Baculovirus DNA replication

M. Kool

Research output: Thesisinternal PhD, WU


<p>Baculoviruses are attractive biological agents for the control of insect pests. They are highly specific for insects and cause a fatal disease (Granados and Federici, 1986). in addition, baculoviruses are successfully exploited as expression vectors for the production of heterologous proteins for various applications (Luckow and Summers, 1988; Luckow, 1991). In both cases large-scale systems for the production of baculoviruses are important. Production in insect larvae is difficult to scale up and to control. Insect-cell cultures offer an attractive alternative. Moreover, in the case of pharmaceuticals and diagnostics in human and veterinary medicine insect-cell systems have to be applied since such systems are well defined.<p>Due to the great interest in baculoviruses as biological insecticides and expression vectors for foreign genes, the molecular genetic aspects of especially the <em>Autographa californica</em> multiple nucleocapsid nuclear polyhedrosis virus (AcMNPV), the type member of the <em>Baculoviridae,</em> have been studied in much detail (Blissard and Rohrmann, 1990). Chapter 2 of this thesis presents, as of March 1994, an overview of the structural and functional organization of the AcMNPV genome. The genomes of AcMNPV (R.D. Possee, pers. comm.) and <em>Bombyx mori</em> MNPV (BmMNPV) (S. Maeda, pers. comm.) have been completely sequenced but are awaiting publication. In contrast to other large DNA viruses such as adenovirus, herpesviruses, and vacciniavirus (Fields and Knipe, 1990), the process of baculovirus DNA replication of AcMNPV is poorly understood. At the start of this study a few genes were found which were thought to be involved in AcMNPV DNA replication such as a helicase and a DNA polymerase. Sequences representing the origin of AcMNPV DNA replication were not known.<p>Baculoviruses can be produced on a large scale in insect-cell cultures using batch (Maiorella <em>et al.,</em> 1988), semicontinuous (Hink and Strauss, 1980) <em></em> and continuous reactors (Kompier <em>et al.,</em> 1988). <em></em> Continuous production of wild-type (wt) AcMNPV and recombinants thereof was achieved in a system consisting of one bioreactor producing insect cells in series with a second bioreactor for virus infection and protein production (Kompier <em>et al.,</em> 1988; <em></em> Van Lier <em>et al.,</em> 1992). <em></em> After a few weeks of continuous operation, however, the productivity decreased to a low level. In the case of wt AcMNPV, the number of polyhedra per cell, the fraction of cells containing polyhedra, and the concentration of extracellular virus were found to be decreased (Kompier <em>et al.,</em> 1988). Continuous production of an AcMNPV recombinant where the polyhedrin gene was replaced by the lacZ gene of <em>Escherichia coli</em> essentially gave the same results (Van Lier <em>et al.,</em> 1992). <em></em> The decrease of virus production was ascribed to a phenomenon known as passage effect (Tramper and Vlak, 1986), <em></em> but the underlying mechanism remained unknown.<p>Analysis of samples obtained from continuous bioreactor systems (Chapter <em>3)</em> showed that with ongoing production a mutant AcMNPV became dominant. This mutant lacked about <em>43%</em> of the original genome. The deleted DNA included the polyhedrin gene and several genes essential for DNA replication. The replication of the mutant appeared to be dependent on the presence of an intact helper AcMNPV. The passage effect in the continuous system is thus thought to be the result of interference between the deletion mutant and helper virus. These so-called defective interfering particles (DIPs) can only accumulate when the concentration of the intact virus is high enough to support the replication of these DIPs. Thus, for a successful continuous production of baculoviruses low multiplicities of infection should be used to avoid the accumulation of DIPs.<p>One of the regions of the AcMNPV genome putatively involved in the generation of the DIPs is located in the EcoRI-C fragment of AcMNPV. Deletion mutants often lacked a considerable portion of <em>Eco</em> RI-C, but also maintained a consistent segment of this fragment that may be essential for replication and/or encapsidation. To investigate the genetic functions of the EcoRI-C fragment in the defective genomes and their possible role in the generation of these genomes, the nucleotide sequence of a <em>7.3</em> kilobase pair region of the right part of the <em>Eco</em> RI-C fragment was determined (Chapter 4). <em></em> Eight putative open reading frames (ORFs) were identified and their respective amino acid sequences compared with a number of data libraries, The product of ORF 1227 <em></em> corresponded with GP41, <em></em> a virion protein, and its predicted protein sequence was found to be 55 amino acids longer at its C-terminus than reported previously (Whitford and Faulkner, 1992). The majority of ORF 1227, including the additional 55 amino acids, moreover, showed a high degree of homology with protein P40 of <em>Helicoverpa zea</em> SNPV, also a structural virion protein (Ma <em>et al.,</em> 1993). Three other ORFs in the analyzed AcMNPV region showed homology with ORF's in the HzSNPV sequence, indicating that the general organization of this region is similar in both viruses, and possibly between MNPVs and SNPVs. However, no sequences have yet been identified within this region that may play a role in the generation and/or encapsidation of the DIPs.<p>The generation and characterization of DIPs was further investigated in Chapter 5. Three small separate regions, representing only 5 % of the original AcMNPV genome, were found to be retained in DNA of defective genomes after 40 serial passages in insect cells with undiluted inocula. Independently, Lee and Krell (1992) showed that after 80 serial passages of AcMNPV, DIPs were found which contained tandem repeats of DNA, mainly derived from a small region of the AcMNPV genome, located in the <em>Hin</em> dIII-K fragment. Since all these defective genomes were still able to replicate in insect cells, although only with the help of intact virus, they must have retained essential cis-acting elements necessary for DNA replication. Therefore, a replication assay was developed to study whether these regions, retained in the defective genomes, contained <em>cis</em> -acting elements such as an origin ( <em>ori</em> ) of DNA replication. Transfection of <em>Spodoptera frugiperda</em> cells with plasmids containing these sequences followed by superinfection with intact helper AcMNPV resulted in amplification of these plasmids, as demonstrated by the <em>Dpn</em> I sensitivity <em></em> assay. In order to demonstrate replicating activity of these plasmids, it appeared essential to transfect the cells well (24 h) before superinfection with helper virus, and for an optimal replication result the multiplicity used for superinfection had to be I or lower (Chapters 5 and 6). Using this assay seven putative origins of DNA replication were identified in the AcMNPV genome (Chapters 5, 6, and 7).<p>Six of the seven putative <em>ori's</em> were found in the homologous regions <em>hr</em> 1, <em>hr</em> 2, <em>hr</em> 3, <em>hr</em> 4a, <em>hr</em> 4b, and <em>hr</em> 5 of AcMNPV (Chapter 6), which are interspersed along the genome (Cochran and Faulkner, 1983; Guarino <em>et al.,</em> 1986). Recently, another <em>hr</em> region, <em>hr</em> 1a, has been identified in the AcMNPV genome, that could also serve as <em>ori</em> in a replication assay (Leisy and Rohrmann, 1993). Initial studies demonstrated that the <em>hr</em> regions function as enhancers for transcription, when placed in <em>cis</em> to the promoter of early baculovirus genes (Guarino <em>et</em><em>al.,</em> 1986; Guarino and Summers, 1986). <em></em> Rodems and Friesen (1993) demonstrated that <em>hr</em> regions also function as enhancers <em>in vivo</em> . These results together with the data of this thesis imply that all <em>hr's</em> in AcMNPV may be bifunctional <em>in vivo</em> , i.e. have both enhancer and <em>ori</em> activity. Sequence analysis has shown that <em>hr's</em> contain two to eight 30 bp imperfect palindromes, interspaced by other repeated sequences, and that each palindrome contains a naturally occurring <em>Eco</em> RI site at its core (Guarino <em>et al.,</em> 1986; Guarino and Summers, 1986). <em></em> One copy of such a palindrome appeared to be sufficient for either enhancer function or <em>ori</em> activity (Guarino et al., 1986; Pearson <em>et al.</em> , 1992).<p>In addition to the seven <em>hr's,</em> the <em>Hin</em> dIII-K fragment of AcMNPV was also found to carry a putative <em>ori</em> , although this fragment does not contain an <em>hr</em> region (Chapter 6). <em></em> The <em>Hin</em> dIII-K ori had a complex structure (Chapter 7), resembling those of other large DNA viruses. This <em>ori</em> contained several regions, some of which were found to be essential for its activity, whereas others contain auxiliary sequences, that enhance <em>ori</em> activity. Sequence analysis of these regions identified several structures often found in other viral replication <em>ori's</em> , such as palindromes and other repeated motifs (DePamphilis, <em>1993).</em> Recently an <em>ori</em> , also with a complex structure, but different from AcMNPV <em>hr's,</em> has been identified in another baculovirus, <em>Orgyia pseudotsugata</em> MNPV (OpMNPV) (Pearson <em>et al.,</em> 1993).<p>The individual role of all these <em>ori's</em> during viral DNA replication, and whether they are all active simultaneously <em>in vivo</em> , is unclear. Deletion of <em>hr</em> 5 from the AcMNPV genome or the closely related <em>Bombyx mori</em> MNPV (BmMNPV) genome had no effect on the replication of these viruses (Rodems and Friesen, 1993; Majima <em>et al.,</em> 1993) <em>.</em> Also from the experiments with DIPs generated by serial passaging it can be deduced that not all the <em>ori's</em> are necessary for replication of the genome. After 40 serial, undiluted passages three small segments of the genome were predominantly found to be retained, harbouring only the <em>hr</em> 1, <em>hr</em> 3, and <em>hr</em> 5 <em></em> regions (Chapter 5). Deletion of all <em>hr's</em> would indicate the importance of these regions for virus replication <em>in vivo.</em><p>The importance of the <em>ori</em> in the <em>Hin</em> dIII-K fragment <em></em> is supported by sequence data of the corresponding region in the closely related BmMNPV (Kamita <em>et al.,</em> 1993). Although most of the auxiliary sequences of this <em>ori</em> were found to be deleted in the BmMNPV genome, the essential part of this <em>ori</em> , containing the palindromes and the A/T rich region, was retained suggesting that these elements could not be deleted. These sequence data and the observation that after prolonged serial passage of AcMNPV (80 passages) large replicating DNA molecules are found in which repeated sequences from the <em>Hin</em> dIII-K fragment accumulate (Lee and Krell, 1992), may be a reflection of the importance of this region as genuine <em>ori</em><em>in vivo</em> (Chapter 7).<p>The occurrence of multiple <em>ori's</em> is not unique for baculoviruses, but has also been reported for herpesviruses and Chilo iridescent virus (CIV). The genome of herpes simplex virus I (HSV-1) contains three <em>ori's</em> , <em>ori</em><sub><font size="-2">L</font></sub> , and two copies of <em>ori</em> , (for review, see Fields and Knipe, 1990) and it has been shown that the presence of a single <em>ori</em> , independent which one, is sufficient for replication (Longnecker and Roizman, 1986; Polvino-Bodnar <em>et al.</em> , 1987; Igarashi <em>et al.</em> , 1993). In CIV at least six putative <em>ori's</em> have been identified (Handermann <em>et al.</em> , 1992). It remains to be seen whether in the case of baculoviruses each of the eight putative <em>ori's</em> is necessary for viral replication. When the <em>ori's</em> are indeed functionally redundant, the presence of multiple <em>ori's</em> in the viral genome may increase the frequency of initiation and thus increase the speed of DNA replication. Analysis of intermediates of DNA replication may shed more light on the nature of <em>in vivo ori's</em> .<p>The experiments in Chapter 6 also supported the view that a circular topology is a prerequisite for replication of <em>ori</em> -containing plasmids. Linear DNA, even if it contained an <em>ori</em> , did not replicate. These results are in line with the circular nature of baculovirus DNA and suggest a model for baculovirus replication involving a theta structure or a rolling circle. The latter model is supported by data of Leisy and Rohrmann (1993), who demonstrated that replicating plasmids form large concatemeric molecules. In addition, the finding of defective genomes with many reiterations (concatemers) of a 2.8 kbp segment, mainly mapping in the AcMNPV <em>Hin</em> dIII-K fragment (Lee and Krell, 1992), supported also a rolling circle as model for DNA replication.<p>Not only <em>cis</em> -acting elements, but also <em>trans</em> -acting factors are important for DNA replication. Chapters 8 and 9 describe the functional mapping of AcMNPV genes required for DNA replication. A transient complementation assay was employed, in which, instead of AcMNPV infection, four co-transfected cosmid clones, encompassing almost the entire genome, provided all the essential <em>trans</em> -acting factors for plasmid DNA replication. No replication of plasmids occurred when one of the cosmids was omitted from the transfection mixture. This result indicated that this assay was a valid and powerful approach to identify the AcMNPV replication genes. The assay was first used to define essential regions in the four cosmids (Chapter 8). Six essential regions were retrieved and these were further subcloned and tested (Chapter 9). Initially in this assay, plasmid replication appeared to be independent of the presence, in <em>cis</em> , of a viral <em>ori</em> , when cloned genes or viral DNA were used instead of complete virus to supply essential <em>trans</em> -acting factors (Chapter 8). However, this was caused by employing high gene copy numbers in the transfections (Chapter 9). As a consequence, a relative abundance of proteins is produced, which may lead to a saturation of specific origins with these proteins. The excess of proteins thus can bind to other originlike structures, even when the affinities are low, and hence cause replication of any plasmid.<p>Nine genes involved in DNA replication were identified in the AcMNPV genome (Chapter 9). Six genes, specifying <em>helicase, dna pol, ie-1, lef-1, lef-2,</em> and <em>lef-3</em> , were found to be essential, while three genes, <em>p35, ie-2</em> , and <em>pe38</em> , stimulated DNA replication. No stimulation was observed by the <em>pcna</em> -like protein gene. Two of the three identified stimulatory genes, <em>ie-2</em> and <em>pe38</em> , are known as transactivators for transcription (Carson <em>et al.</em> , 1988; Lu and Carstens, 1993), whereas the third stimulating gene, <em>p35</em> , has previously been identified as inhibitor of virus-induced apoptosis in <em>S. frugiperda cells</em> (Clem <em>et al.</em> , 1991). However, the observation that infection with a <em>p35</em> deletion mutant in <em>Trichoplusia ni</em> cells did not result in a reduction of virus production (Clem <em>et al.</em> , 1991) suggests that the stimulating effect of <em>p35</em> in the transient replication assays is not based on activation of the replication process, but is due to inhibiting apoptosis, which may be induced by the expression of one or more of the replication genes.<p>Of the six essential AcMNPV DNA replication proteins, putative functions could only be attributed for the helicase and DNA polymerase, based on their homology with other known helicases and DNA polymerases (Lu and Carstens, 1991; Tomalski <em>et al.</em> , 1988). Studying other viral systems, a number of striking similarities was noticed between <em>Baculoviridae</em> and <em>Herpesviridae</em> . Although these two viral families have traditionally been separated based on their different morphology and host specificity, they both have a large double stranded DNA genome, which replicates in the host cell nucleus, and has a circular form in at least one stage of their replication cycle. Their genomes may also replicate in a similar manner as transfection of origin-containing plasmids into infected cells resulted in large concatemers of input plasmid DNA (Leisy and Rohrmann, 1993; Hammerschmidt and Mankertz, 1991). Most strikingly, the number of essential replication genes is similar for both baculoviruses and herpesviruses. An attempt was made to relate the other four, hitherto unassigned, baculovirus replication proteins, IE-1, LEF-1, LEF-2, and LEF-3 with proteins involved in herpesvirus DNA replication (Chapter 10).<p>Firstly, the sequences of replication proteins of five different herpesviruses were aligned, which resulted in the identification of a number of conserved motifs in these proteins. Many of these conserved motifs showed (distant) homology with the four baculovirus replication proteins and, most importantly, in the same linear spatial organization as in their putative herpesvirus homologues. Using these conserved motifs as markers the four replication proteins IE-1, LEF-1, LEF- 2, and LEF-3 of AcMNPV were aligned with herpesvirus homologues (Chapter 10). These alignments suggest that <em>ie-1</em> codes for a single stranded DNA binding protein (SSB), <em>lef-1</em> for a primase-associated protein, <em>lef-2</em> for a DNA polymerase processivity factor, and <em>lef-3</em> for a primase. The assignment to <em>ie-1</em> to code for a SSB was further supported by the finding of a conserved known single stranded DNA binding sequence motif in six baculovirus IE-1, proteins, which is also found in many other prokaryotic and eukaryotic SSBs (Chapter 10). Further computer-assisted examination and biochemical analysis has to be done to confirm the suggested functions for the four baculovirus replication proteins.<p>The similarity between <em>Baculoviridae</em> and <em>Herpesviridae</em> in DNA structure and mechanism of DNA replication, added to the employment of an identical kind and amount of essential replication genes, poses the question whether these two groups of viruses share a common lineage. On the basis of the mutation rate of the conserved baculovirus polyhedrin genes as compared to the insect species in which they occur (Vlak and Rohrmann, 1985) it has been postulated that baculoviruses are ancient viruses that have evolved along with the insects. The relationship among replication genes could imply that herpesviruses have evolved from baculoviruses along with their invertebrate hosts towards vertebrates. Alternatively, the emergence of herpesviruses may be the result of an independent, parallel evolutionary event in ancient vertebrates. Since viral DNA replication in nuclear environments is a conserved process, conserved host replication genes may have been. independently transduced into different ancient viral genomes.
Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Goldbach, R.W., Promotor, External person
  • Tramper, J., Promotor, External person
  • Vlak, Just, Promotor
Award date17 Jun 1994
Place of PublicationS.l.
Print ISBNs9789054852629
Publication statusPublished - 1994


  • baculovirus
  • nuclear polyhedrosis viruses
  • molecular genetics
  • replication
  • translation
  • protein synthesis

Fingerprint Dive into the research topics of 'Baculovirus DNA replication'. Together they form a unique fingerprint.

Cite this