Detached Egg Surface Sterilization: Noting that most animals require a microbiome to facilitate the breakdown and assimilation of nutrients, initial studies focused in establishing the proof-of-concept that the phloem-feeding nature of whiteflies, combined with endosymbiotic bacteria, is sufficient to satisfy the nutritional requirements in a ‘communal axenic’ tissue culture environment. Whitefly eggs collected by gently brushing the underside of cabbage leaves (Fig. 1G) were subjected to surface sterilization. This approach was successful in producing emerging adults on meristem-propagated tissue cultured Ipomoea batatas L. (sweet potato) albeit at very low yield – likely due to a delicate balance of egg age and maturity. The axenic nature of the whiteflies was verified by absence of growth from macerated whitefly adults onto permissive R2A microbial growth agar plates (Reasoner and Geldreich 1985).
The life cycle of whiteflies at the growth temperature of 25oC is roughly 28 days (Aregbesola et al. 2020). Therefore, eggs laid by the primary adults that were produced from the surface-sterilized whitefly eggs require nearly a month of leaf viability to hatch and progress through the 4 nymphal stages to eclosion. Sweet potato is a vine that grows relatively quickly but was observed to excise lower leaves in tissue culture giving insufficient time for whitefly development. To try to overcome this limitation and allow for more efficient transfer of whiteflies to new plants, a culture device was fabricated by silicon gluing together two GA7 (i.e., clear, 3”x3”x4” polycarbonate) culture vessels with a 2-inch pass-through hole. Combining this with GA7 couplers, created an L-shaped culture device where new plants could be alternately replaced (see Fig. S1, Additional File 1). This methodology did provide proof-of-principle with whiteflies being maintained for over a year; however, the population of whiteflies was very low, and their appearance was sporadic. To initiate more efficient axenic whitefly establishment, an empty, inverted ‘upper’ GA7 culture vessel was installed via coupler above an upright GA7 ‘bottom’, containing a young cabbage plant. Whiteflies were collected in large numbers (100+) by disrupting the young cabbage leaves (see supporting video in Additional File 5). This collection approach was enhanced by placing bright yellow tape on the bottom of the (clear) upper GA7 to attack emerging whitefly adults.
Attached leaf whitefly sterilization: To expedite the initiation of large numbers of whiteflies, a GA7 of non-aseptic whiteflies was captured and coupled atop a lower GA7 containing four tissue cultured meristems of sweet potato and allowed to lay eggs. Egg laying was limited to 1–3 days by the appearance of contamination from the non-axenic whitefly honey-dew excrement. Since the whitefly eggs are rather securely attached to the leaf with a pedicel (Buckner et al. 2002), the sweet potato meristems could be uprooted from the agar and surface-sterilized using the same bleach and sodium-thiosulfate procedure by swirling the meristems in (cut-off) ultra-wide mouth 1L Erlenmeyer flasks (see supporting video in Additional File 2). Achieving re-sterilization of the contaminated tissue culture while maintaining viability of the eggs was constrained by systemic fungal infection that took about a week to emerge – often within the agar from the base of the cultured sweet potato meristem (see Fig. S2-A, Additional File 1). This fungus was consistent in appearance, but not identified; shown on potato dextrose agar (PDA (Fig. S2-B, Additional File 1)).
A final key improvement to this procedure was to grow the sweet potato transfer host phototrophically on sugar-free media in a 2% CO2-supplemented growth incubator. This combination of reduced contamination during the initial egg lay, with greatly reduced rate of systemic proliferation of the fungal contaminant, allowed for the 4 weeks needed for the next generation of whitefly emergence – although issues of contamination intermittently persisted, presumably by re-infection by feeding of emerging adults.
Facile, Abundant Axenic Whitefly Subculture: Noting that our desired tomato plant host is particularly prone to fungal infections (Timmer 1963), secondary emerging adults were transferred to 2-week-old cabbage seedlings initiated from surface sterilized seeds (Fig. 1F). This resulted in a dramatic proliferation of whiteflies with dozens apparent at the top of the upper coupled GA7 (Fig. 1E) in less than a month and no obvious fungal infections. Subsequent monthly subculture of whiteflies on 2-week-old cabbage provided hundreds of whiteflies over the next month to screen other plant species for their effectiveness as a maintenance host plant (MHP). Maceration of axenic whiteflies onto permissive microbial growth media confirmed the absence of culturable microorganisms, including the absence of slow-growing plant endosymbionts (Trauger et al. 2022). In screening alternative axenic plant hosts, three different assessments were made: (1) ease of PTC for given tissue culture vessel, (2) the effectiveness of the host for whitefly maintenance based on easily recoverable whitefly for transfer, and (3) observations of whitefly life cycle indicating feeding, egg lay and nymph development to adult (Fig. 2).
Initial MHP screening: Plants such as corn, common bean, and soybean grow rapidly to very large size, such that the month-long culture period (necessitated by the whitefly life cycle) was difficult to contain, even in a coupled 8-inch-tall tissue culture vessel. Cowpea was more manageable, with pea providing a very convenient and compact growth habit. Okra displayed a strange behavior of pressing itself out of the agar, resulting in most of the roots out of the agar (even when germinated submerged), and sometimes inverting the plant. Tomatillo displayed a highly variable response, where its growth habit was similar to Ipomoea in that it rapidly grew tall with lower leaf excision. In an attempt to reduce tomatillo height, a plant in a GA7 was grown on a lighted gyratory shaker to impose thigmomorphogenesis-mediated growth height reduction (Jaffe and Forbes 1993). Although reduction of height was minimal, there was a notable increase in leaf retention (see Fig. S3, Additional File 1), and more consistent production of adult whiteflies.
Numbers of seedlings in a tissue culture vessel was a significant consideration; for example, more than a single tomato seedling would result in highly crowded and epinastic growth. Recommended seed numbers are provided in Additional File 3 along with pictures of the plants at early stage (of whitefly inoculation) and late stage (i.e., after a month of proliferation) to give insight into their potential utility as experimental systems for whitefly-host viral transmission studies. Cucumber and radish both provided a compact plant with robust cotyledons; radish – which is also Brassicaceae – displayed overall replication numbers similar to broccoli and cabbage, both of which resulted in relatively high populations of adult whiteflies. Consistent with observations of tomato var. Lanai, another tomato variety also displayed virtually no recoverable whiteflies. An effort to obtain a ‘reduced trichome mutant’ of tomato(Ki et al. 2021) was unfortunately not successful due to COVID disruption of breeding programs (personal communication, Dr. Dani Zamir). Other reduced tomato trichome mutants would be a high future priority for testing (Fonseca et al. 2022).
Ultimately, six different prospective MHPs proceeded for further assessment of whitefly production: cabbage, Ipomoea, tomatillo, tomato, and two species of the viny monocot yam, Dioscorea cayenensis and D. rotundata. The two Dioscorea species and tomato were the intended targets for virus transmission studies; while cabbage, Ipomoea, and tomatillo were chosen to represent a diverse species group. Sweet potato (was cultured from meristems and all other were grown from surface sterilized seed.
Quantitative MHP Evaluation by Accumulation Rates: After an initial inoculation of 20 ± 2 adult whiteflies and a 2-week incubation period, adult whiteflies started to emerge and fly to the upper GA7. These were captured daily to monitor accumulation rates over a six-week period (see Fig. 3A). All accumulation curves experience a ‘lag’ period of more than 3 weeks, corresponding to the egg lay and nymph development period, followed by accumulation rates that varied dramatically for different MHPs. This assessment is a useful quantification due to the need to efficiently collect and transfer WF in research applications.
Tomatillo achieved a significantly higher production of whitefly when it was grown on a shaker for enhanced leaf retention due to thigmomorphogenesis. Our intended targets for virus transmission studies, Dioscorea and tomato, displayed essentially no whitefly capture over the entire 6-week period. As we and others have measured proliferation rates of whitefly on tomato, this observation is clearly an artifact of the PTC environment and/or the accumulation method. The compact trichome density in tissue cultured tomato is a possible explanation of poor proliferation as well as simply avoidance of leaving the underside of the tomato leaves (see Harvest results below). As is clear from Fig. 2, whiteflies clearly proliferated on these hosts, with prolific eggs, nymphs and exoskeletons of 4th instars nymphs indicating emerged adults (see Fig. S4, Additional File 1 for additional MHP screen photos).
Post-Accumulation Harvest: Noting that accumulation via the upper coupled GA7 is a functional measure of the utility for experimental colony maintenance—rather than a direct measure of proliferation—a destructive harvest was undertaken for cabbage, tomato, D. cayenensis, and sweet potato (Ipomoea), and acyl-sugar knockout (ASKO) Nicotiana benthamiana (N.b.) (Fig. 3B). At the end of the 6-week accumulation period, all but one of the (4) experimental replicates were harvested for examination of total whitefly proliferation. After placement of the GA7 culture vessel in a cold room (4°C) to immobilize all remaining whiteflies, the contents (plants, black-sand coated agar, and walls) were carefully examined to count the total final adult whitefly harvest. Both cabbage and sweet potato revealed an additional ~ 100 whiteflies present at the end of the 6-week cultivation period for a total of about 300 adult whiteflies and corresponding developing nymphs. Figure 3B summarizes the overall total of proliferated whiteflies over the 6-week period including accumulated, harvested, and an estimate of viable whiteflies at the time of harvest (see below). Cabbage and Ipomoea provided the most overall quantity of whiteflies, with cabbage providing more consistent proliferation between replicates. It is not surprising that Ipomoea can act as a host since the common name of our vector is ‘sweet potato whitefly’. These overall proliferation totals suggest the maximum capacity of this culture system when whitefly proliferation is non-MHP-limited. This final harvest approach provided additional insight into the limitation presented by tomato and D. cayenensis, where both display considerable proliferation beyond the inoculum that were not recoverable via accumulation capture. For the wild-type N. benthamiana (Fig. 4B); while the whitefly quantity more than doubled from the initial inoculum number, the vast majority of the whiteflies remained associated with the plant and the culture vessel. Notably, many whiteflies were stuck onto the leaf upper surface that was not apparent for the ASKO N.b. This observation supports the conclusion of the physical role of acyl-sugar stickiness in capturing whiteflies.
Viability Assay: For further assessment of MHPs, the health of the whiteflies at the end of the accumulation time course was assessed using a quantitative fecundity bioassay on cabbage as described elsewhere (Thompson et. al. 2022). In brief, the reserved GA7 tissue culture vessel (i.e., subset not harvested) was introduced into a screen-cage to release the whiteflies onto cabbage to allow proliferation on this favorable host plant. After an additional six weeks of proliferation, the whitefly amplification was assessed using image analysis, and then the initial viable whiteflies needed to give rise to the bio-assay quantified whitefly count was back-calculated using our previously parameterized predator-prey model for this experimental condition. This approach provided for a quantitative estimate of the corresponding number of flies that were healthy and viable on the different MHPs at the end of the accumulation period (Fig. 3B). These results clearly confirm that substantial numbers of axenic whitefly proliferated, which are not necessarily proportional to the whitefly recovery rates, from the tissue culture vessels.
MHP Whitefly Preference Comparisons: With a fully axenic colony, direct comparisons of insects’ preferred MHP are straightforward to implement. D. cayenensis and rotundata were subcultured into the same GA7 (see Methods); whiteflies were then added and after a 30-day proliferation period, quantifiable comparisons of nymphs and emerged exoskeletons that had developed on each plant species were possible (Fig. 5A-B). A further test comparing D. cayenensis to Ipomoea within the same GA7 confirmed accumulation study results of the preference of Ipomoea (Fig. 5D-E). The comparison of N. benthamiana wild-type to the acyl-sugar knockout (Fig. 4) is particularly interesting because of the hypothesized insecticidal roles of acyl-sugars as either mechanical (stickiness of trichomes) or chemical (metabolic disruption) (Feng et al. 2021). An end-point harvest was performed after the 6-week accumulation study as above (Fig. 4B). Consistent with a several-fold higher accumulation and final harvest, N. benthamiana ASKO displayed far fewer whiteflies ‘stuck’ to the top surface of the leaves compared to wild type N. benthamiana.
Axenic Virus Transmission: A major motivation for establishing an axenic whitefly culture was the goal for a biocontained experimental system for the study of plant viruses – in this case, whitefly mediated virus transmission (VT) of a begomovirus. Axenic whiteflies fed on virus-infected ASKO N.b. plants for a 72-hour minimum acquisition access period (AAP) and were moved to fresh axenic ASKO N.b. After a one-week inoculation access period (IAP) on the fresh plants, successful ToMoV transmission confirmed a P-I-V/VT efficacy. The presence of the ToMoV virus in the target plant was confirmed by gel extraction and subsequent Sanger sequencing of the PCR amplicon (Fig. 6).
aP-I-V/VT Limitations: Although the Agrobacterium virus launch into tissue cultured plants provided a proof-of-principle for the transmission of ToMoV and achieved a highly biocontained procedure for virus introduction into the axenic whitefly colony, it has limitations for extension to other viruses. We therefore sought to develop alternative approaches to introducing filter sterilized intact virus to the whitefly. Filter sterilization of the virus extract was accomplished using a 0.22 µm PTFE syringe filter. Many viruses can be isolated in encapsidated form which facilitates feeding to the whitefly through the ‘sachet method’ of droplets sandwiched between layers of parafilm stretched thin (Zhou et al. 2017). We demonstrated that the wax film could be sterilized with a 10Gy dosage gamma irradiation treatment from a Cobalt-60 Dry cell gamma irradiator for 7 hours. Since the gamma irradiation treatment imparted brittleness to the parafilm it was necessary to stretch the film prior to irradiation. This aseptic sachet method was successful in transferring ToMoV into the axenic whiteflies to initiate transmission of the virus to plant tissue cultured ASKO N. benthamiana as confirmed by PCR. Additionally, artificial feeding of whiteflies through autoclavable membranes has been reported in a brief description (Davidson et al. 2000). Since this approach would be more broadly accessible, we identified a readily available autoclavable PFTE asymmetric membrane and confirmed whitefly feeding based on uptake of a fluorescent feeding solution containing 150 g/L sucrose and 10 µM Texas Red dye. The feeding device details, including implementation into a GA7 plant tissue culture lid, are described in Additional File 4.