deletions | additions       diff --git a/section_Discussion_textit_F_tularensis__.tex b/section_Discussion_textit_F_tularensis__.tex  index fd2110a..37c48f4 100644  --- a/section_Discussion_textit_F_tularensis__.tex  +++ b/section_Discussion_textit_F_tularensis__.tex 
...
Like wax moth larvae, OS cockroaches are commercially-available from a large number of suppliers that produce them for the pet industry (primarily used as food for captive reptiles). However, whereas wax moths are relatively difficult to rear in captivity, maintenance of an OS cockroach breeding colony in the laboratory is simple and straightforward. Compared to other species, OS cockroaches are relatively docile and easy to handle. Given adequate access to water, OS cockroaches can be maintained at temperatures above 37°C (up to 40°C in our studies) indefinitely. OS cockroaches undergo incomplete metamorphosis, with each developmental stage (or instar) lasting between 20 and 45 days. In total, it takes approximately 6 months for OS cockroaches to reach adulthood. The juvenile cockroaches used in this study were infected during the 6th instar (next to last), leaving between 30 and 60 days of experimental observation before they would have molted into the adult stage. In contrast, we often observed that significant percentages (>25\%, data not shown) of wax moth larvae pupated during a 7 day experiment. Importantly, deaths of OS cockroaches were extremely rare in our PBS-only or no-injection control groups and the LD_{50} of a non-pathogenic bacteria, \textit{Escherichia coli} DH5$\alpha$, was determined to be nearly 10^{7} CFU (\textbf{Table 1}), indicating that OS cockroaches are capable of mounting protective immune responses to non-pathogenic microorganisms.   Intrahemoceol injection of \textit{F. tularensis} LVS resulted in dose-dependent lethality (\textbf{Figure 1}) with an estimated LD_{50} of 1.7 to 3.5 x 10^4 CFU (with overlapping confidence intervals) for two different laboratory stocks (\textbf{Table 1}). Temperature is known to regulate expression of \textit{F. tularensis} virulence factors \cite{18842136}. One of the advantages of insect models, in comparison with mammalian models, is the ability to experimentally manipulate the temperature at which the host-pathogen interactions occur. When we varied the temperature of incubation following cockroach infection, we observed that as the temperature decreased, so did \textit{F. tularensis} LVS lethality (\textbf{Figure 2, Table 1}). Since \textit{F. tularensis} can be spread by environmental arthropods \cite{24057273,21529386,20482589,18950590,22530023,21612530,20885922, 20810833, 21845949}, temperature may provide an important environmental cue that allows \textit{F. tularensis} to dampen virulence pathways that would otherwise kill these vectors before they have an opportunity to transmit the bacterium to another host mammal.  Host immune function is not static, rather it often varies dramatically across developmental stages \cite{25730277}. We therefore sought to determine if OS cockroach susceptibility to \textit{F. tularensis} LVS varied by developmental stage. We determined the killing kinetics and LD_{50}s of \textit{F. tularensis} LVS againstlate-stage  juvenile, adult female, and adult male OS cockroaches. The susceptibility pattern of juveniles (which we used for all other experiments reported here) and adult females were highly similar. In comparison, adult males showed enhanced susceptibility, with a shorter median time-to-death (\textbf{Figure 4}) and a lower LD_{50} (\textbf{Table 1}). The reason for this difference is currently unknown and future experiments aimed at uncovering the mechanistic differences in immune responses between these groups could identify important anti-\textit{F. tularensis} host pathways. In order to begin to define the genetic requirements for \textit{F. tularensis} virulence in our OS cockroach model, we examined the virulence of a small panel of \textit{F. tularensis} LVS mutants that are attenuated in other model systems (\textbf{Table 1}). All four of the mutants examined here showed significant attenuation in OS cockroaches, which correlates with previous findings in mice and chick embryos. These findings support the idea that \textit{F. tularensis} uses a similar virulence mechanisms to evade immune clearance and cause disease among extremely diverse host organisms.   Since \textit{F. tularensis} is considered a facultative intracellular pathogen, we sought to determine the proportion of bacteria that were located in intracellular and extracellular compartments throughout infection of OS cockroaches. As seen in \textbf{Figure 5}, intracellular bacteria can be recovered as early as six hours post injection. The intracellular population continues to grow throughout the infection process, as does the total bacterial population. Initially, we were surprised that the majority of the bacterial population at each time point was located in the extracelllular environment. However, our results are similar to what others have observed for \textit{F. novicida} in \textit{D. melanogaster} \cite{20865166}. Since hemocoel-injected gentamicin rescued OS cockroaches from lethality (\textbf{Figure 7}), the extracellular \textit{F. tularensis} population likely is most responsible for appear to be essential to  the negative outcomes of infection, infection process,  as has been  recently suggested elsewhere  \cite{22795971}. While the intracellular phase of \textit{F. tularensis} pathogenesis is well-appreciated, our findings suggest that the OS cockroach may be a useful model for elucidating the mechanisms by which \textit{F. tularensis} survives, grows, and moves within the extracellular environment. Finally, we tested the ability of five different antibiotics to protect OS cockroaches from \textit{F. tularensis} LVS infection. Doxycycline is readily absorbed orally and was able to protect OS cockroaches from infection when delivered by either route (\textbf{Figure 7B}). This protection was specific to antibiotics with anti-\textit{Francisella} activity since azithromycin, to which \textit{F. tularensis} LVS is resistant, failed to protect from lethality (\textbf{Figure 7C}).  Streptomycin and gentamicin are aminoglycoside antibiotics with poor oral bioavailability. bioavailability in mammals.  Interestingly, these antibiotics only protected OS cockroaches when delivered by systemic injection and not when provided orally (\textbf{Figure 7, panels B and C}). In contrast, ciprofloxacin 7D  and doxycycline are readily absorbed orally and show equivalent protective activity when delivered by either route (\textbf{Figure 7, panels E and F}). 7E}).  These findings indicate that oral absorption of antibiotics is similar in both mammals and insects and that OS cockroaches provide a preliminary screening platform for identification of new antibiotics with anti-\textit{Francisella} activity. As an example, we examined the ability of resazurin, which has been shown to have potent anti-\textit{F. tularensis} LVS activity \textit{in vitro} \cite{24367766}, to rescue OS cockroaches from lethality. Despite this \textit{in Despite\textit{in  vitro}antibacterial  activity, resazurin failed to protect OS cockroaches when delivered by  either orally or systemically. route.  Thus, we hypothesize that further modifications of the resazurin chemical backbone are required in order to develop effective anti-\textit{Francisella} drugs based upon the resazurin scaffold. The results presented here indicate that the OS cockroach is susceptible to infection by \textit{F. tularensis} LVS. Further, we demonstrated that this model system can be used: (1) to identify environmental (i.e., temperature) and genetic factors that contribute to \textit{Francisella} virulence; (2) to characterize the extracellular phase of \textit{Francisella} virulence; and (3) to discover new antibiotics with favorable oral bioavailability profiles and anti-\textit{Francisella} activity. Together with the logistical advantages discussed above, these opportunities suggest that the OS cockroach is an extremely important addition to our host model repertoire.