Polyprotein processing in the expression of the cowpea mosaic virus genome

Research output: Thesisinternal PhD, WU

Abstract

In our study on the proteolytic processing of the 200K polyprotein encoded by CPMV B-RNA we first examined the types of cleavage sites present in this polyprotein. Previously Zabel etal . (1984) had shown that VPg is released from its 60K precursor by cleavage between a glutamine-serine dipeptide sequence, and the question was whether all cleavages in the 200K polyprotein occurred at glutamine-serine sites or not. The determination of partial amino acid sequences of isolated B-RNA- encoded proteins and alignment of these sequences with the open reading frame in B-RNA revealed that three types of cleavage sites are used to process the 200K polyprotein namely glutamine-serine (2x), glutamine-           glycine and glutamine-methionine amino acid pairs (Chapter 3). A glutamine-methionine and glutamine-glycine site are also present in the M-RNA-encoded polyprotein, as revealed by partial amino acid sequencing of the capsid proteins (Franssen etal ., 1986). A common feature of the sequences surrounding the different cleavage sites is that they have alanine at position -4 and alanine or proline at position -2, but beyond that there is no obvious homology among the cleavage sites. Since there occur several glutamine-glycine, glutamine-serine and glutaminemethionine dipeptide sequences in the polyproteins which are not cleaved, probably the secondary and tertiary structure of the polypeptide chains are also important factors in determining the cleavage sites involved In the processing.

The processing model of the 200K polyprotein, based on analysis of B-RNA-encoded proteins found invivo and invitro translation studies, postulated the formation of a 24K protein as a final cleavage product (Rezelman etal ., 1980; Franssen etal ., 1984a). To demonstrate the actual presence of such protein in infected cells we have used antibodies raised against a synthetic peptide with an amino acid sequence corresponding to part of this hypothetical protein. The antibodies indeed reacted with a 24K protein in CPMV infected protoplasts and also with the 84K, 110K and 170K precursors which contain the sequence of the 24K protein (Chapter 4). In view of the results of processing of the invitro translation products of B-RNA and its homology to the picornavirus encoded proteases it was previously suggested that the 24K protein possesses proteolytic activity. The protease activity of the 24K protein, was examined by expressing a hybrid cDNA construct, containing the 24K coding region linked to the coding region of both capsid proteins, in E . coli using the T7 promoter/polymerase system of Tabor and Richardson (1985) (Chapter 5). This resulted In the production of several virus-specific proteins which were characterized using specific antibodies. Pulse-chase experiments showed that two primary products were produced (the smaller one was probably the result of internal initiation of translation) which underwent faithful cleavage at two glutamine-glycine sites. One of the cleavage products represented the small capsid protein VP23. When a construct was used in which the 24K coding sequence contained a small deletion only two large proteins could be detected. These results unequivocally indicate that the 24K protein catalyzes the cleavages at the glutamine-glycine sites in the CPMV polyproteins (Chapter 5).

Recent experiments described by Verver etal . (1987) have shown that the 24K protein (or proteins containing the 24K sequence) is also able to cleave one of the glutamine -serine sites in the 200K polyprotein. Furthermore, Franssen etal . (1984b) had obtained evidence that the B-RNA-encoded 32K protein was involved in the cleavage at the glutamine- methionine site in the M-RNA-encoded polyprotein. Therefore it was proposed that the 24K protein is probably catalyzing the cleavages at all glutamine-serine and glutamine-glycine sites, while on the other hand the 32K protein would be involved in the cleavage of both glutaminemethionine sites in the CPMV polyproteins. Indeed, serine and glycine are similar amino acids with respect to their side groups (small polar) whereas methionine is very different (large non-polar side group), further supporting the idea that two different proteases would be necessary to cleave all cleavage sites. However, recently Vos etal . (1987a) have definitely shown that the 24K protein is able to catalyze all cleavages in the CPMV polyproteins but that for the cleavage of the glutamine-methionine site in the M-polyprotein the 32K protein is essential as a cofactor (see Vos etal ., 1987a).

Another question we have addressed in our study on the expression of the CPMV RNAs Is the expression of M-RNA invivo . Invitro , M-RNA is translated into two carboxy-terminal overlapping polyproteins (105K and 95K) the smaller one as a result of initiation at an internal AUG codon (Vos etal ., 1984). These proteins are cleaved by a B-RNA-encoded activity into 58K and 48K proteins and a 60K capsid protein precursor (Franssen etal ., 1982). At the other hand in CPMV-infected cells the two capsid proteins VP23 and VP37 are the only M-RNA-encoded proteins readily detectable. To elucidate the expression mechanism of M-RNA invivo we have searched for the 60K capsid precursor and for the 58K and 48K proteins in CPMV-infected cells. Using antibodies against the capsid proteins and ZnCl 2 the 60K capsid precursor protein was detected in CPMV-inoculated protoplasts incubated In the presence of ZnCl 2 (Chapter 6). Zn ++ions are known to inhibit the proteolytic processing of several viral polyproteins and probably this has caused the 60K protein to accumulate in these cells.

Using antibodies against synthetic peptides, corresponding to the common carboxy-terminus of the 48K and 58K proteins, a 48K protein was detected in the membrane fractions of infected cells. A viral 58K protein could not be detected (Chapter 6). The presence of the 48K and 60K proteins in infected cells links the invitro translation results with the invivo situation and shows that also invivo CPMV M-RNA is expressed via proteolytic processing of a polyprotein. As sofar direct evidence for the occurrence of a M-RNA-encoded 58K polypeptide invivo is lacking it remains unknown whether invivo M-RNA, besides being translated into a 95K polypeptide starting at the initiation codon at position 524, is also expressed by translation, starting at the AUG codon at position 161, into a 105K polypeptide. Invitro the 105K protein is usually produced in considerable smaller amounts than the 95K protein and it seems plausible that this also occurs invivo and perhaps the amount of 58K protein is very low and remains below the present level of detection. Another possibility is that the 58K protein is unstable and rapidly degraded in infected cells. It is a striking fact that the M-RNAs of all comoviruses studied sofar produce two polyproteins upon invitro translation (Goldbach and Krijt, 1982), supporting the idea that the presence of two AUG initiation codons in the same reading frame has some functional significance. In our opinion it seems therefore likely that invivo some 58K protein will be produced.
Previously it was suggested that the M-RNA-encoded 48K (and 58K) protein has a function in cell to cell transport of the virus (Rezelman etal ., 1982). We have now shown that the 48K protein is present in the membrane fraction of infected cells and is excreted in the incubation medium of CPMV infected protoplasts (Chapter 6). These findings are consistent with a possible function of the 48K protein in virus transport.
Infection of cells with CPMV is accompanied with the production of membranous vesicles and electron-dense amorphous material. We have demonstrated that the electron-dense material contains (if not represents) non-structural B-RNA-encoded proteins. This was established by immunocytochemical labeling of sections of CPMV infected protoplasts with specific antibodies and protein Agold (Chapter 7). The data collected correlate well with the observation that during a CPMV infection a considerable amount of non-structural proteins is produced (Goldbach and Van Kammen, 1985). Apparently these proteins are located in the cytophatic structure as a part of the electron-dense material. The membranous vesicles have been implicated with viral RNA replication (De Zoeten etal ., 1974) but the question whether the electron-dense material also has a function In virus multiplication remains to be answered.
With the detection of the 60K, 48K and 24K proteins in infected cells probably all major proteins expressed from the CPMV RNAs have now been identified. It is also clear now that all these proteins are produced by proteolytic processing of two polyproteins between three different pairs of amino acids and that the 24K protein As the protease catalyzing these cleavages. Future research will concentrate on the functions of these proteins. The synthesis of infectious transcripts from complete cDNA clones of both M- and B-RNA as recently described by Vos etal . (1987) opens the way to introduce specific mutations in the RNAs and this will, together with the techniques which have been described in this thesis, form the basis for further research on the elucidation of the replication mechanism of CPMV and its pathogenic action in the host plant.

Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
Supervisors/Advisors
  • van Kammen, A., Promotor, External person
  • Goldbach, R.W., Promotor, External person
Award date13 Nov 1987
Place of PublicationWageningen
Publisher
DOIs
Publication statusPublished - 13 Nov 1987

Keywords

  • cowpea mosaic virus
  • genetic engineering
  • recombinant dna
  • rna viruses

Fingerprint

Dive into the research topics of 'Polyprotein processing in the expression of the cowpea mosaic virus genome'. Together they form a unique fingerprint.

Cite this