Genetic Variability of the Whitefly Bemisia Tabaci and Its Secondary Endosymbionts in the Arabian Peninsula

Genetic variability of the whitefly Bemisia tabaci and its secondary endosymbionts in the Arabian Peninsula

Thesis by Ala'a Ragab

In Partial Fulfillment of the Requirements For the Degree of Master of Science

King Abdullah University of Science and Technology Thuwal, Kingdom of Saudi Arabia

May 2013

EXAMINATION COMMITTEE APPROVALS FORM

The thesis of Ala'a Ragab is approved by the examination committee.

Committee Chairperson Prof. Dr. Nina Fedoroff

Committee Co-Chair Dr. Ali Idris

Committee Member Prof. Dr. Arnab Pain

ABSTRACT Genetic variability of the whitefly Bemisia tabaci and its secondary endosymbionts in the Arabian Peninsula Ala'a Ragab

The whitefly Bemisia tabaci species complex has been well documented as one of the most economically important emergent plant virus vectors, through serious feeding damage to its broad range of plant hosts and transmission of plant viruses to important agricultural crops. It has been shown to have associations with endosymbionts which have significant effects on the insect fitness. The purpose of this study was to provide information for the biotype and secondary endosymbiont distribution for B. tabaci populations in the relatively unstudied Arabian peninsula. The geographical localization and variation in endosymbiont populations across the region were identified using a sequence-driven analysis of the population genetics of the secondary endosymbiont. Live field specimens were collected from 22 different locations in the region and preserved in

70% ethanol for genetic studies. Previously established procedures were used to extract and purify total insect DNA from 24-30 individual whiteflies for each location (Frohlich et al., 1999; Chiel et al., 2007). Specimens were subjected to PCR amplification using the respective 16S rDNAprimers for the Rickettsia, Hamiltonella, and Wolbachia to amplify endosymbiont DNA. PCR was run with primers for the highly conserved whitefly mitochondrial cytochrome oxidase subunit I (COI) gene for biotyping. Samples were sequenced using the Sanger method and the data analyzed to correlate the presence, prevalence and geographical distribution of endosymbionts in B. tabaci. Phylogenies 5 were constructed to track evolutionary differences amongst the endosymbionts and insects and how they have influenced the evolution of the regional populations. Samples were characterized by differences in the genomes and endosymbionts of common whitefly ‘biotypes’ that have different host plant preferences, vector capacities and insecticide resistance characteristics. It was found that the B biotype is the predominant haplotype, with no evidence of the Q biotype. All endosymbionts were detected, with

Hamiltonella as the most predominant. Several instances of co-infection by two or more endosymbionts were observed. Samples from the geographically isolated and mountainous region of Fayfa demonstrated higher genetic variability compared to the other locations, leading to the possible identification of a new haplotype, as well as the first time identification of the A biotype in the region.

ACKNOWLEDGEMENTS I would like to thank my committee chair, Professor Nina Fedoroff, and my committee members, Dr. Ali Idris and Professor Arnab Pain, for their guidance and support throughout the course of this research.

My appreciation also goes out to my friends for making my time at King Abdullah

University of Science and Technology a great experience. I would especially like to thank the friends who helped and supported me throughout: Sarah Al Mahdali, Alan Itakura,

William Bass, Jesse Cochran, Octavio Salazar, Alex Barthel and Mariam Awlia.

Finally, my heartfelt gratitude is extended to my parents for their encouragement and support.

TABLE OF CONTENTS Page EXAMINATION COMMITTEE APPROVALS FORM ...... 2 COPYRIGHT PAGE ...... 3 ABSTRACT ...... 4 ACKNOWLEDGEMENTS ...... 6 TABLE OF CONTENTS ...... 7 LIST OF ABBREVIATIONS ...... 8 LIST OF ILLUSTRATIONS ...... 9 LIST OF TABLES ...... 10 1. INTRODUCTION ……………………………………………………………….. 11 1. 1. Biotypes…………………………………………………………… ..... 11 1.2. Virus Transmission………………………………………………….. ... 14 1.3. Endosymbionts ……………………………………………………….. . 14 1.4. Pest Management Strategies………………………………………...... 19 1.5. Current Research…………………………………………………….. ... 20 2. MATERIALS AND METHODS……………………………………………...... 24 2.1. Sample Collection…………………………………………………… ... 24 2.2. DNA Extraction……………………………………………………...... 25 2.3. Primer Design………………………………………………… ...... 25 2.4. Biotype Identification………………………………………………… . 25 2.5. Endosymbiont characterization……………………………………….. . 26 2.6. Gel Electrophoresis………………………………………………….. ... 26 2.7. Cloning…………………………………………………………...... 27 2.8. PCR Product Cleaning……………………………………………….. .. 27 2.9. Sequencing, Sequence Assembly and Data Analysis………………… . 28 3. RESULTS …………………………………………………...... 29 3.1. Identification of B. tabaci biotypes and genetic subgroups………… .... 29 3.2. Detection and identification of the secondary endosymbiotic bacteria .. 29 3.3. Phylogenetic Analyses……………………………………………….. .. 35 4. DISCUSSION………………………………………………………………….. .. 40 5. CONCLUSION & FUTURE WORK………………………………………….. .. 45 REFERENCES…………...... …………………………………………..…... 47 APPENDICES………………………………………………………………….…... 57

LIST OF ABBREVIATIONS mtCOI Mitochondrial cytochrome oxidase subunit I SS Secondary endosymbiont ddH2O Double distilled water PCR Polymerase chain reaction dNTP Deoxyribonucleotide triphosphate DMSO Dimethylsulfoxide rDNA Ribosomal DNA CI Cytoplasmic incompatibility PI Parthenogenesis induction TYLCV Tomato yellow leaf curl virus CLCuV Cotton leaf curl virus IRM Insecticide-resistance management CAPS Cleaved amplified polymorphic sequences bp Base pair PAUP Phylogenetic anaylsis using parsimony

LIST OF ILLUSTRATIONS

Figure 1. Map of Sample Locations ...... 24 Figure 2. Gel of Polymerase Comparison...... 29 Figure 3. Gel of Wolbachia Clone Digest...... 30 Figure 4. Secondary Endosymbiont (SS) Infection Frequency For All Locations...... 33 Figure 5. Secondary Endosymbiont (SS) Infection Frequency in the East and South..... 34 Figure 6. Maximum Likelihood Tree For Mitochondrial Cytochrome Oxidase I (mtCOI) ...... 37 Figure 7. Maximum Likelihood Tree For Wolbachia...... 38 Figure 8. Maximum Likelihood Tree For Hamiltonella...... 39 Figure 9. Histogram Of Endosymbiont Infection Frequency...... 62

LIST OF TABLES Table 1. PCR Primers Used in the Study...... 27 Table 2. Sample Locations with Endosymbiont Distribution...... 31 Table 3. Sample Locations with Endosymbiont (SS) Infection and Co-infection Frequency...... 32 Table 4. PCR Conditions...... 57 Table 5. PCR Reaction Components for Each Primer...... 57 Table 6. Ligation Components...... 58

1. Introduction The whitefly, Bemisia tabaci, (Gennadius; Hemiptera: Aleyrodoidea) is a phloem-feeding global agricultural pest and plant virus vector with great genetic diversity (Dinsdale et al.

2010; De Barro et al. 2011). In 1957, 19 previously described whitefly species (Gill,

1992) were collectively categorized as the single species B. tabaci (Russell, 1958). First recognized in Greece as the tobacco whitefly (Gennadius, 1889), it has been found across the globe, in the US, Africa, Middle East, the Orient, Russia, mainland and Southeast

Asia and South America (Brown et al., 1995). Its geographical diversity and broad host range gave rise to several different common names associated with host plants: i.e. the tobacco, cotton, or sweet potato whitefly. While it causes some damage to crops through direct feeding and promotes the growth of mold by excreting honeydew (a sugar-rich liquid secreted by phloem-feeding insects), the whitefly's major impact on agriculture is through the transmission of plant viruses (Riley and Palumbo, 1995; Palumbo et al.,

2000, Oliviera et al., 2001). This problem is exacerbated by the whitefly's polyphagous nature. It has a broad feeding and/or reproductive plant host range, with more than 600 different host plant species that include tomato, cucumber, hibiscus, beans, tobacco, cotton, pepper, cassava, okra, sweet potato, lettuce and eggplant. (Brown et al., 1995;

Secker et al., 1998; Oliveira et al., 2001).

1. 1. BIOTYPES

B. tabaci is considered a cryptic species complex or sibling species group (Brown, 2010), which means that it is a group of closely related species referred to as biotypes or haplotypes (Mound, 1963; Rosell et al., 1997; Calvert et al., 2001; Gill and Brown,

2010). While morphologically indistinguishable, members of this species complex differ 12 on a molecular level and exhibit full or incomplete reproductive isolation due to reproductive incompatibility (Oliveira et al., 2001; De Barro et al., 2011). B. tabaci is haplodiploid, which means that unfertilized eggs develop into male progeny and fertilized eggs produce the female progeny. In this case, reproductive incompatibility is indicated by a failure to produce uninfected female progeny or a substantial reduction in fecundity and/or the proportion of infected females in the progeny. The species complex has been found to differ in host range (Zang et al. 2006; Xu et al. 2011), insecticide resistance

(Costa et al. 1993; Horowitz et al. 2005; Luo et al. 2010), behavior (Liu et al. 2007;

Crowder et al. 2010; Wang et al. 2010), virus transmission (Bedford et al. 1994; Li et al.

2010; Liu et al. 2010) and interactions with viruses and host plants (Colvin et al. 2006;

Liu et al. 2009; De Barro and Bourne 2010).

The topic of categorizing B. tabaci has been a matter of debate for several decades, with different methods of classification proposed (Russell, 1958; Mound and Halsey, 1978; De

Barro et al., 2011). Further confusion has arisen due to the abundance of biotype designations which are based on genetic markers rather than on biological data that distinguish between biotype boundaries. This has led to a misuse of the term "biotype" and designation of many more biotypes (e.g. Downie (2010) designated at least 36 biotypes) than there actually are based on reproductive incompatibility.

Most recently, Liu et al. (2012), following up on work by Dinsdale et al. (2010) and Hu et al. (2011), summarized available genetic and behavioral data on reproductive incompatibilities derived from genetic studies and observations on mating behavior. They identified 28 biotypes for B. tabaci based on the phylogenetic analysis of mitochondrial 13 cytochrome oxidase I (mtCOI, a highly conserved gene), though field surveys conducted by Chowda-Reddy et al. (2012) appear to indicate more putative species to be added. The biotypes may differ in virus transmission and efficiency, development rate, host utilization, fecundity, dispersal ability, infestation symptoms, insecticide resistance, and endosymbiont populations (Oliveira et al., 2001; Sseruwagi et al., 2005).

Two of the best-known biotypes are B and Q and are the most globally widespread and invasive of the biotypes, while others are generally confined to specific geographical locations. These two biotypes have been responsible for the whitefly’s rising profile as a pest in the past two decades (Brown et al. 1995; Liu et al. 2007; De Barro et al. 2011).

Despite their geographically close origins, the whitefly B biotype (Middle East and Asia

Minor) and the Q biotype (Mediterranean Basin) rarely interbreed due to pre- and post- zygotic reproductive barriers (De Barro et al., 2005; El Baz et al., 2010). They are characterized by different phenotypes. The B biotype (also known as B. argentifolii) is associated with higher fecundity and competitive abilities and has been shown to induce silver leafing (Cohen et al., 1992), whereas the Q biotype has a greater degree of pesticide resistance (Horowitz et al., 2005). Silverleafing describes a disorder that occurs due to whitefly feeding on leaves. Salivary toxins lead to a silvery appearance to leaves

(due to the upper epidermis separating from the lower cells), which can lead to stunted plants with reduced yield and fruit quality (Jimenez et al., 1995; Brown et al., 1994).

The A biotype, indigenous to the New World and previously a serious problem in the southwestern US and Mexico (about 1981), was by 1991 displaced by the B biotype.

Later this biotype was introduced to new locations and became as a serious pest 14 worldwide (Costa et al., 1993; Brown, 1994; Brown et al., 1995; Brown et al., 1992).

While the A and B biotypes are almost equally sensitive to the organophosphate profenphos, the B biotype is more resistant to permethrin (Costa et al, 1993).

1.2. VIRUS TRANSMISSION

B. tabaci is a well known plant virus vector, transmitting distinct virus genera in the families Geminiviridae, Closteroviridae, Comoviridae, Flexiviridae, and Potyviridae, and

Luteoviridae (Duffus 1987, 1996; Brown, 1994; Jones, 200). Of these, the geminiviruses

(Genus Begomovirus) and closteroviruses (Genus Crinivirus) are the most economically significant in terms of the damage they cause, especially the tomato-, bean- and cassava- infecting begomoviruses. Tomato yellow leaf curl begomovirus (TYLCV) causes massive crop yield and quality losses for tomatoes, the world's highest value vegetable crop. The fact that the global tomato market is over $30 billion (FAO, 2007) sheds light on just how economically damaging these viruses can be. Other examples of viruses transmitted by B. tabaci include Bean calico mosaic virus (BCMoV) (Brown et al., 1999) and African cassava mosaic virus (ACMV) (Harrison and Robins, 1988). B. tabaci is the only vector for begomoviruses (Gottlieb et al., 2010) due to a chaperone GroEL protein that binds the virus particles and protects them from degradation in gut and hemolymph (Hobenhout et al., 2008).

1.3. ENDOSYMBIONTS

Phloem-feeding insects (aphids, whiteflies, psyllids, scales and mealybugs) harbor bacterial endosymbionts with both obligate (primary) and facultative (secondary) relationships (Moran, 2001). Obligate primary endosymbionts (P-endosymbiont) 15 synthesize essential non-dietary metabolites and are maternally vertically transmitted.

The P-endosymbiont in the whitefly is Portiera aleyrodidarum and it is localized in specialized bacteriocyte cells in the bacteriome of the insects (reviewed in Baumann,

2005).

A range of secondary endosymbionts is also associated with phloem-feeding insects, seven of which have been identified in B. tabaci: Rickettsia, Wolbachia, Hamiltonella defensa, Cardinium, Arsenophonus, Fritschea bemisia (reviewed in Baumann, 2005;

Gottlieb et al., 2006) and most recently Hemipteriphilus asiaticus (Bing et al., 2012).

Secondary endosymbionts are not necessarily localized in the bacteriocytes; they may be found throughout the host (Moran and Telang, 1998). Transmission is mainly vertical, though horizontal transmission can occur through contact with other infected individuals

(Buchner, 1965; Baumann et al., 2006; Clark et al., 2010).

While not essential for survival, endosymbionts play an important role in the ecology and evolution of the insects by affecting important characteristics such as insecticide resistance, viral transmission (Morin et al., 1999; Gottlieb et al., 2010) and reproduction and development. Reproductive manipulation has been documented for Wolbachia,

Cardinium (Werren et al., 2000; Engelstädter & Hurst, 2009) and Arsenophonus (in wasps, not documented yet in B. tabaci, Werren et al., 1991).

Endosymbionts may play a role in the evolution of their hosts, which is particularly theorized in cases of reproductive manipulators that cause reproductive isolation and speciation. Considering its current status as a species complex, and the debate 16 surrounding its taxonomy, this seems to be the ongoing case with B. tabaci. (Costa et al.,

1995; Zchori- Fein & Brown, 2002; Nirgianaki et al., 2003).

RICKETTSIA

Rickettsia was first identified in B. tabaci by Gottlieb et al in 2007. It is found in both B and Q biotypes and, unlike other symbionts, can be found in most of the whitefly's tissues and organs as opposed to being confined to the bacteriome. It has been associated with improved whitefly fitness through increased fecundity, faster development times, greater survival into adulthood and an increased proportion of females (Himler et al., 2011).

Rickettsia has additionally been shown to confer heat tolerance (Brumin et al., 2011), but increased susceptibility to some insecticides where high densities of Rickettsia come at a physiological cost to the whitefly (Kontsedalov et al., 2008). In fact, the highest susceptibility to acetamiprid, thiamethoxam, sporimesifen and pyripoxen was found in double infected (Rickettsia-Arsenophonus or Rickettsia-Wolbachia) strains (Ghanim and

Kontsedalov, 2009). This increased susceptibility has been demonstrated in both B (Chiel et al., 2007) and Q biotypes (Ghanim and Kontsedalov, 2009).

WOLBACHIA

Wolbachia pipentis (ɑ-Proteobacteria) of the B division of the genus Wolbachia is maternally transmitted through the cytoplasm of eggs as well as being transmitted horizontally between arthropod species. Wolbachia is localized in its hosts' reproductive organs (Laven, 1967; Merçot, 1998), and has been well documented as a reproductive manipulator. Stouthamer et al. (1999) reviewed the four main reproductive abnormalities that have been attributed to it: 17

1. Cytoplasmic incompatibility (CI): The most common effect, it is a crossing incompatibility between infected males and uninfected females, which results in embryo mortality. This can also be bidirectional, with an incompatibility developing between individuals harboring different Wolbachia strains, and is phenotypically observed as mortality of offspring or with male-biased sex ratios. CI acts in favor of affected females by increasing their relative fitness to their uninfected counterparts, leading to the spread of this maternally transmitted infection (Callaini et al., 1997; Laven and Aslamkhan,

1970).

2. Parthenogenesis induction (PI): infected virgin females produce daughters (O’Neill et al., 1997).

3. Feminization: infected genetic males reproduce as females. Additionally, the bacteria act to increase fecundity and fertility of hosts as well as their pathenogenicity (Merçot,

1998).

4. Male killing: male embryos die while female embryos survive, developing into infected females ( O'Neil et al., 1997).

W. pipentis has been shown to have a number of different roles in bedbugs (provides nutrients, Hosokawa et al., 2010), leaf-mining moth (increases fitness, Kaiser et al.,

2010), uzi flies and mosquitoes (Dobson et al., 2002; Guruprasad et al., 2011), drosophila, Culex quiquefasciatus, Aedes aegypti (suppress replication of RNA viruses,

Hedges et al., 2008; Teixeira et al., 2008; Glaser & Meola, 2010; Moreira et al., 2009).

Though its role in B. tabaci has been relatively unstudied, Wolbachia has been identified in B. tabaci (Zchori-Fein & Brown, 2002; Nirgianaki et al., 2003; De Barro, 2005) and 18 has been recognized as a cause of CI amongst its biotypes (De Barro and Hart, 2000). In

Israel, it has been found in the Q biotype and not the B biotype (Chiel et al., 2007).

A recent study by Xue et al. (2012) investigated the role of Wolbachia in whitefly development and reproduction by inactivation with an antibiotic (rifampicin) and comparing the biology of infected and uninfected insects. They demonstrated that

Wolbachia is associated with increased fitness of whitefly hosts, playing a role in development duration, reproduction and defense against parasitism. The presence of W. pipentis promoted faster development, and increased survival and larger size of infected progeny, possibly through some sort of nutritional or immune benefits.

HAMILTONELLA

Hamiltonella defensa is a mutualistic symbiont with a host-dependant metabolism, relying on its host obligate endosymbionts for nutritional requirements (Degnan et al.,

2009). It has been found to facilitate plant virus transmission in the case of tomato yellow leaf curl virus (TYLCV) (Gottlieb et al., 2010) through the production of GroEL protein, which interacts with the TYLCV coat proteins (in B biotype only). H. defensa also confers protection against parasites, encoding toxins, effector proteins and virulence factors. Its presence in hosts is positively correlated to parasitoid presence, with an increased infection rate (Degnan et al., 2009). In Israel, it has been found in the B biotype, not the Q (Chiel et al., 2007).

ARSENOPHONUS

Arsenophonus (γ- Proteobacteria) has also been documented in several other arthropods including whiteflies, aphids, psyllids and louse flies (Baumann, 2005; Dale et al., 2006). 19

It plays a role in virus transmission in B. tabaci as shown by Rana et al. (2012) who found that the GroEL protein of Arsenophonus interacted with the coat protein of Cotton leaf curl virus (CLCuV). In Israel, it has been found in the Q biotype, not the B (Chiel et al., 2007).

CARDINIUM

Cardinium hertigii, like Wolbachia, has been documented as causing CI (Weeks et al.,

2003). Though they both share some proteins with the potential to interfere with eukaryotic cell cycle regulation, there does not appear to be a common evolutionary origin for CI in the two bacteria, with different molecular mechanisms for CI (Penz et al.,

2012).

FRITSCHEA

Candidatus Fritschea tabaci (Chlamydiales) was first identified in 2003 by Thao et al., and unlike the other endosymbionts, has thus far only been documented in the gut of B. tabaci (but not in the whitefly B biotype). Not much is known about its phenotype

(Everett et al., 2005).

HEMIPTERIPHILUS

Hemipteriphilus asiaticus is the most recently identified endosymbiont in whitefly, identified by Bing et al. (2012) in the China 1 B. tabaci species.

1.4. PEST MANAGEMENT STRATEGIES

The use of chemical insecticides is a basic component in current strategies to manage B. tabaci as a pest (Castle et al., 1996; Dennehy et al., 1996; Palumbo et al., 2001). One of 20 the most effective insecticides is imiacloprid, a chloronicotinyl systemic insecticide

(Elbert et al., 1990; Mullins, 1993; Palumbo et al., 2001). However, this has driven the evolution of widespread insecticide resistance, which is a feature of the Q biotype (also a begomovirus vector) and thus it is essential to adopt insecticide-resistance management

(IRM) programs (Castle et al., 1996; Dennehy et al., 1996; Ellsworth et al., 1996a,

1996b; Prabhaker et al., 1996, 1998; Naranjo et al., 1998). IRMs can include the use of insect growth regulators, which delay or reduce resistance, or alternating chemicals and restricting the number of application per crop season (Palumbo et al., 2001). Current strategies are insecticide-driven, but efforts are being made to integrate cultural, biological and non-chemical methods; the combination of controlled strategies has been the most effective (Ellsworth and Martinez-Carrillo, 2001). Other options that could be integrated include natural enemy augmentation and conservation (Gerling et al., 2001;

Naranjo, 2001) and insect exclusion and reflective materials that would physically repel insects. The latter method has been seen to be partially effective by Hilje et al. (2001).

The current goal in the management of B. tabaci is creating sustainable and ecologically based management systems.

1.5. CURRENT RESEARCH

The biotype-specific distribution of endosymbionts is significant due to their effect on the phenotypic characteristics and fitness of B. tabaci. For example, it has been shown that the Q biotype is more insecticide resistant compared to the B biotype, with Rickettsia being linked as a reason to why the B biotype is less resistant (Ghanim and Kontsedalov,

2009). With the increasing insecticide resistance, which has thus far been the main defense against the insects, and as endosymbiont presence in whiteflies has been shown 21 to modify the host phenotype and influence its evolution, it is important to study bacterial endosymbionts' associations. Unfortunately, research into the relationship between B. tabaci and its endosymbionts has been limited due to difficulties in in vitro culturing of the majority of the endosymbionts (Houk & Griffiths, 1980; Douglas, 1989; Wilkinson,

1998). Investigations of their functions has thus been limited to in vivo examination using antibiotic treatments which can reduce or even eliminate the bacteria of interest

(Wilkinson, 1998) and to studies that simply characterize the endosymbiont populations of different biotypes and host associations, such as this current study.

In light of the considerable effect secondary endosymbionts have been demonstrated to have on host fitness, they are the focus of increasing research, in the search for possible pest management strategies. The whitefly has been the subject of many population genetics publications examining its full genetic composition, heterozygosity, and most recently, transcriptomic level for different regions around the world including China (Xie et al., 2012), Israel (Chiel et al., 2007), and Jordan (Atteyat et al., 2009). Fewer studies exist that focus specifically on secondary endosymbionts, though a recent study by Bing et al. (2012) revealed the diversity of symbiont association between native and invasive whiteflies in China. Gnankiné et al. (2012) provided biotype distribution and symbiotic bacteria community composition data across three West African countries (Burkina Faso,

Benin and Togo).

There is very little data available for Saudi Arabia and the surrounding region, where tomato is a very important crop and the whitefly is an emergent agricultural pest. The purpose of this study was to provide information for the biotypes and secondary 22 endosymbiont populations for B. tabaci in the relatively unstudied Arabian peninsula.

The geographical localization and variation in endosymbiont populations across the region were identified using a sequence-driven analysis of the population genetics of the secondary endosymbiont.

Live field specimens were collected from 22 different locations in the region (Figure 1) and preserved in 70% ethanol. Previously established procedures (Frohlich et al., 1999;

Chiel et al., 2007) were used to extract and purify total insect DNA from at least 24-30 individual whiteflies from each location. The respective 16S rDNA primers for the

Rickettsia, Hamiltonella, and Wolbachia endosymbionts were used to amplify endosymbiont DNA from each population by PCR. These primers were designed to match primers used in the literature so that results would be comparable, as well as the fact that these designs have been proven to effectively amplify the specific sequences for each endosymbiont. Additionally, PCR was run with primers for the whitefly mitochondrial cytochrome oxidase subunit I (COI) gene which acts as a molecular marker for within-species genetic diversity and thus can be used for biotyping. Samples were sequenced using the Sanger method and the data analyzed to correlate the presence, prevalence and geographical distribution of endosymbionts in Bemisia. Phylogenies were constructed to track evolutionary differences amongst the symbionts and insects and how they have influenced the evolution of the regional populations.

Based on the sequences obtained for each location, population samples were characterized into different biotypes and their respective endosymbiont populations were identified. It was found that the B biotype is most prevalent in the region, with a minor 23 presence of A and Q-like biotypes. Samples from the mountainous Fayfa region near the

Yemeni border showed higher genetic variability those from the other locations, forming a separate subclade. All locations were shown to harbor at least one of the endosymbionts screened, with several incidences of co-infection by two or more endosymbionts. The most prevalent endosymbiont was found to be Hamiltonella defensa.

2. Materials and Methods 2.1 Sample collection

B. tabaci samples (n=24-31) were collected from wild and greenhouse populations from

22 localities in Saudi Arabia and neighboring countries from the region (Kuwait, Qatar,

Bahrain, Sudan) throughout 2011 to 2012 (Figure 1 and 2) . The whiteflies were stored in 70% ethanol at room temperature until DNA extraction.

Figure 1. Map of Sample Locations. A map of the 22 locations sampled in the study. Image retrieved from www.ephotopix.com.

2.2. DNA extraction

Total DNA was extracted from each individual adult specimen using the method described in Frohlich et al. (1999). For each location, at least eight individual insects (in triplicate, i.e. twenty four samples) were selected randomly and DNA was isolated from each separately by crushing in 30 µl of freshly prepared lysis buffer (See Appendix) after ethanol was rinsed off with double distilled H2O (ddH2O). DNA extraction was completed with a thermocycle run at: 65°C for 20 min / 95°C for 10 min (C1000

Thermocycler, Bio-Rad, California, USA). The extract was spun down at 3000 rpm for 5 min (Eppendorf Centrifuge 5810R, Germany) and the lysate supernatant was stored at -

20°C for use in subsequent reactions.

2.3. Primer Design

Primers were designed for the mitochondrial cytochrome oxidase 1 (mtCO1) molecular marker region using the cleaved amplified polymorphic sequences (CAPS) method by

Khasdan et al., 2005 due to its reliability and relative simplicity. The primers were compared to other widely used primer designs and used to identify B and Q biotypes in several laboratory and population dynamic experiments (Khasdan et al., 2004; Horowitz et al., 2005) since. They were shown to be B and Q biotype specific. All primers were between 18-30 nucleotides long and previously described and confirmed to be specific as shown in Table 1.

2.4. Biotype Identification

To confirm the integrity of DNA extraction and for biotyping, the mtCOI gene was amplified using specific primers MD10 and MD12 using Taq polymerase (Table 1). The 26 cycling conditions were: an initial denaturation at 95 ºC for 2 min, followed by 34 cycles of 95 ºC for 30 s, 52 ºC for 30 s and 72 ºC for 2 min and a final extension of 72 ºC for 10 min.

2.5. Endosymbiont characterization

Specific primers were used to amplify Wolbachia, Hamiltonella and Rickettsia (Table 1).

All PCR analyses were conducted using either Taq or Pfu polymerase. The cycling conditions are outlined in the table below, though generally they were: an initial denaturation at 95 ºC for 2 min, followed by 37 cycles of 95 ºC for 30 s, 45-55 ºC for 30 s and 72 ºC for 2 min and a final extension of 72 ºC for 10 min, with variations on the appropriate annealing temperature for each primer.

Arsenophonus, Cardinium and Fritschea primers (Table 1) were also run with a sampling of insects; however no positive control could be established to confirm if samples were negative or the reaction was simply not optimized.

2.6. Gel Electrophoresis

All PCR products were resolved by gel electrophoresis and stained with ethidium bromide. A mixture of 5 µl PCR product, 1 µl 10x loading buffer and 4 µl ddH2O were loaded into a 1% agarose gel and running in TAE buffer at 100 V for 40 min or 120 V for

30 min.

DNA was subsequently isolated from either a gel using the QIAgen MinElute Gel

Extraction Kit (Qiagen, Hilden, Germany) or directly cleaned using the QIAgen PCR

Purification Kit (Qiagen, Hilden, Germany), according to manufacturer’s instructions

(See Appendix). 27

Table 1. PCR Primers Used In The Study A summary of the target genes and their primers used in the study, with annealing temperatures and expected product sizes shown. a. Khasdan et al., 2005, b. Gottlieb et al., 2006, c. Zchori-Fein & Brown, 2002, d. Heddi et al., 1999, e. Thao & Baumann, 2004, f. Weeks et al., 2003, g. Everett et al., 2005.

Annealing temp. Target gene Primer Sequence (5' - 3') / Product size

B. tabaci MD 10 TTGATTTTTTGGTCATCCAGAAGT a 52°C / ~800 bp MtCOI MD 12 TCCAATGCACTAATCTGCCATATTA Rickettsia Rb-F GCTCAGAACGAACGCTATC b 60°C / ~900 bp 16S rDNA Rb-R GAAGGAAAGCATCTCTGC Hamiltonella Ham-F TGAGTAAAGTCTGGAATCTGG c 60°C / ~700 bp 16S rDNA Ham-R AGTTCAAGACCGCAACCTC Wolbachia Wol16S-F CGGGGGAAAAATTTATTGCT d 55°C / ~700 bp 16S rDNA Wol16S-R AGCTGTAATACAGAAAGTAAA Arsenophonus e Ars23S-1 CGTTTGATGAATTCATAGTCAAA 23S rDNA 60°C / ~600 bp Ars23S-2 GGTCCTCCAGTTAGTGTTACCCAAC Cardinium CFB-F GCGGTGTAAAATGAGCGTG f 58°C / ~400 bp 16S rDNA CFB-R ACCTMTTCTTAACTCAAGCCT Fritschea GATGCCTTGGCATTGATAGGCGATGAAG g U23F 23S rDNA GA 60°C / ~600 bp 23SIGR TGGCTCATCATGCAAAAGGCA

2.7. Cloning To confirm the identity of the PCR products, six samples that were positive for

Wolbachia were cut from a gel, eluted and cloned into the pGEM T-Easy plasmid vector

(Promega) and transformed into E. coli. For each bacterium, three colonies were randomly picked and sequenced using the Sanger method (ABI-3730xl DNA analyzer).

The resulting sequences were compared to known sequences in the Genbank database using the BLAST algorithm in NCBI. The direct PCR products were sequenced alongside the clones. Results were equivalent, therefore the rest of the samples were simply confirmed using the PCR product for sequencing (See Appendix for more details on the cloning process).

2.8. PCR Product Cleaning PCR products were cleaned through a column and submitted for Sanger sequencing.

Depending on whether they appeared as single or multiple bands, they were either loaded 28 onto a gel and then cut and cleaned using the QIAgen Gel Extraction kit or loaded directly onto columns and cleaned with the QIAgen PCR Purification Kit as described by the manufacturer (See Appendix). Generally, samples were cleaned directly with no need for gel extraction using the PCR purification kit.

2.9. Sequencing, Sequence Assembly and Data Analysis

The direct PCR products for positive samples (determined from the bulk screening) were sequenced in both forward and reverse directions. The PCR primers were used as sequencing primers. Between one and seven samples per whitefly location were selected, thus a consensus sequence was generated for each location. Additionally, variants to the consensus were selected for further investigation. The obtained nucleotide sequences were assembled using SeqMan software (available in the Lasergene, software package,

DNASTAR Inc., Madison, WI, USA) and used to search the GenBank database for closely related variants.

The sequences, along with selected reference sequences from the GenBank search results, were aligned using the Clustal V method in MegAlign and included in the phylogenetic analyses. The aligned sequences were used to reconstruct the phylogenetic trees using maximum likelihood (ML) options available in PAUP* software (phylogenetic anaylsis using parsimony, Swofford 2002) as described before (Idris and Brown, 2004). Similarly for endosymbiont characterization, assembled sequences were aligned with reference sequence drawn from previously published data on GenBank to confirm the identity of each endosymbiont and for phylogenetic analysis.

3. Results

3.1. Identification of B. tabaci biotypes and genetic subgroups

Pfu enzyme was compared to the Taq enzyme with the COI primer, and as shown in

Figure 1, the results were comparable. Taq was used for several of the primers (See Table in Methods), including COI. PCR products generated were sequenced in both directions and the phylogenetic tree in Figure 5 was generated.

Figure. 2. Gel of Polymerase Comparison Pfu and Taq polymerase were compared by running a PCR for mtCOI under the conditions described in the Appendix. The gel is 1% agarose, run at 100 V for 40 min and stained with ethidium bromide with 1 kb ladder (Lane 1). The product band appears at between 1000 and 1500 bp, slightly higher than the expected product size of 800 bp. Lanes 2-5 are the Pfu products and a negative control, while Lanes 6-9 are the Taq products with a negative control. Primer dimers can be seen at the bottom of the gel.

1 kb ladder 1 2 3 4 5 6 7 8 9

1500 bp 1000 bp 700 bp

500 bp

3.2. Detection and identification of the secondary endosymbiotic bacteria (SS)

Figure 2 shows the miniprep EcoRI digests for Wolbachia cloned samples, which were sequenced alongside their direct PCR product counterpart. The vector can be seen at 3.0 kb, while the WOL clone band appears at about 500 bp, slightly lower down than its expected size of ~700 bp, though sequences returned were about 592 bp.

Figure 3. Gel of Wolbachia Clone Digest The gel shows the product of EcoRI digestion of Wolbachia clones. The gel is 1% agarose, run at 100 V for 40 min and stained with ethidium bromide with 1 kb Plus ladder (Lane 1). The digest band appears at 500 bp, while the empty vector appears between 2.0 and 5.0 kb, with an expected size of 3.0 kb. Lanes 2-12 are four samples selected from Bahrain cucumber 1.

1 kb Plus ladder 1 2 3 4 5 6 7 8 9 10 11 12

5000 bp 2000 bp

650 bp 500 bp 400 bp 200 bp

Table 2 is a summary of the different locations included in the study, along with their respective plant host and biotype. The association of each endosymbiont to each location is shown. No correlation was apparent between plant host and the SS population distribution.

Almost all locations were found to have at least one positive sample for all three SS, some of the them showing co-infections as well. All but one location (Abha) were positive for Hamiltonella. Hofuf (eggplant), Hofuf (cucumber) and Dammam (eggplant) are all negative for Wolbachia, though positive for Hamiltonella and Rickettsia. Fayfa and Salama are negative for Rickettsia, but positive for the others. Abha is the only location with just one type of SS (Wolbachia) present. Triple co-infections were quite rare, while the R-H double infection was the most common symbiont combination.

Table 2. Sample Locations With Endosymbiont Distribution. The table lists all the sampled locations, along with their respective host plant. The biotype and endosymbiont presence was determined throughout the study, with an endosymbiont considered present if at least one of the samples was positive for it .R, Rickettsia; H, Hamiltonella; W, Wolbachia.

Endosymbiont Population Host Plant Biotype R H W KAUST, KSA Tomato B + + + Om Al Ard, KSA Cucumber B + + + Riyadh, KSA Tomato B + + + Hofuf 1, KSA Cucumber B + + - Hofuf 2, KSA Eggplant B + + - Najran, KSA Pepper B + + + Guaz, KSA Cucumber B + + + Fayfa, KSA Cucumber Q-like - + + Fayfa, KSA Cucumber A-like - + - Salama, KSA Eggplant B - + + Abha, KSA Zucchini B - - + Al Ahsa, KSA Eggplant B + + + Dammam, KSA Cucumber B + + + Dammam, KSA Eggplant B + + - Dammam Al Hussain, KSA Hibiscus B + + + Qatar Eggplant B + + + Qatar Tomato B + + + Bahrain Eggplant B + + + Bahrain 1 Cucumber B + + + Bahrain 2 Cucumber B + + + Kuwait Tomato B + + + Kuwait Wafra Tomato B + + + Al Gezirah, Sudan Cotton B + + +

Table 3 shows a more detailed view of the distribution of infection frequencies for the locations. The endosymbiont columns refer to total presence of each of the SS; whereas the combinations are a break down into co-infected individuals. Interesting results can be seen with Fayfa, one of the few locations where Rickettsia is absent, and Najran, which shows 100% infection frequency with Hamiltonella amongst sampled individuals.

Bahrain and Qatar both show the highest percentage of Rickettsia infections. As previously noted from Table 2, co-infections are rare, though the R-H combination is the most common of these. Abha has the lowest infection status of all the locations, with only

1 sample (3%) of the total 30 individuals positive for Hamiltonella. Wolbachia was the most prevalent species in three locations: Sudan (with the highest incidence at 71%), 32

followed by KAUST (68%) and Riyadh (44%). This is more clearly illustrated in Figure

Table 3. Sample Locations With Endosymbiont (SS) Infection and Co-infection Frequency. The table lists all sampled locations with respect to their infection frequency status. The "Endosymbiont" columns refer to the total percent of individuals infected with the respective SS, including those co-infected. "Combination " represents individuals with different types of co-infections only. N, Number of individuals, R, Rickettsia; H, Hamiltonella; W, Wolbachia. Standard error for 95% confidence intervals. Percentages have been rounded up from two decimal values.

Species Combination Location N R H W RH only RW only HW only RHW Unique None Total

ABHA 30 0% 0% 3% ±7 0% 0% 0% 0% 3% ± 7 97% ± 7 1% ± 2 AHSA 30 40% ± 18 83% ± 14 3% ± 7 33% ± 18 0% 3% ± 7 0% 53% ± 18 10% ± 11 42% ± 10 BAHRAIN 30 53% ± 18 20% ± 15 43% ± 18 7% ± 9 23% ± 16 3% ± 7 0% 50% ± 18 17% ± 14 39% ± 10 CUCUMBER 1 BAHRAIN 30 50% ± 18 100% ± 0 13% ± 13 50% ± 18 0% 13% ± 13 0% 37% ± 18 0% 54% ± 10 CUCUMBER 2 BAHRAIN 28 68% ± 18 21% ± 16 7% ± 10 18% ± 15 4% ± 7 0% 0% 54% ± 19 25% ± 17 32% ± 10 EGGPLANT DAMMAM 27 22% ± 16 22% ± 16 7% ± 10 7% ± 10 0% 0% 0% 37% ± 19 56% ± 19 17% ± 8 CUCUMBER DAMMAM 24 16% ± 14 42% ± 20 0% 8% ± 12 0% 0% 0% 42% ± 21 50% ± 21 19% ± 9 EGGPLANT DAMMAM 24 29% ± 18 25% ± 18 8% ± 11 0% 0% 0% 0% 63% ± 20 38% ± 20 21% ± 10 HIBISCUS FAYFA 25 0% 60% ± 20 20% ± 16 0% 0% 16% ± 15 0% 48% ± 20 36% ± 20 27% ± 10 GUAZ 29 13% ± 13 17% ± 14 28% ± 17 3% ± 7 0% 0% 3% ± 7 41% ± 19 52% ± 19 20% ± 8 HOFUF 1 29 40% ± 18 30% ± 17 0% 13% ± 13 0% 0% 0% 43% ± 18 43% ± 18 23% ± 9 HOFUF 2 26 7% ± 11 50% ± 20 0% 0% 0% 0% 0% 58% ± 19 42% ± 20 19% ± 9 KAUST 31 29% ± 16 23% ± 15 68% ± 17 0% 6% ± 9 16% ± 14 3% ± 6 65% ± 17 10% ± 11 40% ± 10 KUWAIT 24 33% ± 20 8% ± 11 4% ± 8 4% ± 8 4% ± 8 0% 0% 29% ± 19 63% ± 20 15% ± 8 TOMATO KUWAIT 28 39% ± 18 11% ± 12 25% ± 17 0% 10% ± 12 0% 4% ± 7 43% ± 19 43% ± 19 25% ± 9 WAFRA NAJRAN 24 17% ± 15 100% 21% ± 17 4% ± 8 0% 8% ± 12 13% ± 14 75% ± 18 0% 46% ± 12 OM AL ARD 28 39% ± 19 21% ± 16 29% ± 17 7% ± 10 11% ± 12 7% ± 10 0% 39% ± 19 36% ± 18 30% ± 10 QATAR 29 63% ± 18 63% ± 18 27% ± 16 30% ± 17 10% ± 11 0% 10% ± 11 43% ± 18 7% ± 9 51% ± 11 EGGPLANT QATAR 24 17% ± 15 21% ± 17 13% ± 14 0% 0% 0% 0% 50% ± 20 50% ± 21 17% ± 9 TOMATO RIYADH 30 7% ± 10 22% ± 16 44% ± 20 0% 3% ± 7 7% ± 9 0% 47% ± 19 33% ± 18 24% ± 9 SALAMA 30 0% 17% ± 14 17% ± 14 0% 0% 0% 0% 33% ± 18 13% ± 18 11% ± 7 SUDAN 24 33% ±19 42% ± 20 71% ± 19 0% 21% ± 17 29% ± 19 0% 46% ± 21 8% ± 8 49% ± 12

TOTAL 604 28% ± 3 36% ± 3 21% ± 3 8% ± 2 4% ± 2 5% ± 2 2%± 1 45% ± 4 35% ± 3 28% ± 2

Figure 9 (See Appendix) is a histogram that further illustrates the endosymbiont distribution amongst the sampled locations. It is obvious from it that the presence of

Hamiltonella was dominant for most of the samples except for those locations referred to within this text.

Figure 4. Secondary Endosymbiont (SS) Infection Frequency For All Locations. The bar chart summarizes the different endosymbiont combinations infection frequency within individuals of B. tabaci collectively for all locations represented in the study. R, H and W refer to the total percent of individuals infected with the respective SS, including those that are co-infected. The combinations represent the individuals with different types of co-infections only. R, Rickettsia; H, Hamiltonella; W, Wolbachia (n=604).

Frequency of symbiont combinaons

15 Infecon Frequency % 10

0 R H W RH RW HW RHW None Symbiont combinaons

When considering the locations by region (classified as Eastern and Southern), several differences were apparent. The majority of individuals from both regions were generally free of infection. However, when comparing those that are infected, the east showed a high frequency of infection and greater variation in endosymbiont combinations of infection and co-infection (Figure 4a). The presence of Rickettsia was very low in the 34 southern region (Figure 4b). Both regions generally showed the same trend, but the higher Rickettsia count in the east allowed for the co-infection with Hamiltonella to dominate in the co-infection category.

Figure 5. Secondary Endosymbiont (SS) Infection Frequency in the East and South. Bar charts summarizing the different endosymbiont combinations infection frequency in the southern and eastern region represented in the study.a. Infection Frequency of SS in the Eastern Region b. Infection Frequency of SS in the Southern Region. R, Rickettsia; H, Hamiltonella; W, Wolbachia. a. 70.0 QATAR EGGPLANT 60.0 50.0 KUWAIT TOMATO 40.0 BAHRAIN CUCUMBER 1 30.0 BAHRAIN EGGPLANT 20.0 OM AL ARD

Infecon Frequency % 10.0 BAHRAIN CUCUMBER 2

0.0 HOFUF 2 R H W RH RW HW RHW None HOFUF 1 Endosymbiont combinaon

b 100.0 90.0 80.0 70.0 GUAZ 60.0 50.0 FAYFA 40.0 SALAMA 30.0 ABHA

Infecon Frequency % 20.0 10.0 NAJRAN 0.0 R H W RH RW HW RHW None

Endosymbiont combinaon

3.3. Phylogenetic Analyses

Figure 5 shows the maximum likelihood (ML) phylogenetic tree generated from mtCOI sequences for the all the locations. Reference sequences from Japan, Morocco, Korea and

Jatropha were also added as representatives of three biotypes (A, B and Q). The tree splits into three clades for each of the respective biotypes, along with an outgroup i.e. a group that is related to the other clades but not as closely as they are related to each other.

The B and Q biotypes are rooted from the same point. The samples for the locations included in this study were all of the B biotype except for the samples from Fayfa, which comprise their own clade between the A and Q biotype leading from the branch that divides into the B and Q. The B biotype clade is made up of several subclades, with all the other samples equally divided amongst them, showing an even distribution of samples for each location and no apparent genetic distinction for any one sample location.

Figure 6 shows the ML tree generated from the 16S rDNA sequences of Wolbachia of B. tabaci along with references from India and other species (Diaphorina, Nasonia,

Mesaphorura) as representatives of other clades. The tree splits into different clades, with most of the locations from the study confined to one clade. Two samples from Fayfa are included, one branching from the clade that represents the majority of the locations and close to the references from India. The other sample branches from a very different clade amongst other species. The samples from Najran are also distinct from the rest of the samples, sharing a subclade with a non-B biotype B. tabaci sample from China.

Figure 7 shows the ML tree generated from the 16S rDNA sequences of H. defensa along with references from different locations including China, Bangladesh, South Korea and 36

France. The tree splits into four clades, one containing the majority of the samples including all of the reference sequences. The other three clades are outgroups formed by samples from Dammam (eggplant) and Kuwait (tomato), Fayfa and Najran. No reference sequences could be retrieved from Genbank for these clades due to their low substitution rate.

As mentioned in the Materials and Methods section, PCR products were cleaned and directly sequenced in the forward and reverse direction using the PCR primers. Sanger sequencing was chosen for this study due to the read length of generated sequences.

Since all of the regions targeted for amplification were between 400-900 bp (reads with

Sanger sequencing are routinely 800-1000 bp) assembled sequences were more reliable using this method. However, sequences generated by the Sanger method generally show poor quality in the first 15-40 bases (primer binding in this region) and after 700-900 bases. Therefore, sequences were trimmed whilst being assembled using the SeqMan software. These assembled sequences were used to search the GenBank database for closely related variants and to select reference sequences to align with the Clustal V method, a fast and reliable multiple sequence alignment platform. Aligned sequences were used to construct maximum likelihood phylogenetic trees These trees were created by the software generating a number of different phylogenetic trees and selecting the tree that gave the greatest probability of observing the sequences in the data set.

Figure 6. Maximum Likelihood Tree For Mitochondrial Cytochrome Oxidase (mtCOI). This is a maximum likelihood tree showing phylogenetic analysis of the cytochrome c oxidase subunit I (876 bp) of Bemisia tabaci from Arabian Peninsula using heuristic search option available in phylogenetic analysis using parsimony (PAUP v4.0.0b8). The heuristic search for 10 replicates of random stepwise-addition of taxa and tree-bisection-reconnection branch-swapping option was implemented. Reference sequences accession numbers are in brackets. Green, A Biotype, Blue, B Biotype, Red, Q Biotype.

Figure 7. Maximum Likelihood Tree For Wolbachia. This is a maximum likelihood tree showing phylogenetic analysis of 16S rDNA (619 bp) of Wolbachia of Bemisia tabaci from Arabian Peninsula using heuristic search option available in phylogenetic analysis using parsimony (PAUP v4.0.0b8). The heuristic search for 10 replicates of random stepwise-addition of taxa and tree-bisection-reconnection branch-swapping option was implemented. Reference sequences include Hamiltonella from other species and their accession numbers are in brackets. Blue, samples from study, black, references.

Figure 8. Maximum Likelihood Tree For Hamiltonella. This is a maximum likelihood tree showing phylogenetic analysis of 16S rDNA (592 bp) Hamiltonella defensa of Bemisia tabaci from Arabian Peninsula using heuristic search option available in phylogenetic analysis using parsimony (PAUP v4.0.0b8). The heuristic search for 10 replicates of random stepwise-addition of taxa and tree-bisection- reconnection branch-swapping option was implemented. Reference sequences accession numbers are in brackets.

4. DISCUSSION

Given the nature of Saudi Arabia and the region's physical environment (drought and high temperatures), there are very few weeds, so agricultural crops serve as the only host to B. tabaci with long dead seasons, except for those grown in greenhouses. Additionally, there is very heavy pesticide use, so the management strategy for the insects in the region includes both chemical and physical barriers. Pesticide use may be currently effective given the prevalence of the vulnerable B biotype and the presence of Rickettsia, which was associated with increased pesticide susceptibility (Chiel et al., 2007), but does not protect against the pesticide-resistant Q biotype (Ghanim and Kontsedalov, 2009).

However, this may be a clue as to the reason why Rickettsia was found to be almost absent from the southern region of the peninsula, despite it being known to have heat- tolerance and thus should be quite well-adapted for the high temperatures. Perhaps the usage of pesticides is much higher in that location than in the rest of the region, and thus infection frequency is low.

It was interesting to observe the large proportion of uninfected individuals, which averaged to about 35% for all the locations combined, but could be as high as 96%

(Abha). Again, this differs with the findings from Chiel et al. (2007), as the proportion of uninfected individuals was generally much lower than in this study.

Despite Wolbachia having a lower infection profile than the other two endosymbionts, during the random selection of insects for DNA extraction, it was observed that all the populations were predominantly female. This seems to indicate some sort of reproductive manipulation, but for unknown reasons.. Since Wolbachia was unexpectedly low despite 41 the apparent skew in sex-ratio in favor of females, this may be an indication that

Cardinium, the other well-known reproductive manipulator, may have a high infection frequency. This can be further investigated by continuing the study and screening for

Cardinium.

On the other hand, the most well-represented endosymbiont was Hamiltonella, found even in the Fayfa samples, which are non-B biotype. In the context of TYCLV, this is worrying since H. defensa has been associated to more efficient and effective transmission of this virus. Additionally, tomatoes are an important crop in Saudi Arabia, so this finding indicates the necessity for an effective pest management strategy in the region.

Chiel et al. (2007) investigated B. tabaci in Israel, showing that Hamiltonella was exclusively distributed in the B biotype, Wolbachia exclusively to the Q biotype, and

Rickettsia was evenly spread between both. Contrary to these results, there was no such exclusivity in distribution of Wolbachia in this study, as all three endosymbionts were observed in the B biotype studied. The three outgroup clades shown in the tree generated for H. defensa appear to be either localized in Arabia or simply have not been reported yet.

Fayfa

The most noteworthy of findings from the study arise from the data collected from Fayfa.

Fayfa, a geographically isolated mountainous location with cooler temperatures near the

Yemen border, appeared to have the highest genetic distance and diversity compared to the all the other locations included in the study. 42

Phylogenetically, most of the samples were found to share a root with the B and Q clades.

However, only one sample was found to actually be of the A biotype. This was very significant, as the A biotype has generally not been as widespread as the B and Q biotypes. It has stayed fairly localized in the southern US, though even there it has been displaced by the B biotype as the predominant biotype. Regarding endosymbiont distribution, Rickettsia was noticeably absent from the Fayfa location. Fayfa remained distinct from other locations in its placement on the phylogenetic trees generated for the other two endosymbionts. This further underlines its uniqueness in the region and the interest in further investigating it.

Management Strategies

Due to the presence of the B biotype and the Rickettsia associated with some locations

(as previously stated, these are associated with insecticide-sensitivity in B. tabaci), it appears that the current management system in the Arabian peninsula relying mainly on pesticides is sufficient. However, this would not be a sustainable method of control in cases where new biotypes are introduced or emerge in the region, especially the insecticide-resistant Q biotype.

As shown in this study, diversity in biotypes has already been observed in Fayfa. This has also been previously observed by Horowitz et al. (2003) in which it was found that over a period of a year monitoring B. tabaci populations in Israel, the Q biotype emerged and slowly displaced the B biotype in areas that were predominantly B biotype. This is one of the main concerns for the future whitefly status in the Arabian peninsula, and the best 43 course of action would to avoid or at least slow down the emergence of Q biotype would be to switch to greenhouse-based agriculture that would provide physical barriers to the insects, as well as introducing other management strategies concomitantly with the insecticides such as biological controls using predators and parasitoids as well as physical barriers such as greenhouses. Horowitz et al. (2005) showed that under non-pesticide conditions , B biotype is more competitive, so adopting more non-chemical control methods would also help to prevent the emergence of the Q biotype. (CHECK

KHASDAN FoR REFS) Degnan et al. (2009) showed positive correlation to presence of

Hamiltonella and parasitoid infection for whiteflies, since parasitoids are used in whitefly management, this could be another possible target for future management strategies

(Gerling et al., 2001).

Inconsistency in Wolbachia Amplification

The screening reactions for Wolbachia did not proceed as smoothly as the other primers, which generally amplified sequences successfully on the first attempt. The Wol16S F/R primers yielded much more inconsistent results. Several explanations were proposed and tested, including inactive enzyme, poor primer design, problems with the dNTPs, reagent concentration and contamination, annealing/extension/denaturation temperatures, reaction volume, DNA concentration and DNA coiling. Jeyaprakash and Hoy (2000) and

Li et al. (2007) are just two of several studies that have documented this apparent inconsistency or frequent false negatives, using ftsz- and wsp-based methods in standard

PCR for arthropods, including whitefly. Jeyaprakash and Hoy (2000) used it as a basis to investigate the efficacy of standard PCR versus long PCR in screening for Wolbachia. 44

They theorized three main reasons for it:

1) Sub-optimal DNA template : primer ratio due to small amounts of Wolbachia DNA with large amounts of arthropod host DNA.

2) Breakage of template and depurination of bases due to prolonged template denaturation and low buffer pH (8).

3) "Megaprimers" (incomplete/mismatched synthesized DNA strands) being formed due to Taq polymerase's lack of DNA editing abilities and to it being more error-prone compared to other polymerases (introducing one base-pair substitution for every 9000 bases amplified). These megaprimers could interfere with primer annealing and may also sequester Mg2+ ions from the reaction buffer.

Attempts to optimize this reaction included trials with different enzymes and primer designs, as well different PCR programs. Ultimately, screening was done using a premixed PCR mix, which generated faint but consistent bands that were subsequently cleaned and sequenced.

5. CONCLUSION & FUTURE WORK

This study aimed to identify the B. tabaci haplotypes present in Saudi Arabia and the region and to study the presence and distribution of secondary endosymbionts in the populations. Bacterial diversity and prevalence was expected between the regional populations. It was found that the B biotype is the predominant haplotype, with no evidence of the Q found. The geographically isolated region of Fayfa was shown to be more genetically distinct than other locations, with the first identification of the A biotype in the region, and a possible identification of a new haplotype. Thus, it appeared to be consistently distinct from other locations as phylogenetic trees for the Hamiltonella and

Wolbachia endosymbionts placed samples from Fayfa apart from other locations.

Overall, Hamiltonella, Wolbachia and Rickettsia were all identified in most if not all of the locations studied, either as single or co-infections. Hamiltonella was the most widespread endosymbiont. No correlation of host plant to endosymbiont distribution was observed.

These observations lay the groundwork for future work in the region, as well as providing as a reference data set for future changes in pest status, such as the introduction of new biotypes or the increase/decrease or appearance of new symbionts. Additionally, host plant samples could be expanded to further investigate possible endosymbiont to host plant associations. There may be associations that are not apparent since the study only examines eight different types of plants, some of which are similar (for example tomato and cucumber are Solanacea).

A possible focus for of future studies would be determining the phenotypic effects of the bacterial symbionts. Additional studies at the transcriptomic, metabolomic and proteomic 46 levels would also be of relevance in determining further physiological changes in the insects with differing endosymbiont combinations and the added survival advantages to them that could be used in improving insect control methods.

Further work should be done to investigate the Fayfa samples, with more biological work with live insects to observe their behavior, preference of plant host and their mating. This is necessary to determine whether they represent a new biotype or simply an outgroup.

REFERENCES

Ateyyat, M. A., Shatnawi, M., & Awi, M. A. (2010) Isolation and identification of culturable forms of bacteria from the sweet potato whitefly Bemisia tabaci Genn . (Homoptera : Aleyrodidae ) in Jordan. Turk J Agric For, 34, 225–234. doi:10.3906/tar- 0902-35. Baumann, P. (2005) Biology of bacteriocyte-associated endosymbionts of plant sap- sucking insects. Annual Review of Microbiology 59, 155–189. Baumann, P., Moran, N. A., Baumann, L. (2006) Bacteriocyte-associated endosymbionts of insects. Prokaryotes. 1:403–438. Bedford, I. D., Briddon, R.W., Brown, J. K., Rosell, R. C., Markham, P. G. (1994) Geminivirus transmission and biological characterization of Bemisia tabaci (Gennadius) biotypes from different geographic regions. Annals of Applied Biology, 125, 311-325. Bing, X. L., Yang, J., Zchori-Fein, E., Wang, X-W., Liu, S. S. (2012) Characterization of a newly discovered symbiont in the whitefly Bemisia tabaci (Hemiptera: Aleyrodidae). Appl. Environ. Microbiol. doi:10.1128/AEM.03030-12. Bing, X.-L., Yang, J., Zchori-Fein, E., Wang, X.-W., & Liu, S.-S. (2012) Characterization of a newly discovered symbiont in the whitefly Bemisia tabaci (Hemiptera: Aleyrodidae). Applied and environmental microbiology, (November). doi:10.1128/AEM.03030-12 Brown, J. K, Coats, S., Bedford, I. D., Markham, P. G. (1994) General esterase polymorphisms as genetic markers of Bemisia tabaci Genn. biotypes and evidence for the worldwide distribution of the 'B' biotype. Biochem. Gen. In Press. Brown, J. K, Frohlich, D. R., Rosell, R. C. (1995) The sweetpotato or silverleaf whiteflies: biotypes of Bemisia tabaci or a species complex? Annual Review of Entomology, 40:511-534. Brown, J. K. (1994) The status of Bemisia tabaci Genn. as a pest and vector in world agroecosystems. FAO Bull. 42. In press. Brown, J. K. (2010) Bemisia: Bionomics and Management of a Global Pest. (P. A. Stansly & S. E. Naranjo, Eds.). Dordrecht: Springer Netherlands. doi:10.1007/978-90- 481-2460-2. Brown, J. K., Coats, S., Bedford, I. D, Markham, P. G, Bird, J. (1992) Biotypic characterization of Bemisia tabaci populations based on esterase profiles, DNA fingerprinting, virus transmission, and bioassy to key host plant species. Phytopathology 82: 1 104. 48

Brown, J. K., Frohlich, D. R., & Rosell, R. C. (1995) The Sweetpotato Or Biotypes of Bemisia tabaci or Species Complex ? Annual Review Entymology. 40: 511-534 Brown, J. K., Ostrow, K. M., Idris, A. M., & Stenger, D. C. (1999) Biotic, Molecular, and Phylogenetic Characterization of Bean Calico Mosaic Virus, a Distinct Begomovirus Species with Affiliation in the Squash Leaf Curl Virus Cluster. Virology, 89(4): 273–280. Brown, J. K.., Frohlich, D. R. & Rosell, R. C. "The sweetpotato or silverleaf whiteflies – biotypes of Bemisia tabaci or a species complex". Annual Review of Entomology 40 (1995): 511– 534. Brumin, M., Kontsedalov, S., Ghanim, M. (2011) Rickettsia influences thermo- tolerance in the whitefly Bemisia tabaci B biotype. Insect Sci 18: 57–66. Buchner, P. (1965) Endosymbiosis of animals with plant microorganisms. John Wiley and Sons Interscience, New York, N.Y. 332–338. Callaini, G., Dallai, R., Riparbelli, M. G. (1997). Wolbachia-induced delay of paternal chromatin condensation does not prevent maternal chromosomes from entering anaphase in incompatible crosses of Drosophila simulans. J. Cell. Sci. 110:271– 80. Calvert, L. A., Cuervo, M., Arroyave, J. A., Constantino, L. M., Bellotti, A., Frohlich, D. (2001) Morphological and mitochondrial DNA marker analyses of whiteflies (Homoptera: Aleyrodidae) colonizing cassava and beans in Colombia. Annals of the Entomological Society of America, 94, 512-519. Calvert, L., & Villarreal, N. (1995) Using Molecular Techniques to Analyse Whitefly Species and Biotypes in Latin America. Whiteflies and Whitefly-borne Viruses in the Tropics. p251–262. Castle, S.J., Henneberry, T.J., Prabhaker, N., Toscano, N.C. (1996) Trends in relative susceptibilities of whiteflies to insecticides through the cotton season in the Imperial Valley, CA. In: Dugger, P., Richter, D. (Eds.), Proceedings Beltwide Cotton Conference. National Cotton Council, Memphis, TN, pp. 1032–1035.

Chiel, E., Gottlieb, Y., Zchori-Fein, E., Mozes-Daube, N., Katzir, N., Inbar, M., et al., (2007) Distribution of secondary symbionts in Israeli populations of Bemisia tabaci. Bull Entomol Res 97:407–413. Chowda-Reddy, R. V., Kirankumar, M., Seal, S. E., Muniyappa, V., Valand, G. B., Govindappa, M. R., Colvin, J. (2012) Bemisia tabaci phylogenetic groups in India and the relative transmission efficacy of Tomato leaf curl Bangalore virus by an indigenous and an exotic population. Journal of Integrative Agriculture, 11, 235-248. Clark, E. L., Karley, A. J., Hubbard, S. F. (2010) Insect endosymbionts: manipulators of insect herbivore trophic interactions? Protoplasma 244: 25–51. 49

Cohen, S., Duffus, J. E., Liu, H. Y. (1992) A new Bemisia tabaci biotype in the southwestern United States and its role in silverleaf of squash and transmission of lettuce infectious yellows virus. Phytopathology 82:86–90. Colvin, J., Omongo, C. A., Govindappa, M. R., Stevenson, P. C., Maruthi, M. N., Gibson, G., Seal, S. E., Muniyappa, V. (2006) Host-plant viral infection effects on arthropod-vector population growth, development and behaviour: management and epidemiological implications. Advances in Virus Research, 67, 419-452. Costa, H.S., Brown J K, Sivasupramaniam S, Bird J. (1993) Regional distribution, insecticide resistance and reciprocal crosses between the A and B biotypes of Bemisia tabaci. Insect Science and Its Application, 14, 255-266. Costa, H.S., Westcot, D.M., Ullman, D.E., Rosell, R.C., Brown, J.K. & Johnson, M.W. (1995) Morphological variation in Bemisia endosymbionts. Protoplasma 189, 194–202. Crowder, D. W., Sitvarin, M. I., Carrière, Y. (2010) Mate discrimination in invasive whitefly species. Journal of Insect Behavior, 23, 364-380. De Barro, P. J., Bourne, A. (2010) Oviposition host choice by an invader accelarates displacement of its indigenous competitor. Biological Invasions, 12, 3013-3023. De Barro, P. J., Liu, S., Boykin, L., & Dinsdale, A. B. (2011) Bemisia tabaci: A Statement of Species Status. Annual Review of Entomology, 56, 1-19. doi: 10.1146/annurevento-112408-085504 Degnan, P. H., Yu, Y., Sisneros, N., Wing, R. a, & Moran, N. A. (2009) Hamiltonella defensa, genome evolution of protective bacterial endosymbiont from pathogenic ancestors. Proceedings of the National Academy of Sciences of the United States of America, 106(22), 9063–8. doi:10.1073/pnas.0900194106 Dennehy, T.J., Williams III, L., Russell, J.S., Li, X., Wigert, M., 1996. Monitoring and management of whitefly resistance to insecticides in Arizona. In: Dugger, P., Richter, D. (Eds.), Proceedings Beltwide Cotton Conference. National Cotton Council, Memphis, TN, pp. 743–748. Prabhaker, N., Toscano, N.C., Henneberry, T.J., Castle, S.J., Weddle, D. (1996) Assessment of two bioassay techniques for resistance monitoring of silverleaf whitefly (Homoptera: Aleyrodidae) in California. J. Econ. Entomol. 89, 805–815. Dinsdale, A., Cook, L., Riginos, C., Buckley, Y. M., de Barro, P. J. (2010) Refined global analysis of Bemisia tabaci (Gennadius) (Hemiptera: Sternorrhyncha: Aleyrodoidea) mitochondrial CO1 to identify species level genetic boundaries. Annals of the Entomological Society of America, 103, 196-208. Dobson, S. L., Marsland, E. J., Rattanadechakul, W. (2002) Mutualistic Wolbachia infection in Aedes albopictus: Accelerating cytoplasmic drive. Genetics 160: 1087– 1094. 50

Douglas, A. E. (1989) Mycetocyte symbiosis in insects. Biological Reviews, 64: 409– 434. doi: 10.1111/j.1469-185X.1989.tb00682. Elbaz, M., N. Lahav, and S. Morin. (2010) Evidence for pre-zygotic reproductive barrier between the B and Q biotypes of Bemisia tabaci (Hemiptera: Aleyrodidae). Bulletin of entomological research 100.5: 581. Elbert, A., Overbeck, H., Iwaya, H., Tsuboi, S., 1990 Imidacloprid, a novel systemic nitromethylene analogue insecticide for crop protection. In: Proceedings of the Brighton Crop Protection Conference, Pests and Diseases. BritishCrop Protection Council, Farnham, UK, pp. 21–28. Ellsworth, P.C., Dennehy, T.J., Nichols, R.L. (1996b) Whitefly management in Arizona cotton. Univ. Arizona, Coop. Ext. Publ. No. 196004, IPM Series 3, 2pp. Ellsworth, P.C., Diehl, J.W., Husman, S.H. (1996a) Establishment of integrated pest management infrastructure: a community-based action program for Bemisia management. In: Gerling, D., Mayer, R.T. (Eds.), Bemisia: 1995 Taxonomy, Biology, Damage, Control and Management. Intercept, UK, pp. 681–693. Ellsworth, P.C., Martinez-Carrillo, J.L. (2001) IPM of Bemisia tabaci: A case study from North America. Crop Prot. 20, 853–869. Engelstädter, J. & Hurst, G.D.D. (2009) The ecology and evolution of microbes that manipulate host reproduction. Annual Review of Ecology, Evolution & Systematics 40: 127-149. Everett, K. D. E., Thao, M., Horn, M., Dyszynski, G. E., & Baumann, P. (2005) Novel chlamydiae in whiteflies and scale insects: endosymbionts “Candidatus Fritschea bemisiae” strain Falk and “Candidatus Fritschea eriococci” strain Elm. International journal of systematic and evolutionary microbiology, 55 (Pt4), 1581–7. doi:10.1099/ijs.0.63454-0. Gennadius, P. (1889) Disease of the tobacco plantations in the Trikonia. The aleurodid of tobacco. Ellenike Ga. 5: 1-3 Gerling, D., Alomar, O., Arno, J. (2001) Biological control of Bemisia tabaci using predators and parasitoids. Crop Protection. 20, 779–799. Ghanim, M., & Kontsedalov, S. (2009) Susceptibility to insecticides in the Q biotype of Bemisia tabaci is correlated with bacterial symbiont densities. Pest management science, 65(9), 939–42. doi:10.1002/ps.1795. Gill, R. J., Brown, J. K. (2010) Systematics of Bemisia and Bemisia relatives: can molecular techniques solve the Bemisia tabaci complex conundrum-a taxonomist’s viewpoint. In: Stansly P A, Narahjo S E, eds., Bemisia: Bionomics and Management of a Global Pest. Springer, New York, USA. pp. 5-29. 51

Gill, R. I. (1992) A review of the sweetpotato whitefly in southern California. Pan-Pac. Entomol. 68:144-52 Gil-Salas, F. M., Peters, J., Boonham, N., Cuadrado, I. M., & Janssen, D. (2011) Yellowing disease in zucchini squash produced by mixed infections of Cucurbit yellow stunting disorder virus and Cucumber vein yellowing virus. Phytopathology, 101(11), 1365–72. doi:10.1094/PHYTO-12-10-0343 Glaser, R. L., Meola, M. A. (2010) The native Wolbachia endosymbionts of Drosophila melanogaster and Culex quinquefasciatus increase host resistance to West Nile Virus infection. PLoS ONE 5(8): e11977. doi: 10.1371/journal.pone.0011977. Gnankiné, O., Mouton, L., Henri, H., Terraz, G., Houndeté, T., Martin, T., Vavre, F., et al. (2012) Distribution of Bemisia tabaci (Homoptera: Aleyrodidae) biotypes and their associated symbiotic bacteria on host plants in West Africa. Insect Conservation and Diversity, no–no. doi:10.1111/j.1752-4598.2012.00206.x Gottlieb, Y., Ghanim, M., Chiel, E., Gerling, D., Portnoy, V., Steinberg, S., Tzuri, G., Horowitz, A. R., Belausov, E., Mozes-Daube, N., Kontsedalov, S., Gershon, M., Gal, S., Katzir, N. & Zchori-Fein, E. (2006) Identification and localization of a Rickettsia sp in Bemisia tabaci (Homoptera: Aleyrodidae). Applied and Environmental Microbiology 72, 3646–3652. Gottlieb, Y., Zchori-Fein, E., Mozes-Daube, N., Kontsedalov, S., Skaljac, M., et al. (2010) The transmission efficiency of tomato yellow leaf curl virus by the whitefly Bemisia tabaci is correlated with the presence of a specific symbiotic bacterium species. J Virol 84: 9310–9317. Greathead, A. H., (1986). Host plants. In Bemisia tabaci: A Literature Survey on the Cotton Whitefly with an Annotated Bibliography. CAB International Institutes, Biological Control. Silwood Park, UK, pp. 17–26. Guruprasad, N. M., Mouton, L., Puttaraju, H. P. (2011) Effect of Wolbachia infection and temperature variations on the fecundity of the uzi fly Exorista sorbillans (Diptera: Tachinidae). Symbiosis 54: 151–158. Harrison, B. D., Robinson, D. J. (1988) Molecular variation in vector-borne plant viruses: epidemiological significance. Philos. Trans. R. Soc. Lonoon Ser. B 321:447-62. Heddi, A., Grenier, A.M., Khatchadourian, C., Charles, H. & Nardon, P. (1999) Four intracellular genomes direct weevil biology: Nuclear, mitochondrial, principal endosymbiont, and Wolbachia. Proceedings of the National Academy of Sciences USA 96, 6814–6819. Hedges, L. M., Brownlie, J. C., O’Neill, S. L., Johnson, K. N. (2008) Wolbachia and virus protection in insects. Science 322: 702–702. Hilje, L., Costa, H.S., Stansly, P.A. (2001) Cultural practices for managing Bemisia tabaci and associated viral disease. Crop Prot. 20, 801–812. 52

Himler, A. G., Adachi-Hagimori, T., Bergen, J. E., Kozuch, A., Kelly, S. E., et al. (2011) Rapid spread of a bacterial symbiont in an invasive whitefly is driven by fitness benefits and female bias. Science. 332 (6026): 254–256. Horowitz, A. R., Kontsedalov, S., Khasdan, V., Ishaaya, I. (2005). Biotypes B and Q of Bemisia tabaci and their relevance to neonicotinoid and pyriproxyfen resistance. Archives of Insect Biochemistry and Physiology, 58, 216-225. Hosokawa, T., Koga, R., Kikuchi, Y., Meng, X. Y., Fukatsu, T. (2010) Wolbachia as a bacteriocyte-associated nutritional mutualist. Proc Natl Acad Sci USA 107: 769– 774. Houk, E. J., Griffiths, G. W. (1980) Intracellular symbiotes of the Homoptera. Ann Rev Entomol 25: 161–187. doi: 10.1146/annurev.en.25.010180.001113. Hu, J., De Barro, P. J., Zhao, H., Nardi, F., Wang, J., Liu, S. S. (2011) An extensive field survey combined with a phylogenetic analysis reveals rapid and widespread invasion of two alien whiteflies in China. PLoS ONE, 6, e16061. Idris, A. M., Brown, J. K. (2004) Cotton leaf crumple virus Is a Distinct Western Hemisphere Begomovirus Species with Complex Evolutionary Relationships Indicative of Recombination and Reassortment. Phytopathology. Oct. 94(10):1068-74. doi: 10.1094/PHYTO.2004.94.10.1068. Jeyaprakash, A., & Hoy, M. A. (2000) Long PCR improves Wolbachia DNA amplification: wsp sequences found in 76% of sixty-three arthropod species. Insect molecular biology 9(4), 393–405. Jiménez, D. R., Yokomi, R. K., Mayer, R. T., Shapiro, J. P. (1995) Cytology and physiology of silverleaf whitefly-induced squash silverleaf. Physiological and Molecular Plant Pathology 46 (3): 227-242, ISSN 0885-5765, 10.1006/pmpp.1995.1018. Kaiser, W., Huguet, E., Casas, J., Commin, C., Giron, D. (2010) Plant green-island phenotype induced by leaf-miners is mediated by bacterial symbionts. ProcRSoc Lond [Biol] 277: 2311–2319.

Khasdan, V., Levin, I., Rosner, A., Kontsedalov, S., Morin, S. & Horowitz, A.R. (2004) Development and use of DNA markers to distinguish between B and Q biotypes of Bemisia tabaci in Israel. p. 55 in 2nd European Whitefly Symposium EWSII Abstract Compendium, Cavtat, 5–9 October 2004 Croatia, European Whitefly Studies Network. Khasdan, V., Levin, I., Rosner, A., Morin, S., Kontsedalov, S., Maslenin, L. & Horowitz, A.R. (2005) DNA markers for identifying biotypes B and Q of Bemisia tabaci (Hemiptera: Aleyrodidae) and studying population dynamics. Bulletin of Entomological Research 95, 605–613. Kontsedalov, S., Zchori-Fein, E., Chiel, E., Gottlieb, Y., Inbar, M., et al. (2008) The presence of Rickettsia is associated with increased susceptibility of Bemisia tabaci (Homoptera: Aleyrodidae) to insecticides. Pest Manag Sci 64: 789–792. 53

Laven H, Aslamkhan M. (1970) Control of Culex pipiens pipiens and C. p. fatigans with integrated genetical systems. Pak. J. Sci. 22:303–12. Laven, H. (1967) Eradication of Culex pipiens fatigans through cytoplasmic incompatibility. Nature 261:383–84. Li, M., Hu, J., Xu, F. C., Liu, S. S. (2010) Transmission of Tomato yellow leaf curl virus by two invasive biotypes and a Chinese indigenous biotype of the whitefly Bemisia tabaci. International Journal of Pest Management, 56, 275-280. Li, Z.-X., Lin, H.-Z., & Guo, X.-P. (2007) Prevalence of Wolbachia infection in Bemisia tabaci. Current microbiology, 54(6), 467–71. doi:10.1007/s00284-007-0011-7 Liu J., Zhao H., Jiang K., Zhou X. P., Liu S. S. (2009) Differential indirect effects of two plant viruses on an invasive and an indigenous whitefly vector: implications for competitive displacement. Annals of Applied Biology, 155, 439-448. Liu, S. S., De Barro, P. J., Xu, J., Luan, J. B., Zang, L. S., Ruan, Y. M., Wan, F. H. (2007) Asymmetric mating interactions drive widespread invasion and displacement in a whitefly. Science, 318, 1769-1772. Liu, S., Colvin, J., & De Barro, P. J. (2012) Species Concepts as Applied to the Whitefly Bemisia tabaci Systematics: How Many Species Are There? Journal of Integrative Agriculture, 11(2), 176–186. doi:10.1016/S2095-3119(12)60002-1 Luo, C., Jones, C. M., Devine, G., Zhang, F., Denholm, I., Gorman K. (2010) Insecticide resistance in Bemisia tabaci biotype Q (Hemiptera: Aleyrodidae) from China. Crop Protection, 29, 429-434. Merçot, H., Poinsot, D. (1998) ... and discovered on mount Kilimanjaro. Nature 391:853– 53. Moran, N. A (2001) The coevolution of bacterial endosymbionts and phloem- feeding insects. Ann Missouri Bot Garden 88: 35–44. Moran, N. A., Telang, A. (1998) Bacteriocyte-associated symbionts of insects. Bioscience 48: 295–304. Moreira, L. A., Iturbe-Ormaetxe, I., Jeffery, J. A., Lu, G. J., Pyke, A. T. et al. (2009) A Wolbachia symbiont in Aedes aegypti limits infection with Dengue, Chikungunya, and Plasmodium. Cell 139: 1268–1278. Mound, L. A. (1963). Host-correlated variation in Bemisia tabaci (Gennadius) (Homoptera: Aleyrodidae). Proceedings of the Royal Entomological Society of London (Series A), 38, 171-180. Mound, L. A., Halsey, S. H. (1978). Whitefly of the World: A Systematic Catalogue of the Aleyrodidae (Homoptera) with Host Plant and Natural Enemy Data. John Wiley & Sons, Chichester. 54

Mullins, J.W. (1993) Imidacloprid: a new nitroguanidine insecticide. In: Pest Control with Environmental Safety, American Chemical Society Symposium Series 524. American Chemical Society, Washington, DC, pp. 184–198. Naranjo, S.E. (2001) Conservation and evaluation of natural enemies in IPM systems for Bemisia tabaci. Crop Prot. 20, 835–852. Naranjo, S.E., Ellsworth, P.C, Chu, C.C., Henneberry, T.J., Riley, D.G., Watson, T.F., Nichols, R.L. (1998) Action thresholds for the management of Bemisia tabaci (Homoptera: Aleyrodidae) in cotton. J. Econ. Entomol. 91, 1415–1426. Nirgianaki, A., Banks, G.K., Frohlich, D.R., Veneti, Z., Braig, H.R., Miller, T.A., Bedford, I.D., Markham, P.G., Savakis, C. & Bourtzis, K. (2003) Wolbachia infections of the whitefly Bemisia tabaci. Current Microbiology 47, 93–101. Oliveira, M.R.V., Henneberry, T.J. & Anderson, P. (2001) History, current status, and collaborative research projects for Bemisia tabaci. Crop Protection 20, 709–723. O'Neill, S. L., Hoffman, A. A., Werren, J. H. (1997). Influential Passenger: Inherited Mi- croorganisms and Arthropod Reproduction, pp. 1–214. New York: Oxford Univ. Press. Palumbo, J. C., Toscano, N. C., Blua, M. J., & Yoshida, H. A. (2000) Impact of Bemisia Whiteflies (Homoptera: Aleyrodidae) on Alfalfa Growth, Forage Yield, and Quality. Journal of Economic Entomology 93(6):1688-1694. doi: http://dx.doi.org/10.1603/0022-0493-93.6.1688 Palumbo, J.C., Horowitz, R., Prabhaker, N. (2001) Overview of insecticidal control and resistance management for Bemisia tabaci. Crop Prot. 20, 739–765. Penz, T., Schmitz-Esser, S., Kelly, S. E., Cass, B. N., Müller, A., Woyke, T., Malfatti, S. a, et al. (2012). Comparative genomics suggests an independent origin of cytoplasmic incompatibility in Cardinium hertigii. PLoS genetics, 8(10), e1003012. doi:10.1371/journal.pgen.1003012 Perring, T.M. (2001) The Bemisia tabaci species complex. Crop Protection. 20, 725–737. Prabhaker, N., Toscano, N.C., Henneberry, T.J., (1998) Evaluation of insecticide rotations and mixtures as resistance management strategies for Bemisia argentifolii (Homoptera: Aleyrodiae). J. Econ. Entomol. 91, 820–826. Rana, V. S., Singh, S. T., Priya, N. G., Kumar, J., & Rajagopal, R. (2012). Arsenophonus GroEL interacts with CLCuV and is localized in midgut and salivary gland of whitefly B. tabaci. PloS one, 7(8), e42168. doi:10.1371/journal.pone.0042168 Riley, D. G., Palumbo, J. C.(1995) Interaction of Silverleaf Whitefly (Homoptera: Aleyrodidae) with Cantaloupe Yield. Journal of Economic Entomology 88(6) 1726- 1732(7). 55

Rosell R. C., Bedford I. D., Frohlich D. R., Gill R. J., Brown J. K., Markham P. G. (1997). Analysis of morphological variation in distinct populations of Bemisia tabaci (Homoptera: Aleyrodidae). Annals of the Entomological Society of America, 90, 575-589. Russell, L. M. (1958). Synonyms of Bemisia tabaci (Gennadius) (Homoptera: Aleyrodidae). Bulletin of the Brooklyn Entomological Society, 52, 122-123. Secker, A.E., Bedford, I.A., Markham, P.G., William, M.E.C., 1998. Squash, a reliable field indicator for the presence of B biotype of tobacco whitefly, Bemisia tabaci. In: Brighton Crop Protection Conference-Pests and Diseases. BritishCrop Protection Council, Farnham, UK, pp. 837–842. Swofford, D. L. (2002) PAUP* 4.0 b10: Phylogenetic Analysis Using Parsimony. Version 4. Sinauer Associates, Sunderland, Massachusetts. Teixeira L, Ferreira A, Ashburner M (2008) The bacterial symbiont Wolbachia induces resistance to RNA viral infections in Drosophila melanogaster. PLoS Biol 6: 2753–2763. Thao, M. L., Baumann, L., Hess, J. M., Falk, B. W., Ng, J. C. K., Gullan, P. J., & Baumann, P. (2003). Phylogenetic evidence for two new insect-associated Chlamydia of the family Simkaniaceae. Current microbiology, 47(1), 46–50. doi:10.1007/s00284-002- 3953-9 Wang, P., Ruan, Y. M., Liu, S. S. (2010). Crossing experiments and behavioral observations reveal reproductive incompatibility among three putative species of the whitefly Bemisia tabaci. Insect Science, 17, 508-516. Weeks, A.R., Velten, R. & Stouthamer, R. (2003) Incidence of a new sex-ratio-distorting endosymbiotic bacterium among arthropods. Proceedings of the Royal Society Series B 270, 1857–1865. Werren, J.H. & Windsor, D.M. (2000) Wolbachia infection frequencies in insects: evidence of a global equilibrium? Proceedings of the Royal Society Series B 267, 1277– 1285.

Wilkinson, T. L. (1998). The elimination of intracellular microorganisms from insects: an analysis of antibiotic treatment in the pea aphid (Acyrthosiphon pisum). Comp. Biochem. Physiol 119, 871–. Xie, W., Meng, Q., Wu, Q., Wang, S., Yang, X., Yang, N., Li, R., et al. (2012). Pyrosequencing the Bemisia tabaci transcriptome reveals a highly diverse bacterial community and a robust system for insecticide resistance. PloS one, 7(4), e35181. doi:10.1371/journal.pone.0035181. Xu J., Lin K. K., Liu S. S. (2011). Performance on different host plants of an alien and an indigenous Bemisia tabaci from China. Journal of Applied Entomology, 135, 771- 779. 56

Zang, L. S., Chen, W. Q., Liu, S. S. (2006). Comparison of performance on different host plants between the B biotype and a non-B biotype of Bemisia tabaci from Zhejiang, China. Entomologia Experimentalis et Applicata, 121, 221-227. Zchori-Fein, E. & Brown, J.K. (2002) Diversity of prokaryotes associated with Bemisia tabaci (Gennadius) (Hemiptera: Aleyrodidae). Annals of the Entomological Society of America 95, 711–718.

APPENDICES PCR Reaction Details

Table 4. PCR Conditions. An overview of the PCR program used for each primer.

Temperature - time Step COI RB HAM WOL

Initial 95°C - 2 min 95°C -2 min 95°C - 3 min 95°C - 3 min denaturation

Denaturation 95°C - 30 s 95°C - 30 s 95°C - 20 s 95°C - 20 s

Annealing 45°C - 30 s 55°C - 30 s 57°C - 45 s 50°C - 45 s

Extension 72°C - 1 min 72°C - 1 min 72°C - 1 min 72°C - 1 min

35 cycles 35 cycles 35 cycles 35 cycles

Final 72°C - 10 min 72°C - 10 min 72°C - 10 min 72°C - 10 min Extension

Hold 4°C - forever 4°C - forever 4°C - forever 4°C - forever

Table 5. PCR Reaction Components For Each Primer. An overview of the PCR reaction reagent concentrations and volumes used for each of the primers for a 25 µL. *DNA concentration was found to generally be between 50 - 100 ng/µl. 2 ul (about 100 - 200 ng). COI RB HAM & WOL Reaction component For 1 Final For 1 Final For 1 Final reaction(µl) conc. reaction(µl) conc. reaction(µl) conc. Reaction buffer 2.5 1x 2.5 1x 2.5 1x (10X)Invitrogen dNTP(2.5 1.5 150 µM 1.5 150 µM 2 200 µM mM)Invitrogen MgCl (50 2 - - - - 0.75 1.5 mM mM)Invitrogen Primer F(20 µM) 0.5 0.4 µM 0.5 0.4 µM Primer R(20 µM) 0.5 0.4 µM 0.5 0.4 µM 0.3 Enzyme (5 U/ul) 0.125 0.625 U 0.5 2.5 U 0.3 1.5 U Template* 2 - 2 - 2 - ddH2O 17.875 - 17.5 - 18.45 - TOTAL 25 µl 58

Lysis Buffer Composition (1 ml) Proteinase K 1 µg Tris pH 8.0 5 µl EDTA 1 µl NP40 5 µl ddH2O 984 µl

Cloning and Miniprep

Plating

35 g LB Agar was mixed with ddH20 to make 1 l, which was then autoclaved. Plates

were prepared by adding about 20 ml of the LB Agar (ampicillin added) into each petri

dish and set. After solidifying, 45 µl Xgal was added to each plate.

Ligation

The ligation reaction was prepared as shown below using the pGEM-T Easy Vector

System (Promega, A1360, US) and left to incubate overnight at 4°C. The DNA to be

ligated was extracted from a 1% agarose gel and cleaned with the QIAgen MinElute Kit

(28604).

Table 6. Ligation Components. An overview of the ligation components used in the reaction.

Background Reagents Standard Reaction Positive Control Control

2x Rapid Ligation Buffer, 5 µl 5 µl 5 µl T4 DNA Ligase pGEM-T Easy Vector (50 ng) 1 µl 1 µl 1 µl PCR product 3 µl - - Control Insert DNA - 2 µl - T4 DNA Ligase (3 Weiss units/µl) 1 µl 1 µl 1 µl ddH2O to a final volume of 10 µl 10 µl 10 µl

Transformation

100 µl competent E.coli cells were thawed on ice for 20-30 min. 5 µl (50ng) of ligation

DNA were added to the cells, with a further 10 min incubation on ice. The tubes with

DNA and E.coli were heat shocked in a water bath at 42°C for 45 seconds, then put back on ice for 5-6 min to recover. 200 µl of SOC media were then added followed by shaking at 37°C for 1 hour. Two dilutions were prepared for each sample culture, 100 µl and 50

µl, which were spread on the LB/Ampicillin/Xgal plates. The plates were incubated overnight at 37°C. Transformed cells (white colonies) were then picked and added to 5 ml LB (ampicillin added), then incubated at 37°C at 250 rpm overnight.

Miniprep (Qiaprep Spin Miniprep Kit, 27104)

The bacterial culture was pelleted by centrifugation at >8000 rpm for 3 min at 25°C.

Pellets were resuspended in 250 µl Buffer P1 and transferred to microcentrifuge tubes.

250 µl Buffer P2 was added and mixed by inverting until the solution was clear for no more than 5 min. 350 µl Buffer N3 was added, mixed by inverting and then the samples were centrifuged for 10 min at 13,000 rpm (Eppendorf Centrifuge 5424, Germany). The supernatant was loaded onto the QIAprep spin column by pipetting and centrifuged for

60 s at 13,000 rpm. The flow through was discarded, and the column was washed with

750 µl Buffer PE. Following another 60 s centrifugation step and flow through discarding, the column was centrifuged once more to remove residual buffer. Finally, columns were placed into clean 1.5 ml microcentrifuge tubes, and 50 µl of pre-warmed

Buffer EB was applied to the columns and left standing for 4 min. A final centrifuge step followed to elute the purified DNA. 60

Digestion

A 15 µl reaction was prepared to digest the clones: 5 µl DNA, 1.5 µl Reaction Buffer 3, 1

µl EcoRI, 7.5 µl ddH2O. The reaction was incubated for 1 h at 37°C, after which the whole volume is loaded onto a 1% agarose gel and run at 100 V for 40 min. Bands were cut and cleaned using the QIAquick Gel Extraction Kit (28704) and eluted in 20 µl. 10 µl were sent for Sanger sequencing.

Gel Extraction (QIAquick Gel Extraction Kit, 28704)

Bands of interest were cut from a 1% agarose gel using a clean scalpel, after which three volumes of Buffer QB was added. Samples were then incubated for 10 min at 55°C until completely dissolved with periodic mixing. One volume of 100% isopropanol was added and then the sample was loaded onto the QIAquick column and centrifuged for 1 min at

13,000 rpm (Eppendorf Centrifuge 5424, Germany), flow through being discarded (if the color of the mixture was darker then yellow, 10 µl of 3 M sodium acetate was added).

The column was then washed with 750 µl Buffer PE (with ethanol added), and centrifuged twice at 1 min for 13,000 rpm, with a flow discarding step to ensure complete drying of the column. Finally, columns were placed into clean 1.5 ml microcentrifuge tubes, and 30 µl of pre-warmed Buffer EB was applied to the columns and left standing for 4 min. A final centrifuge step followed to elute the purified DNA.

Samples were eluted into 20 - 30 µl of Elution Buffer. DNA concentration was quantified by running 1 µl of each sample in the Nanodrop, with Elution Buffer as a blank. Samples were stored at -20°C or submitted directly to sequencing. 61

PCR Purification (QIAquick PCR Purification Kit (28106))

To bind the sample to the Qiaquick column, 5 volumes of Buffer PB pH 5.0 were added to 1 volume of the PCR reaction and mixed (100 µl added to 20 µl). If the color of the mixture was darker then yellow, 10 µl of 3 M sodium acetate was added. The sample was loaded onto the QIAquick column and centrifuged for 1 min at 13,000 rpm (Eppendorf

Centrifuge 5424, Germany), with the flow through being discarded. The column was then washed with 750 µl Buffer PE (with ethanol added), and centrifuged twice at 1 min for

13,000 rpm, with a flow discarding step to ensure complete drying of the column. Finally, columns were placed into clean 1.5 ml microcentrifuge tubes, and 30 µl of pre-warmed

Buffer EB was applied to the columns and left standing for 4 min. A final centrifuge step followed to elute the purified DNA.

Figure 9. Histogram Of Endosymbiont Infection Frequency. This histogram shows the plot of the different endosymbionts in total for each location. Co-infections and uninfected individuals are not represented. RB, Rickettsia; HAM, Hamiltonella; WOL, Wolbachia.