Isolation and assessment of gut bacteria from the Formosan subterranean termite, Coptotermes formosanus (Isoptera: Rhinotermitidae), for paratransgenesis research and application.

Paratransgenesis targeting the gut protozoa is being developed as an alternative method for the control of the Formosan subterranean termite (FST). This method involves killing the cellulose-digesting gut protozoa using a previously developed antiprotozoal peptide consisting of a target specific ligand coupled to an antimicrobial peptide (Hecate). In the future, we intend to genetically engineer termite gut bacteria as "Trojan Horses" to express and spread ligand-Hecate in the termite colony. The aim of this study was to assess the usefulness of bacteria strains isolated from the gut of FST as "Trojan Horses." We isolated 135 bacteria from the guts of workers from 3 termite colonies. Sequencing of the 16S rRNA gene identified 20 species. We tested 5 bacteria species that were previously described as part of the termite gut community for their tolerance against Hecate and ligand-Hecate. Results showed that the minimum concentration required to inhibit bacteria growth was always higher than the concentration required to kill the gut protozoa. Out of the 5 bacteria tested, we engineered Trabulsiella odontotermitis, a termite specific bacterium, to express green fluorescent protein as a proof of concept that the bacteria can be engineered to express foreign proteins. Engineered T. odontotermitis was fed to FST to study if the bacteria are ingested. This feeding experiment confirmed that engineered T. odontotermitis is ingested by termites and can survive in the gut for at least 48 h. Here we report that T. odontotermitis is a suitable delivery and expression system for paratransgenesis in a termite species.

Assessment of genetically engineered Trabulsiella odontotermitis as a 'Trojan Horse' for paratransgenesis in termites.

BACKGROUND: The Formosan subterranean termite, Coptotermes formosanus is an invasive urban pest in the Southeastern USA. Paratransgenesis using a microbe expressed lytic peptide that targets the termite gut protozoa is currently being developed for the control of Formosan subterranean termites. In this study, we evaluated Trabulsiella odontotermitis, a termite-specific bacterium, for its potential to serve as a 'Trojan Horse' for expression of gene products in termite colonies. RESULTS: We engineered two strains of T. odontotermitis, one transformed with a constitutively expressed GFP plasmid and the other engineered at the chromosome with a Kanamycin resistant gene using a non- disruptive Tn7 transposon. Both strains were fed to termites from three different colonies. Fluorescent microscopy confirmed that T. odontotermitis expressed GFP in the gut and formed a biofilm in the termite hindgut. However, GFP producing bacteria could not be isolated from the termite gut after 2 weeks. The feeding experiment with the chromosomally engineered strain demonstrated that T. odontotermitis was maintained in the termite gut for at least 21 days, irrespective of the termite colony. The bacteria persisted in two termite colonies for at least 36 days post feeding. The experiment also confirmed the horizontal transfer of T. odontotermitis amongst nest mates. CONCLUSION: Overall, we conclude that T. odontotermitis can serve as a 'Trojan Horse' for spreading gene products in termite colonies. This study provided proof of concept and laid the foundation for the future development of genetically engineered termite gut bacteria for paratransgenesis based termite control.

Protozoacidal Trojan-Horse: use of a ligand-lytic peptide for selective destruction of symbiotic protozoa within termite guts.

For novel biotechnology-based termite control, we developed a cellulose bait containing freeze-dried genetically engineered yeast which expresses a protozoacidal lytic peptide attached to a protozoa-recognizing ligand. The yeast acts as a 'Trojan-Horse' that kills the cellulose-digesting protozoa in the termite gut, which leads to the death of termites, presumably due to inefficient cellulose digestion. The ligand targets the lytic peptide specifically to protozoa, thereby increasing its protozoacidal efficiency while protecting non-target organisms. After ingestion of the bait, the yeast propagates in the termite's gut and is spread throughout the termite colony via social interactions. This novel paratransgenesis-based strategy could be a good supplement for current termite control using fortified biological control agents in addition to chemical insecticides. Moreover, this ligand-lytic peptide system could be used for drug development to selectively target disease-causing protozoa in humans or other vertebrates.

Development of microsatellites for population genetic analyses of Tabanus nigrovittatus (Diptera: Tabanidae).

The greenhead horse fly, Tabanus nigrovittatus Macquart (Diptera: Tabanidae), is frequently found in coastal marshes of the Eastern United States. The females are autogenous (i.e., able to develop eggs without a bloodmeal),but they become a considerable pest to both humans and animals when they pursue a source of blood protein to produce additional eggs. In this study, we identified microsatellite markers to provide first insight into the population genetic structure of this notorious pest species. Because no prior genomic information was available for T. nigrovittatus, we used direct shotgun pyrosequencing technology to characterize microsatellite loci. Approximately 10% of the 105,634 short sequence reads generated from random genome sampling contained microsatellites with at least four repeats ofdi-, tri-, tetra-, penta-, and hexamers. Primers were designed for 36 different microsatellite loci with di-, tri-, and tetramer repeat units. After optimization, 20 primer pairs yielded consistent PCR products and were validated for population genetic application in six populations in Western Louisiana Ten loci were polymorphic with 2-9 alleles per locus and an average observed heterozygosity of 0.20 across populations. The horse fly populations from different trap sites (distance 50-144 km) or years of collection (2010 vs 2011) were genetically distinct from each other (FST = 0.05-0.39) and genetically diverse (gene diversity: 0.24-0.37) but considerably inbred (FIS: 0.22-0.47), with high mean relatedness among individuals (r = 0.27), suggesting the capture of a high percentage of sisters at the same trap location who were progeny of incest.

Transcriptome profiling of female alates and egg-laying queens of the Formosan subterranean termite.

Termites are known to have an extraordinary reproductive plasticity and capacity, but the underlying genetic patterns of termite reproductive biology are relatively understudied. The goal of this study was to identify genes for which expression levels differ between dealated precopulatory females (virgins) and egg-laying queens of the Formosan subterranean termite, Coptotermes formosanus Shiraki. We constructed a normalized polyphenic expressed sequence tag (EST) library that represents genomic material from most of the castes and life stages of the Formosan subterranean termite. Microarrays were designed using probes from this EST library and public genomic resources. Virgin females and queens were competitively hybridized to these microarrays and differentially expressed candidate genes were identified. Differential expression of eight genes was subsequently confirmed via reverse transcriptase quantitative PCR (RT-QPCR). When compared to virgins, queens had higher expression of genes coding for proteins related to immunity (gram negative binding protein), nutrition (e.g., termite-derived endo-beta-1,4-glucanase), protein storage, regulation of caste differentiation and reproduction (hexamerin, juvenile hormone binding protein). Queens also had higher transcript levels for genes involved in metabolism of xenobiotics, fat, and juvenile hormone (glutathione-S-transferase-like proteins, and cytochrome P450), among others. In particular, hexamerin, juvenile hormone binding protein, and a cytochrome P450 from the 4C subfamily are likely to be involved in initiating the inactive period during the reproductive cycle of the queen. Vice versa, virgins had higher expression than queens of genes related to respiration, probably due to recent flight activity, and several genes of unknown function.

Dual origin of gut proteases in Formosan subterranean termites (Coptotermes formosanus Shiraki) (Isoptera: Rhinotermitidae).

Cellulose digestion in lower termites, mediated by carbohydrases originating from both termite and endosymbionts, is well characterized. In contrast, limited information exists on gut proteases of lower termites, their origins and roles in termite nutrition. The objective of this study was to characterize gut proteases of the Formosan subterranean termite (Coptotermes formosanus Shiraki) (Isoptera: Rhinotermitidae). The protease activity of extracts from gut tissues (fore-, mid- and hindgut) and protozoa isolated from hindguts of termite workers was quantified using hide powder azure as a substrate and further characterized by zymography with gelatin SDS-PAGE. Midgut extracts showed the highest protease activity followed by the protozoa extracts. High level of protease activity was also detected in protozoa culture supernatants after 24 h incubation. Incubation of gut and protozoa extracts with class-specific protease inhibitors revealed that most of the proteases were serine proteases. All proteolytic bands identified after gelatin SDS-PAGE were also inhibited by serine protease inhibitors. Finally, incubation with chromogenic substrates indicated that extracts from fore- and hindgut tissues possessed proteases with almost exclusively trypsin-like activity while both midgut and protozoa extracts possessed proteases with trypsin-like and subtilisin/chymotrypsin-like activities. However, protozoa proteases were distinct from midgut proteases (with different molecular mass). Our results suggest that the Formosan subterranean termite not only produces endogenous proteases in its gut tissues, but also possesses proteases originating from its protozoan symbionts.

Testing protozoacidal activity of ligand-lytic peptides against termite gut protozoa in vitro (protozoa culture) and in vivo (microinjection into termite hindgut).

We are developing a novel approach to subterranean termite control that would lead to reduced reliance on the use of chemical pesticides. Subterranean termites are dependent on protozoa in the hindguts of workers to efficiently digest wood. Lytic peptides have been shown to kill a variety of protozoan parasites (Mutwiri et al. 2000) and also protozoa in the gut of the Formosan subterranean termite, Coptotermes formosanus (Husseneder and Collier 2009). Lytic peptides are part of the nonspecific immune system of eukaryotes, and destroy the membranes of microorganisms (Leuschner and Hansel 2004). Most lytic peptides are not likely to harm higher eukaryotes, because they do not affect the electrically neutral cholesterol-containing cell membranes of higher eukaryotes (Javadpour et al. 1996). Lytic peptide action can be targeted to specific cell types by the addition of a ligand. For example, Hansel et al. (2007) reported that lytic peptides conjugated with cancer cell membrane receptor ligands could be used to destroy breast cancer cells, while lytic peptides alone or conjugated with non-specific peptides were not effective. Lytic peptides also have been conjugated to human hormones that bind to receptors on tumor cells for targeted destruction of prostate and testicular cancer cells (Leuschner and Hansel 2004). In this article we present techniques used to demonstrate the protozoacidal activity of a lytic peptide (Hecate) coupled to a heptapeptide ligand that binds to the surface membrane of protozoa from the gut of the Formosan subterranean termite. These techniques include extirpation of the gut from termite workers, anaerobic culture of gut protozoa (Pseudotrichonympha grassii, Holomastigotoides hartmanni,Spirotrichonympha leidyi), microscopic confirmation that the ligand marked with a fluorescent dye binds to the termite gut protozoa and other free-living protozoa but not to bacteria or gut tissue. We also demonstrate that the same ligand coupled to a lytic peptide efficiently kills termite gut protozoa in vitro (protozoa culture) and in vivo (microinjection into hindgut of workers), but is less bacteriacidal than the lytic peptide alone. The loss of protozoa leads to the death of the termites in less than two weeks. In the future, we will genetically engineer microorganisms that can survive in the termite hindgut and spread through a termite colony as "Trojan Horses" to express ligand-lytic peptides that would kill the protozoa in the termite gut and subsequently kill the termites in the colony. Ligand-lytic peptides also could be useful for drug development against protozoan parasites.

Colony breeding system influences cuticular bacterial load of Formosan subterranean termite (Isoptera: Rhinotermitidae) workers.

The goal of this study was to test whether the breeding system and/or the degree of inbreeding of field colonies of the Formosan subterranean termite, Coptotermes formosanus, Shiraki (Isoptera: Rhinotermitidae) influences bacterial load on the cuticle of foraging workers. We enumerated bacterial load on the cuticle of groups of workers foraging in 20 inground monitoring stations surrounding the French Market in New Orleans, LA, and identified bacteria species using 16S rRNA gene sequencing. We used microsatellite genotyping to assign the 20 worker groups to seven simple family colonies (headed by a single pair of reproductives) and four extended family colonies (headed by multiple inbreeding reproductives) with a wide range of degrees of inbreeding. Workers from extended family colonies had a higher bacterial load than those from simple family colonies; however, bacterial load was not significantly correlated to the degree of inbreeding, possibly because of confounding factors in colony life history, such as age and/or size of colonies. Colonies with high bacterial load did not have a higher proportion of entomopathogens, and thus, bacterial load is not necessarily an indicator for disease risk. The majority of bacteria cultured from the cuticle of termites were soil bacteria with no known pathology against termites.