Embryonic expression of shuttle craft, a Drosophila gene involved in neuron development, is associated with adult lifespan

Reversions

A control line (control) with the genotype w1118 and a line with the stc mutation, w1118; P{SUPor-P}stcKG01230 (stcP, Figure 1A) were used in this study. Four lines with reversions of the w+ marker phenotype (rev1, rev3, rev4, rev5) were obtained from stcP using standard substitution crosses with balancer chromosomes and the delta 2-3 source of P element transposase [10]. For each line, PCR with primers surrounding the site of the initial P{SUPor-P} insertion were used to assess the nature of reversions. PCR fragment sizes were identical in the control line and in all lines with reversions (Figure 1B), indicating precise excisions of the P{SUPor-P} construct. This was further confirmed by sequencing PCR fragments: all sequences were identical to the standard gene sequence (http://flybase.org). To confirm that excisions did not cause regional aberrations, two additional PCR reactions with primers that amplified approximately 2.5 kb on both sides of the insertion site were made for each line; no deviations from the control line were observed in any line (Figures 1C, D). Thus, in all four lines, the vector construct was precisely excised from the insertion site and complete restoration of the original gene structure was achieved. Negative results were obtained for all six lines (control, stcP, rev1, rev3, rev4, rev5) in a test for the presence of Wolbachia, a Drosophila symbiont known to affect life history traits [11].

Lifespan

We assessed the effect of the stc^KG01230 mutation on lifespan relative to control and reversion lines in unmated and mated females and males. For each experiment, we calculated the following parameters: mean lifespan, median lifespan, minimum and maximum lifespan, lifespan lower and upper quartiles, lifespan of the 10th and 90th percentiles; variance, standard deviation, and standard error for the mean lifespan (Table 1). Figures 2, 3 show survival curves.

A significantly increased lifespan was detected in unmated mutant females compared to control females and to females of reversion lines rev1, rev4 and rev5 (Table 1, Figure 2A). To verify these results, the experiment was repeated twice over two years. In the second and third experiments, the lifespan of mutant females was significantly higher than the lifespan of control females and females of all four reversion lines (Table 1, Figures 2B, 2C). The combined results of the three experiments (Table 1, Figure 2D) clearly demonstrated a difference in lifespan between control and mutant unmated females and between mutant and reversion unmated females. Additional log rank tests confirmed this conclusion (P=0.000001 for each comparison). Survivorships of the four reversion lines were homogenous (P=0.1962, Kruskal-Wallis test). After adjusting for multiple testing, survivorships of rev1 and rev3 lines were not different from survivorship of the control line, whereas survivorships of rev4 and rev5 lines were (P=0.1123, P=0.0491, P=0.0001, P=0.0001, respectively, Kruskal-Wallis test). Consequently, survival curves were not homogenous among control females and females of all reversion lines (P=0.0003, Kruskal-Wallis test), though mean and median lifespans of control females and females of all reversion lines were almost identical and varied between 53 and 55 days (Table 1). The average positive effect of mutation was approximately 10% of the control lifespan. Comparison of survival curves indicated that the mutation slowed aging slightly.

No difference was found between the lifespans of mutant and control males (Table 1, Figure 2E). The lifespan of males in two of four reversion lines, rev1 and rev3, was significantly different from the lifespan of mutant males (Table 1, Figure 2E) and control males according to the Kolmogorov-Smirnov test (P = 0.0063 for rev1; P = 0.0243 for rev3). These differences did not reflect an increase or a decrease in lifespan but fluctuations in the shape of survival curves (Figure 2E). Over the next two years, two other experiments confirmed the lack of difference in lifespan between mutant and control unmated males (Table 1, Figures 2F, 2G).

A significantly decreased lifespan was detected in mated mutant females compared to control females and females of the four reversion lines (Table 1, Figure 3A). To verify these results, the experiment was repeated and the lifespan of mutant females was again shown to be significantly lower than the lifespan of control females and females of all four reversion lines (Table 1, Figures 3B, 3C). Additional log rank tests for the combined results of the two experiments confirmed this conclusion (P=0.000001 for each comparison). Survivorships were homogenous among the four reversion lines (P=0.1198, Kruskal-Wallis test). Survivorships of all reversion lines were not different from survivorship of the control line (P=0.2542, P=0.3366, P=0.6245, P=0.4784, respectively, Kruskal-Wallis test), consequently, survivorships were homogenous among all five lines (P=0.2110, Kruskal-Wallis test). Experiments with mutant and control mated females were repeated four more times over three years. The last experiment was made simultaneously with reproduction assays. In all cases, the lifespan of mutant females was significantly lower than the lifespan of control females (Table 1, Figures 3D-G). In the last experiment, the difference between the lifespan of mutant and control females was greater than in the earlier experiments. Our results demonstrated a difference in lifespan between control and mutant mated females and between mutant and reversion unmated females. The average negative effect of mutation was approximately 15% of the control lifespan. Comparison of survival curves indicated that the mutation slightly accelerated aging.

After adjusting for multiple testing, the lifespan of mutant mated males was not different from the lifespan of control males and males with all four reversions (Table 1, Figure 3H). To verify these results, the experiment was repeated and identical results were obtained (Table 1, Figures 3I, 3J). Three other experiments confirmed the lack of difference between the lifespan of mutant and control mated males (Table 1, Figures 3K, 3L, 3M).

Locomotion

In both mutant and control unmated females, locomotor activity reached a maximum at 20 days of age and then decreased (Figure 4A). No difference in locomotion was detected in 1 day old or 20 day old unmated females, whereas locomotion of 40 day old and 60 day old mutant females was significantly higher than locomotion of 40 day old and 60 day old control females (Table 2, Figure 4A). Locomotion was also measured in 40-day-old unmated females from the rev3 line. No difference was detected compared to controls (Table 2).

Flies naturally tend to move against gravity. We investigated if the stc mutation altered fly geotaxis independently of locomotion per se. Results were similar regardless of whether vials were fixed in horizontal and vertical positions, i.e. horizontal or vertical motion was primarily measured (Figures 4A, 4B). We did not use mechanical stimuli to induce movements, which might explain the lack of difference between the results. Thus, we considered the two variations of the experiment as replicates.

Reproduction

The number of eggs laid by mutant females was significantly lower compared to controls at 3 and 40 days of age. The visual difference in the number of eggs laid by 20 day old females was substantial but not significant. The number of adult progeny was significantly lower from mutant females at 20 and 40 days of age compared to controls (Table 3, Figure 5). Almost no progeny were obtained from any 60 day old females, and the difference between mutant and control females was not detectable. Egg-to-adult viability of progeny from mutant females at 3, 40 and 60 days of age was also lower than controls (Table 3, Figure 5). Overall, reproductive ability appeared to be reduced in mutant females compared to controls.

stc transcript amounts

To understand the molecular basis of differences in lifespan, locomotion and reproduction caused by mutation, we assessed the effect of stc^KG01230 on stc transcript amounts (Table 4). Four annotated and two experimentally confirmed stc transcripts (RA and RB) that differ by 21 nucleotides due to alternative splicing are reported (http://flybase.org). We estimated total stc transcript amounts in 20 day old and 60 day old unmated females and males. No difference was found between control and mutant females (Table 4, Figure 6A) or control and mutant males (Table 4). The same results were obtained when comparing amounts of the sole RB transcript in 20 day old flies (Table 4, Figure 6A). Based on these results and the technical impossibility of measuring the sole RA transcript amount, we measured total stc transcript amounts in all further experiments, which were primarily restricted to females since stc^KG01230 did not affect male lifespan.

stc is predominantly expressed in embryos. According to modENCODE Temporal Expression Data (http://flybase.org), the level of stc mRNA is rather steady throughout embryogenesis, with very high amounts observed at early stages, and high amounts observed up to 18 hours of development. Higher amounts of stc mRNA in early embryos may be explained by the presence of maternally derived transcripts [9]. In preliminary experiments, we evaluated stc transcript amounts in mutant and control 0 to 12 hour and 12 to 18 hour embryos. In both cases, amounts of stc transcripts appeared to be lower in stcP embryos, with the effect being less pronounced at early stages (Table 4). To confirm the difference between the lines, we measured stc transcript amounts in mutant and control 0 to 18 hour embryos. This large interval allowed us to obtain the overall characteristics of control and mutant embryos and to offset the possible uneven contribution of eggs of different stages. Amounts of stc transcripts were significantly lower in mutant embryos compared to controls (Table 4, Figure 6B).

We then measured stc transcript amounts in mutant and control wandering stage III female larvae, 5 to 6 hour old females, 20 day old and 60 day old unmated and mated females, ovaries of 20 day old and 50 to 60 day old unmated and mated females, brain of stage III female larvae, and heads of 5 to 6 hour old females. No differences were observed between control and mutant lines in larvae, 5 to 6 hour old virgin females, and in young or old adult females, both unmated and mated (Figures 6C, 6A, 6D). stc transcripts are abundant in ovaries, presumably because of high demand in embryos [9]. However, no differences were seen in stc transcript amounts between the ovaries of mutant and control females, regardless of age and mating status (6E, 6F). The presence of equal amounts of maternally derived transcripts in control and mutant embryos may mask a difference in embryonic transcription, which explains why stc transcript amounts are less different between control and mutant embryos at early embryonic stages. Considering that stc is required for development of the nervous system, we measured stc transcript amounts in brains of stage III larvae and heads of mutant and control females. No differences were found (Table 4, Figure 6G).

Differences in amounts of stc transcripts were found only in embryos and these differences disappeared as early as in female larvae. We repeated experiments with embryos and larvae and confirmed this result (Table 4, Figure 6H). Finally, we measured stc transcript amounts in embryos of a reversion line rev3and found they were significantly higher than in mutant embryos and did not differ from stc transcript amounts in control embryos (Table 4, Figures 6B, 6H).

Synaptic activity

We investigated whether stc^KG01230 affected neuronal functions after the embryonic stage. No differences were observed in the number of synaptic active zones between female larvae that were mutants (124 ± 8) or controls (142 ± 12) (P = 0.2069 for Kruskal-Wallis test; Figure 7). Substantial differences in synaptic structure were also not detected.

Drosophila lines and crosses

The y¹ w^67c23; P{SUPor-P}stc^KG01230; ry mutant line and the y¹ w^67c23; ry control line were from the Bloomington Stock Center and used in our previous work [5]. Standard substitution crosses with balancers and delta 2-3 source of P element transposase [10] were used to obtain lines with reversions of the stc^KG01230 mutation. Four lines with reversions of the marker w⁺ phenotype were obtained from three males with the active transposase. In crosses, chromosomes 2 from control and mutant lines and chromosomes 2 with reversions were substituted into the X and chromosome 3 genetic background of w¹¹¹⁸ line to reduce the number of unnecessary mutations. In all experiments, flies were kept at 25°C on a standard medium of semolina, sugar, raisins, yeast and agar with nipagin, propionic acid and streptomycin.

PCR and sequencing

DNA was extracted from 20 flies of each genotype using a standard phenol-chloroform method [33]. DNA was used in PCR reactions with primers pstc1 5′-GAACCGTTGCAGTACATTTAAC-3′ and pstc2 5′-GGAACAATCTCGAACTGCCC-3′ (expected product size 555 bp, Fig. 1A, B); pstc0 5′-CTAATTGGAAGGCGGAGCTC-3′ and pstc02 5′-CATTGAGAGTCCGGTGCTGT-3′ (2508 bp, Fig. 1A, C); pstc3 5′-ACACGTGTCTGGAGCTTTTCC-3′ and pstc4 5′-TCCGCTCTGTTACATAGCTGC-3′ (2562 bp, Fig. 1A, D). PCR products were sequenced with Big Dye Terminator V. 3.1. Kit (Applied Biosystems), according to the manufacturer's protocol on a ABI PRIZM 310 Genetic Analyser (Applied Biosystems). Six additional primers for sequencing were: 5′-TCCAACCAGACTGTCAAGTCAAATTAC-3′, 5′-TTCAATTAGCATGATCCAAGG-3′, 5′-AGACGTTGCTCTCGATCAGC-3′, 5′-AGACCACTCCCCGAAAACTG-3′, 5′-ATGTCAGCCCCTGTATGTGC-3′, 5′-AGAATCCAATCAGAGTGCGTC-3′.

Tests for Wolbachia

Wolbachia was detected via real time quantitative PCR with primers to 16S rRNA gene, 5′-CATACCTATTCGAAGGGATAG-3′ and 5′-AGCTTCGAGTGAAACCAATTC-3′ [34].

Lifespan assays

To assess the longevity of unmated flies, five virgin flies of the same genotype and sex, all collected on the same day from cultures with moderate density, were placed in replicate vials. To assess the longevity of mated flies, three virgin males and three virgin females of the same genotype were placed together in replicate vials. Flies were transferred to vials with fresh food containing approximately 5 mL of standard medium without live yeast on the surface weekly (virgin flies) or two times a week (mated flies). Dead flies were recorded daily. Experiments comparing fly lifespans were conducted simultaneously. Sample sizes were 60 to 100 flies/sex/genotype. All experiments were repeated three to six times. Lifespan was estimated for each fly as number of days alive from day of eclosion to day of death. Mean and median lifespan and survival curves were primarily used to characterize lifespan.

Locomotion assays

Flies were collected and maintained as for lifespan assays but without recording deaths. Locomotion was measured in unmated females at age 1, 20, 40 or 60 days at the same time each day. Experiments comparing locomotion were conducted simultaneously. Sample sizes were 30 to 100 flies/genotype/age. One day before measurements, five flies of the same age and genotype were placed in replicate vials. To measure locomotor activity, vials were placed in a Drosophila Population Monitor (TriKinetics), either horizontally or vertically. Fly movement along the walls or in the middle of the vial interrupts infrared beam rings along the length of the vial. Beam interruptions are detected and counted electronically and totals were reported every five minutes to a host computer. Two measurements for five minutes were made for each vial. Locomotion was characterized as mean and median number of beam interruptions per vial.

Fecundity and viability assays

Females aged 3, 20, 40 or 60 days were used. Sample sizes were 79 to 200 females/genotype/age. Fertilized females were placed in replicate vials, allowed to lay eggs for 12 hours and removed. Eggs were counted and transferred to fresh vials for development. Pupa and adult flies were counted in each replicate vial. Mean and median number of eggs and imagos per female were used to characterize fecundity. Mean and median egg-to-pupa and egg-to-imago viability per vial were used to characterize progeny viability.

Larva dissection, immunostaining and microscopy

Third-stage larvae were dissected in phosphate buffered saline (PBS) and fixed in 4% formaldehyde at room temperature for 20 minutes. Preparations were incubated overnight at 4°C with primary antibodies NC82 (DSHB, USA) against Bruchpilot (BRP), a protein specific to active synaptic zones, and then incubated with secondary antibodies labeled by indocarbocyanine (Jackson Immunoresearch, USA) and with anti-HRP (HorseRadish Peroxidase) antibodies with fluorophore Alexa 647 (Jackson Immunoresearch, USA) at room temperature for two hours; placed in a medium for immunofluorescence VectaShield (Vector Labs, USA). Neuromuscular junctions were analyzed using the fourth muscle of the third and fourth abdominal segments of larvae using a confocal laser scanning microscope (LSM 510, Zeiss, Germany) at 63x magnification, 633 and 543 nm wavelength, and confocal slice thickness fixed at 1 micron. ImageJ and LSM Image Browser (Zeiss, Germany) were used to determine the number of synaptic active zones. Sample sizes were 13 to 15 larvae/genotype. Mean number of synaptic active zones was used to characterize synapse activity.

Real-time RT-qPCR

Total RNA for real-time reverse transcription quantitative PCR (RT-qPCR) was extracted from 50 embryos aged 0–18 hours, from 20 female stage III female larvae, from 20 whole bodies of 5 to 6 hour old females, from 20 whole bodies of 20 day old and 60 day old virgin males and virgin and mated females, and from 50 pairs of ovaries of 20 day old and 50 day-old virgin and mated females, from 50 brains of stage III female larvae, from 30 heads of 5 to 6 hour old females using TRIzol reagent (Invitrogen) and DNase I Kit (TURBO DNA-free, Ambion) according to the manufacturers' instructions. Two to ten independent extractions/sex/mating status/genotype/age/tissue were made.

First-strand cDNA was synthesized using SuperScript II Reverse Transcriptase (Invitrogen) with oligo(dT) primers according to the manufacturer's instructions. Amounts of cDNA were determined by RT-qPCR using SYBR Green I/Rox in Chromo4 Real-Time PCR Detector (Bio-Rad).

Gdh and Adh housekeeping genes, characterized by relatively low expression comparable to stc expression were used as reference genes to normalize for differences in total cDNA between samples. Forward and reverse primer sequences were stcRA+RB: stc-rt1 5′-AACAGGCACAGCAACAACAA-3′ and stc-rt 2 5′-CCAGGGAGAAGTTAGTGTAGC-3′ (Figure 1A); stcRB: stc-RB1 5′-GGAGCCTTTGGACTGAACCC-3′ and stc-RB2 5′-ATTCGGAGATTGATGACTCAC-3′ (Figure 1A); Gdh: Gdh1 5′-TATGCCACCGAGCACCAGATTCC-3′ and Gdh2 5′-GGATGCCCTTCACCTTCTGCTTCTT-3′; Adh: Adhd3: 5′-CGGCATCTAAGAAGTGATACTCCCAAAA-3′ and Adhr3: 5′-TGAGTGTGCATCGAATCAGCCTTATT-3′.

MJ Opticon Monitor Analysis Software V. 3.1. 32 (Bio-Rad laboratories Inc., 2004-2005) was used to evaluate C(t) value. Intra-run calibrations were used for each panel of experiments, given the fact that the experiments were conducted for several years and two Bio-Rad PCR Detectors were used. Relative stc mRNA amount was considered as a measure of stc transcription level.

Statistical analyses

To compare control, mutant and revertant lines, the Student's t-test was used for initial analysis of lifespan, locomotion, fecundity, viability, and transcript amount. Nonparametric, distribution-free Kruskal-Wallis test and Kolmogorov-Smirnov tests were used for further comparisons of all traits in lines of different genotypes. Bonferroni and false discovery rate [35] corrections for multiple analyses were used when appropriate.