Leiden Institute of Chemistry, Department of Biochemistry, Gorlaeus Laboratories, Leiden University, PO Box 9502, 2300 RA Leiden, The Netherlands1
Author for correspondence: Jan van Duin. Fax +31 71 527 4340. e-mail j.duin{at}chem.leidenuniv.nl
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Usually, epitopes are expressed as a fusion with a virus capsid protein, whose structure is known at high resolution, allowing rational design of foreign inserts on the surface of the virus particle. In contrast, filamentous DNA bacteriophages such as M13 are generally employed to construct large peptide and synthetic antibody libraries that can be used to identify known or unknown ligands (Zwick et al., 1998 ).
RNA phages are attractive candidates to display epitopes because their capsid structures are known at atomic resolution (Valegrd et al., 1990
). MS2 is a single-stranded RNA bacteriophage of 3569 nucleotides that is able to infect male E. coli bacteria via adsorption to F-pili. Upon infection, each cell produces about 103104 virions. These have icosahedral symmetry and contain 180 copies of the major coat protein and one copy of the maturation protein, which is required for infection (reviewed by van Duin, 1988
).
Virions of the RNA phages are highly antigenic and some groups have examined the possibility of displaying peptides on the surface of empty capsids (Pushko et al., 1993 ; Mastico et al., 1993
; Peabody, 1997
; Heal et al., 1999
). For this purpose, the sequences of choice were inserted after amino acid 2 or 11 of the major coat protein. Both of these positions lie in loops exposed at the capsid surface and the insert would merely enlarge the size of this loop (Valeg
rd et al., 1990
). The octapeptide inserted by Peabody (1997)
at either position inhibited capsid formation, but half of the pentapeptides inserted by Pushko et al. (1993)
at position 2 were compatible with the formation of capsids.
No reports have appeared on inserting peptides in viable RNA phage. Clearly, if one could display peptides on the surface of live phages, this would potentially expand the technique with selection from libraries or with natural in vivo selection. To test the feasibility of this approach, we took a mutant studied by Pushko et al. (1993) in which five amino acids (ASISI) were inserted at position 1 of the coat protein of fr, a close relative of MS2. We chose this mutant because its capacity to form (empty) capsids had remained intact.
Compared with these ex vivo studies on capsid formation, the in vivo approach faces several additional difficulties. For example, the peptide insertion must be compatible with the formation of an infectious phage. This requires the correct incorporation of the maturation protein in the now-modified virion. Furthermore, the new coat protein must still be able to act as a translational repressor by binding the replicase operator (Witherell et al., 1991 ). Another problem facing peptide display on live phages is that the extra RNA encoding the peptide can come under selection pressure for various reasons. If it cannot adopt a proper secondary structure, it may fall prey to E. coli endonucleases (Arora et al., 1996
; Olsthoorn & van Duin, 1996
; Klovins et al., 1997
; van Meerten et al., 1999
; Steege, 2000
). On the other hand, the 15 extra nucleotides encoding the five amino acids are likely to change the local RNA secondary structure at the coat start and this might affect translational yield significantly (de Smit & van Duin, 1990a
).
The degeneracy of the genetic code allowed us to make various mutants, all encoding the same ASISI insert in the coat protein. Several of these mutants were unstable and suffered deletions that reduced the insert to two or three amino acids. Other mutants were fully stable. Thus, it is the sequence of the nucleic acid rather than that of the protein that determines the genetic stability of the mutants. These differences cannot be ascribed to differential translational yields, but they must be a result of different RNA structures.
![]() |
Methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Plasmid pPLc2822-K33 (de Smit & van Duin, 1993 ) was derived from pPLc2833 (Remaut et al., 1981
) and carries the MS2 cDNA from position 1303 to 2057. It contains an engineered unique NsiI site at position 1336. Plasmid pMS2000
, previously called pD (Licis et al., 1998
), was derived from the infectious clone pMS2000 (Olsthoorn et al., 1994
) by replacing the nt 13031901 MS2 cDNA fragment by a short linker. Plasmid pMS2000
confers kanamycin resistance. pMS2 contains the MS2 cDNA from position 103 to 2057 and has been described by Berkhout et al. (1987)
. pMS23 carries the MS2 cDNA from position 103 to the 3' end and has been described previously (Berkhout & van Duin, 1985
). Both plasmids pMS2 and pMS23 have unique XbaI (1303) and AflII (1901) sites in the MS2 cDNA and carry the ampicillin-resistance gene. All MS2 cDNAs in these plasmids are under the control of the PL promoter.
Construction of MS2 cDNA mutants.
The mutants were constructed by ligating complementary oligonucleotides in plasmid pPLc2822-K33 cleaved by NsiI (1336) and SalI (1365). The sequences of all oligonucleotides were checked after insertion in the plasmid.
Each set of complementary oligonucleotides was hybridized by mixing them in a 1:1 molar ratio, heating to 70 °C and slow cooling to room temperature. From the mutant plasmid, the XbaI (1303)AflII (1901) fragment was cloned in plasmid pMS2000 to produce the complete MS2 cDNA sequence. This procedure was followed for ASISI 1 and ASISI 3. Because of the high virulence of full-length MS2 cDNA, it is often difficult to transform cells stably with such a plasmid, as there is a strong selection against its presence. Thus, we sometimes used a two-plasmid system, where the genetic information for the phage is divided over two compatible replicons in an overlapping manner (Olsthoorn, 1996
). In such a case, an RNA or DNA recombination event is required to obtain the contiguous sequence. Mutants ASISI 2, ANISI 1 and ANISI 2 were obtained in this way. The mutations were introduced in pMS23 and complemented with pMS2000
. When this combination of plasmids contained wild-type sequences, the supernatant of an overnight culture contained 109 p.f.u./ml. This is 100-fold less than from the contiguous sequence pMS2000. All titres mentioned in the Results are normalized to the one-plasmid value.
Phage generation, titre estimation and phage evolution.
Bacteria carrying the wild-type or mutant MS2 cDNA plasmid(s) were grown overnight at 28 °C in LC in the presence of antibiotics. The supernatant (cycle 1) of the culture was plated on a lawn of KA797 (F+) cells and incubated overnight at 37 °C. Plaques (cycle 2) were counted and individual plaques were taken and amplified overnight on KA797 cells in liquid culture (cycle 3). Phages were passaged for two more cycles of infection by growing them overnight at 37 °C in 2 ml liquid cultures containing KA797 cells (cycles 4 and 5). Routinely, at several evolutionary stages, the sequence between nt 1300 and 1365 was determined by RTPCR. For some pseudorevertants, it was necessary to extend the sequenced region to nt 1551 (see below).
RTPCR and sequence analysis.
Phages from individual plaques or from 1 µl of an overnight-infected E. coli culture were suspended in 6 µl water and heated at 92 °C for 2 min. One µl was used for RTPCR according to standard procedures recommended by the suppliers (Promega and Eurogentec). The primers used were biotin-labelled DUI715 (identical to nt 11821205) and unlabelled DUI114 (complementary to nt 18941878). PCR fragments were sequenced with DUI59 (complementary to nt 14221409) after separation and purification of the strands using Dynabeads (Dynal).
Coat-protein expression measurements.
To measure the coat-protein yield of mutants and pseudorevertants, the phage RNA was amplified by RTPCR as described above and the XbaI (1303)BstXI (1551) fragment was cloned into the expression plasmid pMS2 carrying the partial MS2 cDNA (nt 1032057) behind the PL promoter. E. coli M5219 cells harbouring the expression plasmids were grown in LC at 28 °C to an OD650 of 0·2, after which transcription was induced by heating to 42 °C. Samples for visualization of coat-protein synthesis by Western blot were taken 1 h after induction (de Smit & van Duin, 1990b ). An antiserum raised against SDS-denatured MS2 coat protein was used for detection.
Reintroduction of pseudorevertant sequences in infectious MS2 cDNA plasmid combinations.
cDNA of ASISI 3.1, which is a pseudorevertant of ASISI 3, was prepared by ligating the appropriate complementary oligonucleotides covering the sequence 13361365 (NsiISalI) in pPLc2822-K33. The mutant sequence was then incorporated in pMS23 (position 103 to the 3' end) as an XbaIAflII fragment (nt 13031901) and complemented with pMS2000 for the production of phages.
The XbaIBstXI fragments (nt 13031551) from pseudorevertants ASISI 1.1, ASISI 1.3, ASISI 2.3 and ASISI 2.5 were prepared from phage RNA by RTPCR and then used to replace the wild-type fragment in pMS2. The complete insert (13031551) was sequenced at the plasmid level after its introduction in pMS2. No mutations were found in addition to those reported in Results, except for a UC transition in ASISI 2.3 at position 1440. For phage production, pMS2 containing the pseudorevertant sequence was complemented with pMS2000
(lacking fragment 13031901). A double cross-over is required to obtain phages. Using wild-type sequences, the plasmid combination yielded 105 p.f.u./ml overnight culture.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Choice of the peptide insert
For the present study, we used the pentapeptide ASISI, inserted after amino acid 1 of the MS2 coat protein, to yield the sequence (M)ASISIASNF..., where the extra amino acids are in bold. The methionine that is normally split off from the mature protein is shown in parentheses. [Note that the insert can also be considered as an insertion after amino acid 3 (ISIAS) (underlined)]. Pushko et al. (1993) reported that this pentapeptide did not interfere with capsid formation.
To simplify insertion at the required position, we made use of an MS2 mutant that contained a unique NsiI site just ahead of the coat start (Fig. 1). There is a large number of possibilities to encode these five new amino acids. Our choice was guided by several principles. Firstly, we avoided codons for which there is little tRNA, as this might affect translation efficiency (Spanjaard & van Duin, 1988
). Secondly, the extra nucleotides should not add too many extra base-pairs to the initiator hairpin, as the resulting stabilization could inhibit coat expression (de Smit & van Duin, 1990b
). On the other hand, the insert should not be present as a mere extension of the hairpin loop, since we have observed previously that very large hairpin loops are excised in RNA phages (Olsthoorn & van Duin, 1996
). In Fig. 1
, we show the three insertion mutants used in this study.
|
|
|
In general, the properties and fate of the ASISI 2 insert were similar to those of the previous series. The titre was extremely low and the return to high titre could apparently only be achieved by a deletion, followed by several adaptive mutations, the advantages of which are only partly understood.
Third nucleotide choice for the pentapeptide sequence
We next constructed ASISI 3, in which we returned to the 9 nt loop but included more purines in it than in ASISI 1 (Fig. 4). This choice was based on the loop composition found in the pseudorevertants of ASISI 1. These revertants, at least ASISI 1.1 and 1.2, show also that a large loop of 10 nt does not have to be a problem. In two independent transformations, the ASISI 3 construct yielded about 104 p.f.u./ml overnight culture. This construct is therefore about 1000 times better than ASISI 1.
|
In order to understand better the deletions and substitutions in all of our pseudorevertants, it was necessary to identify the evolutionary pressure that led to their selection. Is it the sequence of the peptide insert, the RNA hairpin-loop structure or the efficiency of coat gene translation that adversely affects viability in our starting mutants?
The inserts have only a marginal effect on coat-protein gene translation
The effect of the insertion on coat-protein yield can be readily measured by using expression vectors carrying the coat-protein gene sequences of the wild-type, the initial mutants and their pseudorevertants. Western blots such as the one shown in Fig. 5 revealed that the differences in coat-protein yield between wild-type, mutants and pseudorevertants were rather small, at most a 3- to 4-fold decrease for ASISI 2 relative to wild-type. Such differences, however, have never been seen to cause changes in titre of 10 orders of magnitude. For instance, in a previous study, initiation mutants with a 30- to 1000-fold decrease in coat-protein yield still showed a titre that was only decreased by three to four logs (Olsthoorn et al., 1994
). Moreover, ASISI 1 produced about the same amount of coat protein as wild-type but still suffered the same 10 logs drop in titre as ASISI 2.
|
Selection is against the inserted RNA, not the inserted peptide
Although the two revertants ASISI 3.1 and 3.2 tolerated the insert, it is clear that the starting sequence of ASISI 3 is in need of improvement. The suppressor mutations selected in ASISI 3.1 and 3.2 prescribe a different protein insert sequence, thus raising the question whether the evolutionary pressure for the changes originated from a non-functional protein or from an unfavourable RNA structure. To distinguish between the two possibilities, we started from the evolved phage ASISI 3.1, carrying a single substitution G8A that turned serine into asparagine (A of the AUG is residue 1). Either serine is a bad amino acid at this position in the protein, or G8 is somehow deleterious at the RNA level. To answer this question, we made two new MS2 cDNA constructs. In one, ASISI 3.1, the single G8A suppressor mutation was cloned back in the infectious clone. It is now called ANISI 1 because of the Ser-to-Asn change (Fig. 6). This sequence should stay unaltered in the emerging phage and the titre of the construct should be high. In the other construct, ANISI 2, we left the new amino acid sequence unchanged but made three substitutions in wobble positions (circled italics in Fig. 6
).
|
With construct ANISI 2, we tested whether the amino acid or the nucleotide sequence was the target of selection. ANISI 2, encoding the same pentapeptide as ANISI 1, had a titre that was 105 lower than that of ANISI 1, and it formed very small plaques. It is thus clear that it is the RNA sequence and not the resulting amino acid sequence that hampered phage growth. Five plaques derived from ANISI 2 were sequenced. We found the U11G substitution in four of them and U11C in one. Again, these substitutions are in the loop and suggest that, from the point of view of fitness of the phage, there was something wrong with the loop in ANISI 2. The suppressor mutations corrected this defect. That the mutations changed the peptide sequence at the same time was probably a coincidence.
The phenotypic reversions are the result of the recorded suppressor mutations
A priori, we cannot exclude that the phenotypic reversion of our pseudorevertants was due to mutations outside the sequenced region. Therefore, it was necessary to clone back the sequenced section in an otherwise wild-type MS2 cDNA background and to show maintenance of the revertant phenotype. ANISI 1 is an example. The cloned-back sequence extended only from 1336 (NsiI) to 1365 (SalI). Therefore, the four logs difference in titre between ASISI 3 cDNA and ANISI 1 cDNA could be fully ascribed to the single G8A suppressor mutation, which constitutes the only difference between the constructs.
A similar analysis was carried out for pseudorevertants ASISI 1.1, ASISI 1.3, ASISI 2.3 and ASISI 2.5. Here, the revertant region 13031551 (XbaIBstXI) was cloned back in pMS2 (1032057) and, after complementation with pMS2000 (
13031901), the titres of overnight cultures were compared to those of the wild-type equivalent. The original mutant sequences ASISI 1 and ASISI 2 produced no phages in this plasmid combination, but all pseudorevertant sequences showed titres that were the same as that found for wild-type (see Methods).
The cloned-back cDNA of the four pseudorevertants was sequenced (nt 13031551). No mutations were found outside the initiator hairpin except for a UC transition at position 1440 in one of the four sequenced clones (ASISI 2.3). We do not know whether this mutation is part of the phenotype or was introduced by the PCR. At any rate, for the three other pseudorevertants, the sequence changes in the initiator hairpin accounted fully for the phenotype.
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
There are about 1000 ways to encode a pentapeptide and we restricted our choice somewhat by excluding rare codons. In our first mutant (ASISI 1), we made the 9 nt loop of the initiator stem highly pyrimidine rich because there is a preference for these in MS2 RNA hairpin loops (Olsthoorn & van Duin, 1996 ). Coat-protein expression from this mutant was not much different from wild-type, but its fitness was reduced dramatically, as reflected in a drop in titre of about 10 orders of magnitude. Reversion to high titre was achieved by a deletion of 6 or 9 nt (two or three amino acids) followed by some adaptive mutations in the loop of the initiator hairpin.
In our next construct (ASISI 2), two other serine codons were tested. Their presence predicts the formation of an initiator hairpin with a more-common four-membered loop. The properties of this mutant were, broadly speaking, the same as those of ASISI 1. The titre was down by about 10 logs and the cure to high titre was the deletion of 9 nt followed by several adaptive mutations, some of which were in the loop of the initiator hairpin and others in its stem. The third try (ASISI 3) was successful. Here, the 9 nt loop of the initiator hairpin contained more purines than in ASISI 1. This choice reflected the loop composition in the revertants from ASISI 1 and, to some extent, from ASISI 2. The full insert was still present in the majority of plaques. Only some adaptive mutations in the loop had occurred and these changed the amino acids of the insert. Were these adaptations selected to improve the protein or the RNA structure? The answer to this question came from further study of revertant ASISI 3.1, which had a G8A (SerAsn) substitution in the hairpin loop.
We recloned the relevant part of the revertant sequence (ASISI 3.1) in the MS2 infectious cDNA (ANISI 1) and made one derivative in which the new phage-approved amino acid sequence was encoded differently (ANISI 2). As expected, the recloned pseudorevertant sequence (ANISI 1) had a high titre and the sequence stayed unchanged in phages derived from it. In contrast, ANISI 2 gave a low titre and needed a suppressor mutation in the loop to boost its fitness.
The conclusion must be that it is the nucleic acid sequence of the loop that determines the fitness and genetic stability of the insert. This is particularly clear for ASISI 3.1 and 3.2 and for the progeny of ANISI 2, where a single mutation in the loop was apparently sufficient to make a stable phage (ANISI 2.1).
The analysis carried out here is reminiscent of another study from this lab, in which a non-coding hairpin loop was enlarged by 26 nt (Olsthoorn & van Duin, 1996 ). As seen here, there was a drop in titre and there were two kinds of adaptations. Sometimes, the big loop was resected to an acceptable size. In other pseudorevertants, a few substitutions in the loop sufficed to remove most of the genetic burden and raise the fitness.
We tentatively suggest that these big hairpin loops create a target for RNase E, one of the major endonucleases of E. coli. Suppressor mutations in the large loop such as those found in ASISI 3.1 and 3.2 possibly induce resistance. This may result from interactions between loop nucleotides, reminiscent of what has been found for GNRA and UNCG loops (Cheong et al., 1990 ; Heus & Pardi, 1991
).
Why is the drop in viability of the insertion mutants so large?
An unsolved question is why the insertions ASISI 1 and ASISI 2 have such a dramatic effect on the titre. In our previous study, where we introduced 26 single-stranded nucleotides in a non-coding loop (Olsthoorn & van Duin, 1996 ), the titre dropped by only three orders of magnitude. Even considering that, in an evolutionary sense, it is simpler to remove nucleotides from a non-coding sequence than from a reading frame, it is difficult to account for the dramatic drop in titre. We cannot rigorously exclude that the mutations affect replication, although it is hard to envisage why certain loop sequences would support replication while others would not. Unfortunately, unlike results obtained for Q
phage, the replicase of MS2 cannot be isolated in a straightforward, reproducible way (Federoff, 1975
) and the influence of the mutations on replication therefore cannot be tested easily.
Deletions versus substitutions
Another question remaining is why ASISI 1, for instance, with a 9 nt loop, is not rescued by substitutions and suffers a deletion, whereas one or more substitutions suffice for the ASISI 3 and ANISI 2 mutants.
The likely answer is that ASISI 1, with its all-pyrimidine loop, would need more simultaneous substitutions than can be found in the quasispecies pool. For instance, the chances of finding a revertant with three transversions can be estimated to be about (10-5)3 or 10-15. A deletion can then be the more likely event to save the virus. Here, we would have to assume that the three substitutions in ASISI 3.2 have accumulated sequentially. This idea, that deletions are more likely than triple or double substitutions, is supported by a recent study of Licis et al. (2000) , where it was shown that a single stop codon in the MS2 lysis frame was repaired by a substitution to sense, but a double stop codon could not be cured in this way. Instead, the region containing the stop codons was excised, allowing rescue of the phage (at the expense of a shorter lysis protein).
![]() |
Acknowledgments |
---|
![]() |
Footnotes |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Berkhout, B. & van Duin, J. (1985). Mechanism of translational coupling between coat protein and replicase genes of RNA bacteriophage MS2. Nucleic Acids Research 13, 6955-6967.[Abstract]
Berkhout, B., Schmidt, B. F., van Strien, A., van Boom, J., van Westrenen, J. & van Duin, J. (1987). Lysis gene of bacteriophage MS2 is activated by translation termination at the overlapping coat gene. Journal of Molecular Biology 195, 517-524.[Medline]
Borisova, G., Borschukova Wanst, O., Mezule, G., Skrastina, D., Petrovskis, I., Dislers, A., Pumpens, P. & Grens, E. (1996). Spatial structure and insertion capacity of immunodominant region of hepatitis B core antigen. Intervirology 39, 16-22.[Medline]
Burke, K. L., Dunn, G., Ferguson, M., Minor, P. D. & Almond, J. W. (1988). Antigen chimaeras of poliovirus as potential new vaccines. Nature 332, 81-82.[Medline]
Cheong, C., Varani, G. & Tinoco, I. (1990). Solution structure of an unusually stable RNA hairpin, 5'GGAC(UUCG)GUCC. Nature 346, 680-682.[Medline]
Dedieu, J. F., Ronco, J., van der Werf, S., Hogle, J. M., Henin, Y. & Girard, M. (1992). Poliovirus chimeras expressing sequences from the principal neutralization domain of human immunodeficiency virus type 1. Journal of Virology 66, 3161-3167.[Abstract]
de Smit, M. H. & van Duin, J. (1990a). Secondary structure of the ribosome binding site determines translational efficiency: a quantitative analysis. Proceedings of the National Academy of Sciences, USA 87, 7668-7672.[Abstract]
de Smit, M. H. & van Duin, J. (1990b). Control of prokaryotic translational initiation by mRNA secondary structure. Progress in Nucleic Acid Research and Molecular Biology 38, 1-35.[Medline]
de Smit, M. H. & van Duin, J. (1993). Translational initiation at the coat-protein gene of phage MS2: native upstream RNA relieves inhibition by local secondary structure. Molecular Microbiology 9, 1079-1088.[Medline]
Federoff, N. (1975). Replicase of the phage f2. In RNA Phages, pp. 238258. Edited by N. D. Zinder. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
Heal, K. G., Hill, H. R., Stockley, P. G., Hollingdale, M. R. & Taylor-Robinson, A. W. (1999). Expression and immunogenicity of a liver stage malaria epitope presented as a foreign peptide on the surface of RNA-free MS2 bacteriophage capsids. Vaccine 18, 251-258.[Medline]
Heus, H. A. & Pardi, A. (1991). Structural features that give rise to the unusual stability of RNA hairpins containing GNRA loops. Science 253, 191-194.[Medline]
Klovins, J., van Duin, J. & Olsthoorn, R. C. L. (1997). Rescue of the RNA phage genome from RNase III cleavage. Nucleic Acids Research 25, 4201-4208.
Koo, M., Bendahmane, M., Lettieri, G. A., Paoletti, A. D., Lane, T. E., Fitchen, J. H., Buchmeier, M. J. & Beachy, R. N. (1999). Protective immunity against murine hepatitis virus (MHV) induced by intranasal or subcutaneous administration of hybrids of tobacco mosaic virus that carries an MHV epitope. Proceedings of the National Academy of Sciences, USA 96, 7774-7779.
Kratz, P. A., Böttcher, B. & Nassal, M. (1999). Native display of complete foreign protein domains on the surface of hepatitis B virus capsids. Proceedings of the National Academy of Sciences, USA 96, 1915-1920.
Licis, N., van Duin, J., Balklava, Z. & Berzins, V. (1998). Long-range translational coupling in single-stranded RNA bacteriophages: an evolutionary analysis. Nucleic Acids Research 26, 3242-3246.
Licis, N., Balklava, Z. & van Duin, J. (2000). Forced retroevolution of an RNA bacteriophage. Virology 271, 298-306.[Medline]
Lomonossoff, G. P. & Johnson, J. E. (1996). Use of macromolecular assemblies as expression systems for peptides and synthetic vaccines. Current Opinion in Structural Biology 6, 176-182.[Medline]
Mastico, R. A., Talbot, S. J. & Stockley, P. G. (1993). Multiple presentation of foreign peptides on the surface of an RNA-free spherical bacteriophage capsid. Journal of General Virology 74, 541-548.[Abstract]
Olsthoorn, R. C. L. (1996). Structure and evolution of RNA phages. PhD Thesis, Leiden University, The Netherlands.
Olsthoorn, R. C. L. & van Duin, J. (1996). Random removal of inserts from an RNA genome: selection against single-stranded RNA. Journal of Virology 70, 729-736.[Abstract]
Olsthoorn, R. C. L., Licis, N. & van Duin, J. (1994). Leeway and constraints in the forced evolution of a regulatory RNA helix. EMBO Journal 13, 2660-2668.[Abstract]
Peabody, D. S. (1997). Subunit fusion confers tolerance to peptide insertions in a virus coat protein. Archives of Biochemistry and Biophysics 347, 85-92.[Medline]
Pushko, P., Kozlovskaya, T., Sominskaya, I., Brede, A., Stankevica, E., Ose, V., Pumpens, P. & Grens, E. (1993). Analysis of RNA phage fr coat protein assembly by insertion, deletion and substitution mutagenesis. Protein Engineering 6, 883-891.[Abstract]
Remaut, E., Stanssens, P. & Fiers, W. (1981). Plasmid vectors for high-efficiency expression controlled by the PL promoter of coliphage lambda. Gene 15, 81-93.[Medline]
Skripkin, E. A., Adhin, M. R., de Smit, M. H. & van Duin, J. (1990). Secondary structure of the central region of bacteriophage MS2 RNA. Conservation and biological significance. Journal of Molecular Biology 211, 447-463.[Medline]
Smith, A. D., Geisler, S. C., Chen, A. A., Resnick, D. A., Roy, B. M., Lewi, P. J., Arnold, E. & Arnold, G. F. (1998). Human rhinovirus type 14:human immunodeficiency virus type 1 (HIV-1) V3 loop chimeras from a combinatorial library induce potent neutralizing antibody responses against HIV-1. Journal of Virology 72, 651-659.
Spanjaard, R. A. & van Duin, J. (1988). Translation of the sequence AGGAGG yields 50% ribosomal frameshift. Proceedings of the National Academy of Sciences, USA 85, 7967-7971.[Abstract]
Steege, D. A. (2000). Emerging features of mRNA decay in bacteria. RNA 6, 1079-1090.
Taniguchi, T., Palmieri, M. & Weissmann, C. (1978). Q DNA-containing hybrid plasmids giving rise to Q
phage formation in the bacterial host. Nature 274, 223-228.[Medline]
Valegrd, K., Liljas, L., Fridborg, K. & Unge, T. (1990). The three-dimensional structure of the bacterial virus MS2. Nature 345, 36-41.[Medline]
van Duin, J. (1988). Single-stranded RNA bacteriophages. In The Viruses , pp. 117-167. Edited by H. Fraenkel-Conrat & R. Wagner. New York:Plenum Press.
van Meerten, D., Zelwer, M., Régnier, P. & van Duin, J. (1999). In vivo oligo(A) insertions in phage MS2: role of Escherichia coli poly(A) polymerase. Nucleic Acids Research 27, 3891-3898.
Witherell, G. W., Gott, J. M. & Uhlenbeck, O. C. (1991). Specific interaction between RNA phage coat proteins and RNA. Progress in Nucleic Acid Research and Molecular Biology 40, 185-220.[Medline]
Zuker, M., Mathews, D. H. & Turner, D. H. (1999). Algorithms and thermodynamics for RNA secondary structure prediction: a practical guide. In RNA Biochemistry and Biotechnology, pp. 1143. NATO ASI Series. Edited by J. Barciszewski & B. F. C. Clark. Amsterdam: Kluwer Academic.
Zwick, M. B., Shen, J. & Scott, J. K. (1998). Phage-displayed peptide libraries. Current Opinion in Biotechnology 9, 427-436.[Medline]
Received 27 November 2000;
accepted 8 March 2001.
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
INT J SYST EVOL MICROBIOL | MICROBIOLOGY | J GEN VIROL |
J MED MICROBIOL | ALL SGM JOURNALS |