Role of Critical Size on Adult life History Traits in Drosophila Melanogaster.

Silver-spoon hypothesis suggests that tness of individuals is high under good adult conditions provided their development itself has been in good conditions and those who have grown in resource-poor conditions are at a permanent disadvantage. Using two types of Drosophila melanogaster populations grown under two conditions we tested the validity of silver-spoon hypothesis. Three populations were selected for faster pre-adult development as a result of which they had access to food for a shorter duration while the three control populations had access to food for longer duration as growing larvae. In the second set-up the access to food was curtailed immediately on attainment of critical size. We assessed biomolecule levels, copulation latency, copulation duration, life-time realized oviposition and longevity to validate the silver-spoon hypothesis.


Background
Early life nourishment is known to affect the adult physiology and various life-history traits in Drosophila melanogaster [1][2][3] . For example, nutrient acquisition in the form of dietary sugars (source of C) during the larval stage has contribution in egg provisioning during the early adult life of holometabolous females 4 .
In D. melanogaster, early life is divided into pre-adult duration consisting of mobile and voraciously feeding larval phase and non-mobile pupal phase. Further, the larval phase is marked into two stages-(i) pre-critical stage and post-critical (terminal growth period) stage, separated by critical size time point 5,6 .
Larval critical size commit the larva to an irreversible process of metamorphosis, and starvation post attainment of critical size does not affect time course to undergo metamorphosis 5,7 . The role of critical size as "physiological switch" is well established and is marked by ecdysone pulse 7 . Recently, critical size is reported to also act as "Energy allocation switch" in various species of Drosophila 8,9 . In another dipteran species, Aedes egypti, it has been previously reported that threshold amount of energy reserves are pre-requisite to the process of metamorphosis in addition to ecdysone level during last larval instar, thus implying signi cance of energy budgeting in the dipteran species 10 . 'Developmental threshold model' states that there is always a minimum size or condition that must be surpassed before the life-history transition occurs in a wide variety of species undergoing metamorphosis 11,12 . While larger threshold causes a negative relationship between age of transition and growth conditions, the smaller threshold would result in a comparatively positive relationship for higher growth rate. In fast-developing individuals, once this threshold is passed then the excess of energy or resource is translated to overhead threshold and invested in fecundity 11 .
In Drosophila melanogaster, this minimum developmental threshold is represented by critical size/time point 5,6,13 . Being an inhabitant of rotting fruits and vegetables, it is under direct selection for faster preadult development due to over-crowding and food limitation. Selection for faster pre-adult development is known to exhibit reduced development time and subsequently results in the smaller adult body size [13][14][15][16][17][18] .
In D. melanogaster, environmental conditions during larval life are suggested to affect its adult size 19 and their physiology 19,20 which in turn affect various life-history traits 1,3,[21][22][23][24] . Further, if the larval nutritional environment or developmental diet is rich then they tend to emerge with larger adult body size and attain reproductive maturity at an early age that has a positive implication on tness 11,25,26 . In general though not universal 18 "bigger is better" idea prevails with larvae spending more time in weight (equal to resource/energy) gain under good nourishment conditions, eventually emerging as larger adults with higher tness. Also 'silver spoon hypothesis 25 ' posits that individuals born in good conditions have tness or performance advantages in later life with many examples 25 and those born in poor conditions are at a permanent disadvantage 27 . Varying the quality or quantity of diet during developmental in D. melanogaster affects adult body size and its associated life history hence it is known to serve as a tractable model to study dietary manipulation during pre-adult and adult stages 24,28,29 .
The inverse relationship between time to maturity (metamorphosis) and size is expected in species that live in ephemeral habitat 12 . A physiological change from resource-dependent (pre-critical) to resource independent (post-critical duration) rate of development occurs at critical size or threshold time point 7 . Further, the energetically costly process of reproduction is favoured over survival under rich (nutritional) environment than in dietary restriction condition where longevity (somatic maintenance) is favoured 30-32 . However, dietary restriction (DR) is also known to result in short term adaptive response where in females that were deprived of yeast diet as larvae had reduced fecundity but mortality rate was unaltered 24 suggesting that longevity-fecundity trade-off may not be universal.
In the present study, we tested the applicability of 'silver spoon hypothesis' 27, 33 using six outbred populations of D. melanogaster. Three of the six populations were under conscious selection for faster pre-adult development and extended reproductive longevity while the other three were their ancestral control populations. The selected populations were internally driven to stop feeding in about 17-hour post attainment of critical size thus curtail food intake, while the control populations fed for 42-hours post attainment of critical size. Critical size time is an important switch point that commits the organism to irreversible metamorphosis process 5,7 . Further, we also tested the hypothesis under food curtailment post attainment of critical size in both the control and selected populations. The realized life-time oviposition of the populations selected for faster pre-adult development thus under food restriction was not signi cantly different from those of the control populations, refuting the silver spoon hypothesis.

Longevity and Survival Probability
There is no signi cant effect of selection and diet curtailing post attainment of critical size on average longevity of both male and female ies. Overall CS females had signi cantly higher median longevity than NS ies. However, there was no signi cant effect of selection on median longevity. The maximum lifespan of critical-sized males was signi cantly altered (Additional Table S1).
Average, median and maximum longevity are descriptive statistics that can at best provide overall comparison but not compare the progression of the biological process of aging 34 . We compared the survival probabilities of the different type of ies to ascertain the progress of the biological process.
There was a signi cant effect of selection and y type on survival probabilities of both male and female ies (see Table 1, Fig. 2c and 2d).  ies.

Discussion
Recently, it has been reported that glycogen metabolism is required during third larval instar for normal body size growth and even the developmental delay would not rescue the arrest of body size due to reduced glycogen levels in the larval stage 35 . Further, it has been evidenced that defects in glycogen metabolism are known to affect larval physiology and hence the adult tness 35 . While the fat body, muscles and CNS (Central Nervous System) in larva act as the site for glycogen storage 36 , glycogen synthesis occurs during late larval life in fat body 37 lending support to our interpretation. Furthermore, fat bodies act as a peripheral system for ecdysone metabolism 38 . Despite smaller larval size than controls 39 , the selected populations had comparable levels of glycogen (Fig. 1c) throughout the third instar suggesting that glycogen is the primary source of energy that is possibly driving the physiological processes leading to early expression and release of ecdysone in selected populations 39 that might, in turn, facilitate faster development.
When we compared the trend of lipid content during larval life, it is comparable at rst (48 h posthatching of synchronized eggs), 2nd (early L3) and 3rd larval time point (Critical size time pointdevelopmental threshold/metamorphosis commitment time point) suggesting the utilization of lipids for metabolic processes during these periods. The lower level of lipid post-critical size attainment at 4th and 5th larval time point is perhaps due to decrease in the post-critical growth duration of the selected populations 13 . It is known that lipids are majorly stored in fat body to be utilized further during the pupal duration and early adult life 23  The major macromolecules and energy results indicate that our selected populations have evolved mechanisms to maintain the requisite levels of molecules and energy till attainment of critical size-a time point at which the organisms commit to an irreversible process of metamorphosis. This is in agreement with Hironaka et al. who proposed that the investment value in larval tissue is more during the pre-critical period than terminal growth period (a.k.a. post-critical duration) 8 . Further, supporting the view that critical size in Drosophila melanogaster is acting as an optimal switch for energy allocation during larval life 8,9 . Furthermore, following the model for developmental threshold-with the age of maturity (metamorphosis-in the present context) and size at transition, it is likely that selection for faster pre-adult development evolved smaller developmental threshold (Critical size-in this case) through higher growth rate during second larval instar 13 thus accumulating su cient energy reserves to sustain the development without compromising their metamorphosis duration 11,13 .
Owing to short post-critical duration-a period during which most of the growth and weight increase occurs 13 the selected populations have signi cantly reduced protein and lipid content during the mid and late-L3 stages. Although the selected populations are under the physiological trigger to complete development due to higher ecdysone levels throughout L3 38 they can accumulate adequate proteins and lipids to maintain their phenotypic integrity albeit emerge as small adults 13,38 . Despite small size due to curtailed post-critical duration their copulation latency (i.e., time to attain sexual maturity) and copulation duration were unaffected supporting the claim that sexual maturity and copulation duration are speciesspeci c traits 18 . However, copulation latency and duration were affected when Drosophila melanogaster was starved during early adult life 41 . The differences in the two ndings might be due to diet restriction at different life stages ascertained in the two studies. Two studies in the past have reported decoupling of body size and copulation duration 18,42 .
A developmental dietary history is known to in uence adult physiology 1 . In general, the trade-off between longevity and lifetime oviposition are well documented in Drosophila melanogaster 28,[44][45] and other holometabolous insects like Speyeria mormonia 46 . Thus following various studies on dietary restriction, CS ies had lower oviposition (Fig. 2b). However, exception to larval dietary manipulation has also been reported. For example in holometabolous Lepidopteran butter y, Speyeria mormonia, there was no independent effect of semi-starvation on realized egg laying 46 suggesting an indirect effect of larval dietary restriction on oviposition. In our study, conscious selection for faster pre-adult development that resulted in small-sized adults ( Fig. 2e and 2f) did not affect the life-time realized oviposition (Fig. 2b) although the selected ies had signi cantly smaller ovaries 18 . Despite being small, they had realized oviposition comparable to their ancestral controls, suggesting that they Taken together, tness (measured as longevity and life-time realized oviposition) of our selected populations are comparable to their ancestral control thus refuting the widely accepted 'bigger is better' hypothesis 49,50 as well as the silver spoon effect 27,33 .

Conclusion
Overall, our study provides insight into the role of (gradually evolved) critical size on adult life-history in Drosophila melanogaster populations. The populations that are under conscious selection for faster preadult development and thus under curtailed food intake have not compromised on their Darwinian tness thus refuting the silver spoon hypothesis 27 .

Fly husbandry
Two kinds of laboratory Drosophila melanogaster populations were used in this study. The Control (C) populations were on 21 days egg to egg discrete generation cycle, while the Selected (S) populations were derived from the controls by direct selection for faster pre-adult development and indirect selection for extended longevity. Detailed protocols adopted in rearing and maintenance of C and S populations are explained previously 13 . Brie y, each of the three C populations were cultured in 40 vials with 6 ml standard banana-jaggery media (SM) at a density of 40-50 eggs per vial and incubated at SLC for full 12 days 13 . At the end of 12 days, all emerging ies from the 40 vials were transferred to pre-labeled plexiglass population cages and provided with fresh food every alternate day till day 18. On day 18, fresh food plate was supplemented with live yeast-acetic acid paste. Eggs for initiating the next generation were collected on day 21 from the previous egg collection day. Each of the three S populations was derived from the three C populations by collecting 160 vials of 60-80 eggs per 6 mL banana-jaggery food vial. The vials were incubated at SLC. The early emerging 15-20 ies (as ascertained by empty pupal cases) from each vial were transferred to pre-labeled population cages. Two sister cages were maintained per S population to avoid adult crowding. The S population cages too were provided with fresh food plates every alternate day till 50% mortality was noticed in any of the cages, at which point all cages were provided fresh food plates supplemented with live yeast-acetic acid paste for three days following which eggs for starting next generation were collected. The eggs obtained from the two sister cages of a given population were mixed and redistributed into 160 vials.
Originally, the Control populations (also called as JB populations) were derived from IV populations 51 and are described in detail in Prasad et al. 17 . The selected and control populations had been through 134 and 242 generations respectively at the time of being used in this study. To remove non-genetic parental effects which might appear due to the differences in maintenance regime, both the S and C populations were run through common rearing conditions for 1 generation at a moderate density of 50 eggs per 6 mL banana-jaggery media vial and 40 vials per population before being used in this study. The egg collection from the S and C populations were staggered by their developmental time difference to synchronize the emergence of adults. All adults emerging from each of the 40 vials of a given population were transferred to pre-labeled population cage with fresh SM plate. These populations are referred to as standardized ies 13,17 . Embryos for all the experiments were obtained from these standardized ies (SF).
Generation of Critical size adults and Normal-sized adults Synchronized eggs were collected from SF, evenly spread on agar-agar plates and incubated at SLC. Freshly hatched larvae (~22 h post-egg-laying) were harvested using a ne camel hair brush and transferred to Petri-plates (5.5 mm diameter, Tarson) with 2000 µL of LSM (Liquid Standard Media) at a density of 30 larvae per plate 13 . Twenty such plates per population were incubated at SLC. The larvae from all the 20 plates were re-harvested from LSM plates after 64 and 72 h (post-egg-lay) for S and C populations respectively, washed with RO (Reverse osmosis) water, rolled on tissue towel and randomly transferred (25 larvae per vial) to vials containing 6 mL non-nutritive agar-agar or SM and incubated at SLC 13 . At every 6 hour interval, the emerging ies from these vials were collected, sorted according to their gender and held as virgins in pre-labeled holding vials with 6mL SM till use in further assays. The adults that emerged from vials containing non-nutritive agar are referred to as critical-sized (CS) adults, while those that emerged from vials containing SM are referred to as normal-sized (NS) adults.

Macromolecule quanti cation during larval life
Glycogen and lipid content were estimated using Van Handel's method 52  1. Glycogen estimation Five randomly chosen larvae were homogenized in 400 µL of 2% Na 2 SO 4 . 80 µL of homogenate was aliquoted into 5 mL Eppendorf tube, to which 184 µL of Na 2 SO 4 and 3736 µL of the (fresh) mixture of chloroform and methanol (1:1) was added. The tubes with the mix were centrifuged (Eppendorf, 5430R) at 14000 r.p.m. for 10 minutes at 4 °C. The supernatant was discarded and the pellet was air-dried for 10 minutes. The pellet was resuspended in 2000 µL Anthrone reagent and heated at 90 °C in water-bath for 10 min. Aliquots were kept on ice for 5 min following which absorbance was measured at 625 nm on ELISA plate reader (ECIL micro scan, MS5605A). The assay was repeated in triplicate per population and means of the triplicate measures were used in statistical analysis.

Protein estimation
Precipitation assay was done before quanti cation of protein followed by BCA method of protein quanti cation 53 . Five randomly chosen larvae were homogenized in 400 µL of 2% Na 2 SO 4 . Then 80 µL of homogenate was aliquoted and 500 µL of 0.15% Deoxycholate was added to the aliquot. After the incubation period of 10 min on ice, 1000 µl 3M Trichloroacetic acid (TCA) was added. The aliquots were centrifuged at 8500 r.p.m. (Eppendorf, 5430R) for 15 min at 4 °C. Protein was precipitated at the base of each aliquot. Pellets were washed with HCl and air-dried. BCA reagent was added to each of the pellets, resuspended and heated in water-bath at 60 °C for 10 min. Absorbance was recorded at 562 nm using ELISA plate reader (ECIL micro scan, MS5605A). The assay was repeated in triplicate per population and means of the triplicates were used in statistical analysis. for 10 min, cooled on ice for 5-7min., following which 1000 µL of Vanillin reagent was added and incubated at room temperature for 30 min. Absorbance was taken at 525 nm using ELISA plate reader (ECIL micro scan, MS5605A). The assay was repeated in triplicates per population and means of triplicate were used in statistical analysis.

Lipid estimation
All the biomolecules converted to their energy equivalents 54 and compared.
Adult life-history traits 1. Copulation latency and copulation duration Copulation latency (the time lag between the time of emergence and initiation of copulation) and copulation duration (the time difference between initiation and termination of copulation) were ascertained by pairing freshly emerged ies Zero-day ies with 3-day old mature ies of the opposite gender 2. Life-time oviposition and longevity One day old virgins from holding vials were used in this assay. The ies were anaesthetized using CO 2 and a female and male pair were transferred to fresh vials with 3 mL SM. A total of 20 pairs per treatment were set up. Any y that did not wake up within 1 h of transfer was replaced with a new y of the same gender. Further, any y that died within the rst 24 h of set up was also replaced by a fresh y. The pair of ies were transferred to fresh 3 mL SM vials every 24 hours. The eggs laid in the preceding 24 h were counted under stereo zoom microscope (Carl Zeiss Binocular stereozoom microscope, Stemi 305) and recorded. Census records were also maintained till the death of all ies. The average life-time oviposition and life-span were estimated from this primary data (Fig. 2a).

Data analysis
In all cases except survival probability function, univariate analysis of variance, under general linear model (GLM) using SPSS v. 22 was carried out and population means were used as units of analysis with selection, larval growth stage and y type as xed variables and replication as a random variable 55-57 . Hence, only xed-factor effects and interactions could be tested for signi cance 13,17 . The signi cance of adult survival probability curves was analyzed using Kaplan-Meier log-rank test 58