Reciprocal costs of infection and reproduction in D. melanogaster

Trade-offs occur when an organism has to allocate limited resources into multiple biological processes. How organisms allocate their resources and whether one trait gets priority over another is poorly understood. Prior work has shown that reproductive investment reduces the capacity of D. melanogaster to mount an effective immune response against subsequent bacterial infection. However, it has not been tested whether the observed trade-off was unidirectional with reproductive fitness given primacy over immunity, or whether it might also occur in the reciprocal direction with an active prior immune response reducing reproductive output. In the present work, we delivered bacterial infection to female D. melanogaster prior to mating and tested whether reproductive capacity became reduced. We found that infected females produced the same number of eggs as uninfected females, but the eggs from infected females exhibited lower survivorship to adulthood. Additionally, we found that mating destabilizes chronic bacterial infections, stimulating additional host death and increasing variance in pathogen burden. Together, our results suggest the cost of reproduction and infection in Drosophila females is reciprocal, regardless of the order in which they occur.

Methods
Group of 10 male and 10 female Drosophila melanogaster flies from Canton S (CS) strain were reared at room temperature (22-23°C) with 12h:12h light:dark cycle, and allowed to lay eggs. Adult offspring were sorted by sex within 8 hours of emergence from the pupal case and were maintained separately before mating. The female offspring were divided into different treatment and control groups (listed below) and were then subjected to an infection with a Gram-negative bacterium Providencia rettgeri at 5 to 7 days after eclosion:

Tretament groups:

i) M24I,  flies allowed to mate 24 hours prior to infection

ii) I0M,  flies infected and immediately mated

iii) I24M,  flies were infected and mated 24 hours later

iv) I120M,  flies infected then mated 5 days later

Control groups:

 i) M24- , mated, held for 24 hours, then given sterile wound to act as controls for the M24I files

 ii) -0M, given a sterile wound and immediately mated as a control for the I0M and I24M flies

 iii) M0, females mated on the day of infection but were not wounded as a control for the I0M

 iv) -120M, given a sterile wound and then mated 5 days later as a control for the I120M flies

 v) unmated virgin flies given a sterile wound.

Survival assay

Survival assay was carried out in three different experimental blocks with 5-7 vials per treatment per block. Each vial contained 5 females. All treatment and control groups listed above were used for survival assay. Survival of the female flies was observed over 10 days post infection. The flies that survived at the end of the experiment were censored. Survival analysis was carried out using the 'survival' package (v3.2-11) in R, for which a mixed effect model with treatment as fixed effect and experimental block as random effect. In order to compare survival between treatment groups, same Cox proportional hazard model within the emmeans() function from the package emmeans (v1.8.3) was used. 

Egg and offspring count assay

Egg and offspring count was carried out for all treatment and control conditions except foe unmated virgin flies across three experimental blocks. For which, 5 female flies were housed with 8 males in order to ensire mating and allowed to lay eggs. Females were transferred to new food vials every day and eggs were counted manually under a light microscope immediately. The vials were left at room temperature until adult offspring emerged. Adult offspring were transferred to empty vials 24 hours after the emergence of first adult offspring and frozen until they were counted. To estimate per-fly egg and offspring count, the total number of eggs and offsrping were divided by the total number of females in each treatment group within each experimental block. Egg-to-offspring ratio was calculated by dividing the total number of offspring by the total number of eggs in each experimental block. Two mixed effect models one with experimental block as a random effect and another one without experimental block were compared to test the effect of block using lme4 package in R. Multiple pairwise tests were then carried out with Bonferroni correction to evlauate paiwise differences between treatments.

Bacterial load assay

Bacterial load assay was carried out to test whether mating has any impact on previously established pathogen burden. Individual females were anesthetized with CO2 and homogenized in 500 μl sterile 1X phosphate buffered saline (PBS). Fifty microliters of the homogenate were plated on an LB plate using a WASP2 spiral plater (Microbiology International). The plates were incubated overnight at 37°C and the bacterial colonies that developed were counted using the ProtoCOL plate counting system (Microbiology International). There were 13 plates that had colony densities too high to be resolved by the counting software. Those plates were assigned the highest measurable value (112500 colony forming units (CFU) per fly) observed across all three experimental blocks. All of the analysis were carried out in R. CFU per fly was first calculated and then natural log transformed for subsequent analysis. Since some of the CFU per fly values were zero, 1 was added to all the values prior to the log transformation. Two mixed effect models with and without experimental block were compared to test the effect of experimental block using lme4 package in R. Both models had mating status as fixed effect  Further, all three blocks were pooled together and Levene's test was used to compare the variance between treatments. Because treatments did not have equal variances, Welch’s t-test was performed to compare between means of mated and unmated treatments. Similarly, Fisher’s exact test was used to compare between the number of dead individuals between two treatment groups.

当生物体需要将有限资源分配至多项生命活动时，便会出现权衡现象。目前学界对于生物体如何分配资源，以及某一性状是否优先于另一性状，仍知之甚少。既往研究表明，生殖投入会降低黑腹果蝇（Drosophila melanogaster）对后续细菌感染产生有效免疫应答的能力。然而，此前尚未验证这一观察到的权衡是单向的——即生殖适合度优先于免疫，还是先前的主动免疫应答降低生殖产出，从而在反向也存在这种权衡。在本研究中，我们在黑腹果蝇雌蝇交配前对其进行细菌感染，以此检测其生殖能力是否下降。我们发现，感染组雌蝇的产卵量与未感染组并无差异，但感染组雌蝇所产的卵发育至成虫的存活率更低。此外，我们还发现交配会使慢性细菌感染状态失衡，导致宿主额外死亡，并增加病原体负荷的变异程度。综上，我们的研究结果表明，果蝇雌蝇的生殖与感染成本是双向的，与二者发生的先后顺序无关。

方法
将Canton S（CS）品系的10只雄果蝇与10只雌果蝇置于室温（22~23℃）、12h光照:12h黑暗的环境中饲养，使其产卵。成虫后代在蛹壳羽化后8小时内按性别分拣，并在交配前单独饲养。将雌性后代分为不同处理组与对照组（详见下文），并在羽化后5~7天时用革兰氏阴性菌雷氏普罗威登斯菌（Providencia rettgeri）进行感染：

处理组：
i) M24I组：交配24小时后进行感染的果蝇
ii) I0M组：感染后立即进行交配的果蝇
iii) I24M组：感染24小时后进行交配的果蝇
iv) I120M组：感染5天后进行交配的果蝇

对照组：
i) M24-组：交配后静置24小时，随后进行无菌伤口处理，作为M24I组的对照
ii) -0M组：接受无菌伤口处理后立即交配，作为I0M与I24M组的对照
iii) M0组：在感染当日进行交配但未造成伤口的果蝇，作为I0M组的对照
iv) -120M组：接受无菌伤口处理后5天进行交配，作为I120M组的对照
v) 未交配处女蝇组：接受无菌伤口处理的处女蝇

生存测定
生存测定在3个独立实验批次中开展，每个处理组在每个批次中设置5~7个培养管。每管放置5只雌果蝇。上述所有处理组与对照组均参与生存测定。我们对感染后10天内的雌果蝇存活率进行观测，实验结束时仍存活的果蝇视为截尾数据。生存分析采用R语言中的survival包（版本3.2-11）完成，该分析以处理方式为固定效应、实验批次为随机效应构建混合效应模型。为比较不同处理组间的存活率差异，我们采用emmeans包（版本1.8.3）中的emmeans()函数构建相同的Cox比例风险模型。

产卵与后代计数测定
除未交配处女蝇组外，我们在3个实验批次中对所有处理组与对照组开展产卵与后代计数实验。将5只雌果蝇与8只雄果蝇共同饲养以确保交配成功，随后使其产卵。每日将雌果蝇转移至新的食物管中，并立即在光学显微镜下手动计数产卵量。将食物管置于室温下直至成虫后代羽化。首批成虫羽化后24小时，将所有成虫转移至空管中并冷冻保存，待后续计数。为计算每只雌蝇的产卵量与后代数，我们将每个实验批次中各处理组的总产卵量与总后代数除以该组的雌蝇总数。卵-成虫转化率通过每个实验批次中的总后代数除以总产卵量计算得到。我们采用R语言的lme4包，通过比较包含与不包含实验批次作为随机效应的两个混合效应模型，检验实验批次的影响。随后采用Bonferroni校正进行多重两两比较，以评估处理组间的差异。

细菌载量测定
细菌载量测定用于检测交配对已建立的病原体负荷是否存在影响。将单只雌果蝇用二氧化碳麻醉后，在500μl无菌1×磷酸盐缓冲液（phosphate buffered saline, PBS）中匀浆。取50μl匀浆液使用WASP2螺旋涂布器（Microbiology International公司）涂布于LB平板上。将平板置于37℃下过夜培养，随后使用ProtoCOL菌落计数系统（Microbiology International公司）统计生长出的细菌菌落数。有13个平板的菌落密度过高，无法通过计数软件识别，我们将这些平板的菌落数赋值为所有3个实验批次中观测到的最高可测量值：每只果蝇112500个菌落形成单位（colony forming unit, CFU）。所有分析均在R语言中完成。首先计算每只果蝇的CFU数，随后进行自然对数转换以用于后续分析。由于部分CFU值为0，我们在对数转换前将所有数值加1。我们采用lme4包，通过比较包含与不包含实验批次的两个混合效应模型检验实验批次的影响，两个模型均以交配状态为固定效应。进一步将3个批次的数据合并，采用Levene检验比较不同处理组间的变异程度。由于各组方差不齐，我们采用Welch t检验比较交配组与未交配组的均值差异。类似地，我们采用Fisher精确检验比较两个处理组间的死亡个体数。