCC 6803 was transformed with plasmid KpOPNblA, as previously described (15), and the resulting transformants were streaked on plates and cultured in liquid medium under selective pressure of antibiotics until complete segregation was confirmed by DNA sequence analysis.

Spectroscopic analysis.

Whole-cell absorbance spectra of cyanobacterial cultures from 550 to 750 nm were recorded on a Shimadzu UV-2450 spectrophotometer. Spectra were normalized to the optical density at 750 nm, which was used as an index of cell concentration.

Electron microscopy.

The specimens for electron microscopy were prepared as previously described (25). Cyanobacterial cells were collected and fixed with glutaraldehyde. The fixed cells were harvested, dehydrated, and embedded. Ultrathin sections were observed with a JEOL 1230 electron microscope.

Nucleotide sequence accession number.

The complete genome sequence of PaV-LD has been deposited in GenBank under accession no. HQ683709.

RESULTS

Feature of PaV-LD genome.

The PaV-LD genome is a linear, double-stranded DNA molecule of 95,299 bp without terminal repeats. The average G+C content of the genome is 41.5%. Computer-assisted analysis reveals 142 potential genes or open reading frames (ORFs), which encode proteins ranging in size from 35 to 1,584 amino acids. All available annotation information for the 142 ORFs is provided in a detailed overview in Table S1 in the supplemental material. The genome is densely packed in terms of coding sequences, with an average coding density of 1.49 genes per kilobase pair of nucleotide sequence. Figure 1 shows the detailed genomic organization map of PaV-LD, in which the locations, orientations and sizes of the predicted genes are all indicated (Fig. 1). Table 1 lists the putative 53 genes (ORFs) that can be matched with homologues in other phages or microbes. Among the 53 matched genes, 29 can be assigned by sequence homology to the known function proteins (Table 1; Fig. 1). They include genes coding for some important enzymes, such as replicative DNA helicase (007R), cytosine-5-methyltransferase (010R), thymidylate kinase (021L), acetyltransferase (032L), protein phosphatase (037R), protein kinase (044L), nuclease (098R), flavin-dependent thymidylate synthase (115L), N-6-adenine-methyltransferase (117R), and crossover junction endodeoxyribonuclease (119L), which are involved in the processes of the viral life cycle, including DNA replication, transcription, and modification. Moreover, the PaV-LD genome carries a nonbleaching protein A (NblA) gene (022L) that is a homologue in cyanobacteria. The data indicate that the PaV-LD genome is different from the known cyanophage genomes and suggest that a unique genomic strategy might exist in the cyanophage.

Fig 1.

Genomic organization of PaV-LD. Arrows indicate the size, position, and orientation of 142 ORFs (potential protein-coding genes): 66 are on the positive (right) strand, and 76 are on negative (left) strand. The 29 known function proteins are shown above the arrows.

Table 1.

Fifty-three putative ORFs that can be matched with homologues of other phages or microbes in the PaV-LD genome

ORF	Nucleotide position	No. of amino acids	Mol mass (kDa)	Best alignment with predicted protein (accession no./related phage or microbe/E value)a
1. 005R	1904–3067	387	43.44	Phage related replication protein (YP_001285918.1/Lactobacillus phage LL-H/3e−10)
2. 007R	3659–5398	579	63.73	Replicative DNA helicase (ZP_01618944.1/Lyngbya sp. strain PCC 8106/0.0)
3. 010R	6399–7439	346	38.63	Putative cytosine-specific DNA methyltransferase (YP_004063483.1/Ostreococcus tauri virus 2/4e−30)
4. 012R	7872–8192	106	12.44	Hypothetical protein PCC7424_5627 (YP_002380933.1/Cyanothece sp. strain PCC 7424/2e−17)
5. 018R	10405–11538	377	42.93	Integrase (NP_490644.1/Pseudomonas phage phiCTX/1e−12)
6. 020R	11844–12731	295	33.28	Phosphoadenosine phosphosulfate reductase (CBL14553.1/Ruminococcus bromii L2-63/2e−41)
7. 021L	12800–13387	195	22.44	Thymidylate kinase (YP_002250503.1/Dictyoglomus thermophilum H-6-12/3e−28)
8. 022L	13384–13548	54	6.54	Phycobilisome degradation protein nblA (ZP_03273971.1/Arthrospira maxima CS-328/3e−09)
9. 023R	13731–13940	69	7.97	Hypothetical protein S7335_1188 (ZP_05040220.1/Synechococcus sp. strain PCC 7335/6e−04)
10. 028L	16158–16676	172	19.80	Hypothetical protein (YP_001965479.1/Cylindrospermum sp. strain A1345/3e−10)
11. 030L	17318–17536	72	8.36	Protein of unknown function UPF0150 (YP_002371724.1/Cyanothece sp. strain PCC 8801/6e−30)
12. 031L	17533–17760	75	8.95	Conserved hypothetical protein (ZP_03273963.1/Arthrospira maxima CS-328/3e−26)
13. 032L	17897–18490	197	23.11	Putative acetyltransferase (YP_002380473.1/Cyanothece sp. strain PCC 7424/2e−52)
14. 033R	18485–19240	251	27.28	ABC transporter-like (YP_321753.1/Anabaena variabilis ATCC 29413/1e−95)
15. 034R	19296–19589	97	10.73	Hypothetical protein Ava_4592 (YP_325084.1/Anabaena variabilis ATCC 29413/8e−16)
16. 035R	19586–19957	123	14.17	Hypothetical protein alr4134 (NP_488174.1/Nostoc sp. strain PCC 7120/1e−33)
17. 036R	20071–21201	376	41.76	Protein phosphatase 2C-like (ZP_01290001.1/deltaproteobacterium MLMS-1/9e−45)
18. 037R	21207–22025	272	30.67	Protein kinase (ZP_01290002.1/deltaproteobacterium MLMS-1/6e−73)
19. 044L	24263–25288	341	35.01	Phage-related protein (YP_146396.1/Geobacillus kaustophilus HTA426/7e−17)
20. 046L	25849–27576	575	61.43	Hypothetical protein Npun_R2579 (YP_001866076.1/Nostoc punctiforme PCC 73102/4e−17)
21. 049L	27977–30343	788	88.60	Hypothetical protein Npun_F2576 (YP_001866073.1/Nostoc punctiforme PCC 73102/5e−07)
22. 053L	31517–33121	534	60.57	Putative phage terminase large subunit (YP_784949.1/Bordetella avium 197N/8e−90)
23. 057R	35268–37256	662	74.82	Putative portal protein (ABY40530.1/Burkholderia phage Bups phi1/5e−17)
24. 062R	40907–41488	193	21.93	Hypothetical protein AmaxDRAFT_1344 (ZP_06381368.1/Arthrospira maxima CS-328/2e−05)
25. 064R	41786–42583	265	30.11	KilA to -N domain-containing protein (YP_003986509.1/Acanthamoeba polyphaga mimivirus/2e−08)
26. 066R	45076–45588	180	18.58	Hypothetical protein L8106_07531 (ZP_01622095.1/Lyngbya sp. strain PCC 8106/2e−09)
27. 068R	45858–46394	178	19.59	Hypothetical protein (YP_001965479.1/Cylindrospermum sp. strain A1345/3e−13)
28. 069R	46543–46872	139	16.20	G protein (ZP_04754446.1/Actinobacillus minor NM305/2e−13)
29. 071R	47352–48875	507	57.97	Putative head protein (ADF83466.1/Lactobacillus phage LBR48/1e−16)
30. 073R	49731–50684	317	35.83	Putative capsid protein (ADF83468.1/Lactobacillus phage LBR48/6e−07)
31. 075R	51170–51505	111	12.16	Hypothetical protein Saro_2726 (YP_497996.1/Novosphingobium aromaticivorans DSM 12444/1e−07)
32. 078R	52141–53619	492	55.00	Minor head protein (YP_003084174.1/cyanophage PSS2/0.002)
33. 079R	53749–55524	585	65.19	Hypothetical protein AplaP_06772 (ZP_06381368.1/Arthrospira platensis Paraca/2e−05)
34. 080R	55630–56724	364	41.28	Hypothetical protein Dtox_4230 (YP_003193520.1/Desulfotomaculum acetoxidans DSM 771/6e−13)
35. 081R	56721–57875	379	43.09	Phage head morphogenesis protein (EFQ53877.1/Lactobacillus oris PB013-T2-3/8e−05)
36. 085R	59123–60205	360	39.23	Minor head protein (YP_003084174.1/cyanophage PSS2/5e−06)
37. 088R	61293–66047	1584	171.28	Putative tail tape measure protein (YP_003432551.1/Hydrogenobacter thermophilus TK-6/3e−20)
38. 090L	67046–67480	144	16.13	Hypothetical protein all7031 (NP_490137.1/Nostoc sp. strain PCC 7120/2e−29)
39. 091L	67477–67737	86	10.01	Hypothetical protein M23134_03908 (ZP_01690521.1/Microscilla marina ATCC 23134/1e−19)
40. 095L	69187–70020	277	31.29	DNA-damage-inducible protein D (YP_655040.1/Mycobacterium phage Llij/3e−21)
41. 098R	70954–72074	374	41.85	Nuclease (ZP_06012604.1/Leptotrichia goodfellowii F0264/1e−45)
42. 100R	72400–72960	186	20.91	KilA, N-terminal domain protein (YP_001870065.1/Nostoc punctiforme PCC 73102/5e−07)
43. 105L	74659–75186	175	19.71	Hypothetical protein RCCS2_13854 (ZP_01750712.1/Roseobacter sp. strain CCS2/1e−20)
44. 115L	81974–82249	91	10.24	Thymidylate synthase, flavin-dependent (YP_001866026.1/Nostoc punctiforme PCC 73102/1e−32)
45. 116L	82242–82526	94	10.64	Hypothetical protein sce4398 (YP_002377620.1/Sorangium cellulosum “So ce 56”/1e−11)
46. 117R	82652–83203	183	21.40	Dam methylase (YP_003925.1/Enterobacteria phage T1/1e−04)
47. 119L	83587–84066	159	17.64	Crossover junction endodeoxyribonuclease (ZP_01264237.1/“Candidatus Pelagibacter ubique” HTCC1002/1e−25)
48. 120L	84063–84380	105	12.31	Conserved hypothetical protein (ZP_03276335.1/Arthrospira maxima CS-328/6e−15)
49. 122L	84775–85386	203	22.15	Conserved hypothetical protein (ZP_01677308.1/Vibrio cholerae 2740-80/3e−12)
50. 123L	85383–85976	197	21.06	Membrane protein (ZP_01040846.1/Erythrobacter sp. strain NAP1/2e−28)
51. 127L	87003–99430	475	52.12	FtsK/SpoIIIE family protein (YP_003445339.1/Streptococcus mitis B6/6e−05)
52. 140L	93956–94276	106	12.06	Hypothetical protein EUBVEN_00477 (ZP_02025248.1/Eubacterium ventriosum ATCC 27560/5e−10)
53. 142L	94392–95195	267	30.28	Hypothetical protein L8106_22821 (ZP_01624657.1/Lyngbya sp. strain PCC 8106/2e−70)

^aAn ORF was considered to be a match with its homologue at an E value of <0.001.

Major structural proteins of PaV-LD.

The purified PaV-LD virions were subjected to SDS-PAGE, and the separated proteins were identified by matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS). As shown in Fig. 2, a total of 13 proteins, ranging in size from about 172 kDa to 27 kDa, have been identified, and 10 of them, including tail tape measure protein (TMP), portal protein, capsid protein, virion protein, FtsK/SpoIIIE protein, capsid morphogenesis protein, minor virion protein, major capsid protein, DNA-damage-inducible protein, and ABC transporter are, respectively, encoded by the 088R, 057R, 071R, 078R, 127L, 081R, 085R, 073R, 095L, and 033R genes in the PaV-LD genome. However, the remaining three proteins are not encoded by any genes of the PaV-LD genome, and they were revealed to be homologues of the glucose-inducible outer membrane proteins (OprBs) in cyanobacteria Arthrospira platensis, Arthrospira maxima, and Nostoc punctiforme (Fig. 2; see Table S2 in the supplemental material).

Fig 2.

SDS-PAGE and mass spectrometry analysis of structural proteins in the purified PaV-LD virions. Virion proteins were separated by 15% SDS-PAGE and identified by matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS). The gene (ORF) symbols and approximate molecular masses (M) in kDa of the 13 major structural proteins are indicated.

Among these structural proteins, the ABC transporter protein (033R), DNA damage-inducible protein (095R), and FtsK/SpoIIIE protein (127L) are the known structural proteins. Portal protein (057L), two capsid proteins (071R and 073R), and two minor virion proteins (078R and 085R) are structural components of the known phages. 088R is the only ORF that encodes a phage tail tape measure protein (TMP), which has been demonstrated as being involved in determination of phage tail length (41). In addition, the PaV-LD TMP shares less than 22% amino acid identity with the known tailed phages (Fig. 3). Moreover, in comparison with other known genomes of the tailed phages, the PaV-LD genome indeed lacks all of the essential genes that encode major tail structure proteins, such as tail fiber protein, neck protein, sheath protein, tube protein, and baseplate protein (Table 2). Thus, it is possible that PaV-LD represents an independent phylogenetic branch of cyanophages. Furthermore, we analyzed the phylogenetic relationships among the ubiquitous capsid proteins from 20 phages. As shown in Fig. 4, the 19 phages are divided into 3 major groups based on Podoviridae, Siphoviridae, and Myoviridae, and PaV-LD represents an independent branch, suggesting that PaV-LD might possess a quite divergent evolutionary course from other tail phages.

Fig 3.

Amino acid sequence alignment of MT3 motifs in tail tape measure proteins (TMPs). The conserved amino acids are marked by boldface letters. The aligned motif sequences are from the following: PaV-LD (088R; HQ683709;); TM4, Mycobacterium phage TM4 (gp17; AAD17585.1); Catera, Mycobacterium phage Catera (gp94; ABE67842.1); Halo, Mycobacterium phage Halo (gp16; YP_655533.1); Che8, Mycobacterium phage Che8 (gp14; AAN12412.1); Wild, Mycobacterium phage Wildcat (gp38; ABE67643.1); and Bxzl, Mycobacterium phage Bxz1 (gp95; AAN16752.1).

Table 2.

Tail-related proteins shared by or lacking from PaV-LD and other tail phages

Phage	Presence of tail proteina:
	Tape measure	Fiber	Neck	Sheath	Tube	Baseplate
PaV-LD	+
Siphoviridae
lambda	+	+	+
PSS2	+	+
S-CBS2	+	+
S-CBS3	+	+
TP901-1	+	+	+
HK022	+	+
A2	+		+
T1	+	+
Myoviridae
T4		+	+	+	+	+
S-PM2			+	+	+	+
P-SSM2		+		+	+	+
P-SSM4		+		+	+	+
Syn9		+	+	+	+	+
S-RSM4		+	+	+	+	+
S-SM1		+	+	+	+	+
S-SSM5		+	+	+	+	+
S-ShM2		+	+	+	+	+
KVP40		+		+
Podoviridae
T7		+			+
P-SSP7		+	+		+
Pf-WMP3		+	+		+
Pf-WMP4		+	+		+
Syn5		+			+
P60		+	+		+
P-SSP7		+			+
phiYeO3-12		+	+		+

^aThe tail proteins discussed in this article were retrieved from the website http://www.ncbi.nlm.nih.gov/genomes. + indicates the shared tail proteins.

Fig 4.

Phylogenetic tree based on major capsid proteins from 20 the known phages (bacteriophages and cyanophages). The major capsid protein sequences were retrieved from public GenBank databases: PaV-LD, Panktothrix agardhii cyanophage PaV-LD (HQ683709); P60, Synechococcus phage P60 (NP_570347.1); Syn5, Synechococcus phage Syn5 (YP_001285448.1); P-SSP7, Prochlorococcus cyanophage P-SSP7 (YP_214206.1); T7, Enterobacteria phage T7 (NP_041998.1); KP32, Klebsiella phage KP32 (YP_003347548.1); KP34, Klebsiella phage KP34 (YP_003347636.1); PT2, Pseudomonas phage PT2 (ABY71003.1); LKD16, Pseudomonas phage LKD16 (YP_001522824.1); A2, Lactobacillus phage A2 (NP_680487.1); phi3626, Clostridium phage phi3626 (AAL96776.1); Cherry, Bacillus phage Cherry (YP_338137.1); Gamma, Bacillus phage Gamma (YP_338188.1); RB69, Enterobacteria phage RB69 (NP_861877.1); T4, Enterobacteria phage T4 (NP_049787.1); RB49, Enterobacteria phage RB49 (NP_891732.1); P-SSM2, Prochlorococcus phage P-SSM2 (YP_214367.1); syn9, Synechococcus phage syn9 (YP_717802.1); S-PM2, Synechococcus phage S-PM2 (YP_195142.1); and S-RSM4, Synechococcus phage S-RSM4 (YP_003097339.1). The scale bar represents 0.1 fixed mutation per amino acid position. Bootstrap, 1,000.

Electron microscopy observation of tailless PaV-LD.

To establish the tailless nature of PaV-LD, we performed electron microscopy for observation and comparative study of some tail phages isolated from Lake Donghu, China. As shown in Fig. 5, numerous tailless virus particles with identical sizes and morphologies were observed in the infected Planktothrix agardhii HAB0637 filaments by PaV-LD (Fig. 5A), and the negatively stained PaV-LD particles are hexagonal and have an icosahedral symmetry (Fig. 5B). In comparison with the short-tailed phages (Fig. 5C and D) that belong to Podoviridae, the contractile tail phage (Fig. 5E) that belongs to Myoviridae and the long noncontractile tail phage (Fig. 5F) that belongs to Siphoviridae, the novel cyanophage particles are obviously tailless. Therefore, PaV-LD is indeed different from other tail phages.

Fig 5.

Comparative electron micrographs of PaV-LD and several tail phages. (A) Ultrathin section of HAB0637 filament cells 48 h after infection with PaV-LD. Numerous tailless virus particles with identical sizes and morphologies are observed in the infected cells. Scale bar, 200 nm. (B) Negatively stained PaV-LD particles isolated from the infected HAB0637 filament cells. Scale bar, 50 nm. (C and D) Negatively stained short-tailed phages that belong to Podoviridae. (E) Negatively stained contractile tail phage that belongs to Myoviridae. (F) Negatively stained long noncontractile tail phage that belongs to Siphoviridae. The tail phages were isolated from the water samples of Lake Donghu, China. Scale bars of panels C, D, E, and F, 50 nm.

Structure and function of PaV-LD NblA.

Intriguingly, the PaV-LD nonbleaching protein A (NblA) encoded by 022L also contains the highly conserved LTMEQ motif in the N terminus and two amino acid residues (Leu50 and Arg52) in the C terminus (Fig. 6A), which are responsible for binding to the phycobilisome. In addition, the PaV-LD NblA has high amino acid identities of 44.1% to 30.9% with the homologues in cyanobacteria, whereas only 19.1% identity (Fig. 6A) exists with the homologue in cyanophage Ma-LMM01, which belongs to the Myoviridae (64). In cyanobacteria, NblA has been demonstrated to be a key protein of thylakoid and phycobilisome degradation and to result in a color change of cyanophage cultures from normal blue-green to chlorotic yellow (2, 7, 9). To reveal the function of PaV-LD NblA, we expressed the NblA fusion protein and prepared the anti-NblA antibody. Western blot detection showed that PaV-LD NblA could be expressed in the infected host cells at 36 h postinfection and continued to increase until 60 h postinfection (Fig. 6B). This indicates that PaV-LD 022L indeed encodes NblA, and its expression increases steadily with extension of the infection time. Moreover, the nonbleaching phenomenon also occurred in the PaV-LD-infected filamentous cyanobacterium cultures, in which the blue-green color of the normal cyanobacterium cultures (Fig. 6C, normal) became a translucent light yellow after the PaV-LD infection (Fig. 6C; PaV-LD infected). In addition, the typical absorbance peak of phycocyanobilin (PCB), the major pigment protein in phycobilisomes, gradually decreased in the PaV-LD-infected filamentous cyanobacterium cultures from 24 h to 66 h postinfection (Fig. 6D).

Fig 6.

Structural and functional analyses of PaV-LD NblA and overexpression analysis of PaV-LD 022L in the model cyanobacterium Synechocystis sp. strain PCC 6803. (A) Amino acid sequence alignments of PaV-LD NblA and other NblA homologues. The highly conserved LTMEQ motif and two amino acid residues, Leu50 and Arg52, are shaded. The NblA homologues are from the following: PaV-LD (HQ683709); Ma-LMM01, Microcystis aeruginosa phage (phage; YP_851019.1); PCC 6803-1, Synechocystis sp. (cyanobacterium; NP_441275.1); PCC 6803-2, Synechocystis sp. (cyanobacterium; NP_441274.1); PCC 8106-1, Lyngbya sp. (cyanobacterium; EAW33710.1); PCC 8106-2, Lyngbya sp. (cyanobacterium; ZP-01624303.1); PCC 7806, Microcystis aeruginosa (cyanobacterium; CAO88905.1); and PCC 7120, Nostoc sp. (cyanobacterium; BAB76216.1). (B) Western blot detection of NblA expression in the PaV-LD-infected cyanobacterium Planktonthrix agardhii. (C) Color change in the PaV-LD-infected filamentous cyanobacterium cultures. Uninfected cultures are blue-green (Normal), and cyanophage-infected cultures are chlorotic yellow or light yellow (PaV-LD infected). (D) Spectroscopic images of phycocyanobilin (PCB) in the PaV-LD-infected filamentous cyanobacterium Planktonthrix agardhii HAB637 cultures. The absorbance peaks of PCB at 630 nm and chlorophyll a (Chla) at 680 nm were observed. A gradual decrease in the peak contents of PCB appear at 0 h, 24 h, 48 h, and 66 h postinfection. (E) Spectroscopic images of the normal (Wild) and PaV-LD 022L-transferred (+NblA) model cyanobacteria. A significant decrease in the peak contents of PCB appears after the 022L is transfected. (F) Phenotype feature of the untransfected control (Wild) and the 022L-transfected (+NblA) model cyanobacteria showing the color change in the transfected (+NblA) cyanobacterium cultures. (G) Electron microscope images of the untransfected control (Wild) and 022L-transfected (+NblA) model cyanobacteria.

Subsequently, the transformation and expression of PaV-LD 022L were performed in a model cyanobacterium strain, Synechocystis PCC 6803, by using a neutral genomic platform. Compared with the wild-type strain, the typical absorbance peak of phycocyanobilin (PCB) absorption at 620 nm completely disappeared from the PaV-LD 022L-transfected strain (Fig. 6E), and significant nonbleaching color changes also occurred in the PaV-LD 022L-transfected (+NblA) cyanobacterium (Fig. 6F). The significant correlation between PaV-LD 022L expression and PCB decline suggests that PaV-LD NblA should be intimately linked to the phycobilisome degradation during phage infection of cyanobacterial host. Moreover, the wild-type strain and the PaV-LD 022L-transfected (+NblA) strain were observed by transmission electron microscopy, respectively. In comparison with the normal structure of the thylakoids in the control cyanobacterium (Fig. 6G), the thylakoids (and phycobilisomes on the thylakoid membranes) were completely disrupted in the PaV-LD 022L-transfected cyanobacterium, and some vacuoles also occurred along with the disruption (Fig. 6G; +NblA). The data confirmed that the nonbleaching phenomenon was really caused by the PaV-LD NblA.

DISCUSSION

The size of PaV-LD genome (95.3 kbp, which contains 142 ORFs) is much smaller than those of the myoviruses, as reported previously for 14 ordinary cyanophage genomes (with sizes ranging from 174 to 196 kb and numbers of ORFs ranging from 198 to 292) and two bigger cyanophage genomes (S-SSM7, with a size of 233 kb containing 324 ORFs and P-SSM2, with a size of 252 kb, containing 330 ORFs) (50). However, there are exceptions: the size of the PaV-LD genome is larger than the genome of the short-tailed cyanophage (e.g., Syn5, about 46 kb contains 61 ORFs) (44). This may indicate some degree of correlation between cyanophage genome size and contractile tail morphology. Additionally, cyanophage genomes show large variation in GC content. Previous reports suggested that cyanophage genomic GC contents might be more closely related to their hosts, and the host adaptability might enhance cyanophage infection in different hosts (12, 24, 28). The PaV-LD genome has an average GC content of 41.5%, which is significantly different from that of cyanophages isolated from Prochlorococcus (37.2% ± 1.0%) and Syn5 isolated from Synechococcus (55%) (44, 50).

Tail-associated proteins have an important influence on structure and function in cyanophages and phages (1, 43, 55). Tailed cyanophages and phages have been revealed to contain several genes closely related to tail components, such as tail fiber protein, neck protein, sheath protein, tube protein, and baseplate protein (23, 26), and some tail-associated proteins and their similarities have been proposed to be a critical element in phage classification (56). Unlike the tailed cyanophages which have a contractile tail (50), PaV-LD 088R is the only ORF that encodes a phage tail tape measure protein (TMP), and its nucleotide sequence similarity is much less than those of other tailed cyanophages, as described previously (42, 43). The data show that PaV-LD may not constitute a routine tail of cyanophage, and a thorough observation by electron microscopy has confirmed that PaV-LD is tailless and is a novel cyanophage different from other tail cyanophages and phages.

Major capsid protein (MCP) is the primary unit from which the head capsid is assembled. Cyanophage (phage) capsids are widely classified based on their structure and their interactions with their hosts (8, 59, 67). Some of them have developed more complicated structures or even affect the efficiency of host bacterium infection (57, 63). The major capsid protein is highly similar in the tailed cyanophage species from the same family of Myoviridae, Siphoviridae, or Podoviridae. Moreover, two minor virion proteins and three capsid proteins have been identified in PaV-LD by genome analysis and structure protein comparison with the known cyanophages or phages. In addition, phylogenetic analyses of MCP genes clearly demonstrated that PaV-LD formed an independent phylogenetic branch that differed from the known tailed cyanophages. This implies that PaV-LD may be divided into a new subgroup.

High genetic diversity and high proportions of various genes have been revealed in the sequenced cyanophage genomes (11, 50). Many cyanophages have been demonstrated to carry different host genes, and some of these genes can even subsidize redirection of cyanobacterial host carbon metabolism (53). These host-like genes code for proteins involved in the interaction between cyanophage and cyanobacterium, which may benefit viral infection and proliferation. As documented for other tailed cyanophages, some host-like genes have been reported to be involved in the coevolution of viruses and their hosts (19, 45, 53). Similar to genomes of many cyanophages, PaV-LD also contains a high proportion (60%) of genes that are unique (see Table S1 in the supplemental material). Among the predicted 53 ORFs of PaV-LD that encode proteins with best hits in the database, 21 show amino acid sequence similarity to proteins of cyanobacterial origin and 18 are homologs of bacterial cellular proteins. In addition to essential proteins for replication and transcription during the PaV-LD life cycle, there are some homologues of cellular proteins with the characterized functions.

Intriguingly, NblA encoded by PaV-LD 022L is a small polypeptide and is present in all organisms containing phycobilisomes (20). In cyanobacteria and red algae, phycobilisome degradation can improve survival under high light conditions and during nutrient limitation. Previously, the nblA gene had been revealed to be a captured host gene in a cyanophage (Ma-LMM01) genome (64), but its function has not yet been proven. Our current results show that PaV-LD 022L encodes a homologue of host NblA and the gene expression indeed induces a color change from normal blue-green to chlorotic yellow in cultures of the model cyanobacterium strain Synechocystis PCC 6803. These findings are the first evidence that cyanophage nblA has an essential function in phycobilisome degradation and PaV-LD nblA may be evolved from the host cyanophage. These data are also in agreement with the hypothesis that cyanophage NblA is necessarily recruited for host phycobilisome degradation, in which it may provide a way to hijack some materials for packing into the offspring cyanophage particles and to contribute to the pathogenesis and subsequent entry. Unexpectedly, in the purified PaV-LD virions, the remained three proteins (OprB_AP, OprB_Am, and OprB_Np) lack the corresponding genes in the PaV-LD genome. They are most closely related to cyanobacterial carbohydrate-selective porin OprBs and might be supposed to come from the host cells. However, further studies are needed to determine whether the OprBs' origin is related to PaV-LD NblA.

In conclusion, the current study describes the PaV-LD genome, which significantly differs from the genomes of the known cyanophages. The PaV-LD genome does not carry the genes coding for the fiber, neck, sheath, tube, and baseplate proteins that are required for tail formation in cyanophages, except for the putative tape measure protein gene 88R. Also, its major capsid protein is quite divergent from the known cyanophages. In addition, a gene (022L) encoding NblA has been identified in PaV-LD genome. By integrating 022L into a neutral platform of the model cyanobacterium, the expression of this gene was verified to cause a nonbleaching color change in the transfected cyanobacterial cultures. This is the first evidence that cyanophage carries the gene coding for a type of NblA, which indeed has a typical nonbleaching effect in the transgenic cyanobacteria. All of these studies create some concern over the conclusion that the PaV-LD genome is a unique cyanophage genome. The current study has opened a new window into the morphological and genetic diversity of cyanophages as well as interactions between cyanophages and their hosts.