- Email: [email protected]
Reverse Genetics of dsRNA Bacteriophage Φ6
ADVANCES IN VIRUS RESEARCH, VOL 53
REVERSE GENETICS OF dsRNA BACTERIOPHAGE a6 Leonard Mindich Department of Microbiology The Public Health Research Institute New York. New York 10016
I. 11. 111. IV. V.
Introduction Life Cycle of Q6 Infection In Vitro Genomic Packaging with Bacteriophage Q6 Reverse Genetics by in Vim Transcript Acquisition Conclusion References
I. INTRODUCTION a 6 is a bacteriophage that infects the plant pathogen Pseudomonas syringae pv phaseolicola (Mindich, 1988; Vidaver et al., 1973). It has a unique structure and life cycle in that it is the only bacteriophage with a genome of dsRNA and the only phage that has a lipid-containing membrane that surrounds a polyhedral nucleocapsid (Mindich, 1988; Fig. 1). The virion contains three segments of dsRNA. The packaging of RNA by the virus is extremely precise, with each particle having one of each of the three genomic segments. The efficiency of plating of the virus is very close to 1 (Day and Mindich, 1980). Each genomic segment contains four or five different genes packed rather closely together, with noncoding regions of several hundred bases a t the 5’ and 3’ termini. The segments are designated L, M, and S , having 6374, 4063, and 2948 bp, respectively (Fig. 2). The L segment contains five genes, termed 14,7,2,4,and 1,ofwhich four code for proteins necessary for RNA synthesis. Gene 14 is probably involved in control of P7 synthesis. The proteins P1, P2, P4, and P7 form a polyhedral procapsid (Fig. 3) (Butcher et al., 1997; Mindich and Abelson, 19801, and these four proteins comprise the polymerase complex of the virus. Protein P2 shows sequence similarity t o other viral RNA polymerases (Koonin et aZ., 1989). P4 is a n NTPase (Gottlieb et al., 19921, and P7 plays an accessory role in genomic packaging and polymerization (Van Dijk et al., 1995; Gottlieb et al., 1990). P1 constitutes the primary structural framework of the procapsid. The procapsid is enclosed in a shell composed of a single protein, P8 (Fig. 1). Oiltside of the P8 shell is an external lipid-containing membrane that carries the proteins necessary 341
Copyright C) 1999 by Academic Press. All rights of reproduction in any form resewed. 0065-3627/99 $30.00
342
LEONARD MINDICH
FIG1. [For color reproduction, see color section.] Diagram of the structure of @6. The three strands of genomic dsRNA are found inside a dodecahedra1 procapsid structure composed of proteins P1, P2, P4, and P7. This is enclosed in a shell of protein P8. Outside of this structure is the membrane, composed of phospholipids and proteins P9, P10, P13, P6, and P3. P5 is a lytic endopeptidase (Caldentey and Bamford, 1992; Mindich, 1988).
for attachment to the pilus of the host organism (Fig. 1).The genome of the virus has been cloned and sequenced (Mindich, 1988). When stripped of its membrane, the virion has strong transcriptase activity, and it is easy to prepare quantities of full-length transcripts of the genomic segments (Emori et al., 1983; Partridge et al., 1979). cDNA copies of the genomic segments have been obtained and inserted into expression vectors. The proteins of the phage can be produced in the normal host strain or in Escherichia coli. Exact transcripts of the cDNA can also be prepared with the use of T7 promoter plasmids (Qiao et al., 1997).
11. LIFECYCLEOF (D6 INFECTION The life cycle of (D6 is shown diagrammatically in Fig. 4. The virus attaches to the host cell pilus; the pilus retracts, bringing the virion into contact with the outer membrane of the host; the viral membrane then fuses with the outer membrane, resulting in the entry of the nucleocapsid into the periplasmic space. A hydrophobic endopeptidase makes a hole in the murein layer, and the nucleocapsid enters the cell, where it loses protein P8 and starts to transcribe (Romantschuk et al., 1988). Translation is primarily limited to the large genomic segment that codes for proteins P1, P2, P4, and P7. These proteins assemble to
343
REVERSE GENETICS OF dsRNA PHAGE (P6
scale (kbp)
1
2
3
4
5
6
I
I
I
I
I
I
gene 2
gene 14 gene 7
gene 4
gene 1
segment M 4063 bp gene10
gene6
gene 3
gene 13
segment S 2948 bp
gene 8
gene 9 gene 12
gene 5
FIG2. Restriction map of the cDNA copies of the genomic segments of @6. Genes are identified below the segments (Casini and Revel, 1994; Mindich, 1988).The sequences necessary for packaging are located in the 5' noncoding regions (left). The sequences necessary for minus-strand synthesis are a t the 3' ends.
form the procapsid (Mindich and Abelson, 1980; Mindich et al., 1988; Sinclair et al., 19751, which then packages and replicates the RNA. Late in infection, transcription and translation of the S and M genomic segments are greatly amplified (Coplin et al., 1976). Then protein P8 covers the filled procapsids, and a membrane is formed that envelopes the nucleocapsid. Cell lysis liberates about 300 particles per cell. The efficiency of plating of the virion is close to 1,and each particle contains one each of the three genomic segments.
344
LEONARD MINDICH
FIG3. Electron micrograph of procapsids isolated from a strain of E. coli carrying a plasmid with a cDNA copy of segment L. The sample is stained with ammonium molybdate (Gottlieb et al., 1990).
111. IN Vim0 GENOMIC PACKAGING WITH BACTERIOPHAGE @6 The expression of genes 1 , 2 , 4 , and 7 in E. coli results in the production of the corresponding proteins and their assembly into empty pro-
REVERSE GENETICS OF dsRNA PHAGE @6
345
FIG4. [For color reproduction, see color section.] The life cycle of @6.The virion attaches to a pilus and is brought into contact with the outer membrane (om).The viral membrane fuses with the outer membrane to place the nucleocapsid in the periplasmic space. The murein (cw) is digested by viral lysozyme P5, and the filled procapsid (fpc) penetrates the inner membrane (im)and enters the cell, leaving P8 behind. The procapsid transcriptase synthesizes complete copies of the three genomic segments. The L message is translated t o produce P1, P2, P4, and P7, which constitute the procapsid (pc). This is filled with dsRNA and continues transcription until it is covered by P8 t o form nucleocapsids (nc). Membrane proteins are placed in the host membrane and then transferred to the virion, (mv) along with host lipids. The membrane formation or translocation is dependent on protein P12.
capsids (Gottlieb et al., 1988). The purified procapsids are capable, in uitro, of packaging plus-strand copies of the genomic segments. The packaging is dependent on nucleoside triphosphate and is not linked to minus-strand synthesis (Gottlieb et al., 1990). In the presence of all four nucleoside triphosphates, minus-strand synthesis takes place to form dsRNA within the procapsids (Gottlieb et al., 1990). Packaging is serially dependent in that segment S is packaged first; segment M packaging depends on prior packaging of S, and segment L packaging depends on prior packaging of segment M (Qiao et al., 1997).We have proposed and supported a model for packaging in which the binding sites for the genomic segments appear on the outside of the procapsid in a programmed sequence that depends on the amount of RNA within the procapsid (Fig. 5). Thus, segment S is packaged by the empty procapsid by binding to its recognition site and positioning its 5’ end so that it enters one of many entry portals. The binding site for segment M appears only when segment S has been packaged. At that time the binding site for S is lost. Similarly, the binding site for L appears and the
LEONARD MINDICH
346 segment S (3kb)
segment M (4.lkb)
".-
segment L (6.4kb) 4%
+
i 4
3
ii"
P ,J,
empty procapsid, ready for segment S
S Is packaged, ready
for segment M
+
M is packaged, ready for segment L
L Is packaged, ready for minus strand synthesis
FIG5. [For color reproduction, see color section.] Diagram of the packaging model. The procapsid shows only binding sites for S at the beginning. After a full size S is packaged, the S sites disappear and M sites appear. After a full size M is packaged, the M sites disappear and L sites appear. After a full size L is packaged, minus-strand synthesis commences. After minus-strand synthesis is completed, plus-strand synthesis commences.
binding site for M is lost after packaging of the M segment. Packaging is dependent on a sequence ( p a c ) of about 200 nucleotides near the 5' ends of the plus strands (Gottlieb et al., 1994). These pac sequences are unique for each of the three segments, and along with the 5' termini, they are necessary and sufficient for packaging. Each plus strand has a sequence of 18 nucleotides at the very 5' end that is identical for all three segments with the exception of the second nucleotide. This sequence is necessary for packaging, and the distance between this sequence and the pac sequence is critical (Qiao et al., 1997). All three segments share a sequence of high similarity at the 3' ends. These sequences fold into a series of hairpins that seem to play a role in stabilizing the genome (Mindich et al., 1994). They are not segment specific and can be exchanged between the segments. The terminal nucleotide sequence CUCUCUCUCU is necessary for polymerase recognition but plays no role in packaging. Most of the genes of a 6 are closely packed on the genomic segments, and there are noncoding regions at the 5' and 3' ends (Mindich and Bamford, 1988). There is no overlap of pac sequences and open reading frames. A consequence of the precise packaging of the @6 genome is that there are limits to the amount of extra RNA that the virus can accommodate and limits on the amount of extra RNA that individual segments can handle.
REVERSE GENETICS OF dsRNA PHAGE (P6
347
Procapsids that have packaged RNA in vitro can be used to infect host protoplasts and produce viable phage (Olkkonen et al., 1990). This has made it possible to study the replication of the genome in great detail and to prepare useful derivatives of the genome by reverse genetics. We prepare plasmid transcripts that have correct 5' and 3' ends as well as the proper pac sites for the three genomic segments. These RNA molecules are packaged by procapsids that have been isolated from E. coli cultures that carry a plasmid with the genes for proteins P1, P2, P4, and P7. The packaged RNA is copied inside the procapsids to make dsRNA. The particles are then treated with purified protein P8 in the presence of calcium to form a shell around the procapsids (Ojalaet al., 1990; Olkkonenet aZ., 1990,1991).The resulting nucleocapsids are capable of infecting spheroplasts of P. syringae. The feasibility of in vitro packaging for reverse genetics depended on the finding in the laboratory of Dennis Bamford that nucleocapsids of a6 are capable of infecting spheroplasts of the host organism (Ojala et al., 1990). In addition, these investigators showed that the P8 shell could be removed in the presence of EGTA and replaced in the presence of calcium. In the absence of the P8 shell, infection does not take place (Olkkonen et al., 1991). It was then rather straightforward to apply purified P8 to procapsids that had packaged RNA in vitro and to show that they could infect spheroplasts (Olkkonen et al., 1990). We consistently found that the yield of plaques was much greater when plus-strand RNA for packaging was produced from transcription reactions using nucleocapsids of purified virus as compared to T7 RNA polymerase transcripts of cDNA plasmids. It appears that the T7 polymerase does not have as much fidelity a s the viral transcriptase. Because of this finding, we often used viral transcripts from nonsense mutants of @6for our preparations, along with a cDNA transcript for the segment that we wished to replace (Johnson and Mindich, 1994). The cDNA transcript competed for packaging with the segment containing the nonsense mutation. Any procapsids that packaged the RNA with the nonsense mutation did not form plaques. Alternatively, we can use plus-strand transcripts from wild-type virus, along with a T7 polymerase transcript of a plasmid construction that contains the desired mutation together with a ZacH gene (Casini and Revel, 1996; Johnson and Mindich, 1994).lacH codes for a portion of the N terminus of P-galactosidase that is somewhat larger than the original a! fragment. This is necessary for a! complementation in pseudomonads. Blue plaques are picked and tested for the desired properties. At the beginning of our work with reverse genetics, we were concerned about the fidelity of our cDNA copies of the genomic segments. Fortunately, we
348
LEONARD MINDICH
were able to test the function of most of the cloned genes by their ability to complement nonsense and ts (temperature-sensitive)mutants of (P6. We found that the process of cDNA cloning did not introduce many mutations that would compromise gene function. Once the transfectionl packaging approach was working, it was clear that all the cloned genes were functional. A gene for kanamycin resistance was inserted into the 3’ noncoding region of segment M (Onodera et al., 1992). This insert is 1.2 kb long, and it is rather stable. Inserts of about 400 bases containing the a! portion of p-galactosidase (La&) have been inserted into the 3‘ noncoding regions of all three segments, and these are very stable (Onodera et al., 1993).If the inserts are bounded by complementary homopolymer sequences that result in hairpin stems of more than 20 bases, the inserts are genetically unstable and are eliminated by copy choice exchange with the 3’ ends of other segments. This instability is due to the hairpin stem’s preventing the entry of the 3’ end of the RNA into the entry portal of the procapsid (Mindich et al., 1992; Onodera et al., 1993). As a consequence, the procapsid is rescued by a heterologous recombination that leads to the copy choice donation of the 3’ end of one of the other segments, resulting in the loss of the reporter gene. Inserts that are larger than 1.2 kb are lost from segment M in the absence of selection even if they have no secondary structure. The same is true for inserts larger than about 500 bases in segments S and L. Larger inserts can be maintained if selection persists. In this manner we have prepared phage with 3 kbp of extra RNA. IV. REVERSEGENETICS BY IN
VIVO TRANSCRIPT ACQUISITION
The second method of reverse genetics is done i n vivo. In this case we prepare a culture of P. syringae that is carrying a plasmid that produces a transcript with imprecise ends in which is embedded a complete copy of one of the genomic segments. These cells are infected with a derivative of (P6 that contains a deletion in the segment to be exchanged. We were able to prepare virus with nonreverting deletions by the i n uitro packaging method (Onodera et al., 1995; Fig. 6). This involved packaging two normal segments plus one containing the desired deletion and transfecting the particles into cells carrying a plasmid that would complement the missing genes. Once we had stocks of the nonreverting deletions, it was possible to develop the transcript acquisition procedure, which is much easier to carry out. If a deletion mutant was
REVERSE GENETICS OF dsRNA PHAGE a6
349
FIG6. Agarose gel electrophoresis of dsRNA isolated from virions. Lane v shows the distribution of normal segments L, M, and S. Lane a shows dsRNA from bacteriophage Q2007 that has a deletion in segment M. Lane b shows RNAfrom a2064 that has normal L, a n MS chimera picked up from plasmid pLM1114, and a normal segment S. Lane c shows RNA from a2323 that contains normal L, the MS chimera shown in h, and a deleted segment S that contains no genes and is only 798 bp. Lane d shows RNA from Q2361 that contains normal L and a chimera of S and M but no normal segment M or S.
plated on the normal host strain, there would be no plaques unless the host carried a plasmid that expressed the missing gene or genes. If the plasmid produced a transcript in which was embedded a complete copy of the segment that suffered the deletion, the phage could pick up this transcript and incorporate it into the genome, replacing the deletion segment with the new modification. Neither the 5' nor the 3' end of the transcript had to be tailored correctly in order for the acquisition to occur at high frequency. In our working model, the transcript was modified by cellular enzymes before being packaged because we have
350
LEONARD MINDICH
found that in uitro packaging of RNA is very stringent with respect to extra nucleotides at the 5’ end and minus strand synthesis is very (but not absolutely) stringent in terms of the sequence at the 3’ end. In many of the eukaryotic viral reverse systems that use transcript acquisition, it is necessary to engineer ribozyme sequences onto the transcripts so that the ends will be identical to the viral ends. In the case of @6,this is not necessary for in uiuo acquisition. We have even found that pickup can take place when the normal 3’ end of the viral sequence is not embedded in the transcript. In these cases, a proper 3’ end is supplied by copy choice recombination, with the use of one of the other segments as the donor. The frequency is diminished by several orders of magnitude in these cases. At any rate, the 5’ and 3’ ends of the acquired transcripts are identical to those of normal genomic segments. The frequency of pickup can be very high despite the fact that the transcript must be modified at both the 5’ and 3’ ends (Onodera et aZ., 1995,1998).As much as 10%of the total phage has been found to carry the acquired transcript. These levels, however, do not necessarily reflect the actual frequency of pickup. It seems likely that the phage that has acquired the new transcript may have a replication advantage over the phages that are being complemented. We have used the in uiuo transcript acquisition method for most of our recent work in reverse genetics. It is very easy to isolate the desired constructions because there is no background from the deletion mutants. We have used this method for the insertion of reporter groups in the 3’ noncoding regions and also for the construction of chimeras of the genomic segments. We have produced a phage with only two genomic segments by joining segment S and segment M together such that the chimeric molecule has the pac site of S (Fig. 6). This construct is stable and appears normal. We have also prepared a virus in which all three segments are joined together into one genomic segment of 14 kbp. In this case, we did not need t o use a deletion phage, as the cells carrying the plasmid produced plaques on their own lawn without the participation of helper phage.
V. CONCLUSION Using these two methods, we have been able to insert novel genes into the 3‘ noncoding regions of the three segments. We have placed genes for kanamycin resistance (aph),for the a portion of P-galactosidase (ZacH),and for GFP (gfp).We have been able to engineer nonsense mutations in genes (e.g., genes 10 and 14) to assess their function
REVERSE GENETICS OF dsRNA PHAGE (P6
351
when we were unable to isolate nonsense mutations by screening after chemical mutagenesis (Casini and Revel, 1996; Johnson and Mindich, 1994). We have also been able t o prepare virus containing genomic segments with small or large deletions or virus with segments joined together so that the number of genomic segments can be two or even one (Onodera et al., 1998). In the latter case, we have prepared a live virus that has the entire genome in a segment of about 14 kbp (Onodera et al., 1998). Although there are significant differences in the structure and mode of replication of (D6 and the members of the Reoviridae, it is possible that the mechanisms found in (D6 may also operate in the eukaryotic dsRNA viruses. ACKNOWLEDGMENT This work is supported by Grants GM31709 and GM34352 from the National Institutes of Health to L.M.
REFERENCES Butcher, S. J., Dokland, T., Ojala, P. M., Bamford, D. H., and Fuller, S. D. (1997l.Intermediates in the assembly pathway of the double-stranded RNA virus 46. EMBO J. 16,4477-4487. Caldentey, J.,and Bamford, D. H. (1992).The lytic enzyme of the Pseudomonas phage @6. Purification and biochemical characterization. Biochim. Biophys. Acta 1159, 44-50. Casini, G., and Revel, H. R. (1994). A new small low-abundant nonstructural protein encoded by the L segment of the dsRNA bacteriophage Q6. Virology 203, 221-228. Casini, G., and Revel, H. R. (1996). Construction and analysis of a bacteriophage @6 gene 14 nonsense mutant. Virology 216, 455-458. Coplin, D. L., Van Etten, J. L., and Vidaver, A. K. (1976). Synthesis of bacteriophage 4% double-stranded ribonucleic acid. J. Gen. Virol. 33, 509-512. Day, L. A,, and Mindich, L. (1980). The molecular weight of bacteriophage @6 and its nucleocapsid. Virology 103, 376-385. Emori, Y., Iba, H., and Okada, Y. (1983). Transcriptional regulation of three doublestranded RNA segments of bacteriophage (1)6 in vitro. J . Virol. 46, 196-203. Gottlieb, P., Strassman, J., Bamford, D. H., and Mindich, L. (1988). Production of a polyhedral particle in E. coli from a cDNA copy of the large genomic segment of bacteriophage @6. J . Virol. 62, 181-187. Gottlieb, P., Strassman, J., Qiao, X., Frucht, A.,and Mindich, L. (1990). In vitro replication, packaging and transcription of the segmented dsRNA genome of bacteriophage 06: Studies with procapsids assembled from plasmid encoded proteins. J . Bacteriol. 172, 5774-5782. Gottlieb, P., Strassman, J., and Mindich, L. (1992). Protein P4 of the bacteriophage @6 procapsid has a nucleoside triphosphate-binding site with associated nucleoside triphosphate phosphohydrolase activity. J . Virol. 66, 622045222. Gottlieb, P., Qiao, X., Strassman, J., Frilander, M., and Mindich, L. (1994). Identification of the packaging regions within the genomic RNA segments of bacteriophage @6. Virology 200, 42-47.
352
LEONARD MINDICH
Johnson, M. D., and Mindich, L. (1994). Isolation and characterization of nonsense mutants in gene 10 of bacteriophage @6. J . Virol. 68, 2331-2338. Koonin, E. V., Gorbalenya, E. E., and Chumakov, K. M. (1989). Tentative identification of RNA-dependent RNA polymerases of dsRNA viruses and their relationship to positive strand RNA viral polymerases. FEBS Lett. 252, 42-46. Mindich, L. (1988). Bacteriophage @6: A unique virus having a lipid-containing membrane and a genome composed of three dsRNA segments. Adu. Virus Res. 35,137-176. Mindich, L., and Abelson, R. D. (1980). The characterization of a 120 S particle formed during a 6 infection. Virology 103, 386-391. Mindich, L., and Bamford, D. H. (1988). Lipid-containing bacteriophages. In “The Bacteriophages” (R. Calendar, ed.), Vol. 2. pp. 475-520. Plenum, New York and London. Mindich, L., Nemhauser, I., Gottlieb, P., Romantschuk, M., Carton, J.,Frucht, S., Strassman, J., Bamford, D. H., and Kalkkinen, N. (1988). Nucleotide sequence of the large dsRNA segment of bacteriophage 0 6 : The genes specifying the viral replicase and transcriptase. J. Virol. 62, 1180-1185. Mindich, L., Qiao, X., Onodera, S., Gottlieb, P., and Strassman, J . (1992). Heterologous recombination in the dsRNA bacteriophage @6.J. Virol. 66, 2605-2610. Mindich, L., Qiao, X., Onodera, S., Gottlieb, P., and Frilander, M. (1994). RNA structural requirements for stability and minus strand synthesis in the dsRNA bacteriophage @6. Virology 202, 258-263. Ojala, P. M., Romantschuk, M., and Bamford, D. H. (1990). Purified @6nucleocapsids are capable of productive infection of host cells with partially disrupted outer membrane. Virology 178, 364-372. Olkkonen, V. M., Gottlieb, P., Strassman, J., Qiao, X., Bamford, D. H., and Mindich, L. (1990). In vitro assembly of infectious nucleocapsids of bacteriophage @6: Formation of a recombinant double-stranded RNA virus. Proc. Natl. Acad. Sci. U.S.A.87, 91739177. Olkkonen, V. M., Ojala, P., and Bamford, D. H. (1991).Generation of infectious nucleocapsids by in vitro assembly of the shell protein onto the polymerase complex of the dsRNA bacteriophage @6.J . Mol. Biol. 218, 569-581. Onodera, S., Olkkonen, V. M., Gottlieb, P., Strassman, J., Qiao, X., Bamford, D. H., and Mindich, L. (1992). Construction of a transducing virus from dsRNA bacteriophage (1.16:establishment of carrier states in host cells, J. Virol. 66, 190-196. Onodera, S., Qiao, X., Gottlieb, P., Strassman, J., Frilander, M., and Mindich, L. (1993). RNA structure and heterologous recombination in the dsRNA bacteriophage @6. J. Virol. 67, 4914-4922. Onodera, S., Qiao, X., Qiao, J., and Mindich, L. (1995). Acquisition of a fourth genomic segment in bacteriophage 96: A bacteriophage with a genome of three segments of dsRNA. Virology 212, 204-212. Onodera, S., Qiao, X., Qiao, J., and Mindich, L. (1998). Directed changes in the number of dsRNA genomic segments in bacteriophage @6. Proc. Nut. Acad. Sci. U.S.A. 95, 3920-3924. Partridge, J. E., Van Etten, J . L., Burbank, D. E., and Vidaver, A. K. (1979). RNA polymerase activity associated with bacteriophage 9 6 nucleocapsid. J. Gen. Virol. 43, 299-307. Qiao, X., Qiao, J., and Mindich, L. (19971. Stoichiometric packaging of the three genomic segments of dsRNA bacteriophage @6.Proc. Natl. Acad. Sci. U S A . 94, 4074-4079. Romantschuk, M., Olkkonen, V. M., and Bamford, D. H. (1988). The nucleocapsid of bacteriophage (P6 penetrates the host cytoplasmic membrane. EMBO J , 7,1821-1829.
REVERSE GENETICS OF dsRNA PHAGE @6
353
Sinclair, J. F., Tzagoloff, A., Levine, D., and Mindich, L. (1975).Proteins of bacteriophage Q6. J . Viral. 16, 685-695. Van Dijk, A. A., Frilander, M., and Bamford, D. H. (1995).Differentiation between minusand plus-strand synthesis: Polymerase activity of dsRNA bacteriophage 06 in a n in vitro packaging and replication system. Virology 211, 320-323. Vidaver, A. K., Koski, R. K., and Van Etten, J. L. (1973). Bacteriophage @6: A lipidcontaining virus of Pseudomonas phaseolicola. J. Viral. 11, 799-805.
REVERSE GENETICS OF dsRNA BACTERIOPHAGE a6 Leonard Mindich Department of Microbiology The Public Health Research Institute New York. New York 10016
I. 11. 111. IV. V.
Introduction Life Cycle of Q6 Infection In Vitro Genomic Packaging with Bacteriophage Q6 Reverse Genetics by in Vim Transcript Acquisition Conclusion References
I. INTRODUCTION a 6 is a bacteriophage that infects the plant pathogen Pseudomonas syringae pv phaseolicola (Mindich, 1988; Vidaver et al., 1973). It has a unique structure and life cycle in that it is the only bacteriophage with a genome of dsRNA and the only phage that has a lipid-containing membrane that surrounds a polyhedral nucleocapsid (Mindich, 1988; Fig. 1). The virion contains three segments of dsRNA. The packaging of RNA by the virus is extremely precise, with each particle having one of each of the three genomic segments. The efficiency of plating of the virus is very close to 1 (Day and Mindich, 1980). Each genomic segment contains four or five different genes packed rather closely together, with noncoding regions of several hundred bases a t the 5’ and 3’ termini. The segments are designated L, M, and S , having 6374, 4063, and 2948 bp, respectively (Fig. 2). The L segment contains five genes, termed 14,7,2,4,and 1,ofwhich four code for proteins necessary for RNA synthesis. Gene 14 is probably involved in control of P7 synthesis. The proteins P1, P2, P4, and P7 form a polyhedral procapsid (Fig. 3) (Butcher et al., 1997; Mindich and Abelson, 19801, and these four proteins comprise the polymerase complex of the virus. Protein P2 shows sequence similarity t o other viral RNA polymerases (Koonin et aZ., 1989). P4 is a n NTPase (Gottlieb et al., 19921, and P7 plays an accessory role in genomic packaging and polymerization (Van Dijk et al., 1995; Gottlieb et al., 1990). P1 constitutes the primary structural framework of the procapsid. The procapsid is enclosed in a shell composed of a single protein, P8 (Fig. 1). Oiltside of the P8 shell is an external lipid-containing membrane that carries the proteins necessary 341
Copyright C) 1999 by Academic Press. All rights of reproduction in any form resewed. 0065-3627/99 $30.00
342
LEONARD MINDICH
FIG1. [For color reproduction, see color section.] Diagram of the structure of @6. The three strands of genomic dsRNA are found inside a dodecahedra1 procapsid structure composed of proteins P1, P2, P4, and P7. This is enclosed in a shell of protein P8. Outside of this structure is the membrane, composed of phospholipids and proteins P9, P10, P13, P6, and P3. P5 is a lytic endopeptidase (Caldentey and Bamford, 1992; Mindich, 1988).
for attachment to the pilus of the host organism (Fig. 1).The genome of the virus has been cloned and sequenced (Mindich, 1988). When stripped of its membrane, the virion has strong transcriptase activity, and it is easy to prepare quantities of full-length transcripts of the genomic segments (Emori et al., 1983; Partridge et al., 1979). cDNA copies of the genomic segments have been obtained and inserted into expression vectors. The proteins of the phage can be produced in the normal host strain or in Escherichia coli. Exact transcripts of the cDNA can also be prepared with the use of T7 promoter plasmids (Qiao et al., 1997).
11. LIFECYCLEOF (D6 INFECTION The life cycle of (D6 is shown diagrammatically in Fig. 4. The virus attaches to the host cell pilus; the pilus retracts, bringing the virion into contact with the outer membrane of the host; the viral membrane then fuses with the outer membrane, resulting in the entry of the nucleocapsid into the periplasmic space. A hydrophobic endopeptidase makes a hole in the murein layer, and the nucleocapsid enters the cell, where it loses protein P8 and starts to transcribe (Romantschuk et al., 1988). Translation is primarily limited to the large genomic segment that codes for proteins P1, P2, P4, and P7. These proteins assemble to
343
REVERSE GENETICS OF dsRNA PHAGE (P6
scale (kbp)
1
2
3
4
5
6
I
I
I
I
I
I
gene 2
gene 14 gene 7
gene 4
gene 1
segment M 4063 bp gene10
gene6
gene 3
gene 13
segment S 2948 bp
gene 8
gene 9 gene 12
gene 5
FIG2. Restriction map of the cDNA copies of the genomic segments of @6. Genes are identified below the segments (Casini and Revel, 1994; Mindich, 1988).The sequences necessary for packaging are located in the 5' noncoding regions (left). The sequences necessary for minus-strand synthesis are a t the 3' ends.
form the procapsid (Mindich and Abelson, 1980; Mindich et al., 1988; Sinclair et al., 19751, which then packages and replicates the RNA. Late in infection, transcription and translation of the S and M genomic segments are greatly amplified (Coplin et al., 1976). Then protein P8 covers the filled procapsids, and a membrane is formed that envelopes the nucleocapsid. Cell lysis liberates about 300 particles per cell. The efficiency of plating of the virion is close to 1,and each particle contains one each of the three genomic segments.
344
LEONARD MINDICH
FIG3. Electron micrograph of procapsids isolated from a strain of E. coli carrying a plasmid with a cDNA copy of segment L. The sample is stained with ammonium molybdate (Gottlieb et al., 1990).
111. IN Vim0 GENOMIC PACKAGING WITH BACTERIOPHAGE @6 The expression of genes 1 , 2 , 4 , and 7 in E. coli results in the production of the corresponding proteins and their assembly into empty pro-
REVERSE GENETICS OF dsRNA PHAGE @6
345
FIG4. [For color reproduction, see color section.] The life cycle of @6.The virion attaches to a pilus and is brought into contact with the outer membrane (om).The viral membrane fuses with the outer membrane to place the nucleocapsid in the periplasmic space. The murein (cw) is digested by viral lysozyme P5, and the filled procapsid (fpc) penetrates the inner membrane (im)and enters the cell, leaving P8 behind. The procapsid transcriptase synthesizes complete copies of the three genomic segments. The L message is translated t o produce P1, P2, P4, and P7, which constitute the procapsid (pc). This is filled with dsRNA and continues transcription until it is covered by P8 t o form nucleocapsids (nc). Membrane proteins are placed in the host membrane and then transferred to the virion, (mv) along with host lipids. The membrane formation or translocation is dependent on protein P12.
capsids (Gottlieb et al., 1988). The purified procapsids are capable, in uitro, of packaging plus-strand copies of the genomic segments. The packaging is dependent on nucleoside triphosphate and is not linked to minus-strand synthesis (Gottlieb et al., 1990). In the presence of all four nucleoside triphosphates, minus-strand synthesis takes place to form dsRNA within the procapsids (Gottlieb et al., 1990). Packaging is serially dependent in that segment S is packaged first; segment M packaging depends on prior packaging of S, and segment L packaging depends on prior packaging of segment M (Qiao et al., 1997).We have proposed and supported a model for packaging in which the binding sites for the genomic segments appear on the outside of the procapsid in a programmed sequence that depends on the amount of RNA within the procapsid (Fig. 5). Thus, segment S is packaged by the empty procapsid by binding to its recognition site and positioning its 5’ end so that it enters one of many entry portals. The binding site for segment M appears only when segment S has been packaged. At that time the binding site for S is lost. Similarly, the binding site for L appears and the
LEONARD MINDICH
346 segment S (3kb)
segment M (4.lkb)
".-
segment L (6.4kb) 4%
+
i 4
3
ii"
P ,J,
empty procapsid, ready for segment S
S Is packaged, ready
for segment M
+
M is packaged, ready for segment L
L Is packaged, ready for minus strand synthesis
FIG5. [For color reproduction, see color section.] Diagram of the packaging model. The procapsid shows only binding sites for S at the beginning. After a full size S is packaged, the S sites disappear and M sites appear. After a full size M is packaged, the M sites disappear and L sites appear. After a full size L is packaged, minus-strand synthesis commences. After minus-strand synthesis is completed, plus-strand synthesis commences.
binding site for M is lost after packaging of the M segment. Packaging is dependent on a sequence ( p a c ) of about 200 nucleotides near the 5' ends of the plus strands (Gottlieb et al., 1994). These pac sequences are unique for each of the three segments, and along with the 5' termini, they are necessary and sufficient for packaging. Each plus strand has a sequence of 18 nucleotides at the very 5' end that is identical for all three segments with the exception of the second nucleotide. This sequence is necessary for packaging, and the distance between this sequence and the pac sequence is critical (Qiao et al., 1997). All three segments share a sequence of high similarity at the 3' ends. These sequences fold into a series of hairpins that seem to play a role in stabilizing the genome (Mindich et al., 1994). They are not segment specific and can be exchanged between the segments. The terminal nucleotide sequence CUCUCUCUCU is necessary for polymerase recognition but plays no role in packaging. Most of the genes of a 6 are closely packed on the genomic segments, and there are noncoding regions at the 5' and 3' ends (Mindich and Bamford, 1988). There is no overlap of pac sequences and open reading frames. A consequence of the precise packaging of the @6 genome is that there are limits to the amount of extra RNA that the virus can accommodate and limits on the amount of extra RNA that individual segments can handle.
REVERSE GENETICS OF dsRNA PHAGE (P6
347
Procapsids that have packaged RNA in vitro can be used to infect host protoplasts and produce viable phage (Olkkonen et al., 1990). This has made it possible to study the replication of the genome in great detail and to prepare useful derivatives of the genome by reverse genetics. We prepare plasmid transcripts that have correct 5' and 3' ends as well as the proper pac sites for the three genomic segments. These RNA molecules are packaged by procapsids that have been isolated from E. coli cultures that carry a plasmid with the genes for proteins P1, P2, P4, and P7. The packaged RNA is copied inside the procapsids to make dsRNA. The particles are then treated with purified protein P8 in the presence of calcium to form a shell around the procapsids (Ojalaet al., 1990; Olkkonenet aZ., 1990,1991).The resulting nucleocapsids are capable of infecting spheroplasts of P. syringae. The feasibility of in vitro packaging for reverse genetics depended on the finding in the laboratory of Dennis Bamford that nucleocapsids of a6 are capable of infecting spheroplasts of the host organism (Ojala et al., 1990). In addition, these investigators showed that the P8 shell could be removed in the presence of EGTA and replaced in the presence of calcium. In the absence of the P8 shell, infection does not take place (Olkkonen et al., 1991). It was then rather straightforward to apply purified P8 to procapsids that had packaged RNA in vitro and to show that they could infect spheroplasts (Olkkonen et al., 1990). We consistently found that the yield of plaques was much greater when plus-strand RNA for packaging was produced from transcription reactions using nucleocapsids of purified virus as compared to T7 RNA polymerase transcripts of cDNA plasmids. It appears that the T7 polymerase does not have as much fidelity a s the viral transcriptase. Because of this finding, we often used viral transcripts from nonsense mutants of @6for our preparations, along with a cDNA transcript for the segment that we wished to replace (Johnson and Mindich, 1994). The cDNA transcript competed for packaging with the segment containing the nonsense mutation. Any procapsids that packaged the RNA with the nonsense mutation did not form plaques. Alternatively, we can use plus-strand transcripts from wild-type virus, along with a T7 polymerase transcript of a plasmid construction that contains the desired mutation together with a ZacH gene (Casini and Revel, 1996; Johnson and Mindich, 1994).lacH codes for a portion of the N terminus of P-galactosidase that is somewhat larger than the original a! fragment. This is necessary for a! complementation in pseudomonads. Blue plaques are picked and tested for the desired properties. At the beginning of our work with reverse genetics, we were concerned about the fidelity of our cDNA copies of the genomic segments. Fortunately, we
348
LEONARD MINDICH
were able to test the function of most of the cloned genes by their ability to complement nonsense and ts (temperature-sensitive)mutants of (P6. We found that the process of cDNA cloning did not introduce many mutations that would compromise gene function. Once the transfectionl packaging approach was working, it was clear that all the cloned genes were functional. A gene for kanamycin resistance was inserted into the 3’ noncoding region of segment M (Onodera et al., 1992). This insert is 1.2 kb long, and it is rather stable. Inserts of about 400 bases containing the a! portion of p-galactosidase (La&) have been inserted into the 3‘ noncoding regions of all three segments, and these are very stable (Onodera et al., 1993).If the inserts are bounded by complementary homopolymer sequences that result in hairpin stems of more than 20 bases, the inserts are genetically unstable and are eliminated by copy choice exchange with the 3’ ends of other segments. This instability is due to the hairpin stem’s preventing the entry of the 3’ end of the RNA into the entry portal of the procapsid (Mindich et al., 1992; Onodera et al., 1993). As a consequence, the procapsid is rescued by a heterologous recombination that leads to the copy choice donation of the 3’ end of one of the other segments, resulting in the loss of the reporter gene. Inserts that are larger than 1.2 kb are lost from segment M in the absence of selection even if they have no secondary structure. The same is true for inserts larger than about 500 bases in segments S and L. Larger inserts can be maintained if selection persists. In this manner we have prepared phage with 3 kbp of extra RNA. IV. REVERSEGENETICS BY IN
VIVO TRANSCRIPT ACQUISITION
The second method of reverse genetics is done i n vivo. In this case we prepare a culture of P. syringae that is carrying a plasmid that produces a transcript with imprecise ends in which is embedded a complete copy of one of the genomic segments. These cells are infected with a derivative of (P6 that contains a deletion in the segment to be exchanged. We were able to prepare virus with nonreverting deletions by the i n uitro packaging method (Onodera et al., 1995; Fig. 6). This involved packaging two normal segments plus one containing the desired deletion and transfecting the particles into cells carrying a plasmid that would complement the missing genes. Once we had stocks of the nonreverting deletions, it was possible to develop the transcript acquisition procedure, which is much easier to carry out. If a deletion mutant was
REVERSE GENETICS OF dsRNA PHAGE a6
349
FIG6. Agarose gel electrophoresis of dsRNA isolated from virions. Lane v shows the distribution of normal segments L, M, and S. Lane a shows dsRNA from bacteriophage Q2007 that has a deletion in segment M. Lane b shows RNAfrom a2064 that has normal L, a n MS chimera picked up from plasmid pLM1114, and a normal segment S. Lane c shows RNA from a2323 that contains normal L, the MS chimera shown in h, and a deleted segment S that contains no genes and is only 798 bp. Lane d shows RNA from Q2361 that contains normal L and a chimera of S and M but no normal segment M or S.
plated on the normal host strain, there would be no plaques unless the host carried a plasmid that expressed the missing gene or genes. If the plasmid produced a transcript in which was embedded a complete copy of the segment that suffered the deletion, the phage could pick up this transcript and incorporate it into the genome, replacing the deletion segment with the new modification. Neither the 5' nor the 3' end of the transcript had to be tailored correctly in order for the acquisition to occur at high frequency. In our working model, the transcript was modified by cellular enzymes before being packaged because we have
350
LEONARD MINDICH
found that in uitro packaging of RNA is very stringent with respect to extra nucleotides at the 5’ end and minus strand synthesis is very (but not absolutely) stringent in terms of the sequence at the 3’ end. In many of the eukaryotic viral reverse systems that use transcript acquisition, it is necessary to engineer ribozyme sequences onto the transcripts so that the ends will be identical to the viral ends. In the case of @6,this is not necessary for in uiuo acquisition. We have even found that pickup can take place when the normal 3’ end of the viral sequence is not embedded in the transcript. In these cases, a proper 3’ end is supplied by copy choice recombination, with the use of one of the other segments as the donor. The frequency is diminished by several orders of magnitude in these cases. At any rate, the 5’ and 3’ ends of the acquired transcripts are identical to those of normal genomic segments. The frequency of pickup can be very high despite the fact that the transcript must be modified at both the 5’ and 3’ ends (Onodera et aZ., 1995,1998).As much as 10%of the total phage has been found to carry the acquired transcript. These levels, however, do not necessarily reflect the actual frequency of pickup. It seems likely that the phage that has acquired the new transcript may have a replication advantage over the phages that are being complemented. We have used the in uiuo transcript acquisition method for most of our recent work in reverse genetics. It is very easy to isolate the desired constructions because there is no background from the deletion mutants. We have used this method for the insertion of reporter groups in the 3’ noncoding regions and also for the construction of chimeras of the genomic segments. We have produced a phage with only two genomic segments by joining segment S and segment M together such that the chimeric molecule has the pac site of S (Fig. 6). This construct is stable and appears normal. We have also prepared a virus in which all three segments are joined together into one genomic segment of 14 kbp. In this case, we did not need t o use a deletion phage, as the cells carrying the plasmid produced plaques on their own lawn without the participation of helper phage.
V. CONCLUSION Using these two methods, we have been able to insert novel genes into the 3‘ noncoding regions of the three segments. We have placed genes for kanamycin resistance (aph),for the a portion of P-galactosidase (ZacH),and for GFP (gfp).We have been able to engineer nonsense mutations in genes (e.g., genes 10 and 14) to assess their function
REVERSE GENETICS OF dsRNA PHAGE (P6
351
when we were unable to isolate nonsense mutations by screening after chemical mutagenesis (Casini and Revel, 1996; Johnson and Mindich, 1994). We have also been able t o prepare virus containing genomic segments with small or large deletions or virus with segments joined together so that the number of genomic segments can be two or even one (Onodera et al., 1998). In the latter case, we have prepared a live virus that has the entire genome in a segment of about 14 kbp (Onodera et al., 1998). Although there are significant differences in the structure and mode of replication of (D6 and the members of the Reoviridae, it is possible that the mechanisms found in (D6 may also operate in the eukaryotic dsRNA viruses. ACKNOWLEDGMENT This work is supported by Grants GM31709 and GM34352 from the National Institutes of Health to L.M.
REFERENCES Butcher, S. J., Dokland, T., Ojala, P. M., Bamford, D. H., and Fuller, S. D. (1997l.Intermediates in the assembly pathway of the double-stranded RNA virus 46. EMBO J. 16,4477-4487. Caldentey, J.,and Bamford, D. H. (1992).The lytic enzyme of the Pseudomonas phage @6. Purification and biochemical characterization. Biochim. Biophys. Acta 1159, 44-50. Casini, G., and Revel, H. R. (1994). A new small low-abundant nonstructural protein encoded by the L segment of the dsRNA bacteriophage Q6. Virology 203, 221-228. Casini, G., and Revel, H. R. (1996). Construction and analysis of a bacteriophage @6 gene 14 nonsense mutant. Virology 216, 455-458. Coplin, D. L., Van Etten, J. L., and Vidaver, A. K. (1976). Synthesis of bacteriophage 4% double-stranded ribonucleic acid. J. Gen. Virol. 33, 509-512. Day, L. A,, and Mindich, L. (1980). The molecular weight of bacteriophage @6 and its nucleocapsid. Virology 103, 376-385. Emori, Y., Iba, H., and Okada, Y. (1983). Transcriptional regulation of three doublestranded RNA segments of bacteriophage (1)6 in vitro. J . Virol. 46, 196-203. Gottlieb, P., Strassman, J., Bamford, D. H., and Mindich, L. (1988). Production of a polyhedral particle in E. coli from a cDNA copy of the large genomic segment of bacteriophage @6. J . Virol. 62, 181-187. Gottlieb, P., Strassman, J., Qiao, X., Frucht, A.,and Mindich, L. (1990). In vitro replication, packaging and transcription of the segmented dsRNA genome of bacteriophage 06: Studies with procapsids assembled from plasmid encoded proteins. J . Bacteriol. 172, 5774-5782. Gottlieb, P., Strassman, J., and Mindich, L. (1992). Protein P4 of the bacteriophage @6 procapsid has a nucleoside triphosphate-binding site with associated nucleoside triphosphate phosphohydrolase activity. J . Virol. 66, 622045222. Gottlieb, P., Qiao, X., Strassman, J., Frilander, M., and Mindich, L. (1994). Identification of the packaging regions within the genomic RNA segments of bacteriophage @6. Virology 200, 42-47.
352
LEONARD MINDICH
Johnson, M. D., and Mindich, L. (1994). Isolation and characterization of nonsense mutants in gene 10 of bacteriophage @6. J . Virol. 68, 2331-2338. Koonin, E. V., Gorbalenya, E. E., and Chumakov, K. M. (1989). Tentative identification of RNA-dependent RNA polymerases of dsRNA viruses and their relationship to positive strand RNA viral polymerases. FEBS Lett. 252, 42-46. Mindich, L. (1988). Bacteriophage @6: A unique virus having a lipid-containing membrane and a genome composed of three dsRNA segments. Adu. Virus Res. 35,137-176. Mindich, L., and Abelson, R. D. (1980). The characterization of a 120 S particle formed during a 6 infection. Virology 103, 386-391. Mindich, L., and Bamford, D. H. (1988). Lipid-containing bacteriophages. In “The Bacteriophages” (R. Calendar, ed.), Vol. 2. pp. 475-520. Plenum, New York and London. Mindich, L., Nemhauser, I., Gottlieb, P., Romantschuk, M., Carton, J.,Frucht, S., Strassman, J., Bamford, D. H., and Kalkkinen, N. (1988). Nucleotide sequence of the large dsRNA segment of bacteriophage 0 6 : The genes specifying the viral replicase and transcriptase. J. Virol. 62, 1180-1185. Mindich, L., Qiao, X., Onodera, S., Gottlieb, P., and Strassman, J . (1992). Heterologous recombination in the dsRNA bacteriophage @6.J. Virol. 66, 2605-2610. Mindich, L., Qiao, X., Onodera, S., Gottlieb, P., and Frilander, M. (1994). RNA structural requirements for stability and minus strand synthesis in the dsRNA bacteriophage @6. Virology 202, 258-263. Ojala, P. M., Romantschuk, M., and Bamford, D. H. (1990). Purified @6nucleocapsids are capable of productive infection of host cells with partially disrupted outer membrane. Virology 178, 364-372. Olkkonen, V. M., Gottlieb, P., Strassman, J., Qiao, X., Bamford, D. H., and Mindich, L. (1990). In vitro assembly of infectious nucleocapsids of bacteriophage @6: Formation of a recombinant double-stranded RNA virus. Proc. Natl. Acad. Sci. U.S.A.87, 91739177. Olkkonen, V. M., Ojala, P., and Bamford, D. H. (1991).Generation of infectious nucleocapsids by in vitro assembly of the shell protein onto the polymerase complex of the dsRNA bacteriophage @6.J . Mol. Biol. 218, 569-581. Onodera, S., Olkkonen, V. M., Gottlieb, P., Strassman, J., Qiao, X., Bamford, D. H., and Mindich, L. (1992). Construction of a transducing virus from dsRNA bacteriophage (1.16:establishment of carrier states in host cells, J. Virol. 66, 190-196. Onodera, S., Qiao, X., Gottlieb, P., Strassman, J., Frilander, M., and Mindich, L. (1993). RNA structure and heterologous recombination in the dsRNA bacteriophage @6. J. Virol. 67, 4914-4922. Onodera, S., Qiao, X., Qiao, J., and Mindich, L. (1995). Acquisition of a fourth genomic segment in bacteriophage 96: A bacteriophage with a genome of three segments of dsRNA. Virology 212, 204-212. Onodera, S., Qiao, X., Qiao, J., and Mindich, L. (1998). Directed changes in the number of dsRNA genomic segments in bacteriophage @6. Proc. Nut. Acad. Sci. U.S.A. 95, 3920-3924. Partridge, J. E., Van Etten, J . L., Burbank, D. E., and Vidaver, A. K. (1979). RNA polymerase activity associated with bacteriophage 9 6 nucleocapsid. J. Gen. Virol. 43, 299-307. Qiao, X., Qiao, J., and Mindich, L. (19971. Stoichiometric packaging of the three genomic segments of dsRNA bacteriophage @6.Proc. Natl. Acad. Sci. U S A . 94, 4074-4079. Romantschuk, M., Olkkonen, V. M., and Bamford, D. H. (1988). The nucleocapsid of bacteriophage (P6 penetrates the host cytoplasmic membrane. EMBO J , 7,1821-1829.
REVERSE GENETICS OF dsRNA PHAGE @6
353
Sinclair, J. F., Tzagoloff, A., Levine, D., and Mindich, L. (1975).Proteins of bacteriophage Q6. J . Viral. 16, 685-695. Van Dijk, A. A., Frilander, M., and Bamford, D. H. (1995).Differentiation between minusand plus-strand synthesis: Polymerase activity of dsRNA bacteriophage 06 in a n in vitro packaging and replication system. Virology 211, 320-323. Vidaver, A. K., Koski, R. K., and Van Etten, J. L. (1973). Bacteriophage @6: A lipidcontaining virus of Pseudomonas phaseolicola. J. Viral. 11, 799-805.