Distribution, Isolation, Host Specificity, and Diversity of Cyanophages Infecting Marine Synechococcus spp. in River Estuaries

Abstract

The abundance of cyanophages infecting marine Synechococcus spp. increased with increasing salinity in three Georgia coastal rivers. About 80% of the cyanophage isolates were cyanomyoviruses. High cross-infectivity was found among the cyanophages infecting phycoerythrin-containing Synechococcus strains. Cyanophages in the river estuaries were diverse in terms of their morphotypes and genotypes.

Viruses are now known to be an important microbial component that could regulate biological production and species composition of bacterial and phytoplankton populations, influence biogeochemical cycling, and mediate gene transfer between microorganisms in aquatic ecosystems (1, 3, 8, 9, 16). Unicellular cyanobacteria of the genus Synechococcus are among the most abundant picophytoplankton and account for 5 to 30% of the primary production in the open ocean (12). The concentrations of viruses or cyanophages which infect specific strains of marine Synechococcus spp. typically range from 10² to 10⁴ ml⁻¹ in nearshore and offshore waters (11, 13) and sometimes can be more than 10⁵ phage particles ml⁻¹ (11). It was found previously that 1 to 3% of marine Synechococcus spp. from a variety of locations contained mature phages (8), and cyanophages may be responsible for ca. 5 to 14% of cyanobacterial mortality on a daily basis (11).

Abundant cyanophages in seawater could be a significant factor in determining the dynamics of Synechococcus populations. In both coastal and open oceans, the distribution of cyanophages is in concordance with the distribution of total Synechococcus cells, and the abundance of cyanophages varies spatially and temporally (11, 13). However, little is known about the effect of salinity on the distribution of cyanophages, particularly in river estuarine areas.

Cyanophages that infect marine Synechococcus spp. in marine environments are diverse in terms of their morphotypes and genotypes. Various tailed cyanophages have been observed in coastal and open ocean waters (10, 13, 14). Different cyanophages propagated in marine Synechococcus sp. strain WH7803 could be differentiated based on their restriction fragment length polymorphisms (14). It was found previously that in the Atlantic Ocean cyanophage diversity was high and changed with depth (15).

This study was initiated to examine the abundance and distribution of cyanophages that infect marine Synechococcus spp. in Georgia estuarine rivers. We also determined the morphology, host specificity, and genetic diversity of cyanophages isolated from estuarine systems.

Water samples.

Water samples were collected from three Georgia coastal rivers (Savannah, Altamaha, and Satilla), by researchers on board the R/V Bluefin between 11 and 17 July 1998 (Fig. 1). The samples were kept at 4°C and were used to determine the titer of cyanophages after the cruise (within 24 h). The virus communities in two coastal samples collected from the pier at the Skidaway Institute of Oceanography were concentrated by the ultrafiltration method described by Chen et al. (2).

FIG. 1.

Sampling sites in the Georgia estuaries. For the Savannah River, sites 1 to 6 were designated stations SV1 to SV6, respectively; for the Altamaha River, sites 1 to 7 were designated stations AL1 to AL7, respectively; and for the Satilla River, sites 1 to 6 were designated stations SL1 to SL6, respectively.

Synechococcus strains.

Eleven Synechococcus strains, including four phycoerythrin-containing (PE) strains (WH7803, WH7805, WH8108, and WH8103) and seven phycocyanin-containing (PC) strains (WH8101, WH8007, WH5701, CCMP1628, CCMP1629, CCMP1630, and CCMP1632) were used in this study. The strains whose designations begin with WH came from the Woods Hole Collection of Cyanobacteria (Biology Department, Woods Hole Oceanographic Institution), and the strains whose designations begin with CCMP were from the Culture Collection at The Bigelow Laboratory. Cultures were grown at 26°C with constant light (20 to 30 microeinsteins m⁻² s⁻¹) in an illuminated room by using SN growth medium (12).

Isolation of cyanophages.

Clonal cyanophages were isolated by the liquid serial dilution method (10). Twenty-microliter portions of water samples were filtered through 0.45-μm-pore-size filters and transferred into exponentially growing cultures of the host suspended in SN growth medium in 96-well tissue culture plates. The controls received no virus. The plates were incubated under the conditions described above and were monitored daily for 7 to 10 days. Lysed cultures were examined by transmission electron microscopy to verify the presence of phages and then purified by another serial dilution procedure.

Estimation of cyanophage and cyanobacterium concentrations.

Cyanophages were quantified by using the most-probable-number technique described by Suttle and Chan (10). The dilutions were monitored for cell lysis daily for 2 weeks. The number of wells in which lysis occurred was determined, and the concentration of infective units and the error associated with the estimates were determined by the method described by MacDonell et al. (7). The number of cyanobacteria in a water sample was determined by the protocol described by Waterbury et al. (12).

Host range.

Purified cyanophage isolates were tested with other strains of marine Synechococcus spp. to determine cross-infectivity. A lysate was filtered with a 0.2-μm-pore-size sterile Acrodisc filter (Gelman Sciences, Ann Arbor, Mich.), and 20 μl of filtrate was added to four 2-ml exponentially growing cultures. The control received no phage. Each host-cyanophage preparation was incubated in the conditions described above and was monitored daily for host cell lysis.

Phage morphology.

Forty microliters of a high-titer phage stock preparation was spotted onto a Formvar-carbon-coated 400-mesh copper grid and allowed to adsorb to the grid for 30 min. The grid was then promptly washed with a few drops of deionized water and stained with 1 drop of a 2% (wt/vol) aqueous uranyl acetate solution for 30 s. The excess stain was removed immediately with filter paper. The cyanophages on the grid were examined with a JEOL-100CX transmission electron microscope at The Microscopic Center, University of Georgia.

Cyanophage DNA extraction.

In order to obtain enough DNA from a specific cyanophage, 500 ml of clonal lysate was treated with chloroform, DNase I, and Rnase by using the protocol described by Wilson et al. (14). Viral particles were centrifuged at 140,000 × g and 4°C for 6 h by using a Surespin ultracentrifuge (Discovery 100S; Sorvall, Newtown, Conn.) with a 630/36 swing bucket rotor (Sorvall). The pellet was resuspended with 0.5 ml of 1× TE buffer (pH 8.0). Viral DNA was extracted by using a Phase Lock GEL kit (5 Prime-3 Prime Inc., Boulder, Colo.) as recommended by the manufacturer. The DNA concentration was measured with DU640 spectrophotometer (Beckman Instruments Inc., Fullerton, Calif.).

Restriction enzyme digestion.

Restriction endonucleases AccI, BamHI, EcoRI, EcoRV, and HindIII (Promega, Madison, Wis.) and restriction endonucleases SmaI, KpnI, and XbaI (Boehringer Mannheim, Indianapolis, Ind.) were used to digest cyanophage DNA according to the protocols described by the manufacturers. Digested DNA was separated by agarose gel electrophoresis (1.5% agarose) and stained with ethidium bromide. Each gel image was captured and analyzed by using a gel documentation system (Alpha Innotech Corp., San Leandro, Calif.).

The concentration of cyanophages infecting WH7803 ranged from 65 to 44,046 phage particles ml⁻¹, and the mean in the three rivers was 8,296 phage particles ml⁻¹. The concentrations of cyanophages infecting WH8101 (PC strain) were determined only for the Savannah River, and the values ranged from undetectable to 147 phage particles ml⁻¹. The cyanophage titers for the coastal sites of rivers were similar to the titers observed in Woods Hole Harbor (13), the Gulf of Mexico (10), and Bermuda samples (14). The average titer of cyanophages infecting WH7803 was 324 times higher than the average titer of cyanophages infecting WH8101, suggesting that the level of cyanophages was host dependent. The concentration of cyanophages which infected marine Synechococcus strains increased with increasing salinity, and the highest titers all occurred near the mouths of rivers, where the salinity was highest (Fig. 2). The highest Synechococcus concentrations also occurred in the most saline water in the Altamaha and Satilla rivers (Fig. 2).

FIG. 2.

Distribution of cyanophages propagated in Synechococcus WH7803 and WH8101 along salinity gradients in three Georgia coastal rivers. For the Savannah River sites the suffix S indicates surface samples collected at a depth of 1 m, and the suffix B indicates bottom samples collected 1 m above the bottom. MPN, most probable number.

It is believed that salinity in the estuarine systems studied has a significant impact on both the cyanophage population structure and the host population structure. We observed that PE strains WH7803 and WH7805 did not grow well when the salinity of the medium was less than 28‰, while PC strains WH8101 and WH8007 grew well at salinities ranging from 18 to 30‰. It appeared that the PE strains have less salinity tolerance than the PC strains. The concentration of WH7803-like Synechococcus strains could decrease rapidly with decreasing salinity in the rivers, although the total Synechococcus counts did not change dramatically. Because the most-probable-number method is a culture-dependent method, a lower concentration of cyanophages infecting WH7803 in the upper parts of rivers does not necessarily mean that the concentration of all cyanophages is lower.

Earlier studies showed that the distribution of cyanophages infecting WH7803 was similar to the distribution of all Synechococcus strains in marine environments (11, 13). Cyanophages infecting marine Synechococcus may require a certain range of salinity for replication, like other marine bacteriophages. Zachary (17) found that in one-step growth studies salinity was required for growth of the bacteriophages of the marine bacterium Beneckea natriegens and suggested that in an area where the salinity is too low for B. natrigens to survive, the phage cannot replicate and decays through a natural inactivation process. It has been reported that the concentration of marine members of the alpha subclass of the class Proteobacteria declines with decreasing salinity in the Savannah and Satilla estuaries (4). It should be noted that the distribution pattern for cyanophages which we report here was limited to the pattern for the cyanophages infecting Synechococcus strains WH7803 and WH8101.

Seventy-two clonal cyanophages which infect marine Synechococcus spp. were isolated from Georgia estuarine water samples. Various morphotypes of cyanophages were found among these isolates (Fig. 3). About 80% of the cyanophage isolates were cyanomyoviruses. The head sizes of the cyanophages ranged from 42 to 92 nm for the cyanomyoviruses. We noticed that the cyanophages isolated from high-salinity water (coastal water) exhibited more diverse morphotypes than those isolated from low-salinity water (upstream river).

FIG. 3.

Thirteen representative strains of cyanophages isolated from Georgia coastal areas exhibited different phenotypes as determined by transmission electron microscopy. (A to O) Cyanophage isolates P3, P5, P6, P14, P16, P39, P49, P53, P71, P72, P73, P79, P81, P82, and P83, respectively. The scale bar applies to all panels.

Cyanophages that lysed the PE strains were detected in almost all of the water samples collected from the three coastal rivers, whereas cyanophages that lysed the PC strains appeared much less frequently. None of CCMP strains or WH5701 could be lysed with the water samples collected from Georgia estuaries. Of the 72 clonal cyanophages, 91% were isolated from PE strains and 9% were isolated from PC strains. Suttle and Chan (10) also found that cyanophages that infected the green strains (PC strains) of Synechococcus were much less common than cyanophages that infected the red strains (PE strains).

Our results also showed that the cyanophages infecting the PE strains have a broad host range compared to the cyanophages infecting PC strains (Table 1). Cyanophage isolates P3, P5, P8, P14, and P16 appeared to have broader host ranges than the other isolates and were able to infect some PE and PC strains. However, none of cyanophages isolated were able to cross-infect WH5701 and the CCMP strains. High cross-infectivity among cyanophages infecting PE strains (marine cluster A) was also found by Waterbury and Valois (13); some of their cyanophages were able to infect 10 to 13 strains of Synechococcus in marine cluster A, but only a few strains could cross-infect PC strain WH8101 (marine cluster B).

TABLE 1.

Host ranges of representative Synechococcus phages

Phage isolate	Host strains
	PE Synechococcus strains				PC Synechococcus strains
	WH7803	WH7805	WH8108	WH8103	WH8101	WH8007
P2	Ia	C	C
P3	C	I	C	C	C	C
P5	I	C	C	C	C
P6	C	I	C			C
P8	C	C			I
P12					I
P14	C	C			I
P16	I	C	C	C	C	C
P39	I	C	C
P49	I	C	C
P53	I	C	C
P56	I	C	C
P68					C	I
P71	C	I	C
P72	C	I
P73	C	I	C
P76						I
P79	C	I	C
P81	C	I	C
P82	C	I	C
P83	C	I	C

^aI, host used for phage isolation; C, host which was cross-infected. 

Why is it difficult to isolate cyanophages from the PC strains? Viral cross-infectivity may reflect the genetic relatedness of the hosts. The phylogenetic relationships of the Synechococcus strains based on the RuBisCO gene sequence revealed that PE strains were closely related, but the PC strains used in this study were distantly related, and some of them were even closely related to the freshwater strains of Synechococcus (Chen, unpublished data). Cyanophages infecting marine Synechococcus were not able to infect freshwater Synechococcus (10). Thus, it is important to understand the phylogenetic relationship of the genus Synechococcus. A recent study based on 16S rRNA gene sequence analysis indicated that the genus Synechococcus is so diverse that it may not be a natural taxon (5).

Although restriction enzymes could be used to differentiate genotypes of cyanophages, not all cyanophage DNA could be digested with the restriction enzymes tested. Of the eight restriction endonucleases tested, AccI and EcoRV could completely digest most of the phage DNA tested, and they appeared to generate more distinguishable band patterns than the other enzymes. The DNA of seven cyanophages which belong to the same family (Myoviridae) but were isolated from various marine environments exhibited seven different digestion patterns with EcoRV (Fig. 4A). Most of the cyanophages included in Fig. 4A were isolated from Synechococcus sp. strain WH7803; the only exception was P3, which was isolated from WH7805. The other seven cyanophages (propagated in strains WH7803 and WH7805) isolated from the Georgia coastal rivers produced five discernible DNA fingerprints with AccI (Fig. 4B). The genome sizes of these cyanophages ranged from ca. 36 to 78 kb. Wilson et al. (14) showed that BamHI and EcoRV gave clearer patterns than the other endonucleases tested. Marine cyanophages or bacteriophages that are resistant to digestion by some endonucleases have been described previously (6, 14). It is believed that the presence of a base modification such as methylation in cyanophages protects the phage DNA from digestion by many endonucleases.

FIG. 4.

Restriction digestion patterns for cyanophage DNA. (A) Cyanophage DNA digested with endonuclease EcoRV. Lanes a to h, molecular marker IV, T4, P60, φ9, φ22, S-PWM1, P3, and φ18, respectively. (B) Cyanophage DNA digested with endonuclease AccI. Lanes a to h, molecular marker IV, P1, P6, P5, P72, P79, P80, and P60, respectively.

In this study, we demonstrated the effect of salinity on the distribution of cyanophages in three of Georgia's estuaries. It is believed that mixing of freshwater and seawater in the estuaries has significant effects on the structures of the populations of cyanophages and their hosts. Cyanophages are diverse not only morphologically but also genetically. A better understanding of the host ranges or specificities of cyanophages in estuaries requires more study of the phylogenetic relationships of Synechococcus strains in marine and freshwater environments.