The reproductive status of the host determines the tolerance and resistance to 1 Mycobacterium marinum infection in Drosophila melanogaster

21 Both the sex and the reproductive status of the host have a major impact on the regulation of the 22 immune response against infection. Due that Drosophila melanogaster has become a powerful model for study such interactions, we wanted to understand whether the sex or the reproductive status has an 24 impact on the tolerance or resistance of the host to the in the model of systemic Mycobacterium 25 marinum infection. We measured gene expression by RT-qPCR of immune genes, diptericin and 26 drosomycin , as well as host survival and the bacillary load at time of death. We also assessed the impact 27 of metabolic (by expression of upd3 and impl2 ) and hormonal (by ecR

Both the sex and the reproductive status of the host have a major impact on the regulation of the 22 immune response against infection. Due that Drosophila melanogaster has become a powerful model 23 for study such interactions, we wanted to understand whether the sex or the reproductive status has an 24 impact on the tolerance or resistance of the host to the in the model of systemic Mycobacterium 25 marinum infection. We measured gene expression by RT-qPCR of immune genes, diptericin and 26 drosomycin, as well as host survival and the bacillary load at time of death. We also assessed the impact 27 of metabolic (by expression of upd3 and impl2) and hormonal (by ecR expression) regulation in the 28 protection against infection. Data showed that resistance increased in actively mating males and 29 females, and also in mated females while reducing the tolerance to infection. The data also suggest the 30 Toll and IMD pathways determine tolerance and resistance, respectively, while basal levels of ecR 31 favors the stimulation of the IMD pathway. A time related role has been found for upd3 expression, 32 linked to the increase or decrease in the mycobacterial load at the beginning and at the end of the 33 infection, respectively. Finally, impl2 expression has been related to the increase of resistance only in 34 non-actively mating males. The results obtained allows further interpretation of the results when 35 assessing the differences between sexes and highlights the role of the reproductive status in D. 36 melanogaster to face infections, since our data demonstrated their importance to determine resistance 37 and tolerance against M. marinum infection. 38 39 40

Introduction 41
Reproduction and immunity are resource-intensive traits in their deployment and maintenance, thus, a 42 trade-off exists between immunity and reproduction when the resources are limited [1]. This trade-off 43 has been documented in both mammals and non-mammals [2][3][4][5]. Its study is important to shed light 44 on the impact of directional selection and evolution in the maintenance or the variation of these traits, 45 thus helping to better characterize the relationship between these two physiological mechanisms [6]. 46 In this regard, Drosophila melanogaster is a powerful model organism to study the interaction between 47 reproduction and immunity due to its already well-studied courtship ritual and its well-characterized 48 immune system [7]. Immunity in D. melanogaster relies solely on innate immune responses, 49 nevertheless, there are a lot of aspects of this innate immunity that are conserved between flies and 50 mammals, such as the organization of the immune response into humoral and cellular arms, and the 51 presence of NF-κB family transcription factors and signal transduction pathways [8]. The humoral 52 response is mainly based on the synthesis of antimicrobial peptides (AMPs) in the fat body of the fly. 53 This process is regulated by two signalling pathways: the Toll and the immune deficiency (Imd) 54 pathways, which activate different NF-κB-like factors and induce the production of Drosomycin and 55 Diptericin (as the main AMP of each pathway, respectively) [9,10]. Flies also have a simpler but 56 complete JAK/STAT pathway, which regulates a wide range of biological processes in the fruit fly, 57 including immunity [11,12]

Fly stocks and husbandry 133
Oregon-R-C wildtype flies were obtained from the Bloomington Drosophila Stock Centre (BDSC,  134 Indiana University) and they belong to stock number #5. Flies were raised on a standard cornmeal 135 medium at 25ºC, 65-70% humidity with a 12h light/dark cycle. Male and female flies were aged for 3-136 5 days before experimentation. 137

Mycobacterial stock preparation and infection 138
Mycobacterium marinum E11 strain resistant to kanamycin (a kind gift from Wilbert Bitter, Vrije 139 Universiteit Amsterdam) was used [60]. The mycobacterial strain was cultured in 7H9 complete 140 medium complemented with 20µg/ml of antibiotic and placed at 30ºC with constant agitation (170rpm) 141 for 10 days (until an OD600nm of 1.5). The cultures were centrifuged for 5min at 4000g, resuspended in 142 phosphate buffered saline (PBS) with 0.2% Tween 80 and centrifuged again for 5min at 500g to remove 143 clumps. Supernatants were transferred to a new tub, centrifuged for 5min at 4000g and resuspended in 144 1ml of 7H9 with 15% glycerol. Mycobacterial cultures were then aliquoted and frozen at -80ºC. Each 145 stock was tittered after being frozen at least overnight. 146 For systemic infections, an aliquot was defrosted and centrifuged at 15000g for 5min. Pellet was rinsed 147 twice with sterile PBS and diluted to the proper concentration. Fourteen nanolitres of the diluted 148 mycobacterial solution were injected systemically employing a nano-injector (Nanoject II, Drummond) 149 into the abdomen of anaesthetized flies. 150

Experimental design 151
Flies were divided into three experimental groups: virgin flies, mated flies that were separated by sex 152 after the infection and flies that were allowed to mate throughout the infection (actively mating). Virgin 153 flies were collected at the 2-3h post-eclosion and kept in same-sex groups before and after 154 experimentation. In all experiments, male and female flies were kept in groups of 30 individuals, either 155 only one sex or equally distributed between males and females. 156 Overall, the experiment consisted of 30 male and 30 female flies per experimental group and 3 157 biological replicates were performed, with a total of 90 males and 90 females per group. Survival was 158 checked daily and the bacterial load upon death (BLUD) was measured. Dead flies were collected 159 every day, washed with 70% ethanol and rinsed twice with PBS. Each fly was then mechanically 160 homogenized into 200µl of sterile PBS, diluted and plated into 7H10 plates complemented with 161 kanamycin, and incubated for 10 days at 30ºC. 162

Tolerance and resistance 163
Previous studies have defined three defensive strategies against infections: qualitative resistance, as 164 the ability to remain uninfected upon pathogen exposure; quantitative resistance, as the ability the 165 reduce the pathogen load; and tolerance, as the ability of the host to maintain the fitness given a certain 166 pathogen load [61,62]. Given our model of study in which the minimum lethal dose of the pathogen is 167 1, meaning that a single mycobacterium is able to kill the host, we did not assess qualitative resistance 168 but focused on tolerance and quantitative resistance.

The effect of reproduction in tolerance and resistance to M. marinum infections is sex-211 dependent 212
Our study revealed no significant differences in the general vigour of males, independently of their 213 reproductive status. Instead, results showed that the main difference was due to males being 214 reproductively active, rather than having previously mated or not. Thus, males kept together with 215 females were less tolerant but more resistant to the infection by M. marinum compared to males kept 216 alone and virgin males (Figure 1a - Table 2). 217 On the other hand, the general vigour was significantly lower in virgin females, when compared with 218 both mated groups. In addition, results suggested that females' response to infection was more 219 conditioned by whether they have mated or not, regardless of whether they were reproductively active 220 at the time of infection or not. Thus, virgin females were more tolerant but less resistant to the infection, 221 compared to both mated groups (Figure 1b -Table 2). 222 When we compared tolerance and resistance of males and females for each reproductive status 223 independently, data showed that virgin flies differed neither in general vigour nor in overall tolerance 224 to the infection, but males were significantly more resistant (Figure 1c - Table 3). The same pattern 225 was observed in flies that were kept together in equal proportions throughout the procedure: both males 226 and females had the same general vigour and the same tolerance levels, but males showed up to be 227 more resistant (Figure 1e - Table 3). On the contrary, when flies had mated but were kept separated by 228 sex after the infection, females showed increased general vigour and resistance, but reduced tolerance 229 to the infection with M. marinum (Figure 1d - Table 3). by the Imd and the Toll pathways, respectively. In males, only those kept together with females showed 236 immediate significant production of Diptericin, while the other groups showed a delayed production 237 of this AMP. Moreover, virgin males presented significantly increased production of Drosomycin 238 immediately after the infection, while in males alone this production was delayed, and no significant 239 changes were observed in males kept together with females. In females, we observed similar expression 240 profiles as males for each reproductive status: females kept together with males had immediate 241 production of Diptericin, but not Drosomycin; virgin females had delayed production of Diptericin, 242 but early production of Drosomycin; females alone showed delayed or null production of both AMPs. 243 These results for the expression levels showed correlation with the phenotypic results obtained 244 previously. We found that the activation of the Toll pathway (production of Drosomycin) correlated 245 with more tolerant but less resistant phenotypes in both males (virgins and alone) and females (virgins). 246 In addition, the more sustained was this activation (males alone) the lower resistance was observed. 247 On the other hand, we also observed that while early activation of the Imd pathway correlated with the 248 more resistant groups (males and females together), late or null activation of this pathway seemed to 249 be linked with reduced resistance with the exception of females alone. These correlations were also 250 observed when comparing the expression levels between males and females for each reproductive 251 status (Supp. Figure 1a). We found that the less resistant groups (virgin females, females kept together 252 7 with males, and males alone) showed significantly increased production of both AMPs late in the 253 infection, while the more tolerant group (males alone) had increased production of Drosomycin. 254 Finally, results also suggest that females kept alone after mating may acquire their resistance through 255 a mechanism that does not involve the humoral innate immunity, as they did not show any activation 256 of the Imd pathway throughout the infection. 257

The metabolic regulation during infection is altered by the reproductive status of the host 258
The link with the metabolism, especially with the insulin/insulin-like signalling pathway (IIS), together 259 with the suggested role for Upd3 and Impl2 as "selfish immune factors" (SIFs), prompted us to 260 investigate whether a differential expression of these molecules within the different reproductive status 261 of the host might also influence the tolerance and resistance levels to the infection by M. marinum 262 (Figure 2b). 263 In males, both virgins and those kept alone showed significant early overexpression of upd3, which 264 was not sustained later in the infection. On the other hand, those males that were kept together with 265 females during the whole procedure presented a significant overexpression of this gene later in the 266 infection. In females, virgins showed a significant early but not sustained overexpression of upd3, 267 while females that have mated presented a significantly increased expression of this gene later in the 268 infection whether they were in the presence or absence of males. Overall, these results suggest a dual 269 role for upd3, as its early expression correlated with more tolerant but less resistant groups while its We also compared the expression levels of both genes between males and females for each 277 reproductive status independently (Supp. Figure 1b). We found that females have an overall increased 278 expression of upd3 compared to males, while impl2 seemed to only play a role in males with its late 279 production correlating with increased tolerance (males alone). 280

but not in females 282
We also evaluated the expression levels of the ecdysone receptor (EcR) due to the tight relationship of 283 the Ecdysone pathway with immunity and reproduction. We measured the relative expression of this 284 gene after the infection and the basal expression levels when injected with PBS for each reproductive 285 status (Figure 3c). 286 We found that virgin males significantly induced the expression of the receptor later in the infection, 287 while males alone did it immediately after. On the other hand, the infection triggered a significant 288 repression of the receptor in those males kept together with females ( Figure 3a). We also found that 289 males presented a significant correlation between the basal level of EcR and the induction of this 290 receptor post-infection, in which lower basal levels translated into higher activation of this gene post-291 infection (Figure 3b and 3c). In females, virgins showed no changes neither in the basal levels nor in 292 the induction of EcR after the infection. Females that had mated showed similar patterns as their males 293 counterparts, while females kept alone showed decreasing induction of EcR when infected, but no 294 8 significant changes on the basal levels (Figure 3a and 3b). However, the correlation between basal 295 levels and induction of EcR after the infection was not statistically significant in females (Figure 3c). 296 The results also showed a link between the basal expression levels of EcR and the resistance to the 297 infection. Those groups with higher basal expression levels (males and females kept together and 298 females alone) were more resistant to the infection, while those groups with lower basal levels (virgin 299 flies and males alone) were less resistant. 300 When comparing both basal and induction levels between males and females (Supp. Figure 2a), we 301 observed that virgin males presented lower basal levels but increased production of EcR after the 302 infection, while flies separated after the infection showed no significant different basal levels but 303 females increased its production after the infection. Finally, when flies were kept together throughout 304 the procedure, no differences at basal levels among sexes were found, but a late induction in females. resistance profiles as well (Figure 4b). In females, only virgins presented a significantly different gene 312 expression profile (Figure 4b), also linking with the phenotypic results. Finally, the analysis of the 313 variable contribution (Figure 4b) for dimension 1 (PC1), revealed that, in males, this difference was 314 mainly driven by the differential expression of Impl2, together with Drosomycin and Upd2, while in 315 females the main differences were driven by Upd3, together with EcR and Impl2 (Figure 4c). 316

Discussion and conclusions 317
Typically, laboratory studies have revealed a significant "cost of mating" to Drosophila females in the 318 form of reduced longevity. However, here we present that virgin females show a significant decrease 319 in general vigour compared to mated females when uninfected, which is not observed in virgin males. 320 This phenomenon has only been described previously for wild-caught flies [69]. Previous studies 321 described that both males and females reduce their resistance to some infections after mating 322 [45,50,70]. In contrast, data presented herein shows that the opposite happens in both sexes during 323 infection with M. marinum. Males increase their resistance to the infection after mating when they are 324 together with females, although at the expense of reducing their fitness, while females increase their 325 resistance to M. marinum infection after mating, independently of the presence or absence of males. 326 Overall, phenotypic results from this study show that the sexual dimorphism observed in D. 327 melanogaster in the outcome of M. marinum infection is highly related to the reproductive status of 328 the host. This probes that when assessing the differences in the tolerance and/or resistance of D. 329 melanogaster to other infections, the hosts' reproductive situation needs to be characterised. 330 Regarding the hormonal levels and their relation with the immune response, our data show that certain 331 EcR levels are required for the flies after the infection, as flies with higher basal levels presented lower 332 induction or repression of the gene post-infection. However, this correlation was only statistically 333 significant in males. Previous studies showed that ecdysone induces the expression of the receptor 334 PGRP-LC and, thus, modulates the Imd pathway [39]. Other studies had related the Imd pathway in 335 the control of resistance to infections [71], while its negative regulation mediates tolerance to infections 336 [72], although any of these were performed in mycobacterial infections. Our data support these 337 9 findings, as those flies with higher basal levels of EcR show higher expression levels of Diptericin 338 when infected and correlate with the more resistant phenotypes, while flies that did not show early 339 production of this AMP were more tolerant to the infection. 340 Several studies performed with a wide range of pathogens have described the Toll pathway as key in 341 determining resistance to infections [32,73,74], and even to be essential for resistance in males, but not 342 in females [29]. However, no role in increasing tolerance to infection has been described for this 343 pathway previously. In our study, we showed that production of the its production late in the infection correlate with increased resistance, probably due to its role in 356 increasing the metabolism of phagocytic cells [23]. In addition, our data also suggest that the increased 357 resistance in females alone might be driven by Upd3 rather than by innate immune pathways. 358 Finally, the expression of Impl2 during M. marinum infections seems to be only relevant in males that 359 are not actively mating and to be related to increased bacillary loads, although further studies should 360 be performed to determine its role during mycobacterial infections in D. melanogaster. 361

Concluding remarks 362
It has been extensively studied that males and females respond differently to the same infections. Many 363 of these studies have focused on establishing the physiological basis for these differences, but very few 364 have studied how the reproductive status of the host affects males and females and the role it plays in 365 the response to infection. This study aimed to show that the intrinsic differences observed between 366 males and females during the infection by M. marinum were tightly related to the reproductive status 367 of the host and that reproduction affects males and females differently and that when assessing the 368 sexual dimorphism of D. melanogaster, the hosts' reproductive situation need to be characterised. 369 Here we show that being actively mating (kept together) increased the resistance but reduced the 370 tolerance to the infection in males and females. The same phenotype was found in mated females 371 regardless if they were kept alone or together after mating. 372 The results obtained in this study also suggest a possible role for the Toll pathway in determining 373 tolerance to infection, while it appears that the IMD pathway is associated with increased resistance. 374 Results also showed a correlation between basal levels of EcR and its induction after the infection, 375 suggesting that a certain amount of this receptor is required upon infection. This phenomenon was 376 observed in both males and females in all reproductive statuses, although only in males was statistically 377 significant. In addition, basal levels of EcR also correlated with higher expression levels of Diptericin. 378 10 Finally, we have also proposed a dual role for Upd3 upon infection with M. marinum. Our results show 379 that an early production of this molecule correlate with higher bacillary loads. This might be linked 380 with the Upd3-mediated lipid droplets accumulation inside phagocytic cells, which favours 381 mycobacterial replication. However, the late production of this molecule is associated with lower 382 bacillary loads, most likely due to its role in increasing phagocytic cells metabolism. Finally, we could 383 only validate the role for Impl2 in increasing the tolerance in non-virgin non-actively mating males. 384  Figure 1| Tolerance and resistance of D. melanogaster to M. marinum infections depending on the reproductive status or the sex of the flies. (a) Males that were not in presence of females (virgins and alone) were more tolerant but less resistant than those kept together with females. (b) In females, virgins were more tolerant but less resistant than mated females, independently of the presence or absence of males. When comparing both sexes for each reproductive status, we observed that males were more resistant when virgins or together with females, although no differences in tolerance were found in these groups (c and e). When flies were kept alone after infection, females showed higher general vigour and resistance, but lower tolerance to infection (d). Lines represent the regression lines fitted for each group and each circle represents an individual. Both survival and bacillary load between the groups were analysed independently for each inoculation dose and were tested for normality. Statistically significant differences were represented as follow: *p≤0.05, **p≤0.01, ***p≤0.001, ****p≤0.0001 (Kruskal-Wallis test).

Figure 2|
Expression of innate immunity (a) and metabolic (b) genes during the infection in males and females depending on their reproductive status. Gene expression relative to the internal control gene rpl32 was quantified in 9 replicate pools of 3 males or females exposed to the infection with M. marinum relative to their expression in uninfected controls. Each group at each time-point was compared to its relative control independently (the line in each graph represents the controls' relative expression with a log2 fold change of 0). Data was analysed for normality and significant differences were represented as follow: *p≤0.05, **p≤0.01 (Welch's corrections for normally distributed data and Mann-Whitney test for not normally distributed data).

Figure 3|
Expression of the ecdysone receptor (EcR) in males (up) and females (down) depending on their reproductive status after the infection (a). Gene expression relative to the internal control gene rpl32 was quantified in 9 replicate pools of 3 males and 3 females each exposed to the infection with M. marinum relative to their expression in uninfected controls. Each group at each time-point was compared to its relative control independently (the line in each graph represents the controls' relative expression with a log2 fold change of 0). Data was analysed for normality and significant differences were represented as follow: *p≤0.05, **p≤0.01 (Welch's corrections for normally distributed data and Mann-Whitney test for not normally distributed data). Basal expression levels of the EcR gene in uninfected flies (b) were calculated using the 2 -ΔCT method with the rpl32 gene for normalization (all values were multiplied by 10 4 for more visual results). Groups were compared independently for each time-point. Data was analysed for normality and significant differences were represented as follow: *p≤0.05, **p≤0.01 (Kruskal-Wallis test). Finally, the correlation between the basal levels and the fold change after infection was performed using the nonparametric spearman correlation test (c).