Introduction

The Asian citrus psyllid, Diaphorina citri Kuwayama (Hemiptera: Liviidae), is the primary vector of the phloem-limited bacterium Candidatus Liberibacter asiaticus (CLas), the putative causal agent of huanglongbing (HLB, citrus greening) disease. HLB is currently the most destructive citrus disease in many parts of the world including Asia, USA, Brazil, and Central America1,2,3. CLas can infect all commercial citrus cultivars causing major reductions in fruit quality, yield and lifespan of infected trees2. HLB has spread quickly in Florida-USA since its first detection in 2005, and is now threatening citrus production in Texas, California and other parts of the USA4.

CLas is transmitted in a circulative-propagative manner by D. citri; it is acquired by psyllid nymphs or adults during feeding on the phloem of infected citrus plants and ingestion of CLas into the psyllid’s gut, followed by its translocation and probable multiplication into various organs including the salivary glands5,6,7,8,9. During feeding/salivation of the vector, CLas presumably moves from the salivary glands with salivary secretions into citrus phloem resulting in pathogen inoculation of susceptible plants9,10. CLas is also transmitted transovarially to the eggs and nymphs from infected mothers but only at a very low rate, ≤ 1% (2–6% in pooled samples of eggs and 3rd- 5th instar nymphs)6. D. citri nymphs acquire CLas from infected citrus plants much more efficiently than adults5,6,8 likely because nymphs feed (ingest) from the phloem more frequently and for longer durations than adults11, although other factors including differences in innate immunity between nymphs and adults may also influence successful acquisition12,13,14.

Diaphorina citri adults that acquired CLas from infected citrus plants as nymphs have higher titers of CLas and are more likely to transmit/inoculate this bacterium than those acquiring the bacterium during the adult stage5,8. However, most of the previous investigations on CLas transmission by D. citri have been conducted using mainly adults for inoculation tests, whether these adults had acquired the bacterium during the nymphal or adult stages5,6,8,15. This was probably done for several reasons: (1) D. citri nymphs are much more fragile than adults and nymphal duration is much shorter than the adult life span16, (2) D. citri nymphs feed/probe (i.e. insert their stylets into host tissue) nearly continuously11, and thus are more difficult to handle or transfer without causing injury to their stylets/mouthparts during transmission experiments, and (3) Nymphs cannot fly and typically remain on the plant from which they were hatched, thus contributing in only a minor way to CLas transmission to other plants. However, there are epidemiological, molecular and/or behavioral questions about CLas transmission by D. citri that could be answered from a more complete understanding of acquisition/transmission rate comparisons between nymphs and adults.

In this work, we tested the inocualtivity of 4th instar nymphs and 1–4 week old adults of D. citri when both stages of the vector were reared on CLas-infected citrus plants for several generations. Our experimental design insured that tested nymphs and adults had access to infected plants at the youngest possible age (early first instar nymphs) without the difficulty of having to handle/transfer the very delicate first instar nymphs. In these tests, we used the excised leaf assay method developed and used earlier by Ammar et al.17 and Raiol-Junior et al.18. Complementary electrical penetration graph (EPG) measurements indicated that feeding behavior of D. citri on excised citrus leaves was indistinguishable from that observed on intact leaves of whole plants11. Our results support the hypothesis that D. citri adults inoculate CLas more efficiently than nymphs following acquisition of CLas by early instar nymphs because: (1) the titer and/or incidence of CLas in the salivary glands of emergent adults are higher than in nymphs, and (2) adults are afforded a longer latent period necessary for multiplication and/or translocation of CLas into the salivary glands of the vector. We propose that CLas uses D. citri nymphs mainly for pathogen acquisition and multiplication, and their adults mainly for pathogen inoculation and disease spread.

Results

CLas-infected D. citri adults are more efficient than nymphs in inoculating CLas into citrus leaves

In Expt. 1, we compared the inoculativity of D. citri nymphs (4th instar) versus that of adult males or females (1–2 week old) that had been reared for several generations on CLas-infected citron plants. The proportion of CLas-infected (qPCR positive) leaves, following a 4-day inoculation period (with 5 insects/leaf), was 2.5% (1/40 leaves) for nymphs and 36.3% (29/80 leaves) for adults, and this difference was highly significant (P = 0.0001 in X2 test). This difference was consistent in three out of four tests of Expt. 1 (Table 1). No significant difference was found between adult males and females with regard to the proportion of CLas positive leaves (32.5 and 40.0% for males and females respectively; P = 0.4839). The Ct value of CLas positive leaves was 25.5 for a single leaf inoculated by nymphs, and 26.6 ± 0.55 in leaves inoculated by adults.

Table 1 Percentage of CLas-infected (qPCR-positive) excised citron leaves that were inoculated by D. citri nymphs or adults for 4 days (5 insects/leaf, 10 leaves/treatment/test), then assayed by qPCR one week later (Expt. 1)*.
Full size table

In this experiment, the proportion of CLas-infected (qPCR positive) psyllids was 79.5% for nymphs (n = 200) and 91.4% for adults (n = 394) (P = 0.01). However, there was no significant difference in infection rate between male and female adults (Supplementary Table S1).

In summary, the results of Expt. 1 indicated that young (1–2 week old) D. citri adults reared on infected citrus for several generations exhibited a higher infection rate, and were more efficient in inoculating CLas into citrus leaves, compared to 4th instar nymphs reared similarly. Furthermore, there were no differences in inoculation efficiency or infection rate between adult males and females reared on infected citrus.

CLas incidence and titer in the head-thorax and abdomen sections are higher in adults than in nymphs of CLas-infected D. citri

The salivary glands of D. citri are found in the head and first two thoracic segments, and we postulated that entry and/or multiplication/accumulation of CLas in the salivary glands are essential for CLas transmission and inoculation to occur. To test this hypothesis, we compared the incidence and titer (as measured by normalized Ct values) of CLas in the head-thorax and abdomen sections of nymphs and adults separately in Expts. 2 and 3.

In Expt. 2, which was comprised of two tests (with 30 nymphs and 40 adults/test), the proportion of CLas-infected (qPCR positive) anterior (head-thorax) sections in the adults (45%, n = 80) was significantly higher than that in nymphs (26.7%, n = 60, P = 0.040). Also, a significantly higher proportion of infected abdomens was found in adults (92.5%) than in nymphs (65%, P = 0.0001). Both nymphs and adults had a significantly higher proportion of infected abdomens than infected head-thorax sections (P = 0.0001 and 0.0002 for adults and nymphs, respectively).

In this experiment, the CLas titer found in infected adults was significantly higher (= lower Ct values) than in nymphs in both the anterior (head-thorax) and posterior (abdomen) sections (Table 2). ANOVA on Ct values indicated that the effect of life stage (nymphs vs. adults) on CLas titer was highly significant (P = 0.0000), but the effect of body section and the interaction between these two factors were not significant (P = 0.992 and 0.992, respectively).

Table 2 Normalized Ct values of CLas in the anterior (head-thorax) and posterior (abdomen) sections of D. citri nymphs and adults that were reared on CLas-infected citron plants (Expts. 2 and 3).
Full size table
Table 3 Normalized Ct values of CLas in the head-thorax and abdomen sections of younger and older D. citri adults that were reared for several generations on CLas-infected citron (Expt. 3).
Full size table

In Expt. 3, we compared 4th instar nymphs vs. 1–2 week and 3–4 week old adults to determine if adult age may affect the infection rate or CLas titer. No differences were found between younger and older adults in the proportion infected or in CLas titer for both the abdomen and head-thorax sections (Table 3). However, the proportion of CLas-infected abdomens was significantly higher in adults (86.7%) than in nymphs (56.7%, P = 0.006). With regard to the head-thorax section, the proportion infected was higher in adults (36.7%) than in nymphs (30%) but this difference was not significant (P = 0.532). Additionally, ANOVA on the data summarized in Table 2 (Expt. 3) indicated that the effect of life stage on CLas titer in both body sections was highly significant (P = 0.0000), the effect of body section was significant (P = 0.0302), but the interaction between these two factors was not significant (P = 0.221).

In summary, the results of Expts. 2 and 3 indicated that D. citri adults reared for several generations on CLas-infected citron plants exhibited higher incidence and titer of CLas than did nymphs reared similarly, when measured either in the head-thorax (which includes the salivary glands) or abdomen sections. Additionally, there were no differences in infection rate or CLas titer in these sections between younger (1–2 week) and older (3–4 week) psyllid adults.

Feeding behavior of D. citri nymphs and adults on excised citron leaves as measured by EPG recordings

Diaphorina citri nymphs have been reported to ingest from, and salivate into, the phloem of healthy or CLas-infected citron plants more frequently and for longer durations than adults11. However, in that study only healthy psyllids (previously unexposed to CLas) were tested on intact leaves of citron plants. In the present work, we investigated the feeding/probing activities of CLas-infected D. citri nymphs and adults using the same EPG equipment, recording duration (42 h) and methodology as that in George et al.11 except for using excised (rather than intact) healthy citron leaves.

Our EPG results (Table 4, Fig. 1) were congruent with those obtained previously by George et al. 11. In short, the duration and frequency of phloem salivation and ingestion phases (E1 and E2 waveforms, respectively) recorded from CLas-exposed D. citri nymphs (4th–5th instar, n = 51) were significantly greater than those recorded from CLas-exposed adults (1–2 week old, n = 52). The percentages of CLas-infected (qPCR positive) nymphs and adults used in this EPG study were 87.5 and 100% respectively, with mean Ct values of 35.97 ± 0.48 and 29.31 ± 0.52 for nymphs and adults, respectively. The difference in the proportion of infected nymphs and adults was not significant (P = 0.07), but the difference in mean Ct values was highly significant (P = 0.001), indicating higher CLas titer in adults than in nymphs as we reported above in Expts. 1–3.

Table 4 Electrical penetration graph (EPG) recordings of feeding/probing activities of CLas-exposed D. citri nymphs (4th–5th instar, n = 51) and adults (n = 52) on excised, healthy, young citron leaves over 42 h: no. of bouts (count), mean duration (minutes) and total duration (count × mean duration) for each waveform (data analyzed by non-parametric Kruskal–Wallis test).
Full size table
Figure 1
figure 1

Graphic presentation of various feeding/probing activities (waveforms) of CLas-exposed D. citri nymphs (a) and adults (b) on excised healthy citron leaves. EPG recordings were performed for 42 h. The total duration for each activity is indicated by percentage in the pie charts. The data represent the average of 51 nymphs and 52 adults.

Full size image

During the 42 h recording period, nymphs performed significantly more phloem ingestion bouts (7 ± 1) than did adults (2 ± 1). On the other hand, adults engaged in more bouts of xylem ingestion (9 ± 1) than nymphs (2 ± 1). Additionally, insect stage had a significant effect on the mean and total duration of all waveforms (Table 4, Fig. 1). The duration of individual bouts of waveforms D, E1 and E2 (all related to phloem feeding activities) were significantly longer for nymphs than adults. Total duration of phloem ingestion (E2) was significantly longer in nymphs (1823 ± 94 min) than in adults (385 ± 44 min), as was the total duration of phloem salivation for nymphs (13.0 ± 1.3 min.) compared to adults (3.2 ± 0.6 min.). Nymphs were engaged in phloem ingestion 5 times longer, on average, than were adults; but the adults spent 9 times longer engaged in xylem ingestion compared with nymphs. During the 42-h recordings, nymphs spent 68% of their time (30.4 h) engaged in phloem ingestion, whereas adults spent only 14% of their time (6.4 h) in phloem ingestion. On the other hand, adults spent 17% of the total feeding time engaged in xylem ingestion, whereas nymphs spent only 2% of their time in xylem feeding activities (Fig. 1).

In summary, our EPG results indicated that feeding behavior of CLas-infected D. citri nymphs and adults on excised leaves was indistinguishable from that described previously for unexposed D. citri on intact citrus leaves11. These results indicate that differences in inoculation rate between CLas infected D. citri nymphs and adults, reported above, are not associated with differences in feeding behavior on excised versus intact citrus leaves.

Discussion

The circulative-propagative mechanism of transmission that occurs among hemipteran-borne plant pathogen vectors has been hypothesized to occur in the CLas/HLB pathosystem7,10. This mechanism is comprised of several phases that include: (a) ingestion of the pathogen and its movement into the vector’s gut during phloem feeding on an infected host plant, (b) translocation of the pathogen from the gut into the hemolymph and various tissues including the salivary glands with probable pathogen multiplication in these tissues, and (c) inoculation of the pathogen by injection within salivary secretions during subsequent feeding on a susceptible host7,10,19. However, prior to inoculation, there is a ‘latent period’, during which the pathogen translocates into, and presumably multiplies in, the salivary glands. The minimum latent period for CLas in D. citri that acquired CLas as nymphs ranged between 7 and 10 days, with a median of 16.8 days8,20. The total nymphal duration of D citri at 25 °C averages 12.8 days16. This is much shorter than the median latent period reported for CLas, and hence it is expected that most of CLas inoculations occur at the adult stage, rather than during the preceding nymphal stage of the vector, as our results here suggest.

In the present work, we used D. citri that were reared for several generations on CLas-infected citrus plants. Since transovarial transmission of CLas in D. citri may occur at a very low rate approaching 1%6, it is likely that the great majority of D. citri nymphs and adults used here had acquired CLas by feeding on infected plants as early instar nymphs. Our results indicate that CLas-infected adults were more efficient than nymphs in subsequently inoculating this pathogen into citrus leaves. Congruently, these adults were characterized by greater CLas incidence and higher CLas titer (lower Ct values), than counterpart nymphs, in the head-thorax body section that included the salivary glands. In D. citri head and first 2 thoracic segments, other than the salivary glands, there are mainly neural tissues (the brain and compound ganglionic mass) which are rarely infected with CLas9,23,26. Although CLas incidence and titer was also higher in the abdomen section of adults, compared to nymphs, this difference was not consequential to the inoculation process, which requires presence of CLas in the salivary glands according to the circulative and circulative-propagative model7,10,19. Additionally, it has been reported that overall CLas titer (in the whole psyllid) is positively correlated with higher inoculation rates on excised leaves or intact citrus plants15,17. Similarly, transmission experiments using field-collected D. citri in Japan suggested that CLas-transmitting insects tend to exhibit higher bacterial densities than do non-transmitting insects, and that a threshold level of CLas density in the psyllid (ca.106 bacterial cells per insect) is required for successful transmission to occur21. In these investigations, CLas titer in the salivary glands was unknown; however, Wu et al.22 reported that when CLas was acquired by D. citri nymphs the proportion of infected hemolymph and salivary glands, as well as their CLas titer, increased significantly 12–18 days following a 3-day acquisition access period. These results suggest that CLas multiplies and/or accumulates in the vector’s hemolymph and salivary glands during the latent period.

The higher CLas incidence (infection rates) and titers measured from the head-thorax sections of adults compared to nymphs is consistent with our hypothesis that greater transmission efficiency observed in adults than nymphs is related to the latent period during which the pathogen must translocate to, and/or multiply in, the salivary glands before inoculation can occur. This latent period may be incomplete during nymphal development, which requires less time to complete than the median reported CLas latent period in this vector16,20. Our results also indicated that the proportion of infected head-thorax sections was considerably lower than that of infected abdomens in both nymphs and adults. This is consistent with previous findings indicating that the infection rates of salivary glands from field-collected or laboratory-infected D. citri adults are lower than those of infected guts or other tissues23,24. Similarly, Cooper et al.25 found that fewer salivary glands of the potato psyllid were infected with Candidatus Liberibacter solanacearum (CLso) compared with the alimentary canals from those same individuals. These results suggest the existence of transmission barriers that may impede CLas and CLso from exiting the gut and/or entering salivary glands in some vector psyllids of these phytopathogens10,26. These barriers are probably more permissive in D. citri nymphs than in adults, which may also explain that CLas multiplies faster in nymphs than in adults5,8,10.

In our inoculativity tests, we used the excised leaf assay method developed and used earlier by Ammar et al.17 and Raiol-Junior et al.18, because it saves considerable time, space and material compared to using whole plants (qPCR results can be obtained in 2 weeks rather than ≥ 6 months post-inoculation). Here, this method was also much more suitable for testing nymphs because they were confined into a much smaller space (in a 50 mL tube) which reduced their chance of getting lost or dropping off the plant during the test. Inoculation rates (% CLas-positive leaves or plants) using these two methods were comparable in two previous studies17,18. Furthermore, the EPG recordings conducted here indicated no differences in feeding behavior of D. citri on excised citrus leaves, compared to those made previously using intact leaves on whole plants11. In both studies, the occurrence (frequency) and duration of both E1 and E2 waveforms, representing phloem salivation and ingestion phases respectively, were much greater in nymphs than in adults. Previous EPG studies indicated that CLas acquisition occurs during phloem ingestion and that CLas inoculation occurs during phloem salivation phases27,28.

In our experiments, there was no differences in infection rates or titers measured from 1 to 2 week versus 3 to 4 week old adults. These results are perhaps unsurprising, since multiplication of CLas in D. citri adults is known to occur much more slowly than in nymphs5,8. Also, there were no gender related differences in the proportion of insects that were CLas-infected, inoculative, or in their CLas titer. Similarly, no gender differences were reported from laboratory reared, CLas-infected D citri adults15,22, but the proportion of infected females was generally higher than that of males in field-collected samples from Florida29. Since D. citri females live much longer than males30 this might affect their rate of infectivity under field conditions, which may bias the proportion of infected psyllids slightly toward females.

Epidemiologically, it is interesting to note that although D. citri nymphs are much more efficient than adults in CLas acquisition from infected plants5,6,8,11,22, adults appear to be more important in CLas inoculation and disease spread because of three factors: (a) their greater mobility compared to nymphs3, (b) longer occurrence within the lifecycle (adult life span vs. nymphal duration)16, and (c) greater efficiency of CLas inoculation by adults compared to nymphs as we reported here. Lee et al.31 proposed the model of ‘flush transmission’ of CLas between infected D. citri adults and their young nymphs via the host plant when these nymphs feed on newly flushing leaves that became recently infected by their parents. However, they did not specify whether CLas inoculation during the second generation (that acquired CLas from the flush) occurs during the nymphal or adult stages. Based on our results, it seems reasonable to assume that most of these 2nd generation psyllids would inoculate CLas mainly during their adult life. This also increases the significance of adults for spreading CLas infection in the field. Thus, in an evolutionary sense, CLas seems to have taken maximum advantage of both stages of its vector D. citri, using the nymphal stage mainly for acquisition and pathogen multiplication in the vector5,8,22, and the adult stage mainly for pathogen inoculation and short or long-distance disease spread. This apparent specialization of the two developmental stages of D. citri may have been aided by differences in innate immunity between nymphs and adults, as suggested by quantitative proteomic and microscopic studies of uninfected and CLas-infected D. citri12,13,14.

Although the rates of CLas inoculation per insect, especially in Florida, is fairly low6,8,10,17 this is more than compensated for by the great reproductive potential and short generation time of D. citri3,16 in addition to the beneficial effects of this bacterium on its psyllid vector32. Our work helps to further elucidate the great degree of adaptation between CLas and its primary vector and the different roles played by nymphs and adults of D. citri in the transmission cycle of CLas. These differences should be taken into consideration when devising methods for combating HLB, the most serious citrus disease worldwide.

Methods

Psyllids and plants used

Diaphorina citri nymphs and adults originated from a healthy (uninfected, non CLas-exposed) laboratory colony maintained for several years in the ARS-USDA Laboratory in Fort Pierce, FL. on healthy (uninfected) orange jasmine [Murraya paniculata (L.) Jack] trees, and more recently on healthy (uninfected) citrus trees (Citrus macrophylla Wester) in the greenhouse. Individual adults from the colony were sampled and qPCR-assayed every three months to ensure that the colony remained CLas-free. CLas-exposed (potentially infected) D. citri nymphs and adults were obtained by rearing uninfected D. citri from eggs on young CLas-infected citron plants (Citrus medica) that tested positive for CLas by qPCR. Routine qPCR was conducted on healthy and infected colony insects and citrus trees using Li primers33. Healthy (uninfected) citron plants were grown from seeds under insect-proof greenhouse conditions, and their excised leaves were used for testing CLas inoculation by D. citri nymphs and adults as described below.

CLas-infected D. citri nymphs and adults

Diaphorina citri has five nymphal instars that can be distinguished morphologically and by size3. Fourth instar nymphs and 1–2 or 3–4 week old adults were selected from known CLas-positive D. citri colonies, with high CLas titers (= lower Ct values, typically Ct values below 28 for single adult samples and an infection rate typically greater than 70%) as evidenced by previous qPCR tests. Fourth instar nymphs were chosen for our experiments because they are hardier than earlier instars, allowing handling with less chance of damage, while allowing at least 4 days of testing prior to molting into adults16. These nymphs were manipulated with a moistened fine paint brush under a stereomicroscope. D. citri nymphs are known to insert their stylets into citrus leaves and feed/probe continuously for several hours11. Thus, in order to avoid injuring their delicate mouthparts, they were “tickled” gently with the paint brush until they stopped probing and began to move. They were then gently picked up with the paint brush and placed on the underside of excised healthy young citron leaves, and then caged within modified conical tubes as described below.

One to 2 or 3–4 week old D. citri adults were collected from the same CLas-positive colony as nymphs for use in the same experiment. Adults were collected with an aspirator and placed at 4° for 10–20 min to briefly immobilize them. They were then examined under a stereomicroscope to determine their sex. In some tests, males and females were separated and placed in tubes with 5 insects/leaf/tube. However, earlier tests showed no significant differences between adult males and females with regard to CLas infectivity or inoculativity (see “Results”); thus in later tests mixed gender adults were used.

Experiment 1: testing inoculativity of D. citri nymphs and adults after rearing on CLas-infected plants

Healthy, young citron leaves were used for testing the inoculativity of D. citri nymphs and adults that had been reared for several generations on CLas-infected citron plants, using a slightly modified excised leaf method developed and used earlier by Ammar et al.17 and Raiol-Junior et al.18. Fully expanded young leaves (5–7 cm long) were selected from healthy (uninfected) citron plants. They were excised by a razor blade with their petioles intact. Leaf petioles were then placed under water and cut again with a razor blade at 45° angle. Leaves were immediately placed in a modified conical base of 50 mL conical tubes with 3 mL of sterile water (Supplementary Fig. S1). The base/bottom part of the tube had been cut earlier to facilitate adding more water when necessary without disturbing the feeding psyllids in the tube. Parafilm membrane was used to maintain contact between the base and remaining part of the tube except when water was added. Five D. citri nymphs (4th instar), or adult males or females (1–2 week old) were placed on each leaf, after which tubes were capped with a screen screw top17. The nymphs or adults placed on an excised leaf were allowed to move freely and to select their preferred feeding sites on that leaf during the 4 days of inoculation. These tube-caged leaves were kept on a lab bench for 4 days under a grow light (24 °C, 14 h of light) before the insects were removed, frozen, then tested later individually for CLas by qPCR. The 4-day inoculation period was chosen so that most of the 4th instar nymphs tested did not molt into adults prior to the conclusion of the test, and 5 insects/leaf was chosen because it yields much higher inoculation rates than using a single insect/leaf17. At the conclusion of each test, leaves in the conical tubes were placed into an illuminated incubator at 25 °C for 7 days, which is known to increase CLas titer in these leaves17. Thereafter, the leaves were washed thoroughly before being tested for CLas with qPCR essentially as described earlier8,17. The main difference was that total nucleic acid extractions containing both DNA and RNA were analyzed using q-RT-PCR (quantitative-reverse transcriptase- PCR) instead of just qPCR. We have found that detection of both DNA and RNA in q-RT-PCR reactions is a more sensitive method of detecting CLas transmission to plant leaves. This analysis was performed using the same cycling parameters as previously described with the exception of using the GoTaq 1-Step RT-qPCR System (Promega, Inc., Madison, WI) and their one-step pre-incubation cDNA synthesis recommendation.

In Expt. 1, the inoculativity test on excised leaves (5 insects/leaf, 10 leaves/ treatment/test) was repeated four times. The proportion of insects that survived at the end of the 4-day tests was 90% for nymphs (n = 200) and 94.8% for adults (n = 400). Also, at the end of the 4-day test, 26.1% of the nymphs used emerged as adults.

Experiments 2 and 3: Testing the anterior (head-thorax) vs. posterior (abdomen) parts of nymphs and adults by qPCR

The two salivary glands of D. citri are located in the head and first two thoracic segments23,24,26. Thus, in order to test the hypothesis that differences in the occurrence and/or titer of CLas in the salivary glands may contribute to differences in CLas inoculativity between D. citri nymphs and adults, we performed qPCR tests on the anterior (head-thorax) and posterior (abdomen) sections of each insect separately. Based on preliminary trials, we estimated that dissecting and handling the tiny and delicate salivary glands of a large number of psyllids, especially nymphs, would be challenging logistically, and some of these tiny salivary glands could be damaged or lost during dissection or processing24. Thus, dissecting the thorax-head part was determined to be a suitable compromise, especially that the main tissues in D. citri head and first 2 thoracic segments, other than the salivary glands, are neural tissues (the brain and compound ganglionic mass) which are rarely infected with CLas9,23,26.

In Expt. 2 we tested 4th instar nymphs vs. 1–2 week old adults (two tests, with 30 nymphs and 40 adults/test), while in Expt. 3 we tested 4th instar nymphs vs. 1–2 or 3–4 week old adults (n = 30 for each category). Fourth instar nymphs and 1–4 week old adults were selected from the same CLas positive colony as described above. Under a stereomicroscope, each insect was carefully cut between the second and third thoracic segments with a clean sharp razor blade (a new blade was used for each insect). For simplicity, we refer to these anterior and posterior sections as ‘head-thorax’ and ‘abdomen’, respectively, although the posterior section included the 3rd thoracic segment in addition to the full abdomen. Each section was placed in a separate 1.5-mL tube, frozen on liquid nitrogen and stored at − 80 °C until extractions were conducted. For DNA extraction, samples were ground with a small tube pestle in liquid nitrogen, and then homogenized further with pestles in 150 µL TE buffer (10 mM Tris–HCL, 1 mM EDTA pH 8.0). Then 150 µL of phenol (equilibrated with 10 mM Tris–HCL pH 8.0) was added and the samples were vortexed. The aqueous phase was removed and DNA was precipitated by addition of 0.5 volumes of 5 M ammonium acetate and 2 volumes of 95% ethanol followed by centrifugation for 15 min at 21,000×g. The DNA pellet was washed with 300 uL of ice-cold 70% ethanol, allowed to air dry for 10 min and then resuspended in 30 µL of sterile deionized water and stored at − 80 °C until used. The qPCR parameters for CLas detection were as described previously8. Normalization of CLas 16S rRNA Ct values to psyllid genome equivalents was performed by simultaneous qPCR detection of the CLas 16S rRNA gene and the D. citri Ribosomal S20 (RPS20) psyllid gene (Genbank accession #DQ673424) used as an internal reference: Dci-S20-L:5′- GCCCAAGGGCCCAATCA -3′, and Dci-S20-R:5′- GGAGTCTTACGGGTGGTTATTCTG-3′. Normalization was performed by subtracting the Ct value for the individual with the lowest Dci-S20 Ct value from all the individual Dci-S20 Ct values and then the resulting number from each individual was subtracted from that individual’s CLas Ct value. For statistical analysis on the results of Expts.1–3, X2 tests were conducted on the proportion of CLas-infected (qPCR positive) leaves or psyllids, while T tests and/or ANOVA were conducted on the Ct values of the individual psyllids tested by qPCR, using the stat. analysis program JMP (v. 10, SAS Inc., 541 Cary, NC). It has been established earlier that Ct values are negatively correlated with CLas titer in the psyllids and leaves tested 8.

Experiment 4: Feeding behavior of CLas-infected D. citri nymphs and adults on excised citron leaves measured by electrical penetration graph (EPG) recordings

Diaphorina citri nymphs (4th instar) and adults (1–2 week old) used in the EPG study were selected from psyllid colonies with high CLas titers (low Ct values). Previous EPG recordings indicated no differences in EPG waveforms produced by adult males or females of D. citri11,34. Therefore, the sex of adults used in this study was not determined. EPG recordings were performed on D. citri nymphs or adults feeding on excised leaves from healthy (uninfected) citron plants grown from seeds. Young leaves (soft, fully expanded, ca. 5 cm long and 3 cm wide) were selected. Leaves with their petioles were excised from the plant; the leaf surface was cleaned and air dried, and the leaf petiole was introduced into a 4 mL glass vial filled with water.

EPG recordings were obtained using a DC-monitor, GIGA-8 model, EPG-Systems, Wageningen, and the Netherlands, adjusted to 50 × gain. The analog signal was digitized through a DI-710 board and displayed using Windaq Lite ver. 2.40 software (Dataq Instruments Inc. Akron, OH, USA) on a Dell desktop computer. The EPG monitoring system was housed in a grounded Faraday cage in an environmentally controlled room under continuous lighted conditions. Temperature was set to 26 °C with 60–65% RH. Psyllid nymphs or adults were collected 4 h prior to the start of the experiment each day and were starved for this period inside glass vials. Adult psyllids were then placed in a freezer (− 4 °C) for 45–60 s to immobilize them, then held by a plastic pipette tip connected to a gentle vacuum supply under a dissecting microscope. Each psyllid nymph or adult was attached to a 25 μm-diam. gold wire (Sigmund Cohn Corp., Mt. Vernon, New York) by a droplet of silver conducting paint (Ladd Research Industries, Burlington, VT) applied to the insect’s pronotum (Supplementary Fig. S2). The gold wire lead was attached to a copper electrode (3 cm × 1 mm diameter) connected to the EPG probe. To complete the electrical circuit, a reference copper electrode (10 cm × 2 mm) was inserted into the water in the vial containing the leaf petiole. Psyllid nymphs and adults were restricted to the abaxial (lower) surface of the leaf, the preferred feeding site for nymphs3,11. The feeding behavior of individual D. citri (51 nymphs and 52 adults) was monitored for a continuous period of 42 h. This is the same recording duration we have used in our previous CLas acquisition study on D. citri nymphs and adults11 and is much longer than most EPG studies for D. citri reported previously.

Following the 42 h recordings, all tested insects were kept individually in vials with 90% ethanol and stored in a refrigerator until DNA extraction and qPCR testing. Characterization of EPG waveforms was accomplished by visually identifying and annotating waveforms based on comparison with prior EPG and histological studies on D. citri11,25,34. Windows Dataq waveform browser (Dataq Instruments Inc., Akron, OH) was used to annotate waveforms. The number and duration of waveform bouts were tabulated in an electronic spreadsheet. The waveforms were visually inspected for frequency patterns and annotated as non-probing (Np), mesophyll intercellular pathway (C), phloem penetration (D), phloem salivation (E1), phloem ingestion (E2) or xylem ingestion (G) phases. Statistical analysis was performed using JMP (v. 10, SAS Inc., Cary, NC). Data analyzed by non-parametric test, Kruskal–Wallis test (n = 51 for nymphs, n = 52 for adults, α = 0.05) without differentiating insect sex (v. 10, SAS Inc., Cary, NC, USA).

Ethics

All experimental protocols were approved by the Shatters Lab., US Horticultural Research Laboratory, Fort Pierce, Florida, USA. All methods were carried out in accordance with relevant guidelines and regulations approved by Agric. Res. Service, USDA.