Thesis (MSc (Biomedical Sciences))--Cape Peninsula University of Technology, 2019 / Background: Baculoviruses have great potential as alternatives to conventional chemical insecticides. The large scale adoption of such agents has however been hampered by the slow killing times exhibited by these bio-insecticides, limitation to single target insect and difficulty of large scale production of these preparations. Trichoplusia ni single nucleopolyhedrovirus (TnSNPV), initially identified in the Eastern Cape region of South Africa, has potential as a biocontrol agent as it possesses a higher speed of kill compared to other baculoviruses. Aims and methods: The main objective of this study was the identification, molecular characterisation and cloning of a structural core gene (polyhedrin) and three auxiliary genes, the inhibitor of apoptosis (iap2 and iap3) and the ecdysteroid UDP-glucosyltransferase (egt) genes, from TnSNPV in order to delineate its phylogenetic relationship to a Canadian isolate of the same virus and to other baculoviruses. In addition, the genes were expressed in an Escherichia coli (E. coli) based system as a prelude to genetic modification to increase the pesticidal property of the virus. Results: The genome size of the South African strain of TnSNPV was estimated at 160 kb and is significantly larger than the Canadian isolate of TnSNPV and may reflect genetic variation as the two strains have adapted to varying environmental conditions. Occlusion bodies of the South African strain of TnSNPV were visualised by Transmission Electron Microscopy and consisted of rod shaped single virions composed of a single enveloped nucleocapsid. Insect bioassays showed that the median lethal time (LT50) of the virus strain averaged 1.8 days which is significantly faster than other baculoviruses. The South African and Canadian strains of TnSNPV share nucleotide similarities greater than 95% for the genes analysed in this study, which indicates that they are closely related. From this analysis, the South African strain of TnSNPV identifies as a Group II NPV with the closest relatives being the Canadian strain of TnSNPV and ChchNPV. The topology of the tree for the polyhedrin protein was better resolved than that of the IAP2, IAP3 and EGT proteins and was comparable to the tree inferred from a concatenated data set consisting of complete polyhedrin/granulin, LEF8, and LEF9 proteins of 48 completely sequenced genomes. For the IAP2, IAP3 and EGT proteins, the separation of the lepidopteran and hymenopteran specific baculoviruses was not evident while the separation of Group I and II Alphabaculoviruses diverged from that observed from the baculovirus core gene polyhedrin as well as the tree inferred from complete polyhedrin/granulin, LEF8, and LEF9 proteins. Five distinct groups relating to IAP-1, 2, 3, 4 and 5 could be distinguished from the tree inferred from all IAP proteins from 48 fully sequenced baculoviruses. From this analysis, the IAP protein from the South African isolate of TnSNPV can be designated as an IAP3 due to sequence homology to other IAP3 proteins. Similarly, the IAP2 can be confirmed as an IAP2 protein as it clusters with other IAP2 proteins. RNA transcripts of the four genes were detected by RT-PCR at one hr after induction with Larabinose in BL21-A1 E. coli and persisted until four hrs post induction. Antisera directed against the C-terminal 6X His tag was able to detect the recombinant proteins at two hours after induction confirming the rapid rise in expression of the proteins which persisted at high levels until four hrs after induction. The discrepancy observed with the predicted molecular mass of the EGT protein and the migration on SDS-PAGE may be due to the absence of posttranslational modification in the E. coli expression system and the hydrophobic residues present in the N-terminal signal sequence. Conclusion: Sequence and phylogenetic analysis suggest that the two isolates of TnSNPV have been exposed to similar evolutionary pressures and evolved at similar rates and represent closely related but distinct variants of the same virus. The difference in genome size between the two strains is likely to reflect actual genetic differences as the strains have adapted to their local environments and hosts and the extent of the differences will only be apparent as more sequencing results become available. Phylogenetic analysis of the IAP and EGT proteins yields a tree that varies from the phylogenetic reconstruction observed for the polyhedrin gene as well as the concatenated data set consisting of complete polh/gran, LEF8, and LEF9 proteins and highlights the risks inherent in inferring phylogenetic relationships based on single gene sequences. The tree inferred from the concatenated data set of polh/gran, LEF8, and LEF9 proteins was a quick and reliable method of identification particularly, when whole genome data is unavailable and mirrors the accepted lineage of baculoviruses. Expression of the recombinant IAP2, IAP3, EGT and polyhedrin was confirmed by RT-PCR and immunoblot analysis and rose rapidly after induction and persisted at high levels. It is as yet unclear if the expressed proteins are functional particularly as post translation modifications are lacking in this system.

The diamondback moth, Plutella xylostella (L.) (Lepidoptera: Plutellidae), is an important pest of cruciferous crops worldwide. The prolonged use of synthetic chemical insecticides as a primary means of control has resulted in the development of resistance in pest populations. In addition, the pest has also evolved resistance to the bacterial insecticidal protein of Bacillus thuringiensis which is also widely used as a method of control. Baculoviruses are considered as effective alternatives to conventional methods of control when incorporated into integrated pest management (IPM) programmes. These viruses target the larval stages of insects, are generally host-specific and are safe for use in the environment. This study aimed to isolate a baculovirus from a laboratory-reared P. xylostella colony, characterise it genetically and then evaluate its virulence against neonate and fourth instar larvae. A laboratory colony of P. xylostella was established using pupae and asymptomatic larvae collected from a cabbage plantation outside Grahamstown in the Eastern Cape province of South Africa. The colony flourished in the laboratory due to prime conditions and availability of food. The duration of development from egg to adult was determined by observation and imaging of the various life stages. The mean developmental time from egg to adult was observed to be 14.59 ± 0.21 days. The population of the insects increased rapidly in number leading to overcrowding of the insect colony, and hence appearance of larvae with viral symptoms. Occlusion bodies (OBs) were extracted from symptomatic larval cadavers and purified by glycerol gradient centrifugation. Analysis of the purified OBs by transmission electron microscopy revealed the presence of a granulovirus which was named PlxyGV-SA. The virus isolate was genetically characterised by restriction endonuclease analysis of the genomic DNA, and PCR amplification and sequencing of selected viral genes. The complete genome sequence of a Japanese P. xylostella granulovirus isolate, PlxyGV-Japan, has been deposited on the GenBank database providing a reference strain for comparison with DNA profiles and selected gene sequences of PlxyGV-SA. BLAST analysis of the granulin gene confirmed the isolation of a novel South African PlxyGV isolate. Comparison of the restriction profiles of PlxyGV-SA with profiles of PlxyGV-Japan and other documented PlxyGV profiles obtained by agarose gel electrophoresis revealed that PlxyGV-SA is a genetically distinct isolate. The data obtained from the sequencing and alignment of ecdysteroid UDP-glucosyltransferase (egt), late expression factor 8 (lef-8) and late expression factor 9 (lef-9) genes with those of PlxyGV-Japan also showed that PlxyGV-SA is a genetically different isolate. In order to determine the biological activity of PlxyGV-SA against neonate and fourth instar P. xylostella larvae, surface dose bioassays were conducted. The median lethal concentration of the virus required to kill 50% (LC₅₀) and 90% (LC₉₀) of the larvae was estimated by feeding insects with a range of doses. In addition, the time to kill 50% of the larvae (LT₅₀) was determined by feeding insects with the LC₉₀ concentration. Larval mortality was monitored daily until pupation. The data obtained from the dose response assays were subjected to probit analysis using Proban statistical software. The time response was determined using GraphPad Prism software (version 6.0). The LC₅₀ and LC₉₀ values for the neonate larvae were 3.56 × 10⁵ and 1.14 × 10⁷ OBs/ml respectively. The LT₅₀ was determined to be 104 hours. The neonate larvae were found to be more susceptible to infection than the fourth instar larvae with the same virus concentration. The concentrations used for the neonate larvae assay did not have a significant effect on the fourth instar as no mortality was recorded. This is the first study to describe a novel South African PlxyGV isolate and the results suggest that PlxyGV-SA has significant potential for development as an effective biopesticide for the control of P. xylostella in the field.

Baculoviruses are lytic insect viruses. Upon internalisation, the viral genome orchestrates a sequential expression process ultimately leading to lysis of the infected cell. Release of progeny capable of infecting other cells during the process completes the infection cycle. Studies of the infection cycle in cell culture are typically conducted by synchronous infection, i.e. near simultaneous infection of all cells, by means of high virus concentrations. The behaviour of the synchronously infected culture, such as the timing of onset of progeny release, is considered representative for the infection progression within individual cells. In reality, however, the synchronously infected culture only reflects the average behaviour of all infected cells. The infection progresses in individual cells display large variability; this is most obvious in the observation that within the same culture some cells undergo cell lysis at two days post infection while others remain viable up to four days post infection. Such variabilities or asynchronies observed in synchronously infected culture is the topic of this thesis. Using a simple phenomenological model, it is demonstrated that cell death and associated intracellular product release is adequately described assuming that the waiting time from infection to cell death follows a Gaussian distribution with a mean of 59 hours post infection (hpi) and a standard deviation of 15hpi. Unlike other deterministic models developed over the last decade (Licari and Bailey 1992; Nielsen 2000), this stochastic model does not make the biologically inconsistent assumption that cells continue to be metabolically active following loss of membrane integrity. While elegant in its simplicity, the model provides no explanation for the underlying stochasticity. Investigations into the cause of this dispersion of cell death highlighted further asynchronies in the specific recombinant protein yield, in viral DNA content, in virus budding rate, and in cell volume increase instead of clarifying the issue. A modelling framework developed by Licari & Bailey (1992) and later Hu & Bentley (2000) incorporates the number of infectious particles each individual cell receives as a possible source of the dispersions in the host cell responses. However, this was found NOT to be the cause of the observed asynchronies under non-substrate limiting conditions. The timing of cell death, cell volume increase, recombinant product yield, viral DNA content, and virus budding rate is identical in Sf9 cell cultures infected at multiplicities of infection of ~5, ~15, and ~45 infectious particles per cell. Cell cycle variation has previously been suggested as a possible cause for observed asynchronies in baculovirus infections (Brown and Faulkner, 1975). The cell cycle phase is indirectly linked to the cell volume, because a G_2-phase cell prior to division is inherently twice the cell volume of a G_1- phase cell after cell division. By the same logic, it is also apparent that a G_2-phase cell possesses twice the number of ribosomes of a G_1-phase cell and thus a doubled protein production capacity. The effect of the cell cycle or cell volume on the baculovirus infection was determined by splitting an exponentially growing Sf9 cell culture into 5 cell size dependent fractions by centrifugal elutriation. The subsequent infection of these fractions showed (1) no dependency of the timing of cell lysis and cell volume increase and (2) approximately twofold increase of a) recombinant protein yield, b) viral DNA concentration, and c) budded virus yield. The recombinant protein yield showed a strong proportionality to the initial cell volume and the total RNA concentration during the late phase of the infection. As argued in chapter 6, these proportionalities suggest that the observed differences in the responses of the cell fractions to the baculovirus infection are more likely caused by the difference in the protein production capacity than by cell cycle specific molecules. This investigation gave also reason to speculate that infected cells cannot progress beyond the G_2/M phase, and cell cycle progression continues undisturbed until ~8hpi when all cellular DNA replication appears to cease. Resuspended, infected Sf9 cells synchronised by centrifugal elutriation showed an identical cell cycle distribution as the non-infected control cultures for the first ~8hpi; G_1 and G_2/M phase cell proportions remained unchanged, whereas S phase cells progress to G_2/M phase. Subsequently, the non-infected control cells resumed normal cycling whereas all infected cells remained at the same cell cycle phase from 8 to 11hpi. The initial cell cycle arrests in G_2/M phase in both infected and non-infected cultures were caused through medium exchange.

Baculoviruses are arthropod-specific DNA viruses that are highly virulent to most lepidopteran insects. Their host specificity and compatibility with IPM programmes has enabled their usage as safe microbial insecticides (biopesticides). Two baculovirus-based biopesticides, Cryptogran and Cryptex, which have been formulated with Cryptophlebia leucotreta granulovirus (CrleGV) have been registered for the control of false codling moth (FCM), Thaumatotibia (=Cryptophlebia) leucotreta (Meyrick) (Lepidoptera: Tortricidae) in South Africa and have been successfully incorporated into IPM programmes. However, several studies have indicated that insects can develop resistance to baculovirus-based biopesticide as was shown with field populations of codling moth (CM), Cydia pomonella (L.), which developed resistance to the biopesticide Cydia pomonella granulovirus (CpGV-M) in Europe. Other studies have shown that, under laboratory conditions, FCM populations differ in their susceptibility to Cryptogran and Cryptex. In order to investigate difference in susceptibility as well as protect against any future resistance by FCM to Cryptogran and Cryptex, a search for novel CrleGV-SA isolates from diseased insects from different geographic regions in South Africa was performed. Six geographic populations (Addo, Citrusdal, Marble Hall, Nelspruit, Baths and Mixed colonies) of FCM were established and maintained in the laboratory. Studies on the comparative biological performance based on pupal mass, female fecundity, egg hatch, pupal survival, adult eclosion and duration of life cycle of the Addo, Citrusdal, Marble Hall, Nelspruit and Mixed colonies revealed a low biological performance for the Citrusdal colony. This was attributed to the fact that FCM populations found in the Citrusdal area are not indigenous and may have been introduced from a very limited gene pool from another region. When insects from five colonies, excluding the Baths colony, were subjected to stress by overcrowding , a latent baculovirus resident in the Addo, Nelspruit, Citrusdal, Marble Hall and Mixed colonies was brought into an overt lethal state. Transmission electron micrographs revealed the presence of GV occlusion bodies (OBs) in diseased insects. DNA profiles obtained by single restriction endonuclease analysis of viral genomic DNA using BamH 1, Sa/1, Xba1 , Pst1, Xh01 , Kpn1, Hindlll and EcoR1 revealed five CrleGV-SA isolates latent within the insect populations. The new isolates were named CrleGV-SA Ado, CrleGV-SA Cit, CrleGV-SA Mbl, CrleGVSA Nels and CrleGV-SA Mix isolates. The novelty of the five CrleGV-SA isolates was confirmed by the presence of unique submolar bands, indicating that each isolate was genetically different. PCR amplification and sequencing of the granulin and egt genes from the five isolates revealed several single nucleotide polymorph isms (SNPs) which, in some cases, resulted in amino acid substitutions. DNA profiles from RFLPs, as well as phylogenetic analysis based on granulin and egt sequencing showed the presence of two CrleGV-SA genome types for the CrleGV-SA isolate. Cryptex and CrleGV-SA Ado, CrleGV-SA Cit, CrleGV-SA Mbl and CrleGV-SA Mix were placed as members of Group one CrleGV-SA, and Cryptogran and CrleGV-SA Nels isolate were placed into Group two CrleGV-SA. In droplet feeding bioassays, the median survival time (STso) for neonate larvae inoculated with Group one and two CrleGV-SA were determined to range from 80 - 88 hours (3.33 - 3.67 days), for all five colonies. LDso values for Group one and two CrleGV-SA against neonates from the Addo, Citrusdal, Marble Hall, Nelspruit and Mixed colonies varied between some populations and ranged from 0.80 - 3.12 OBs per larva, indicating some level of variation in host susceptibility. This is the first study reporting the existence of genetically distinct CrleGV baculovirus isolates infecting FCM in different geographical areas of South Africa. The results of this study have broad-ranging implications for our understanding of baculovirus-host interactions and for the application of baculovirus basedbiopesticides.

The potato tuber moth, Phthorimaea operculella (Zeller), is a major pest of potato crops worldwide causing significant damage to both field and stored tubers. The current control method in South Africa involves chemical insecticides, however, there is growing concern on the health and environmental risks of their use. The development of novel biopesticide based control methods may offer a potential solution for the future of insecticides. In this study a baculovirus was successfully isolated from a laboratory population of P. operculella. Transmission electron micrographs revealed granulovirus-like particles. DNA was extracted from recovered occlusion bodies and used for the PCR amplification of the lef-8, lef-9, granulin and egt genes. Sequence data was obtained and submitted to BLAST identifying the virus as a South African isolate of Phthorimaea operculella granulovirus (PhopGV-SA). Phylogenetic analysis of the lef-8, lef-9 and granulin amino acid sequences grouped the South African isolate with PhopGV-1346. Comparison of egt sequence data identified PhopGV-SA as a type II egt gene. A phylogenetic analysis of egt amino acid sequences grouped all type II genes, including PhopGV-SA, into a separate clade from types I, III, IV and V. These findings suggest that type II may represent the prototype structure for this gene with the evolution of types I, III and IV a result of large internal deletion events and subsequent divergence. PhopGV-SA was also shown to be genetically more similar to South American isolates (i.e. PhopGV-CHI or PhopGV-INDO) than it is to other African isolates, suggesting that the South African isolate originated from South America. Restriction endonuclease profiles of PhopGV-SA were similar to those of PhopGV-1346 and PhopGV-JLZ9f for the enzymes BamHI, HindIII, NruI and NdeI. A preliminary full genome sequence for PhopGV-SA was determined and compared to PhopGV-136 with some gene variation observed (i.e. odv-e66 and vp91/p95). The biological activity of PhopGV-SA against P. operculella neonate larvae was evaluated with an estimated LC₅₀ of 1.87×10⁸ OBs.ml⁻¹ being determined. This study therefore reports the characterisation of a novel South African PhopGV isolate which could potentially be developed into a biopesticide for the control of P. operculella.

Cydia pomonella (Linnaeus) (Family: Tortricidae), the codling moth, is considered one of the most significant pests of apples and pears worldwide, causing up to 80% crop loss in orchards if no control measures are applied. Cydia pomonella is oligophagous feeding on a number of alternate hosts including quince, walnuts, apricots, peaches, plums and nectarines. Historically the control of this pest has been achieved with the use of various chemical control strategies which have maintained pest levels below the economic threshold at a relatively low cost to the grower. However, there are serious concerns surrounding the use of chemical insecticides including the development of resistance in insect populations, the banning of various insecticides, regulations for lowering of the maximum residue level and employee and consumer safety. For this reason, alternate measures of control are slowly being adopted by growers such as mating disruption, cultural methods and the use of baculovirus biopesticides as part of integrated pest management programmes. The reluctance of growers to accept baculovirus or other biological control products in the past has been due to questionable product quality and inconsistencies in their field performance. Moreover, the development and application of biological control products is more costly than the use of chemical alternatives. Baculoviruses are arthropod specific viruses that are highly virulent to a number of lepidopteran species. Due to the virulence and host specificity of baculoviruses, Cydia pomonella granulovirus has been extensively and successfully used as part of integrated pest management systems for the control of C. pomonella in Europe and around the world, including South Africa. Commercial formulations have been typically based on the Mexican strain of CpGV. However due to long-term multiple applications of CpGV and the reliance on CpGV in organic farming practices in Europe, resistance to the CpGV-M strain has developed in a number of field populations of C. pomonella. This study aimed to identify and characterize novel isolates of CpGV in South Africa and compare their virulence with the commercial standard CpGV-M. Secondly, since C. pomonella is difficult to culture on a large scale, an alternate method of CpGV production was investigated in order to determine if CpGV could be produced more efficiently and at a reduced cost without negatively impacting the quality of the product. Several isolates of CpGV were recovered either from field collected larvae or from a laboratory-reared C. pomonella colony. Characterisation of DNA profiles using a variety of restriction enzymes revealed that only a single isolate, CpGV-SA, was genetically different from the Mexican strain of the virus used in the commercially available CpGV based products in South Africa. In dose-response bioassays using CpGV-SA, LC₅₀ and LC₉₀ values for neonate C. pomonella larvae were 3.18 x 10³ OBs/ml and 7.33 x 10⁴ respectively. A comparison of these values with those of CpGV-M indicated no significant difference in the virulence of the two isolates under laboratory conditions. This is a first report of a genetically distinct CpGV isolate in South Africa. The biological activity and novelty of CpGV-SA makes this isolate a potentially important tool for CpGV resistance management in South Africa. In order to justify production of CpGV in an alternative host, studies on the comparative biological performance of C. pomonella and T. leucotreta based on oviposition, time to hatch, larval developmental times and rearing efficiency as well as production costs were performed. Thaumatotibia leucotreta was found to be more fecund and to have significantly shorter egg and larval developmental times. In addition, larval production per unit of artificial diet was significantly higher than for C. pomonella. This resulted in T. leucotreta being more cost effective to produce with implications for reduced insectary space, sanitation practices as well as the labour component of production. Virus yield data generated by inoculation both C. pomonella and T. leucotreta with nine concentrations of CpGV resulted in comparable virus yields, justifying the continuation of the research into production of CpGV in T. leucotreta. It was important to determine the LC and LT values required for mass production of CpGV in late instar T. leucotreta larvae. Dose- and time-response bioassays with CpGV-M were conducted on artificial diet to determine these values. Fourth instar LC₅₀ and LC₉₀ values were 5.96 x 10³ OBs/ml and 1.64 x 10⁵ OBs/ml respectively. LT50 and LT90 values were 81.10 hours and 88.58 hours respectively. Fifth instar LC₅₀ and LC₉₀ values were 6.88 x 10⁴ OBs/ml and 9.78 x 10⁶ OBs/ml respectively. LT₅₀ and LT₉₀ values were 111.56 hours and 137.57 hours respectively. Virus produced in fourth instar T. leucotreta larvae was bioassayed against C. pomonella neonate larvae and compared to CpGV-M to establish if production in the heterologous host negatively affected the virulence of the isolate. No significant difference in virulence was observed between virus produced in T. leucotreta and that produced in C. pomonella. The data generated in the bioassays was used in CpGV mass production trials to evaluate production. All production methods tested produced acceptable virus yields. To examine the quality of the virus product, genomic DNA was extracted from larval cadavers and subjected to REN analysis with HindIII. The resulting DNA profiles indicated that the virus product was contaminated with the homologous virus, CrleGV. Based on the above results, the use of T. leucotreta as an alternate host for the in vivo production of CpGV on a commercial basis is not at this stage viable and requires further investigation before this production methodology can be reliable used to produce CpGV. However, this study has shown that CpGV can be produced in a homologous host, T. leucotreta and significant strides have been made towards developing a set of quality control standards that are essential for further development of successful production methodology. Finally a novel isolate of CpGV has been identified with comparable virulence to the CpGV-M. This is an important finding as it has broad reaching implications for resistance management of CpGV products in South Africa.