Reverse genetics

The cDNA copies of the genomic segments were inserted into a derivative of plasmid pT7T319U (GenBank: U13870.1) that had the T7 RNA polymerase promoter replaced by the promoter of SP6 RNA polymerase so that transcription would start efficiently at the terminal G of segments S and M. The fidelity of the cDNA constructs was tested by their activity in the production of live virus resulting from their electroporation into strain LM3313. LM3313 is a derivative of LM2489 carrying plasmid pLM2790 that expresses the SP6 RNA polymerase. Three different constructs of the L segment were utilized; they contained 5' sequences beginning with the normal ACAA start, or with GACAA or GCAA. Segments M and S were normal (Figs. 2 and 3). The cDNA plasmids are ColE1 derivatives and unable to replicate in pseudomonads. The constructs with ACAA or GACAA produced about six thousand plaques from 10¹⁰ cells while the construct of GCAA produced about one hundred plaques. The amount of transcript with ACAA would be expected to be much lower than that for GACAA, suggesting that its efficiency in plaque production would be greater than that for GACAA on a per molecule basis. While the phage resulting from the L segment with the normal 5' sequence of ACAA behaved identically to that of wild type Φ2954, the phage with the GCAA sequence showed novel transcription behavior. Whereas wild type Φ2954 nucleocapsids transcribe only genomic segments S and M in vitro; the mutant with GCAA instead of ACAA at the 5' terminus transcribes all three genomic segments in vitro (Fig. 5). The differences in template activity for genomic segment L as opposed to S and M is used by most members of the Cystoviridae to effect temporal regulation of transcription. In the case of Φ6 the L segment has the sequence GU at the 5' end of the plus strand while S and M have GG. The polymerase favors G over U as the second nucleotide and it is proposed that the second base is the first to be paired during transcription [13]. A host protein, YajQ, is able to alter this selection so as to enable active transcription of the L segment upon entry into the host cell [4]. A change of GU to GG in segment L of Φ6 results in in vitro transcription of segment L [14]. Although the structure of the polymerase of Φ2954 has not been studied, it seems likely that in this case the terminal nucleotide would be paired first and that G is preferred to A.

The host specificity of Φ2954 is different from that of its close relative Φ12; however it was possible to construct viable phage with a middle segment containing the pac sequence of Φ2954 and the genes 6 and 3 of Φ13. Gene 3 codes for the host attachment protein while gene 6 codes for its membrane bound anchor [15]. The plasmid pLM3575 has the 5' region of Φ2954M up to the SphI site at position 491 and the sequence of Φ13M from SacII at nucleotide 80. The resulting phage, Φ3010 does not plate on the normal host of Φ2954, HB10Y but does plate on strains that have rough LPS such as LM2509 or LM2489.

We have also constructed a plasmid with the pac sequence of Φ2954M and the genes 6 and 3 of Φ6. The resulting phage has the same plating properties as Φ2954 with respect to pilus attachment.

Another test of the functionality of the cDNA copy of segment M was to determine whether bacteriophage Φ12 could acquire the transcript of this plasmid in order to change its host range. Plasmid pLM3497, which carries the cDNA copy of Φ2954 genomic segment M, was electroporated into strain LM3313 before infection with Φ12. These cells were plated along with those of HB10Y and plaques were obtained. These plaques plated on HB10Y but not on a strain missing the type IV pili. The genomic segments of these phages were consistent with the segments L and S of Φ12 and M of Φ2954 (Fig. 6). The finding that Φ12 is able to acquire segment M of Φ2954 is intriguing in that the pac sequences in M are very different for both phages. This is reminiscent of the case of bacteriophage Φ13 acquiring segment M of Φ6 in which case there is again very little sequence similarity in the pac sites [2]. This is so despite the observation that small changes in the pac sequences of Φ6 M or S drastically reduce the ability of Φ6 to acquire these segments [16].

The finding that it is possible to change the host attachment proteins is of special interest in that it shows that the proteins P6 and P3 are able to recognize viral membrane that contains the major membrane protein P9 of distantly related phages of the same family.

Another test of genomic packaging was the production of a genomic segment containing segments S and M joined together. The ApaI to XbaI segment of M was joined to the PstI site that is present in the vector following the 3' end of the cDNA copy of segment S. A plasmid pLM3662 containing this construct was able to donate its transcript to phage Φ3010 that contains segments S and L of Φ2954 and an M segment that has the pac sequence of Φ2954M but the genes of segment M of Φ13, a phage that plates on host cells containing rough LPS.

Electroporating plasmid pLM3695 into strain LM3313 produced a phage with the entire genome contained in a single segment. This plasmid contained the cDNA copies of the complete segment S with the sequence of segment M beginning with the ApaI site at position 34 to the XbaI site following its C terminus with segment L beginning with an MfeI site at position 611 that was converted to XbaI. The observation that phage were produced in high yield from this plasmid is consistent with the previous observations of the preparation of single segment genomes in Φ6 and Φ13. It also suggests that the open reading frames of genes 14 and 15, starting at 243 and 426, are not necessary for phage production.

Bacterial strains, phage and plasmids

LM2489 is a rough derivative of P. syringae pv. phaseolicola HB10Y (HB)[1] and was used as the primary host for plating Φ2954, Φ12 and Φ6.

Plasmid pLM1454 is a derivative of the cloning vector pT7T3 19U (GenBank: U13870.1). It was used for the cloning of cDNA copies of phage DNA produced by RTPCR.

Media

The media used were LC and M8 {Sinclair, 1976 #80}. Ampicillin plates contained 200 mg of ampicillin per ml in LC agar.

Enzymes and Chemicals

All restriction enzymes, T4 DNA ligase, T4 DNA polymerase, T4 polynucleotide kinase, Klenow enzyme, and Exonuclease BAL-31 were purchased from Promega, New England Biolabs and Boehringer Gmbh, Mannheim.

Preparation of pure virions of Φ2954

Bacteriophage Φ2954 was harvested from soft LB agar plates. The soft agar was spun at 7000 rpm for 10 minutes at 4°C. 0.5 M NaCl and 10% PEG-6000 was added the supernatant liquid to precipitate the phage. The suspension was centrifuged; the pellet was resuspended in 0.5 ml of buffer B overnight at 4°C. Buffer B is composed of 10 mM KHPO₄, 1 mM MgCl₂ and 200 mM NaCl, pH 7.5.

The resuspended Φ2954 was then spun at 28,000 rpm for 70 minutes in a zone gradient of 10-30% Renocal in 200 mM Tris-HCl pH8, 200 mM NaCl, 1 mM MgCl2. The phage band was isolated and treated with PEG to precipitate the virions. The pellet was resuspended in 30 μl of the Tris buffer and extracted with phenol, ethanol precipitated and resuspended in 5 μl of DNA buffer.

Preparation of cDNA. PolyA tailing

The RNA was denatured by boiling for 5 minutes and rapidly cooling with dry ice/ethanol. 5× polyA-polymerase buffer was added to the RNA along with ATP and yeast polyA polymerase (Amersham). The mixture was incubated at 30°C for 1 minute and transferred to ice and the reaction stopped with EDTA. The polyA-RNA was then extracted with phenol/chloroform and precipitated and resuspended in water.

First strand synthesis

1 μl of phosphorylated oligo dT was added to 10 μl of polyA-RNA. After 5 minutes at 70°C the sample was cooled on ice for 5 minutes. Then 4 μl of 5× first strand buffer, 3 μl H₂O, 40 u RNase inhibitor (RNasin) and 30 u AMV reverse transcriptase was added and incubated at 42°C for 1 hour. All products needed for the first and second strand synthesis were provided by the Promega cDNA kit (Universal Riboclone cDNA Synthesis System). The reaction products were stored at -70°C overnight.

Second strand synthesis

After thawing the reverse transcribed RNA, 40 μl 2.5 × second strand buffer, 37.6 μl H₂O, 0.8 u RNaseH and 23 u E. coli DNA polymerase I was added. After the second strand synthesis proceeded for 3 hours at 16°C, the E. coli DNA polymerase I was inactivated at 70°C for 10 minutes. Then T4 DNA polymerase was added for 10 minutes at 37°C to blunt the ends of the cDNA. The sample was then treated with phenol/chloroform, ethanol precipitated and resuspended in 2.5 μl H₂O.

Preparation of the vector used for cloning

pLM1454 was cut with HincII, dephosphorylated with shrimp alkaline phosphatase and then purified by electrophoresis, electroeluted, precipitated and resuspended in 20 μl TE buffer. The ligation mixture was composed of 2.5 μl Φ2954 cDNA, 0.5 μl vector, 0.5 μl 10 × ligation buffer, 0.5 μl 10 mM ATP and 2.5 u T4 DNA ligase. All products are provided by the Promega cDNA kit. Incubation was overnight at16°C. The ligation mixture was used to transform super competent Epicurean E. coli (Stratagene). The cells were resuspended in 100 μl SOC medium and plated out on LC plates with 40 μg/ml X-gal (5-bromo-4-chloro-3-indolyl-beta-D-galactopyranoside) and 200 μg/ml Ampicillin. White colonies were picked and small DNA preparations were made. The plasmids were cut with restriction enzyme PvuII and promising candidates were sequenced first with M13 primers and then with oligonucleotides prepared on the basis of the sequence found. At the point where it seemed that the ends of the segments were identified, we prepared cDNA copies by using RTPCR with oligonucleotides having sequences found in the first copies found. Sequencing was done at the New Jersey Medical School Sequencing Facility. The sequences of segments L, M and S were deposited in GenBank with respective accession numbers of [GenBank: FJ608823, FJ608824 and FJ608825].

Preparation of complete cDNA plasmids

The cDNA pieces were assembled to form complete copies of the three genomic segments. In many cases, the connections could be made by using unique restriction sites made evident by the sequencing project. The ends of segments were prepared by using oligonucleotides with convenient restriction sites as primers for PCR reactions. Five plasmids were prepared, pLM3496, pLM3497, pLM3697, pLM3698 and pLM3691. They contain exact complete copies of genomic segments S and M in plasmid pT7T3 19U and three variants of segment L sequence. The sequences start at the first nucleotide of the SP6 RNA polymerase transcript.

In vitro transcription with nucleocapsids

Nucleocapsids of Φ2954 were prepared from purified virions stripped of their lipid-containing membranes by treatment with two percent Triton X-100 [17]. Transcription was performed in magnesium buffers [18, 19]. Labeling was with α-³²P-UTP and products were analyzed by electrophoresis in agarose gels.