Effects of Wolbachia elimination and B-vitamin supplementation on bed bug development and reproduction

Obligate blood feeders, such as Cimex lectularius (common bed bug), have symbiotic associations with nutritional endosymbionts that produce B-vitamins. To quantify the symbiont’s contribution to host fitness in these obligate mutualisms, the symbiont must be eliminated and its absence rigorously confirmed. We developed and validated procedures for complete elimination of Wolbachia (Wb) in bed bugs and quantified development and reproduction in bed bugs with and without Wb and with and without B-vitamins supplementation. Aposymbiotic bed bugs had slower nymphal development, reduced adult survivorship, smaller adult size, fewer eggs per female, and lower hatch rate than bed bugs that harbored Wb. In aposymbiotic bed bugs that were fed B-vitamins-supplemented blood, nymph development time, adult survivorship and hatch rate recovered to control levels, but adult size and egg number only partially recovered. These results underscore the nutritional dependence of bed bugs on their Wb symbiont and suggest that Wb may provide additional nutritional benefits beyond the B-vitamin mix that we investigated.


Results
Wolbachia quantification. The experimental design is described in Fig. 1. The Wb titers were quantified by ddPCR assay in 125 bed bugs comprising third instars, and adult females during the preparation of an aposymbiotic colony and third instars, fifth instars, adult females, and adult males from four experimental treatments. Wolbachia was absent in 12 of 13 females in the antibiotic-fed aposymbiotic preparation colony group. Only one female had a detectable level of Wb but the titer was greatly reduced ( Fig. 2A). Only the offspring from the 12 Wbfemales were used in the next antibiotic-free generation. All 13 third instars and 20 females did not have detectable levels of Wb in the antibiotic-free Wbbed bugs (Fig. 2B).
Qualitative conventional PCR of the 16S rRNA and wsp gene targets of experimental bugs showed high bacteria and Wb in Wb + samples and low or no bacteria or Wb in Wbsamples. However, all samples were positive for the bed bug elf gene (Supplementary Methods and Fig. S1). In quantitative ddPCR, Wb was detected in all 40 Wb + samples and we found no significant difference in the Wb titer in Wb + bed bugs either provisioned with B-vitamins or not (Wb + Vit + vs. Wb + Vit -) (Fig. 2C). However, the Wb concentration was significantly different (p < 0.0001) among the four life stages (Fig. 2D). Wolbachia titers in the Wb + bed bugs were significantly higher in adults than in nymphs (p < 0.05) and fifth instars had nearly tenfold more Wb than third instars (p < 0.05). In contrast, Wb was detected in only one adult female of the 40 Wbbed bugs we tested (Fig. 2E,F). However, the Wb density in this female was 1000-fold lower than in Wb + adult females.
Nymph development. Percentage nymphs fed. The weekly percentage nymphs that fed was variable over time, but not consistently different among the four treatments over 16 weeks (F 3,9 = 0.2569, p = 0.8546). The percentage nymphs that fed decreased in all treatments between weeks 6 and 16 (Fig. 3A). More nymphs fed in the Wb + Vit + treatment than in the Wb -Vittreatment in weeks 1-6 (F 3,116 = 3.4544 p = 0.0188) (Fig. 3B).
Adult emergence and nymph survivorship. Mean adult emergence times were shortest for the two Wb + treatments (Wb + Vit -= 8.06 weeks, Wb + Vit + = 7.03 weeks) and 2 males emerged 1 week earlier than in all other treatments (Fig. 4A). Adult emergence was completed in the Wb + Vit + treatment by week 10, whereas in the Wb + Vitcontrol 3 (4.69%) additional adults emerged between weeks 11-16. In contrast, mean adult emergence times were longest for Wb -Vit -(9.84 weeks) and adults did not start to emerge until week 6, which was one week after all other treatments. In addition, the Wb -Vittreatment continued to emerge every week until the assay was terminated. Therefore, nymphal development was significantly longer for Wb -Vitbed bugs than in the other treatments (W = 46.6316, df = 3, p < 0.0001). Supplementation of blood with B-vitamins caused Wbnymphs to recover normal development with a mean emergence time of 8.19 weeks; both Wb + treatments, and the Wb -Vit + treatment had similar emergence distributions. In the Wb -Vit + treatment, four adults (7.02%) emerged between weeks 11-16, similar to the Wb + Vitcontrol.
By the end of the nymph development assay, on week 16, the total number of adults that emerged from each treatment was unaffected by the absence of Wb or the addition of vitamins (F 3,16 = 1.2070, p = 0.3391), although there was much greater variation among the five Wb -Vitreplicates than in the other three treatments (Fig. 4B). Nymph survival was also unaffected by the absence of Wb or B-vitamin supplementation of the blood meal (F 3,16 = 0.7896, p = 0.5172) (Fig. 4C).
Adult body size, fecundity, hatch rate, and survival. Adult body size. Thorax width in both sexes of the Wb -Vitbed bugs was significantly smaller than in the other three treatments (2-way ANOVA, F 7,163 = 102.954, p < 0.0001) (Fig. 5). Blood fortified with B-vitamins was not sufficient to fully recover the thorax width to that of Wb + bed bugs when Wb was absent, because both sexes of Wb -Vit + bed bugs were still significantly smaller than the respective Wb + bed bugs. In normal C. lectularius, adult females are larger than males, and thorax width of females in the Wb + Vitcontrol was 0.0943 mm (6.4%) larger than in males (t-test, t = 5.7639, www.nature.com/scientificreports/ p < 0.0001). Likewise, thorax width of females in the Wb + Vit + and Wb -Vit + treatment groups was 0.0936 mm (6.5%) and 0.0791 mm (5.7%) larger, respectively, than males, and these differences were statistically significant in both (Wb + Vit + , t = 6.2502, p < 0.0001; Wb -Vit + , t = 5.2895, p < 0.0001). In contrast, thorax width of Wb -Vitfemales was only 0.0107 mm (0.8%) wider than in males (t = 0.6622, p = 0.5114) (Fig. 5). These results indicate that the absence of Wb affects female development and ultimately adult female size more than it does in males.
Moreover, it appears that B-vitamin supplementation can partially, but not completely, restore female development and the size differential between females and males.
Percentage egg hatch. The hatch rate was consistently high in the Wb + Vitand Wb + Vit + treatments during the 45-day experiment ( Fig. 6C) with mean hatch percentages of 90.5 ± 1.4% and 89.5 ± 1.8%, respectively, which did not differ statistically (ANOVA, Tukey's HSD) (Fig. 6D). In contrast, egg hatch rate in Wb -Vitfemales declined by 88.5%, from 41.7% on day 5 to only 4.8% on day 45, while in normal control females it declined only 14.0% Figure 1. Experimental design. A laboratory colony (Wolbachia-present, Wb + ) was fed blood with no added vitamins (Vit -). The colony was split into two lines (1) Wb + and Wolbachia-absent (aposymbiotic, Wb -). The Wb + Vitline was fed antibiotic-fortified but not vitamin-supplemented blood for 15 months. The Wb -Vit + line was generated by concurrently feeding bugs on blood fortified with rifampicin and the K&M mix of B-vitamins for multiple generations for 15 months (2). To clear the antibiotic from the colony, Wb -Vit + bed bugs were reared for 6 weeks (one generation) on rifampicin-free blood supplemented with the B-vitamin mix (3). When adults emerged in the Wb -Vit + treatment group, they were placed in oviposition vials and first instars were collected and split into two treatment groups (4a) Wb -Vitand Wb -Vit + . The Wb + Vitcontrol line was concurrently fed non-supplemented blood, and when adults emerged in the Wb + Vitgroup, they were placed in oviposition vials and first instars were collected and split into control (Wb + Vit -) and vitamin-supplemented (Wb + Vit + ) lines (4b). Nymphs in the four treatment groups were fed weekly and their development recorded for 16 weeks. As they emerged, adult females and males within the same treatment group were paired (5), fed their respective treatment blood every five days for 45 days, and eggs counted every five days. Adults were sampled twice and nymphs were sampled once for ddPCR during the course of generating the Wbline, as indicated. During nymph development of the experimental groups, third instars, fifth instars, and newly emerged adults were sampled from each of five replications for Wb quantification by ddPCR.  6C). Egg hatch rate over the 45-day interval in the Wb -Vittreatment averaged 30.7 ± 3.6%, significantly lower than in the two Wb + treatments (ANOVA, Tukey's HSD, F 3,82 = 55.5504, p < 0.0001). However, the percent hatch per female in Wbfemales was fully recovered to control levels (Wb -Vit + = 86.6 ± 3.1%) by adding B-vitamins to their blood meals (Fig. 6D). Comparison of two B-vitamin mixes. Because we used a different B-vitamin mix (K&M) than previous researchers (L&F) [16][17][18] , we directly compared the two mixes in adult females. We did not include a Wb + Vittreatment because only the vitamin mixes were being compared. Figure 8A shows that Wb + females provided either of the two vitamin mixes produced similar numbers of eggs. Likewise, Wbfemales provided either of the two vitamin mixes produced similar numbers of eggs. However, the Wbfemales supplemented with either of the two vitamin mixes produced significantly fewer eggs than similarly treated Wb + females (ANOVA, Tukey's HSD, F 3,73 = 9.9333, p < 0.0001). Egg hatch was high (range: 91.6 to 96.3%) across all four treatments (ANOVA, Tukey's HSD, F 3,73 = 1.6778, p = 0.1792) (Fig. 8B), consistent with our previous results (Fig. 6D).

Discussion
Our results highlight the importance of Wolbachia as a nutritional mutualist in bed bugs. The elimination of Wb significantly increased nymph development time and reduced adult size and survivorship, eggs laid per female, and hatch rate. When aposymbiotic bed bugs were fed blood supplemented with B-vitamins, nymph development time, egg hatch rate and adult survival were fully recovered to the levels of control bed bugs, but adult size and Figure 2. Wolbachia quantification of adults and third and fifth instar bed bugs by ddPCR. Twelve out of 13 antibiotic-fed females were Wb-free and one female had a significantly lower Wb titer (A). All 20 females and 13 third instar nymphs in the antibiotic-free generation were Wb-free (B). All 20 Wb + adults, and all 20 Wb + third and fifth instars had Wb (C). Adults had significantly higher wsp gene copies than nymphs, and Wb titers were significantly higher in fifth than in third instars (D). Wolbachia was not detected (< 0.5 wsp copies per µl) in all 20 Wbthird and fifth instars, all 10 Wbadult males, and 9 out of 10 Wbadult females (E, F). www.nature.com/scientificreports/ per capita egg production only partially recovered. This is the first study in bed bugs that fully eliminated Wb before and throughout experimentation, and that mitigated the side-effects of antibiotics by discontinuing the use of antibiotics for a generation before the start of experiments. It is also the first study that used a complete design that included normal Wb-present (Wb + ) control bed bugs in all fitness-related assays. Our results support and extend the findings of previous studies that used a range of Wb-free (Wb -) and Wb-deficient bed bugs. Our results also add life history traits that are affected by Wb to previously documented traits.

What motivated this work?
Pioneering studies of the Wb-bed bug symbiosis confirmed that Wb is housed in a gonad-associated bacteriome and is vertically transmitted, and demonstrated that Wb produces biotin and riboflavin that benefit the bed bug host [16][17][18] . Hosokawa et al. 16 examined several fitness-related traits, including the rates of egg and nymph development, and the percentage of nymphs that emerged into adults. Their results showed that Wb was an essential nutritional symbiont for normal growth and reproduction of C. lectularius. Four main reasons motivated us to re-examine these traits and consider additional traits. First, in the previous studies, "aposymbiotic" bed bugs obtained with rifampicin treatments were shown to be Wb-deficient, with titers reduced by 100 to 10,000 wsp gene copies; only 8 of 30 (26.7%) of the "aposymbiotic" bed bugs were shown by qPCR to be free of Wb 16 . Second, in previous work, rifampicin was used throughout the experiments in two treatment arms-"aposymbiotic" bed bugs fed blood and "aposymbiotic" bugs fed blood supplemented with B-vitamins 16 . Adverse effects of the antibiotic might therefore affect life-history traits throughout the experimental period. Third, in some experiments, control bed bugs harboring Wb were not used 17,18 , so it was difficult to determine if B-vitamins fully or partially recovered the traits that were examined. Finally, we felt that some incongruent results for numbers of eggs laid and adult emergence in Hosokawa et al. 16 and Moriyama et al. 18 required closer examination. For example, the number of eggs laid was unaffected by antibiotic treatment in the early study 16 , but the same treatment resulted in complete elimination of egg production in the later study 18 . Likewise, adult emergence was greatly reduced in the 2010 and 2014 reports 16,17 , but the 2015 study 18 showed almost no adult emergence in the Wb-deficient colonies. It is important to emphasize, however, that although our results differ quantitatively from these previous studies, they broadly support the general conclusions of these groundbreaking studies.
To address these four issues, we developed procedures for absolute quantification of Wb titers in bed bugs and used these procedures to confirm that our rifampicin-treated bed bugs were indeed aposymbyotic. Compared to qPCR, droplet digital PCR (ddPCR) has higher sensitivity, specificity, and reproducibility; it can quantify low target numbers 20 (our empirically derived lower detection level was 0.5 gene copies per µl) and it provides the absolute quantification of DNA in a sample 21,22 , making it an appropriate technology for confirming the absence of Wb in bed bugs. Using ddPCR, we showed that 100% of the third instars, fifth instars, and adult males, and There were no significant differences among the four treatment groups in percentage nymphs that fed in weekly feeding sessions (Repeated measures ANOVA, F 3,9 = 0.2569, p = 0.8546) (A), but there was a significant difference among treatments in the percentage that fed each week during weeks 1 to 6 (One-way ANOVA, F 3,116 = 3.4544, p = 0.0188) (B). Week 6 represents when adults began to emerge in all four treatment groups (Fig. 4A) and the feeding rate is based on the nymphs remaining in the vial. The mean is shown as a white circle and white horizontal line and sample size is indicated in parentheses. Percent data were arcsine square root transformed prior to analysis. Groups that do not share letters are significantly different from each other (Tukey's honestly significant difference test, p < 0.05). www.nature.com/scientificreports/ 90% (9 of 10) of the adult females tested in our experiments were Wb-free; the one adult female that contained Wb was highly Wb-deficient with 1000-fold fewer Wb wsp gene copies than normal Wb + females. These procedures ensured that the changes in life history traits attributed to the presence or absence of Wb and B-vitamin supplementation were not confounded by incomplete elimination of the endosymbiont. Persistent antibiotic treatments have many unintended side effects. Yet, many studies on insect-microbe symbiosis administer large doses of antibiotics during the implementation of experiments. If available, chemicals with similar side effects but no antibiotic properties could be used in control treatments. However, this approach . There were no significant differences among the four treatments in the average cumulative percentage of adults that emerged in each treatment after 16 weeks (One-way ANOVA, F 3,16 = 1.2070, p = 0.3391) (B). There were also no significant differences among the four treatments in the average percentage of nymphs that survived the 16-week experiment (One-way ANOVA, F 3,16 = 0.7896, p = 0.5172) (C). Adults that emerged were included as having survived development. For each box plot, the mean is shown as a white circle and a white horizontal line and sample size is indicated in parentheses. www.nature.com/scientificreports/ is complicated and we are not aware of its use in research on insect symbiosis. Instead, we treated a colony of bed bugs with rifampicin for multiple generations (experimental design in Fig. 1), confirmed that they were aposymbiotic by ddPCR, then reared one generation without rifampicin to clear the antibiotic from the colony, and again confirmed by ddPCR that they remained aposymbiotic throughout the experiments. We also included Wb + control groups in all fitness tests to determine if vitamins fully or partially recovered the effect of Wb elimination. Table 1 summarizes the biological traits tested in our study, whether the elimination of Wb affected the trait and if blood meals fortified with B-vitamins could recover the trait to the levels observed in control bed bugs (Wb + Vit -).

Specific effects of aposymbiosis on bed bugs and B-vitamin supplementation.
Our results showed that the elimination of Wb did not affect feeding rate in bed bugs. It also did not affect survival of nymphs and the cumulative emergence of adults. The latter results were notably different from previous findings [16][17][18] . Consistent with earlier findings 16 , however, development of aposymbiotic nymphs was slower and their mean time to adult emergence was 1.8 weeks (22.2%) longer than control (Wb + Vit -) nymphs. Supplementation of the blood meal with B-vitamins restored the rate of development of aposymbiotic nymphs and their mean time to adult emergence was only 0.14 weeks (1.7%) slower than the control nymphs.
Adult size was significantly lower in aposymbiotic females and males than in normal Wb + bed bugs. Interestingly, whereas normal females are significantly larger than normal males, aposymbiotic adults converged on the same smaller size, indicating that the absence of Wb affected female size more than male size. Smaller adult size was also observed in aposymbiotic stinkbugs, Megacopta punctatissma and Megacopta cribraria 23 , but the underlying mechanisms remain unknown. In our study, B-vitamin supplementation recovered the sexual dimorphism in adult size, but body size only partially recovered relative to normal adults. These findings suggest that the recovery may require multiple generations or that Wb provisions the bed bug with other factors, including other nutrients, in addition to B-vitamins.
Our results demonstrated dramatic effects of Wb on reproduction. Aposymbiotic females oviposited 85.8% fewer eggs than control females (Wb + Vit -), despite being fed every 5 days and remated once. Supplementation of the blood meal with B-vitamins significantly increased per capita oviposition by 5.5-fold, but it was only partially recovered to the oviposition rate of normal females. Of interest, however, was the observation that in B-vitamins-fed aposymbiotic females (Wb -Vit + ) oviposition increased over time, reaching within 10.7% (2.15 eggs) of the oviposition rate of normal control females on days 41-45, but control females changed minimally. These observations suggest that the recovery of egg production with B-vitamin supplementation is a gradual process, and it may require more than 45 days for their full recovery.
In contrast, egg hatch rate in aposymbiotic females declined by 88.5% between days 5 and 45, whereas in normal control females it declined only 14.0% during the same period. B-vitamins fully recovered the effect of Wb elimination as early as day 5 (91.6% hatch), and the hatch rate of Wb -Vit + eggs declined only 18% between days 5 and 45, similar to control females. These results are generally consistent with previous findings of low egg development in aposymbiotic C. lectularius 16 . Although vitamin supplementation (Wb -Vit + ) significantly increased thorax width in both sexes compared with Wb -Vitbed bugs, the vitamin-fortified blood was not sufficient to recover thorax width to the level of control bed bugs (Wb + Vit -). The mean is shown as a white circle and a white horizontal line and sample size is indicated in parentheses. Groups that do not share letters are significantly different from each other (Tukey's honestly significant difference test, p < 0.05). www.nature.com/scientificreports/ Effects of adding B-vitamins to Wb + bed bugs. Bed bugs require a blood meal before every molt 9 and have 5 instars. Therefore, we did not expect adults until week 6, and indeed 37.5% of the adult emergence in the Wb + Vitcontrol treatment occurred on week 6. The addition of B-vitamins to normal (Wb + ) bed bugs sped up adult emergence, and the mean adult emergence of vitamin-supplemented Wb + Vit + nymphs was a week earlier than that of normal bed bugs fed non-supplemented blood; two males (of 67 total, 3%) emerged after only 5 weeks. B-vitamin supplementation of normal bed bugs has not been reported previously, so this is the first report of faster nymph development on B-vitamin supplemented blood. B-vitamin supplementation of normal nymphs did not affect their cumulative emergence to adults, nymph survival during the 4-month experiment, or adult size. However, it is interesting that the addition of B-vitamins to the blood meals of normal adult females reduced their per capita egg production and female survivorship over 45 days, but not the egg hatch rate. This was unexpected because B-vitamins are water soluble, animals excrete high levels of excess vitamins, and they store only moderate levels of B-vitamins 8 . It is possible that bed     www.nature.com/scientificreports/ processing of lipids, carbohydrates, and amino acids, is only required in minute doses for insects 24 and dietary overdose of biotin has been shown to be an effective chemosterilant in the house fly, Musca domestica 25 and it reduced fertility in the hide beetle, Dermestes maculatus, by suppressing embryogenesis 26 . Because Wb is known to produce biotin in bed bugs 17 it is possible that the low concentration of biotin (10 ng/mL blood) fed to Wb + bed bugs similarly affected female fecundity. The mechanisms underlying the differential effects of B-vitamin supplementation on development and reproduction are unknown and should be investigated.

Differences among studies and limitations of this study.
In addition to differences in the Wb titers of aposymbiotic bed bugs in this and previous studies, other differences are notable. The bed bug strain used in previous studies [16][17][18] to test the biological effects of Wb was the Japan Environmental Sanitation Center (JESC) from Kanagawa, Japan, which lacks a facultative γ-proteobacterial symbiont [16][17][18] . This symbiont is present in the Harold Harlan strain used in the current study, but we did not test whether it was eliminated by the rifampicin treatment. It is plausible that the γ-proteobacterial symbiont is also a nutritional mutualist and it too was eliminated by our extensive rifampicin treatments. This might explain the only partial recovery of several traits with B-vitamin supplementation. It is also important to note that in all studies using antibiotic treatments, the gut bacterial communities were likely dramatically affected or eliminated. Thus, any life history traits that could not be fully recovered in Wb-free bed bugs with B-vitamin supplementation might have been affected by a disrupted gut microbiome. Moreover, as mentioned before, Nikoh et al. 17 and Moriyama et al. 18 reared one generation of rifampicintreated bed bugs without blood meal supplementation with B-vitamins, to eliminate transgenerational carryover of B-vitamins. Insects do not have a large capacity to store B-vitamins 8 , but it is conceivable that this experimental design depleted certain B-vitamins in the developing nymphs. It is unclear if this difference in design contributed to the differences between our studies.
Finally, we used the Kao and Michayluk B-vitamin mix, as did Fisher et al. 27,28 , which differs substantially from previous studies with aposymbiotic bed bugs [16][17][18] . Nevertheless, direct comparisons of these vitamin mixes showed that both had no effect on egg production in Wb-present (Wb + ) bed bugs and both similarly, but only partially, restored egg production in Wb-free (Wb -) females. Both vitamin mixes equally restored hatch rate of Wbfemales to the high levels of Wb + females.

Outstanding questions on the role of Wolbachia in bed bugs.
There are numerous questions to be addressed about the nutritional fitness-related benefits of Wb in bed bugs. B-vitamins are co-enzymes in various metabolic reactions 8 , so it is not surprising that Wb would contribute to nymph growth and development, egg production and hatch rate. Our study shows the importance of B-vitamins in bed bug development and reproduction, but adult size and egg production results suggest that either our B-vitamin mix was not optimal or that Wb, or other symbionts that were eliminated by rifampicin, contribute nutrients to its C. lectularius host. Our observations also suggest that B-vitamins aid in the digestion of blood meals. We observed that aposymbiotic bed bugs that were not supplemented with B-vitamins retained blood in their gut for a longer duration than in all other treatments. B-vitamin supplementation in aposymbiotic bed bugs partially recovered faster blood meal digestion. In tsetse flies, the endosymbiont Wigglesworthia glossinidia was shown to produce vitamin B-6 (pyridoxine) and elimination of the symbiont showed that Wigglesworthia influences immunity, reproduction and digestion 29 . A closer look at how tsetse flies utilized vitamin B-6 showed that it was critical for proline homeostasis, which tsetse flies rely on to produce ATP 30,31 . These metabolic downstream effects of vitamin provisioning could be present in bed bugs, requiring further investigation.
Furthermore, it remains to be determined whether dramatic effects on fecundity and hatch rate due to Wb elimination, are due exclusively to an effect on females, or whether reproduction is also affected in aposymbiotic males. It has been shown that the blood source fed to males can affect sperm competition and female fecundity 32 but the effect of Wb removal on male reproductive fitness has not been investigated. Both sexes house Wb in gonad-associated bacteriomes 16 , so the elimination of Wb might uniquely affect reproductive traits in females Table 1. Summary of the effects of Wolbachia (Wb) on life-history traits of C. lectularius and whether the addition of B-vitamins to blood meals partially or fully restored the trait to the level of control (Wolbachiapresent, Wb + ) insects. Bold italics indicates no effect of Wb removal on the fitness-related traits, italics indicates traits that were partially recovered by the addition of B-vitamins, and bold indicates fitness-related traits that were fully recovered to control levels with the addition of B-vitamins. www.nature.com/scientificreports/ and males. To date, all experimental designs in bed bugs have allowed males and females to mate within each treatment (e.g., either Wb + or Wb -). A cross-treatment design, involving reciprocal crosses of Wb + × Wb -, should reveal sex-specific effects of Wb.
Finally, Wb has been investigated extensively for its ability to interfere with host reproduction through cytoplasmic incompatibility (CI) 13 which is typically detected as a reduction in egg hatch rate 33 . CI expression in C. lectularius thus would result in unfertilized eggs, which would be resorbed, resulting in lower oviposition with minimal effects on egg hatch. It would be fascinating to discover whether Wb-free bed bugs experience fitnessrelated deficiencies in part due to CI.
In summary, we found ddPCR much more reliable than qPCR for detecting low copy numbers and confirming the absence of Wb in aposymbiotic bed bugs. We therefore urge a careful re-examination of other nutritional symbioses with ddPCR to confirm that the symbiont was fully eliminated. The results presented in this study not only highlight the nutritional importance of Wb in bed bugs but also provide a useful experimental design to examine other endosymbiont-host relationships and methods for confirming the elimination of a symbiont. It is possible that Wb is responsible for providing additional nutritional benefits that contribute to adult size and fecundity in bed bugs. The mechanisms of transfer of B-vitamins from Wb to its host and how B-vitamins influence development, fecundity and survival are still unknown and would be a valuable area for future investigations. In addition, elucidating male vs. female effects on fitness traits would be useful to understand how Wb influences bed bug success.

Materials and methods
Insects. We used the Harold Harlan strain of C. lectularius. This strain was collected in 1973 in Ft. Dix, NJ, USA, and reared on a human host. Since 2008 bed bugs were fed defibrinated rabbit blood at North Carolina State University using an artificial feeding apparatus. The stock colony was used for the Wb-present (Wb + ) treatments, and a second colony was maintained on antibiotic-and vitamin-supplemented blood to eliminate Wb, generating a Wb-absent (Wb -) lineage of C. lectularius. All colony bed bugs, experimental nymphs, adults and eggs were maintained in the same incubator at 27 °C, a 12:12 (L:D) photoperiod, and 45 ± 5% RH, and fed using the same methods. For all treatments, and colony maintenance, bed bugs were fed defibrinated rabbit blood (Hemostat Laboratories, Dixon, CA) through a custom-made glass feeder (Prism Research Glass, Raleigh, NC), water-jacketed and warmed to 35℃ using a thermal circulator water bath (B. Braun Biotech, Inc., Allentown, PA), as previously described 34 .
Generating a Wolbachia-free (Wb -) colony. Two hundred adult virgin females and 100 virgin males 7-10 days after emergence were taken from the stock colony, and placed in a round plastic container (41 mm diameter × 67 mm) with folded filter paper and a 0.3 mm-mesh screened lid through which bed bugs could feed (oviposition jar). Adults were fed for 1 h. Adults remained in the container post-feeding to mate and lay eggs for five days and then all adults were removed and the container with eggs was placed in an incubator for 10 days. Cohorts of fifty 7-10 day-old first instars were transferred to three round plastic containers with filter paper and 0.3 mm-mesh screened lid for feeding (colony jar). Blood was fortified with 10 µl (= 10 µg) of a rifampicin-PBS solution (1 mg rifampicin/1 ml PBS) and 10 µl of an undiluted Kao and Michayluk vitamin solution (K3129, 100-fold concentration of vitamins, Millipore Sigma) per ml blood; generally, 2 ml of blood were provided to each treatment in each feeding session. The fortified blood was fed weekly to nymphs throughout development to rear the Wbcolony. The colony was maintained for 15 months (May 2019 to Sept 2020) on weekly rifampicinand vitamin-supplemented blood (#2 in Fig. 1). To mitigate any side effects of antibiotic treatment, a subset of adults was removed from the rifampicin treatment, placed in a new colony jar and reared for one generation before we initiated the experiments. These adults were fed vitamin-supplemented blood without rifampicin to produce a new "antibiotic-free Wbgeneration" (#3 in Fig. 1), and were the parental stock for the experimental bed bugs. Virgin females (100) and virgin males (50) 7-10 days after adult emergence, from the antibiotic-free Wbcolony, were placed in a fresh oviposition jar and fed vitamin-supplemented blood for 1 h. Adults remained in the jar to mate and oviposit for 5 days and then removed. The first instars that hatched were used for the Wbtreatments (#4 in Fig. 1). For Wb + treatments, a second oviposition jar with the stock colony of the Harold Harlan strain was reared in parallel using the same methods, but the bed bugs were fed blood without vitaminsupplementation, to generate Wb + first instars for the study (#5 in Fig. 1). The stock colony was never treated with antibiotics and never received vitamin supplementation.

DNA extraction.
To confirm the Wb status in bed bugs of the Wb-free colony, adult females (n = 13) from the rifampicin treatment (Wb -) and nymphs (n = 13) and adults (n = 20) from antibiotic-free Wbgeneration were assayed by ddPCR for the presence of Wb. We sampled adults from the Wbcolony before the rifampicin treatment was discontinued (#2 in Fig. 1) and analyzed for the presence of Wb. We also sampled the nymphs and adults that emerged from the first rifampicin-free generation (Wb -Vit + ; #3 in Fig. 1) and tested for Wb prior to experimentation (ddPCR in Fig. 1). Then, during the experimentation period, and throughout nymph development we sampled one third instar (n = 5), one fifth instar (n = 5), one adult male (n = 5) and one adult female (n = 5) from each replicate vial in all four treatment groups for Wb quantification to confirm that experimental bed bugs remained Wb-free.
Bed bug samples were individually surface sterilized with 10% bleach and 70% ethanol, followed by three washes with sterile water. The rationale was to eliminate potential DNA contamination from contact with feces, exuviae, and dead bugs. The samples were homogenized for 60 s using the FastPrep-24 tissue homogenizer (speed: 4 m/s; MP Biomedicals, Solon, OH) followed by DNA extraction using the DNeasy Blood and Tissue kit (Qiagen, Germantown, MD). Briefly, the homogenized samples were incubated at 56 °C in 180 µl of ATL buffer www.nature.com/scientificreports/ and 20 µl of proteinase K mixture for 4 h. Samples were briefly vortexed, 200 µl of AL buffer was added, and the samples were incubated for 10 min at 56 °C. The samples were then mixed with 200 µl of absolute ethanol and centrifuged through spin columns. The spin columns were washed with AW1 and AW2, following the manufacturer's protocol, and DNA was eluted in sterile nuclease-free water. Following the elution, the samples were further purified, concentrated by ethanol precipitation, reconstituted in 100 µl sterile nuclease-free water, and stored at -20 °C until further use. All the samples were amplified for Wb and total bacteria targeting wsp and 16S rRNA genes respectively, and a subset of samples were amplified for bed bug elongation factor ef1α gene using conventional PCR and agarose gel electrophoresis (Supplementary methods). Treatment groups: nymph development and adult emergence. Figure 1 shows the experimental design. Twenty 3-5 day-old first instars from the antibiotic-free Wbcolony, were transferred to each of five 20 ml vials for the Wbwithout vitamin-supplementation treatment group (Wb -Vit -) and five additional vials for the Wbwith vitamin-supplementation treatment group (Wb -Vit + ) (#4a in Fig. 1). Twenty 3-5 day-old first instars from the untreated stock colony were transferred to each of five 20 ml vials for the Wb + Vittreatment group (Wb-present, no vitamins added) and five additional vials for the Wb + Vit + treatment group (Wb-present, vitamins added) (#4b in Fig. 1). Thus, each of the four treatments started with 100 first instar bed bugs and all bed bugs were fed weekly. The Wb + Vit + and Wb -Vit + treatments were fed vitamin-supplemented blood with undiluted Kao and Michayluk vitamin solution (10 µl/ml blood), and the Wb + Vitand Wb -Vittreatments were fed blood without additional vitamins. Vials were placed individually on an artificial feeder for 20 min and the numbers of fed, unfed and dead nymphs were recorded after each feeding for 16 weeks. Newly emerged females and males were removed and paired in the adult fecundity assays, below. The percentage nymphs that fed, cumulative adults that emerged and cumulative dead nymphs were calculated for each vial (replicate) of 20 nymphs (n = 5) per treatment.

Quantification of
Adult fecundity, female survivorship and body size. When adults began to emerge in the nymph development assays, they were removed and paired with an adult of the opposite sex from the same treatment group (#5 in Fig. 1). All the females that emerged from each treatment group were used, but the number of females in each treatment varied. Each pair was placed in a 7.5 ml vial with a strip of filter paper for egg laying and a screened cap for feeding. The filter paper contacted the screen to ensure access to the blood. Paired males and females were fed for 20 min the same blood treatment as their respective nymphal treatment group. If either the male or female was not fully engorged, the vial was placed back on the feeder until both adults were fully engorged. After feeding, vials were placed in the incubator and after 5 days, the male was removed and placed in a 20 ml vial with other males from the same treatment group, and fed weekly. The female was moved to a new 7.5 ml vial and fed individually every 5 days for 45 days to assess egg laying of each female. Females were remated individually on day 20. Vials containing eggs were stored in an incubator for 10 days, and then moved to a -20 °C freezer to preserve eggs and hatched first instars for counting. Eggs and first instars were counted using a dissecting microscope (MZ6, Leica microscopes, Feasterville, PA). The number of eggs and egg hatch per female during 45 days was calculated for each treatment group (Wb + Vit -: n = 24, Wb + Vit + : n = 24, Wb -Vit -: n = 23, Wb -Vit + : n = 15). All adults were placed in a -20 °C freezer after 45 days and the size of the thorax was measured using a digital microscope (Andonstar AD409, Shenzhen, China). If a female died before 45 days, the date was recorded and used to assess survivorship and the female was placed in a -20 °C freezer for measurements.
Comparison of vitamin mixes. Newly emerged adults from the Wb + and Wbcolonies were paired and fed blood supplemented with either the Lake and Friend (L&F) total vitamin mix 35 , which has been used in similar studies 16 www.nature.com/scientificreports/ Twenty Wb + and 20 Wbpairs were placed pairwise in 7.5 ml vials with filter paper for egg laying and screened caps for feeding. The pairs were fed and used only if both sexes were fully engorged after the first feeding. The male was removed after 5 days, and the female was moved to a new vial and fed every 5 days for 15 days. The eggs from each vial were placed in an incubator for 10 days to hatch, then placed in a -20 °C freezer. Egg number and percent hatch were recorded for each pair. We conducted this comparison for only 15 days to avoid the subsequent decline in oviposition in Wbfemales. The results from bed bugs fed the L&F vitamin mix were compared to similarly treated adult pairs that were fed the K&M vitamin solution during the adult fecundity assay.
Data analysis. All analyses were conducted using JMP version 13 (SAS Institute, Cary, NC). ANOVA was used to compare vitamin mixes and male and female thorax width. If the ANOVA results were significant, Tukey's HSD test was used to detect differences among treatments. Prior to using an ANOVA, the data were tested for normal distribution using the Shapiro-Wilk W test and for homogeneity using the O'Brien test, and percent data were arcsine square root prior to analysis. A product-limit (Kaplan-Meier) survival analysis was conducted to examine the distribution of adult emergence and adult survival. The Wilcoxon Chi-square statistic test was used to determine differences among treatments. Repeated measures ANOVA was used to assess the percentage nymphs that fed, cumulative adult emergence, cumulative mortality, total eggs per female and percent egg hatch per female. The F-test was used to detect differences among treatments, and the Wilk's Lambda multivariate test was used to assess differences over time when Sphericity Chi-squared was not significant. If the Sphericity test was significant, then the univariate test G-G Epsilon was used to determine whether there was a significant difference over time. We present summary data as box plots, showing all the replicates, as well as the mean value (white circle with a white horizontal line). The box represents the interquartile interval (50% of the replicates) and the black horizontal line within the box is the median. Whiskers indicate the 10th and 90th percentiles.