
==== Front
101084138
22395
Infect Genet Evol
Infect Genet Evol
Infection, genetics and evolution : journal of molecular epidemiology and evolutionary genetics in infectious diseases
1567-1348
1567-7257

38901621
10.1016/j.meegid.2024.105624
nihpa2006063
Article
Differences in Salmonella Typhimurium infection and excretion among laboratory and field strains of the German cockroach suggest a genomic basis for vector competence
Ismael Bashar a
Wilson Morgan b
Miller Dini b
Pietri Jose E. a*
a University of South Dakota, Sanford School of Medicine, Division of Basic Biomedical Sciences, Vermillion, SD, USA
b Virginia Tech University, Department of Entomology, Blacksburg, VA, USA
* Corresponding author. Jose.Pietri@usd.edu (J.E. Pietri).
9 7 2024
9 2024
18 6 2024
01 9 2024
123 105624105624
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/).
The German cockroach, Blattella germanica, can be a vector of human enteric bacterial pathogens, including Salmonella enterica serovar Typhimurium (S. Typhimurium). Transmission of such pathogens by cockroaches has largely been considered a passive mechanical process, but recent studies have argued against this dogma by demonstrating bacterial proliferation within the cockroach gut and the necessity of specific bacterial genes for successful transmission in the feces, revealing unappreciated biological complexity in the vector-pathogen relationship between cockroaches and S. Typhimurium. However, the influence of naturally occurring variation among cockroach populations on pathogen infection and dissemination has not been investigated. Thus, this study aimed to examine whether distinct strains of B. germanica exhibit differences in their ability to become infected by and disseminate S. Typhimurium. We performed controlled infections of one long-term laboratory strain and three recently field-collected strains reared under identical conditions, then compared bacterial loads in the body and excreta of individual insects. Separately, we also compared rates of necrophagy, a behavior known to contribute to the horizontal spread of S. Typhimurium among cockroaches. Our data show significant differences in infection susceptibility, pathogen shedding in the excreta, and necrophagy between laboratory and field strains as well as between some field strains. These observations represent the first evidence that genomic variation among cockroach populations may influence their ability to become infected by and disseminate pathogens, providing further support for the hypothesis that German cockroaches are active biological vectors rather than passive mechanical vectors of S. Typhimurium. Additional studies are needed to identify the genomic drivers of vector competence for S. Typhimurium in B. germanica.

German cockroach
Blattella
Salmonella
Infection
Transmission
Necrophagy
==== Body
pmc1. Introduction

German cockroaches (Blattella germanica) have been suspected of spreading human enteropathogenic bacteria for over a century (Toda, 1923). Numerous studies have documented the presence of such bacteria (e.g., Salmonella spp. Shigella spp., Listeria spp. Escherichia coli) in association with field-collected cockroaches (Nasirian, 2019), sometimes coinciding with human outbreaks (Graffar and Mertens, 1950; Burgess and Chetwyn, 1981). Others have shown a reduction in human enteric infections concomitant with local implementation of cockroach control measures (Graffar and Mertens, 1950; Tarshis, 1962), further supporting a connection between the insects and pathogen transmission. Although the role of cockroaches in spreading human pathogens is widely accepted, the route(s) and mechanism(s) of transmission are debated and unresolved.

Transmission of bacterial pathogens by cockroaches has been widely considered a mechanical process. That is, the prevailing model has long been that cockroaches become transiently contaminated by bacteria and passively transfer these on their cuticle or in excreta without becoming actively colonized or amplifying the bacteria. This model, however, has been opposed by some investigators in the past (Stek, 1982). More recently, our own molecular studies of the interaction between German cockroaches and the human pathogen Salmonella enterica serovar Typhimurium (S. Typhimurium) have provided new evidence supporting the role of cockroaches as active biological vectors rather than passive mechanical vectors. For example, S. Typhimurium undergoes multiple phases of replication in the cockroach gut (Turner et al., 2022). Moreover, successful passage of S. Typhimurium through the cockroach gut for transmission in the feces requires intact bacterial type III secretion systems (Turner et al., 2022). In addition, S. Typhimurium induces an innate immune response in the form of antimicrobial peptide expression in the cockroach gut (Turner and Pietri, 2022) and interacts with the gut microbiota (Turner et al., 2023).

Under the model of mechanical transmission, negligible differences in the ability of cockroaches from distinct populations to maintain and transmit S. Typhimurium would be expected due to a lack of biological interaction between the pathogen and vector. In contrast, active biological vectors often exhibit interpopulation differences in vector competence, which has been known for decades to have a genomic basis in some species (Beerntsen et al., 2000). For instance, geographic differences in vector competence for West Nile virus have been observed in Culex pipiens mosquitoes from California (Vaidyanathan & Scott 2007). Similarly, Ornithodoros turicata ticks from Texas and Florida differ in their vector competence for the tick-borne relapsing fever agent Borrelia turicatae (Krishnavajhala et al., 2018).

Although our recent studies have provided new mechanistic insight into the interaction between S. Typhimurium and B. germanica, these studies have been conducted exclusively in an inbred laboratory cockroach strain. Thus, the influence of naturally occurring variation among cockroach populations on infection and transmission has remained unknown. Here, we sought to investigate whether recently field-collected strains of B. germanica differ in their ability to become infected by and disseminate S. Typhimurium relative to each other and to an inbreed laboratory strain. Relatedly, we also sought to examine whether these cockroach strains exhibit differences in necrophagy (Gahlhoff et al., 1999; Buczkowski et al., 2008), a behavior known to facilitate horizontal transmission of S. Typhimurium between individual cockroaches (Mond and Pietri, 2023).

2. Materials and methods

2.1. Cockroaches

This study included four different strains of Blattella germanica. The American Cyanamid Orlando normal strain, which has been maintained in laboratory culture for over 40 years, was included for comparison to other strains. Three additional, recently field-collected strains (Douglas, Gilpin, Hopewell) were also examined. Each strain represents a population started from a collection from multiple apartments units within a geographically distinct site (complex). Each site consisted of duplex, four-plex, or eight-plex buildings that included one and two-story, one-to-three-bedroom apartment units. The Douglas strain was collected from a site in Baltimore, MD in fall 2022. The Gilpin strain was collected from a site in Richmond, VA in February 2023. Finally, the Hopewell strain was collected from a site in Hopewell, VA in February 2023. Insects from each of the field-collected strains were shipped to the University of South Dakota insectary facility in February 2023 where ~15 individuals from each were used to establish colonies alongside the Orlando strain. Colony rearing took place in plastic enclosures containing cardboard harborages, dog chow (Purina, St. Louis, MO, USA), and tap water, at 25 ± 1 °C and ~ 40–45% relative humidity with a 12:12 (L:D) hour photoperiod. The experiments described herein were conducted between September 2023 and March 2024, following a > 6-month adaptation period and multiple generations of laboratory reproduction for each field strain. Experiments were carried out using only male cockroaches to minimize physiological variation and preserve females and nymphs for colony health.

2.2. Salmonella typhimurium infections

Groups of adult male cockroaches from each strain were orally infected with a naturally relevant concentration of S. Typhimurium strain 14028s encoding kanamycin resistance and GFP markers, as described in our previous work (Turner et al., 2022; Turner and Pietri, 2022; Turner et al., 2023). To summarize, groups of cockroaches were starved of food and water for three days, which promotes consistent feeding on bacterial culture media in the laboratory. Following the starvation period, a shallow Petri dish containing a stationary-phase culture of S. Typhimurium grown overnight in LB medium at 37 °C and standardized to a concentration of OD600 = 1 was provided as a sole food source for approximately 30 min. Blue #1 food dye was added to the culture to enable visual tracking of fed cockroaches, and any cockroaches that did not feed on the culture were subsequently discarded. Infected cockroaches were provided dog chow and water and maintained under the conditions described for colony rearing until further analyses were conducted.

2.3. Determination of internal S. Typhimurium loads

The internal S. Typhimurium load of individual infected cockroaches was determined by selective culture 24 h post-infection as described in our previous work (Turner et al., 2022). This time point represents an important bottleneck during infection of the cockroach gut identified in our previous studies (Turner et al., 2022). Individual cockroaches were cold anesthetized and surface sterilized by successive rinses in 10% bleach, 70% ethanol, and water. The cockroaches were then mechanically homogenized in sterile PBS using a handheld electronic tissue homogenizer, and the suspension was serially diluted and plated on LB agar containing 100 μg/mL of kanamycin for selective growth of S. Typhimurium. The plates were incubated overnight at 37 °C and colony forming units (CFU) of S. Typhimurium were counted the next day. Raw CFU counts per plate were used to calculate CFU/insect, with a theoretical limit of detection of 500 CFU/insect. Plates that yielded CFUs too numerous to count were conservatively set to 300 CFUs in calculations. Both the prevalence of infection (proportion of insects with detectable CFUs) and the intensity of infection (CFU/insect) were analyzed. Replication consisted of three biological replicates (independent infections), including a total of 34–35 insects per strain. Statistical analysis consisted of Fisher’s exact test for prevalence data and ANOVA with Dunn’s post-hoc test for log transformed CFU data. P-values <0.05 were considered statistically significant and reported. Negative controls consisting of uninfected insects of each strain were plated in each replicate and were always free of CFUs.

2.4. Determination of S. Typhimurium loads in excreta

Early phase excretion of S. Typhimurium by individual infected cockroaches was quantified by selective culture as described in our previous work (Turner et al., 2022). Immediately after infection, single cockroaches were placed into individual wells of a 12-well plate lined with filter paper and maintained under colony rearing conditions for 24 h without food or water. The cockroaches were then removed, and excreta were collected by soaking the filter paper in sterile PBS. The PBS solution was then diluted and plated on LB agar containing 100 μg/mL of kanamycin for selective growth of S. Typhimurium. The plates were incubated overnight at 37 °C and colony forming units (CFU) of S. Typhimurium were counted the next day. Raw CFU counts were used to calculate CFU excreted/insect, with a theoretical limit of detection of 50 CFU/insect. Plates that yielded CFUs too numerous to count were conservatively set to 300 CFUs in calculations. Both the prevalence of excretion (proportion of insects excreting detectable CFUs) and the intensity of excretion (CFU/insect) were analyzed. Replication consisted of three biological replicates (independent infections), including a total of 33 insects per strain. Statistical analysis consisted of Fisher’s exact test for prevalence data and ANOVA with Dunn’s post-hoc test for log transformed CFU data. P-values <0.05 were considered statistically significant and reported. Negative controls consisting of excreta from uninfected insects of each strain were plated in each replicate and were always free of CFUs.

2.5. Quantitation of cockroach feeding on infection medium

To determine whether different cockroach strains consumed significantly different doses (volumes) of S. Typhimurium culture under our free feeding protocol, which could confound the interpretation of infection data, we performed two different experiments. The first experiment followed a protocol modified from methods used to quantify feeding by Drosophila melanogaster (Wong et al., 2009). Groups of adult male cockroaches of each strain were starved of food and water for three days, then provided LB medium containing blue #1 dye as a sole food source and allowed to free feed for 30 min, as done during infections. Following this period, individual cockroaches were immediately placed into tubes of sterile PBS and mechanically homogenized to release the ingested blue #1 dye from the gut into solution. The homogenates were then centrifuged at 20,000 RCF to pellet tissue and cellular debris, and aliquots of the supernatants containing the dye were transferred to individual wells of a 96 well plate in triplicate for each cockroach. Absorbance at 629 nm (A629) on a spectrophotometer was measured for relative quantitation of the concentration of blue #1 dye in each sample, which served as a proxy for the volume ingested after normalization to absorbance values from unfed cockroach samples. This assay included 9–10 individual cockroaches per strain, and A629 data were statistically analyzed by ANOVA with Tukey’s post-hoc test. P-values <0.05 were considered statistically significant and reported.

In the second experiment, both A629 and internal S. Typhimurium CFUs were measured from the same cockroach samples using the above methods 24 h post-infection. This experiment included 59 individual cockroaches across all four strains and a Pearson correlation between CFU and A629 data was calculated to determine if there was a significant correlation between the volume of culture ingested and S. Typhimurium CFUs detected 24 h later.

2.6. Quantitation of cockroach necrophagy

Rates of necrophagy, a behavior known to contribute to the horizontal spread of S. Typhimurium among B. germanica (Mond and Pietri, 2023), were compared between cockroach strains. Twenty-one uninfected adult male cockroaches were transferred from each colony into separate experimental enclosures containing harborage and water but no food source. Immediately, 7 uninfected nymph cadavers of the same strain, obtained by freezing live nymphs overnight, were weighed and placed into each enclosure on a Petri dish as a sole food source. Consumption of the cadavers by the adults over a period of three days was quantified by measuring the mass lost over time, with adjustment for desiccation based on cadavers exposed to the environment in the absence of cockroaches. Mortality in the adults over this time was negligible, but enclosures were checked daily, and any dead adults were promptly removed. Consumption was expressed as total mass consumed as well as the proportion of initial cadaver mass consumed by each cohort of live adults. Three independent biological replicates of this assay were conducted, and data were analyzed by ANOVA with Tukey’s post-hoc test. P-values <0.05 were considered statistically significant and reported.

3. Results

3.1. The prevalence and intensity of S. Typhimurium infection vary among cockroach strains

Significant differences in both the prevalence (Fig. 1A) and intensity (Fig. 1B) of S. Typhimurium infection were observed between the different cockroach strains. Infection prevalence at 24 h was 71% for the Orlando lab strain, 63% for the Douglas field strain, 89% for the Gilpin field strain, and 94% for the Hopewell field strain. Upon statistical analysis (Fisher’s exact test), the prevalence was found to be significantly higher in the Hopewell field strain relative to both the Orlando lab strain (P =0.0234) and Douglas field strain (P = 0.0027). Prevalence was also significantly higher in the Gilpin field strain relative to the Douglas field strain (P = 0.024). The average S. Typhimurium load (intensity) in infected individuals from the Orlando lab strain was 7.81 × 104 CFU/insect, consistent with our previously published work (Turner et al., 2022). In the Douglas field strain, the average load was 4.48 × 104 CFU/insect. The Gilpin field strain and Hopewell field strain harbored higher S. Typhimurium loads of 1.08 × 105 CFU/insect and 9.82 × 104 CFU/insect, respectively. ANOVA with Dunn’s post-hoc test revealed statistically significant differences in infection intensity between the Gilpin field strain and the Douglas field strain (P = 0.0009) and between the Hopewell field strain and the Douglas field strain (P = 0.0103). Together, these data paint a picture of elevated susceptibility to S. Typhimurium infection in the Hopewell and Gilpin field strains and less susceptibility in the Douglas field strain, with the Orlando lab strain having an intermediate phenotype.

3.2. Early phase excretion of S. Typhimurium varies among cockroach strains

Differences in excretion of S. Typhimurium were also observed between cockroach strains (Fig. 2). However, excretion patterns were distinct from internal infection patterns. Fifty-four percent of individuals from the Orlando lab strain that were fed S. Typhimurium excreted detectable CFUs of the bacteria in the 24 h following infection (Fig. 2A). Meanwhile, the prevalence of excretion of detectable CFUs was 24% for the Douglas field strain, 15% for the Gilpin field strain, and only 12% for the Hopewell field strain. Fisher’s exact test determined that the prevalence of detectable S. Typhimurium excretion was significantly lower in all field strains relative to the Orlando lab strain (P < 0.05). When the concentration of S. Typhimurium in the excreta of insects from each strain was determined, slightly different patterns emerged (Fig. 2B). Although the prevalence of excretion in the Douglas field strain was intermediate among the strains, individuals from this strain that did excrete the bacteria, excreted the highest concentration, at 1.02 × 104 CFU/insect. The Orlando laboratory strain, while having the highest prevalence of excretion, excreted an intermediate concentration of 2.58 × 103 CFU/insect, consistent with our prior observations of this strain (Turner et al., 2022). Meanwhile, the Gilpin field strain and Hopewell field strain, which exhibited the lowest prevalence of excretion, also excreted the lowest concentrations of S. Typhimurium, at 4.3 × 102 CFU/insect and 1.38 × 102 CFU/insect, respectively. ANOVA with Dunn’s post-hoc test revealed statistically significant differences in the concentration of S. Typhimurium excretion between the Douglas field strain and Gilpin field strain (P =0.0019) as well as between the Douglas field strain and the Hopewell field strain (P = 0.0097).

3.3. S. Typhimurium infection is not correlated with volume of culture consumed

Two distinct experiments revealed no connection between differential feeding on S. Typhimurium culture and CFU counts 24 h post infection across the four cockroach strains that we examined (Fig. 3). Based on ANOVA with Tukey’s post-hoc test, all field strains consumed significantly less bacterial media on average in the free feeding assay than the Orlando lab strain (0.0001 ≤ P ≤ 0.0147), but there were no significant differences in feeding between any of the three field strains (P > 0.05) (Fig. 3A). Further, when consumption of bacterial solution (A629) and CFUs counts 24 h post-infection measured from the same insects were directly compared, no significant correlation between the values was observed across 59 individuals from all four strains (Fig. 3B, Pearson correlation, R = 0.24, P = 0.063). These results indicate that observed differences in infection prevalence and intensity (Fig. 1) among strains are due to intrinsic differences in susceptibility and rule out variation in bacterial consumption among the strains as a relevant variable.

3.4. Necrophagy varies among cockroach strains

In addition to differences in biological susceptibility to S. Typhimurium infection (Fig. 1), the cockroach strains we examined also exhibited differences in behavioral potential for infection via necrophagy (Fig. 4). Both the Orlando laboratory strain and the Hopewell field strain were highly necrophagic in our assay, while the Douglas and Gilpin strains were minimally necrophagic. Specifically, average consumption of cadavers as a sole food source over three days was 0.103 g (27.86% of available mass) for cohorts of the Orlando lab strain, 0.113 g (35.99% of available mass) for cohorts of the Hopewell field strain, 0.036 g (8.50% of available mass) for cohorts of the Gilpin field strain, and 0.016 g (4.25% of available mass) for cohorts of the Douglas field strain. Based on ANOVA with Tukey’s post-hoc test, necrophagy was significantly lower (P < 0.05) in the Douglas field strain relative to both the Orlando lab strain and Hopewell field strain when comparing total mass consumed as well as the percentage of available mass consumed. When comparing only the percentage of available mass consumed, necrophagy in the Gilpin field strain was also significantly lower than in the Hopewell field strain (P = 0.012). These patterns were readily apparent visually upon inspection of cadavers (Fig. 5). Cadavers provided to the Douglas and Gilpin strains appeared mostly intact after three days whereas those provided to the Orlando and Hopewell strains typically were typically dismembered and missing large portions of their bodies or consumed entirely.

4. Discussion

By conducting controlled laboratory infection assays of multiple strains of B. germanica, we provide for the first time evidence that the ability of these cockroaches to become infected by and disseminate S. Typhimurium is influenced by their hologenomic background (i.e., cockroach strain). Moreover, differences in patterns of infection compared to excretion observed among cockroach strains provide additional evidence that infection of the cockroach gut by S. Typhimurium and its subsequent dissemination in the excreta are decoupled, consistent with our previous work examining S. Typhimurium mutants (Turner et al., 2022). That is, higher or lower levels of internal infection in one cockroach strain compared to another strain did not necessarily correspond to the same pattern in excretion. For example, although the Douglas field strain exhibited the lowest internal levels of S. Typhimurium infection, excretion was intermediate among the strains. Yet, while internal infection of the Orlando laboratory strain was intermediate, it excreted by far the most S. Typhiurium. In contrast, the Gilpin and Hopewell field strains, which were the most heavily infected, excreted S. Typhimurium minimally.

Importantly, we assessed the propensity of each strain to consume the bacterial media used to administer oral S. Typhimurium infection. The purpose of these experiments was specifically to rule out that strain specific differences in feeding on S. Typhimurium would confound infection assays. However, our results revealed no significant difference in feeding on bacterial media among the field strains, along with no correlation between volume of bacterial media ingested and internal levels of bacteria 24 h later when considering data from all strains. These data support the premise that the differences in infection and excretion observed among cockroach strains are due to hologenomic variation.

Additional work is needed to identify the drivers of infection susceptibility and vector competence for S. Typhimurium in B. germanica. These may be genome encoded, epigenetic, or microbiome dependent. For example, we have previously shown that expression of antimicrobial peptides, such as Blattellicins and Attacins, is upregulated in the gut of B. germanica following ingestion of live S. Typhimurium (Turner and Pietri, 2022), suggesting that these peptides play a role in limiting infection. Therefore, polymorphisms that affect the regulation or function of antimicrobial peptides, or other immunity related genes, may contribute to differences in S. Typhimurium infection and excretion among cockroach strains. On the other hand, German cockroaches have a highly diverse gut microbiota that differs between field and laboratory strains, as well as between strains collected from different field sites (Kakumanu et al., 2018). Our previous work in the Orlando laboratory strain demonstrates that the presence of a gut microbiota has pleiotropic effects on infection by ingested enterobacteria, including E. coli (Ray et al., 2020) and S. Typhimurium (Turner et al., 2023). However, the specific bacteria that underlie microbiome mediated effects on infection in cockroaches have yet to be identified, and the role of natural microbiome variation on infection dynamics remains to be explored. The ability to generate gnotobiotic cockroaches will facilitate future studies of this nature.

The finding that necrophagy significantly differs among cockroach strains was particularly surprising and indicates that differences in the potential for S. Typhimurium infection may not only be biological, but also behavioral, as necrophagy serves as a route for horizontal transmission of this bacterium among B. germanica (Mond and Pietri, 2023). There is conflicting evidence in the literature regarding necrophagy by German cockroaches. Both laboratory and field studies have reported the frequent occurrence of necrophagy (Gahlhoff et al., 1999; Buczkowski et al., 2008; Gemeno et al., 2011; Mond and Pietri, 2023), sometimes even when alternative food is available (Gemeno et al., 2011; McPherson et al., 2023). In contrast, several other studies have reported minimal to no necrophagy even in the absence of alternate foods (Tabaru et al., 2003; Appel et al., 2008). Although one hypothesis to explain these conflicting observations is that they may be due to the different environmental conditions under which the assays were carried out, our results comparing necrophagy under identical conditions indicate that this behavior also has a hologenomic basis which should be further investigated.

Together, the results presented here are important steps towards developing a mechanistic understanding of S. Typhimurium transmission by cockroaches. In particular, while controlled studies in the Orlando laboratory strain have begun to illuminate the spatiotemporal complexity of infection and the bacterial factors involved (Turner et al., 2022), the identification of cockroach strains with distinct infection profiles is critical for eventually elucidating the host drivers of infection and transmission. As such, metagenomic and transcriptomic sequencing of different cockroach strains during infection remains an essential future direction. Lastly, the realization that different cockroach populations may have markedly different propensities to become infected by or excrete S. Typhimurium may help to reconcile any discrepancies between studies of this vector-pathogen relationship in the past or future (Ash and Greenberg, 1980; Turner et al., 2022).

Acknowledgements

We thank Matthew Turner at the University of South Dakota for assistance with cockroach rearing.

Funding

This work was funded by the National Institutes of Health, National Institute of Allergy and Infectious Diseases, grant R01AI171014 to JEP.

Data statement

All data are presented in the manuscript and are available upon request.

Data availability

Data will be made available on request.

Fig. 1. Variation in Salmonella infection levels among strains of Blattella germanica. Adult males of different strains of B. germanica were orally infected with S. tTyphimurium in parallel under identical conditions. 24 h post-infection, whole insects were homogenized and cultured on LB plates with kanamycin to assess infection status. (A) Prevalence of infection expressed as the proportion of insects harboring or not harboring detectable colony forming units (CFUs). The limit of detection was 500 CFUs and statistical analysis consisted of Fisher’s exact test. (B) S. Typhimurium loads in infected insects expressed as CFU/insect. Mean and SEM are shown. Statistical analysis consisted of ANOVA with Dunn’s post-hoc test. The data in both (A) and (B) were derived from three independent infection replicates including a total of 34–35 insects per strain.

Fig. 2. Variation in Salmonella excretion among strains of Blattella germanica. Adult males of different strains of B. germanica were orally infected with S. Typhimurium in parallel under identical conditions. Single insects were then housed in individual wells of a 12-well plate for collection of excreta. 24 h post-infection, excreta were collected and cultured on LB plates with kanamycin. (A) Proportion of insects excreting or not excreting detectable colony forming units (CFUs) in the feces. The limit of detection was 50 CFUs and statistical analysis consisted of Fisher’s exact test. (B) S. Typhimurium loads in the feces of individual infected insects expressed as CFU excreted in 24 h. Mean and SEM are shown. Statistical analysis consisted of ANOVA with Dunn’s post-hoc test. The data in both (A) and (B) were derived from three independent infection replicates including a total of 33 insects per strain.

Fig. 3. Salmonella levels in the cockroach are not correlated with the volume of bacterial culture ingested. (A) Differences in ingestion of the LB medium used for S. Typhimurium culture by strains of B. germanica were assessed 30 min after feeding on medium containing blue #1 dye by using spectrophotometry (A629) to measure the concentration of the dye in whole cockroach homogenates. Although the Gilpin and Hopewell strains had higher S. tTyphimurium levels 24 h post-infection, these strains consumed less medium on average than the Orlando lab strain when offered. Each data point represents an individual cockroach. Mean and SEM are shown. Statistical analysis consisted of ANOVA with Tukey’s post-hoc test. (B) 24 h post-infection with S. Typhimurium in LB, the same spectrophotometric assay was used to assess the concentration of blue dye in individual cockroaches. In parallel, internal S. Typhimurium CFUs were assessed in the same insects by culture. Pearson correlation analysis found no significant correlation between S. tTyphimurium levels and the amount of blue dye present. The data in (B) show combined results from all four cockroach strains.

Fig. 4. Variation in necrophagy among strains of Blattella germanica. Groups of male cockroaches from different B. germanica strains were maintained under identical conditions in separate enclosures containing harborages and water but no food source. Cadavers of nymphs from the same strain were added to the enclosures as a sole food source and the consumption of these cadavers over time was assessed by weight. After three days, the total mass of cadavers consumed by the living adults of each strain was measured in (A) grams (g) or (B) as a percentage of available mass, accounting for weight lost due to desiccation. Three independent replicates were conducted, and the mean and SEM are shown. Statistical analysis consisted of ANOVA with Tukey’s post-hoc test.

Fig. 5. Representative photographs of necrophagy variation. (A) Minimal necrophagy of nymph cadavers by the Douglas field strain after three days. (B) Appreciable necrophagy by the Hopewell field strain after three days. Note that only six cadavers are shown for the Hopewell strain, as one was entirely consumed and could not be photographed. These photographs are derived from a single replicate of the necrophagy assay but are representative of results seen in the three biological replicates that were conducted.

Declaration of competing interest

None.

CRediT authorship contribution statement

Bashar Ismael: Writing – review & editing, Writing – original draft, Investigation, Formal analysis, Data curation, Conceptualization. Morgan Wilson: Writing – review & editing, Resources. Dini Miller: Writing – review & editing, Resources. Jose E. Pietri: Writing – review & editing, Writing – original draft, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization.
==== Refs
References

Appel AG , Sims RS , Eva MJ , July 2008. Factors affecting coprophagy and necrophagy by the German cockroach (Dictyoptera: Blattellidae). In: 6th International Conference on Urban Pests, Budapest, Hungary, vols. 13–16 , pp. 139–142.
Ash N , Greenberg B , 1980 Sep 30. Vector potential of the German cockroach (Dictyoptera: Blattellidae) in dissemination of Salmonella enteritidis serotype typhimurium. J. Med. Entomol. 17 (5 ), 417–423.6999154
Beerntsen BT , James AA , Christensen BM , 2000 Mar. Genetics of mosquito vector competence. Microbiol. Mol. Biol. Rev. 64 (1 ), 115–137.10704476
Buczkowski G , Scherer CW , Bennett GW , 2008 Jun. Horizontal transfer of bait in the German cockroach: indoxacarb causes secondary and tertiary mortality. J. Econ. Entomol. 101 (3 ), 894–901.18613592
Burgess NR , Chetwyn KN , 1981. Association of cockroaches with an outbreak of dysentery. Trans. R. Soc. Trop. Med. Hyg. 75 (2 ), 332–333.7029807
Gahlhoff JE , Miller DM , Koehler PH , 1999. Secondary kill of adult male German cockroaches (Dictyoptera: Blattellidae) via cannibalism of nymphs fed toxic baits. J. Econ. Entomol. 92 (5 ), 1133–1137.
Gemeno C , Williams GM , Schal C , 2011. Effect of shelter on reproduction, growth and longevity of the German cockroach, Blattella germanica (Dictyoptera: Blattelidae). Eur. J. Entomol. 108 (2 ), 205–210.
Graffar M , Mertens S , 1950 Nov. Le role des Blattes dans la transmission des salmonelloses [Rôle of Blatta in transmission of salmonellosis]. Ann. Inst. Pasteur. (Paris) 79 (5 ), 654–660.14799998
Kakumanu ML , Maritz JM , Carlton JM , Schal C , 2018 Aug 17. Overlapping community compositions of gut and fecal microbiomes in lab-reared and field-collected German cockroaches. Appl. Environ. Microbiol. 84 (17 ) e01037–18.29959246
Krishnavajhala A , Armstrong BA , Lopez JE , 2018 Oct 17. Vector competence of geographical populations of Ornithodoros turicata for the tick-borne relapsing fever spirochete Borrelia turicatae. Appl. Environ. Microbiol. 84 (21 ) e01505–18.30143510
McPherson S , Wada-Katsumata A , Silverman J , Schal C , 2023 Apr 24. Glucose- and disaccharide-containing baits impede secondary mortality in glucose-averse German cockroaches. J. Econ. Entomol. 116 (2 ), 546–553.36888567
Mond M , Pietri JE , 2023 May 8. Horizontal transmission of Salmonella typhimurium among German cockroaches and its possible mechanisms. Ecol. Evol. 13 (5 ), e10070.37181208
Nasirian H , 2019 Oct 28. Contamination of cockroaches (Insecta: Blattaria) by medically important Bacteriae: a systematic review and meta-analysis. J. Med. Entomol. 56 (6 ), 1534–1554.31219601
Ray R , Potts R , Pietri JE , 2020 Nov 13. The persistence of Escherichia coli infection in German cockroaches (Blattodea: Blattellidae) varies between host developmental stages and is influenced by the gut microbiota. J. Med. Entomol. 57 (6 ), 1964–1971.32516418
Stek M , 1982. Cockroaches and enteric pathogens. Trans. R. Soc. Trop. Med. Hyg. 76 (4 ), 566–567.6926780
Tabaru Y , Mochizuki K , Watanabe Y , 2003. Coprophagy and necrophagy of the German cockroach, Blattella germanica, in the laboratory condition. Med. Entomol. Zool. 54 (1 ), 13–16.
Tarshis IB , 1962 Sep. The cockroach - a new suspect in the spread of infectious hepatitis. Am. J. Trop. Med. Hyg. 11 , 705–711.13993369
Toda T , 1923 Aug. Cholera and the ship “cockroach”. J. Hyg. (Lond.) 21 (4 ), 359–361.20474783
Turner M , Pietri JE , 2022 May 15. Antimicrobial peptide expression in the cockroach gut during enterobacterial infection is specific and influenced by type III secretion. Biol. Open 11 (5 ) bio059414.35611712
Turner M , Peta V , Pietri JE , 2022 Mar-Apr. New insight into the relationship between Salmonella typhimurium and the German cockroach suggests active mechanisms of vector-borne transmission. Res. Microbiol. 173 (3 ), 103920.34954364
Turner M , Van Hulzen L , Pietri JE , 2023 Aug 22. The gut microbiota induces melanin deposits that act as substrates for fimA-mediated aggregation of Salmonella typhimurium and enhance infection of the German cockroach vector. Microbiol. Spectr. 11 (5 ), e0211923.37606369
Wong R , Piper MD , Wertheim B , Partridge L , 2009 Jun 26. Quantification of food intake in Drosophila. PLoS One 4 (6 ), e6063.19557170
