Introduction

The fruit fly Drosophila melanogaster has been indispensable over the past century in research to unravel the biological principles of ageing. This includes our understanding of ageing as underpinned by specific evolutionarily conserved signalling pathways. For instance, Drosophila research has revealed that healthy lifespan can be extended through multiple interventions, including perturbation of insulin and/or TOR signalling (Clancy et al., 2001; Tatar et al., 2001; Kapahi et al., 2004; Proshkina et al., 2015), restriction of specific dietary components (Partridge, Piper and Mair, 2005; Kapahi, Kaeberlein and Hansen, 2017), and tissue-specific interventions such as enhanced neuronal autophagy (Simonsen et al., 2008; Ulgherait et al., 2014; Bolukbasi et al., 2021) or altered intestinal stem cell proliferation (Biteau, Hochmuth and Jasper, 2008; Rera, Clark and Walker, 2012).

These advances, in large part, have been made possible by an ever-expanding genetic toolbox in Drosophila, which permits fine-tuneable, tissue- and cell-type-specific modulation of gene expression, and thus biological function, with relative ease. This toolkit has largely been built upon the backbone of multi-partite expression systems such as the classical GAL4/GAL80/UAS system, which utilises a transcriptional activator (GAL4) and/or its repressor GAL80 placed under the control of specific promoters to mimic the spatial and temporal expression pattern of those promoters when combined with a complementary responder gene (UAS-transgene) (Brand and Perrimon, 1993). Further iterations of this system, namely the temperature-dependent GAL80^ts system (TARGET) (McGuire et al., 2003) and the drug-inducible GeneSwitch system (Osterwalder et al., 2001), have since added conditional control over transgene expression, for example allowing researchers to bypass developmental effects of genetic perturbations while permitting investigation of their role in ageing via adult-onset manipulation (Sun et al., 2013). However, current tools present some limitations for studying ageing in the fly, due to off-target effects of inducers used. For instance, given the poikilothermic nature of Drosophila, temperature can affect metabolic rates and longevity, which can render use of the TARGET system challenging due to possible confounding effects (Miquel et al., 1976; Mołoń et al., 2020). Regarding the GeneSwitch system (He and Jasper, 2014), drawbacks include safety concerns around the antiprogestogen effects of the inducer compound RU-486, “leaky” transgene expression in the absence of RU-486 (Scialo et al., 2016), variability in expression levels achieved due to dose and age differences (Poirier et al., 2008), and issues regarding the palatability of RU-486 influencing feeding and lifespan in a diet-dependant manner (Yamada et al., 2017). Thus, these systems require careful selection of experimental controls for accurate interpretation of results.

A recently-developed system for conditional GAL4 expression is the auxin-based gene expression system (hereafter referred to as AGES), which relies on spatiotemporal modulation of gene expression via synthetic variants of the phytohormone auxin, typically via K-NAA (1-naphthaleneacetic acid potassium salt) (McClure et al., 2022). In brief, this system utilises GAL80 flanked by fused auxin-inducible degron (AID) tags, which, in the absence of auxin, accumulates in cells leading to inhibition of GAL4 transcriptional activity. After auxin ingestion, GAL80 is polyubiquitinated and targeted for degradation, alleviating GAL4 repression and permitting UAS-transgene expression. Crucially, major benefits include compatibility with vast existing libraries of Drosophila GAL4 lines and auxin’s water solubility, cost effectiveness, and high safety profile.

Numerous studies have now successfully used the AGES system in fields including development (Corke et al., 2025), immune signalling (Kuklinski et al., 2026), and behaviour and metabolism (Richhariya et al., 2023, 2025; Fleck et al., 2024). However, the utility of AGES to study ageing has been less well explored, which we sought to redress in this study. Here we evaluate how transgene induction following auxin feeding varies by sex, chromosomal transgene location, auxin dose, and age. In addition, we explore whether AGES can recapitulate a classical anti-ageing intervention of reducing insulin signalling, examining both lifespan and healthspan parameters in this setting.

Materials and methods

Fly stocks and husbandry

The white Dahomey strain of Drosophila melanogaster, positive for the bacterial cytoplasmic endosymbiont Wolbachia (w^Dah+), was the genetic background for all flies in this study. Wild-type Dahomey (Dah) flies were collected in 1970 in Dahomey (now Benin) and have since been maintained in large population cages with overlapping generations. The white Dahomey (w^Dah) stock was derived by backcrossing the w¹¹¹⁸ mutation into this outbred background. Fly stocks used here were da-GAL4 (FBti0013991) (Ikeya et al., 2009); tubP-TIR1-2A-GAL80.AID [‘AGES’] on the second chromosome (a kind gift of C. McClure, Queens University Belfast); tubP-TIR1-2A-GAL80.AID [‘AGES’] on the third chromosome (BDSC #92470, described in (McClure et al., 2022)); UAS-InR^DN [InR^K1405A] (BDSC #8252); and UAS-mCD8::GFP (BDSC #5130). da-GAL4,UAS-mCD8::GFP and da-GAL4,tubP-TIR1-2A-GAL80.AID [‘AGES’] stocks were recombined from their single constituents within the lab for this study.

All transgenes were backcrossed into the outbred w^Dah+ background for six generations but have since been maintained as stocks and likely possess a degree of genetic background variability. For Figures 5 and 6 (lifespan and stress survival assays), however, parental stocks were refreshed into the w^Dah+ background by performing backcrossing for an additional three generations immediately before experimental use.

Our protocol for experimental cultures is described in (Piper and Partridge, 2016). Briefly, all experimental flies were reared at a standardised larval density. Emerging adults were then collected within a 24-hour window and allowed to mate for 48 hours before being separated into single-sex vials at a density of 15 flies per vial, housed in DrosoFlippers (drosoflipper.com). Both stocks and experimental flies were maintained and experiments conducted at 25°C on a 12h:12h light:dark cycle.

Fly media

All stocks and experimental flies were reared on sugar-yeast-agar (SYA) media containing 10% (w/v) brewer’s yeast (#903312, MP Bio), 5% (w/v) sucrose (#BP22010, Fisher Scientific) and 1.5% (w/v) agar (#A7002, Sigma) supplemented with 0.3% nipagin (w/v) (added as 30mL/L from a 10% (w/v) stock solution in ethanol (#H3647, Sigma), and 0.3% (v/v) propionic acid (#149300025, Fisher Scientific) as preservatives (Bass et al., 2007).

For auxin-based and corresponding control media, K-NAA (#GK2088, Glentham Life Sciences) and/or KCl (#26764.260, VWR Chemicals) were added from 2 M stock solutions in dH2O to standard SYA food, thoroughly mixed, and dispensed into individual fly vials to achieve the desired final concentrations as specified for each experiment. For experiments pertaining to Figures 1–3, auxin-based medium was supplemented with KCl, such that the final concentration of potassium across all conditions was 20mM. For experiments related to Figures 4–6, KCl was added to control media only, such that the resultant potassium concentration was 5mM in both control and auxin conditions.

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Whole-Fly GFP Quantification assays

For all GFP-based plate reader assays, flies were maintained on auxin or control food at varying concentrations and starting ages for 7 days as specified for each experiment. Flies (3 per replicate) were then homogenised in Eppendorf tubes containing 240 µL of 1X RIPA lysis buffer (diluted from a 10X stock (#20-188, Millipore) plus protease inhibitors (added from a 7X solution (#11836170001, Roche). Whole-fly homogenates were obtained by motorised homogenisation followed by centrifuging each sample at 4°C and 20,800 × g for 2 minutes.

80 µL of the resulting supernatant was added from each sample to a black, flat bottom 96 well plate (#237105, Fisher Scientific). Standard curves were generated by serially diluting the pooled homogenate of 10 positive control flies (ubiquitously expressing mCD8::GFP) in duplicate. Full plate adjustment was used to optimise the gain setting per run, and fluorescence intensity was measured (488 nm excitation; and 505 nm emission) in a Thermo MultiSkan FC plate reader. Fluorescence intensity values for each sample were then quantified relative to the standard curve values.

Triacylglycerol (TAG) and BCA assays

For all TAG assays, flies were maintained on auxin or control food at varying concentrations from day 2 of adulthood for 7 days. Female flies were then divided into groups of 5 and weighed on a Thermo microbalance to an accuracy of .01 mg. Weighed females were then added to 2mL screw-cap tubes containing 1.4mm ceramic beads (#19-645-3, Cole-Parmer) with 125 µL of TET buffer (10mM Tris (#T6066, Sigma), 1mM EDTA (#10526383, Fisher Scientific), 0.1% (v/v) Triton-X-100 (#648466, Millipore)). Samples were homogenised via an Omni Bead Ruptor Elite bead mill homogeniser at 6.45 (m/s) for one minute and then immediately incubated at 72°C for 15 min in a heat block to inactivate lipases/other degradative enzymes. Sample tubes were then centrifuged at 4°C and 20,800 × g for 5 minutes before adding 2 µL of the supernatant from each sample or glycerol stock (1-0 µg/µL range) to a clear, flat bottom 96-well plate (Corning Life Sciences). 300 µL of TAG reagent (full constituents listed in Supplementary Table 1) was added to each well and absorbance measured at 540 nm in a Thermo MultiSkan FC plate reader. Triacylglycerol levels for each sample were then quantified relative to the standard curve values.

The BCA Protein Assay Kit (#23227, Pierce) was used to assay protein content from frozen sample homogenates as per the manufacturer’s instructions.

Fly tissue dissections and imaging

Adult brains were dissected in PBT (1X phosphate-buffered saline (PBS) containing 0.5% Triton X-100) and fixed in 4% paraformaldehyde (PFA) in PBT for 20 minutes at room temperature. After fixation, samples were washed with PBT three times for 20 minutes at room temperature, incubated in Vectashield with DAPI (#H-1200-10, Vector Laboratories) overnight at 4 °C and subsequently mounted on slides. Whole guts were dissected in 1X PBS and fixed in 4% PFA in PBS for 30 minutes in a 6 well culture plate (Corning Life Sciences). Guts were then incubated in 1 mL PBT overnight at 4 °C. PBT was aspirated from sample wells and samples were washed an additional three times for 20 minutes each before being mounted in Vectashield with DAPI. For fat body dissections, gut, tubules and reproductive anatomy were removed from isolated fly abdomens, leaving abdominal cuticle with adipose tissues attached. Thereafter, samples were fixed, washed and mounted as for gut tissues above. All slides were stored at 4°C in the dark before imaging.

Images were acquired using a 5X objective on a Leica DM5500 B microscope via a K3M Leica camera and equipped with LASX (3.8.1.26810.1) software, where the optical settings were kept constant between samples of a given tissue. The Polygon selection tool in Fiji (ImageJ) (Schindelin et al., 2012) was used to trace regions of interest (ROI’s) which corresponded to either the central brain (chosen due to non-uniformity of the optic lobes from dissection), whole gut, or abdominal fat body, all of which were visualised/traced in the DAPI channel. GFP mean fluorescence intensity (MFI) per ROI was analysed using Fiji to obtain measurements for a total of 6-10 biological replicates for a given condition per tissue group.

Quantitative real-time PCR (qPCR)

Total RNA was isolated from whole flies via TRIzol (#12034977, Invitrogen), as per the manufacturer’s instructions. RNA samples were then subjected to DNA digestion via ezDNase (#15585209, Invitrogen), followed by conversion to cDNA using oligo(dT) primers (#11685581, Invitrogen) and SuperScript IV Reverse Transcriptase (#15307696, Invitrogen). qPCR was performed using 2x QuantiNOVA SYBR Green PCR Master Mix (#208054, Qiagen) in a 7500 Fast Real-Time PCR machine. Relative quantities of transcripts were determined through the relative standard curve method, normalised to the geometric mean of eIF1A and Tub84B reference genes, which have demonstrated good stability in samples from Drosophila across ageing (Ling and Salvaterra, 2011). Primer sequences were as follows:

eIF1A_for: ATCAGCTCCGAGGATGACGC; eIF1A_rev:

GCCGAGACAGACGTTCCAGA; InR_total_for: CGGACGACCGTCAAAGGTTA;

InR_total_rev: GCCTTGACGAGAGGTTCAGT; InR_DN[K1405A]_for:

ATCGCGAGTGTGCCATTGC; InR_DN[K1405A]_rev:

GGAGCAAACACCGAGCAATC; InR_WT_[K1405]_for:

GATCGCGAGTGTGCCATTAA; InR_WT_[K1405]_rev:

GGAGCAAACACCGAGCAATC; Tub84B_for:

TGGGCCCGTCTGGACCACAA; Tub84B_rev:

TCGCCGTCACCGGAGTCCAT.

Starvation and Oxidative Stress assays

For all stress assays, flies were maintained on auxin or control food at the described concentrations from day 2 of adulthood until 14 days of age, with an n of ∼150 flies for a given condition (density of 15 flies per vial). For starvation, flies were transferred to vials containing 1.5% (w/v) agarose (#16500-500, Invitrogen) to permit hydration but restrict nutritional value. Deaths and censors were scored two times per day, with flies transferred to fresh vials three times per week. For oxidative stress assays, flies were transferred to vials containing 5% (v/v) H₂O₂ (#216763, Merck) (constituted in 1.5% (w/v) agar and 5% (w/v) sucrose) and similarly transferred to fresh vials three times per week with deaths and censors scored two times per day.

Lifespan Assays

Lifespan assays were set up as above, with an n of ∼150 flies for a given condition (density of 15 flies per vial). Flies were transferred to fresh vials three times per week where deaths and censors were recorded during each transfer. Microsoft Excel (scoring template available at http://piperlab.org/resources/) was used to calculate survival proportions after each transfer.

Female fecundity assay

Fecundity assays were set up as above, with an n of ∼50 flies for a given condition (density of 5 flies per vial). Flies were transferred to fresh vials three times per week, and the number of eggs laid per vial over a 20-24h period was manually counted once weekly. Data are presented as the cumulative number of eggs laid per female per 24h over the entire measured period (1 through 5 weeks of age), calculated in Microsoft Excel (http://piperlab.org/resources/).

Statistical Analysis

Statistical analyses were either carried out in Microsoft Excel or GraphPad Prism 10.6.1 with tests for a given experiment detailed in each figure legend. Exact lifespan and stress assay deaths and censors (corresponding to the final sample sizes) are detailed in full within the source data. For all statistical tests (specific tests listed in each figure legend), p < 0.05 was considered significant.

Results

Transgene induction via AGES demonstrates sex and insertion-site-specific variability

To investigate applicability of AGES in the same context as our longevity experiments, we characterised transgene induction in both sexes, from different transgene insertion sites, and across an auxin titration (∼0-20 mM) (as per McClure et al., 2022). Since we used the potassium salt of auxin, we added equimolar potassium chloride as a vehicle control throughout (see Methods). To assess transgene expression, we expressed membrane-bound GFP (UAS-mCD8::GFP, hereafter termed UAS-GFP) ubiquitously, via da-GAL4 under control of the Tub84B-AtTIR1-T2A-AID-GAL80-AID transgene (hereafter termed AGES) inserted on either the second or third chromosome. We fed auxin from days 2-9 of adulthood, then quantified GFP fluorescence (Figure 1).

In both sexes and from both insertion sites, GFP fluorescence was significantly increased in the control-fed (0 mM auxin) da-GAL4,AGES>UAS-GFP flies (full genotypes w^Dah;AGES/+;da-GAL4,UAS-mCD8::GFP/+ or w^Dah;+;da-GAL4,UAS-mCD8::GFP/AGES) compared to the negative control groups (containing the da-GAL4 construct alone, full genotype w^Dah;+;da-GAL4/+), suggesting a degree of GAL4 activity even in the absence of auxin feeding (Figure 1A-D). GFP fluorescence was increased by ≥5 mM auxin in females, regardless of AGES transgene insertion site (Figure 1A and C). In males bearing AGES on the second chromosome, we observed significant GFP induction only at 10 mM and 20 mM auxin, while males with AGES on the third chromosome exhibited increased GFP only at 20 mM auxin (Figure 1B and D). When AGES was present, GFP fluorescence did not reach the level of positive controls (da-GAL4>UAS-GFP, full genotype w^Dah;+;da-GAL4,UAS-mCD8::GFP/+), even when using the highest auxin dose (Figure 1A-D). Together, these results suggest that transgene induction depend on sex and, in males, AGES insertion site; and that some residual GAL80 activity may occur in both sexes and from both transgene insertion sites.

In all subsequent experiments, we utilised female flies with the AGES construct on the third chromosome, owing to its strongly inducible expression at 5, 10, and 20 mM auxin as well as its lower leakiness at 0 mM (Figure 1C).

AGES permits controlled transgene induction throughout the fly life course

We next asked whether the AGES system could induce transgene expression effectively and in a stable manner as flies age, given known age-associated changes in efficacy of alternative established systems (Barwell et al., 2017; Delandre, McMullen and Marshall, 2025). Again, using da-GAL4,AGES>UAS-GFP flies, we established young, middle-aged and old cohorts (14d, 35d, and 56d, respectively), then fed auxin (10 mM) for 7 days, and assessed GFP induction (Figure 2A). At all ages, we found that flies fed auxin had significantly increased GFP expression compared to 0 mM control conditions (Figure 2B-D). At 21d, auxin treatment produced a 1.8-fold increase in mean GFP signal compared to the 0 mM group (Figure 2B), whereas at 42d (Figure 2C) and 63d (Figure 2D) this fold increase was 1.7 and 1.5 respectively. However, these induction levels were still significantly less than in positive control (da-GAL4>UAS-GFP) flies (at 21d, the auxin-fed mean GFP signal was only 47%; at 35d only 33%; and at 63d, only 28%). At both 21d and 42d (but not 63d), fluorescence in flies bearing UAS-GFP was greater than negative controls (da-GAL4/+) even in the absence of auxin (Figure 2B-D), indicating that AGES does not completely repress GAL4 activity at these two younger timepoints. Altogether these results suggest that AGES can be used to conditionally induce transgene expression across ages, although induction does not fully recapitulate expression without AGES transgenes, and residual transgene expression is possible in youth and middle age even in the absence of induction by auxin.

Auxin exerts genotype-dependent and dose-dependent effects on metabolic indices

Auxin can alter fly metabolism, for example causing a decrease in triacylglycerol (TAG) levels (Fleck et al., 2024). We therefore tested whether we recapitulated metabolic effects in flies treated with auxin (0-20 mM) for 7 days (days 2-9 of adulthood). We measured body mass, TAG, and protein levels in flies expressing da-GAL4 alone (full genotype w^Dah;+;da-GAL4/+), da-GAL4,AGES alone (full genotype w^Dah;+;da-GAL4/AGES) and da-GAL4,AGES>UAS-GFP (full genotype w^Dah;+;da-GAL4,UAS-mCD8::GFP/AGES). These conditions allowed us to test the effects of auxin in flies lacking AGES transgenes, as well as the effect of auxin-induced GAL4 activity in flies with or without a UAS-trangene.

In all genotypes, body mass was reduced by ≥10mM auxin (Figure 3A-C). TAG levels were not affected by auxin in either da-GAL4 or da-GAL4,AGES (Figure 3D and E), but ≥10 mM auxin decreased TAG in da-GAL4,AGES>UAS-GFP flies (Figure 3F). Protein levels were not affected by auxin in da-GAL4 (Figure 3G), but were increased by ≥5 mM auxin in da-GAL4,AGES and ≥1 mM auxin in da-GAL4,AGES>UAS-GFP (Figure 3H and I). Taken together, these results indicate that high dose auxin affects body mass regardless of the presence of AGES, and that there are further AGES-dependent and/or transgene/GFP-dependent effects on TAG and protein levels. In subsequent experiments, we therefore used 5 mM auxin, aiming to minimise off-target metabolic effects of auxin while still achieving sufficient transgene induction (Figure 1C).

Ubiquitous transgene induction via AGES displays tissue-specific expression variability

Given the reliance on homogenised whole-fly extracts within our fluorescence intensity assays which mask tissue-specific expression levels, we next investigated whether, AGES can induce transgene expression effectively across multiple organs in both young and older flies. To verify this, we fed da-GAL4, AGES>UAS-GFP flies auxin (5 mM) for 7 days starting at day 2 and day 35 of adulthood, representing a young (9d) and older (42d) timepoint respectively, when we examined GFP expression in the brain, whole gut, and abdominal fat body tissues following dissection (Figure 4). Notably, these tissues cover a range of expression strengths for the da-GAL4 driver, which, while strongly expressed in many peripheral tissues, has previously been reported to drive expression at lower levels in the brain than some other ubiquitous or neuron-specific drivers (Legan et al., 2008).

Within our experiments, auxin treatment (5 mM) did not increase GFP fluorescence in the adult brain in either young or old flies (Figure 4A-C and F). However, auxin treatment increased GFP fluorescence in the fat body of young but not old flies (Figure 4A and D) (Figure 4B and G). In youth, this fat body induction was 49% of that displayed by the positive control (da-GAL4>UAS-GFP) condition (Figure 4D). In the gut, auxin increased GFP signal in old and young flies (Figure 4A, B, E and H). At 9d, induction in guts was equivalent to 50% of the positive control (Figure 4E), but at 42d, unexpectedly 24% higher in the auxin-induced group than in the positive control, Figure 4H). Finally, across all tissues and both timepoints, we observed negligible transgene leakiness by these assays (Figure 4A-H). These results suggest the presence of tissue- and age-specific differences in expression with the AGES system, but they also demonstrate that in some tissues such as the gut, AGES permits strong transgene expression even at older ages.

AGES-directed expression of InR^DN only partially reproduces known effects of reduced IIS on lifespan and fecundity

We next tested whether AGES recapitulated the extension of healthy lifespan associated with reduced insulin/insulin-like signalling (IIS), an evolutionarily-conserved phenotype (Mathew, Pal Bhadra and Bhadra, 2017). Using AGES, we expressed a dominant-negative insulin receptor variant (UAS-InR^DN, encoding the kinase-dead InR^K1405A variant) throughout adulthood, mimicking an established lifespan-extending manipulation using GeneSwitch (Slack et al., 2011). We confirmed by qPCR that auxin treatment (5 mM from day 2 to 9 of adulthood) induced InR^DN expression in da-GAL4,AGES>UAS-InR^DNflies (full genotype w^Dah;UAS-InR^DN/+;da-GAL4,AGES/+) using primers that amplify either wild-type-specific, DN-specific, or both sequences (Figure 5A). We found that auxin treatment substantially induced expression of InR^DN, with a small but significant elevation in wild-type InR mRNA level (Figure 5A), potentially representing compensatory up-regulation of InR as reported previously for GeneSwitch-driven InR^DN expression in flies (Graze et al., 2018).

Figure 5.

Given that increased longevity mediated through lower IIS is frequently, but not always, accompanied by a reduction in fecundity (Flatt, 2011; Ibrahim et al., 2026), we quantified fecundity impact of InR^DNexpression under control of AGES (via weekly egg counts over a 34-day period). Consistent with previous work using GeneSwitch (Slack et al., 2011), we found that UAS-InR^DN induction by auxin feeding reduced egg laying, whereas fecundity of control genotypes was not affected by auxin (Figure 5B).

In parallel to fecundity assays, we also measured lifespan of flies fed auxin continuously from day 2 of adulthood. Here, we observed complex, genotype-dependent effects of auxin. First, we found no significant difference in the lifespan of auxin-treated +>UAS-InR^DN flies (full genotype w^Dah;UAS-InR^DN/+;+) (Figure 5C). However, auxin shortened lifespan of da-GAL4,AGES>+ flies (full genotype w^Dah;+;da-GAL4,AGES/+) (Figure 5D: median -8.3%). In da-GAL4,AGES>UAS-InR^DNflies, auxin led to a significant but complex and relatively modest change in longevity (Figure 5E: median lifespan +3.2%, but with reduced survival at other ages). Taken together, these results suggest that auxin does not affect fly lifespan in the absence of the AGES construct, but de-repressing GAL4 by feeding auxin in the presence of AGES can shorten lifespan. When transgenes are recombined to enable auxin-dependent InR^DN expression, AGES can recapitulate established effects of InR^DN expression (decreased fecundity and compensatory up-regulation of wild-type InR for instance), but the interpretation of lifespan data is challenging here given the effects of auxin in control genotypes.

Auxin impairs starvation and oxidative stress resistance

UAS-InR^DN induction in adults via the GeneSwitch system increases resistance to oxidative stress and starvation stress (Slack et al., 2011; Bolukbasi et al., 2017). However, responses to these stresses are related to lipid metabolism, which has been shown to be affected by auxin/AGES (Fleck et al., 2024). We therefore tested whether InR^DN induction by AGES recapitulated the stress resistance phenotypes shown previously following InR^DN induction by GeneSwitch. We fed flies auxin or control media from days 2-14 of adulthood, then transferred them to 5% (v/v) H₂O₂ - containing food for oxidative stress (or agarose-only media for starvation) and measured survival in +>UAS-InR^DNalone, da-GAL4,AGES>+, and da-GAL4,AGES>UAS-InR^DNflies. We found that auxin universally reduced both oxidative stress resistance and starvation resistance across all genotypes (Figure 6A-F), suggesting a detrimental effect of auxin treatment independent of the presence or absence of AGES- or InR^DN-encoding transgenes.

Figure 6.

Discussion

Since its establishment, the spatiotemporal gene expression system AGES has been used to investigate questions in fields from immunity to neuroscience in Drosophila (Richhariya et al., 2023; Corke et al., 2025; Sobrido-Cameán et al., 2025; Kuklinski et al., 2026). Our study tests the additional application of ageing biology. Here, we have characterised the capability of the system to permit transgene induction across conditions typical for Drosophila ageing and longevity experiments, where we observed that transgene expression depends on auxin dose, age, sex, and tissue. Conversely, we have further confirmed that auxin and/or activation of the AGES machinery exert complex effects on fly metabolism which need to be considered in designing studies assessing longevity or other indices of organismal health.

When characterising feeding conditions commonly used for ageing experiments, we found that transgene induction was most optimal in female flies, fed in the 5-20 mM dose range of auxin, with the AGES transgenic construct inserted on the third chromosome (Figure 1C). Regarding sex-dependent differences, it is likely that these are, at least in part, underpinned by known differences in food consumption, where females ingest greater amounts driven by metabolic demands for oogenesis (Wong et al., 2009; Vargas et al., 2010). Notably, however, even at the highest auxin dose, transgene induction never reached the levels displayed by positive control flies lacking the AGES construct (da-GAL4>UAS-GFP), an observation in agreement with the original study describing the AGES system (McClure et al., 2022). Here, it is worth noting that the positive control flies will have been expressing GFP throughout development and adulthood, so the comparison to adult-onset expression via AGES would not necessarily be expected to achieve fully equal expression levels. Other work has also shown that parallel gene expression systems like TARGET can achieve higher transgene expression than that achievable with AGES, despite AGES demonstrating significantly less apparent transgene leakiness (Hawley, Roberts and Fitzsimons, 2023).

We also demonstrated that throughout the fly life course transgene induction remains feasible with AGES, albeit with decreased magnitudes with increasing age (Figure 2). Therefore, AGES may remain a viable option to permit the study of transgene-based manipulations in the aged fly following acute induction periods. However, given our testing of a single ubiquitous driver (da-GAL4), along with recognised GAL4-driver-specific differences in activity with age in other systems like TARGET (Delandre, McMullen and Marshall, 2025), driver selection should be carefully vetted when utilising AGES for ageing experiments in future studies. Our results agree with prior work that a 5 mM auxin dose is sufficient to elicit transgene induction in fat body tissue (McClure et al., 2022), and we further demonstrate this in the gut (Figure 4A and E). However, unlike prior observations at 10 mM auxin (McClure et al., 2022), we found that 5 mM auxin could not induce transgene expression in the brain of flies, regardless of age (Figure 4A-C and F). This could reflect either limitations of our chosen concentration of auxin (5 mM) in crossing the blood-brain barrier, or the relatively low expression of da-GAL4 within the nervous system (Legan et al., 2008). Within the fat body, our results suggest that only younger (but not older) flies can achieve significant transgene expression (Figure 4A-B, D, and G), potentially due to age-related decreases in feeding rates (Carey et al., 2006). Within the gut, we observed relatively strong transgene induction at young ages and a surprising level of induction in aged flies that exceeded that of the positive control (Figure 4B and H). While the mechanism underlying the higher than positive control expression is unclear, it is possible that positive control flies develop long-term adaptive responses to GAL4 and/or GFP expression, for example endogenous turnover mechanisms that could ameliorate established negative consequences of constitutive GFP expression on physiology (Mawhinney and Staveley, 2011). With increasing age, female flies also exhibit a loss of gut barrier integrity (Rera, Clark and Walker, 2012; Regan et al., 2016) that may also underlie the age-specific effects seen here.

Our study also corroborates previous evidence suggesting that auxin feeding exerts detrimental effects on flies independent of genotype. For example, we observed metabolic impacts of auxin doses 10 mM and greater in our feeding regimen, consistent with previous evidence (Fleck et al., 2024). In our hands, high dose auxin supplementation (10-20 mM range) demonstrated significant body mass decreases (3-12% difference in mean) (Figure 3A-C) independent of fly genotype, potentially driven by decreased feeding events in auxin-containing media at this concentration range, as previously observed (Fleck et al., 2024). Even with no difference in the total mass of flies upon treatment at a lower dose of 5 mM auxin (Figure 3A-C), we still observed a stark reduction in stress-related survival for all genotypes of auxin-fed flies under starvation or oxidative stress (Figure 6), which suggests that auxin exerts detrimental effects even in flies where body mass remains constant. These effects may be driven by stress responses mounted against auxin itself, as prior evidence has demonstrated that 10mM auxin feeding induces transcriptomic signatures akin to that of a xenobiotic response with upregulation of detoxification-related transcripts (Fleck et al., 2024). It remains possible that a similar physiological reprogramming is apparent at 5 mM too, which in turn impacts stress survival capabilities independently of overall mass differences.

In addition, our study has identified detrimental effects of activating the AGES machinery independent of auxin feeding per se. For example, we observed subtle changes in protein levels (but not TAG levels) in auxin-fed flies that possess the AGES transgenes, but we did not observe similar effects in flies lacking the AGES transgenes (Figure 3G-H). Here, our findings differ somewhat from those of (Fleck et al., 2024), who observed a >50% decrease in TAG/protein ratios for flies of multiple genotypes exposed to auxin. These differences in the nature and magnitude of metabolic effects may be due to distinctions in auxin feeding regimens (e.g. pre-starved and fed on liquid auxin-containing food in previous studies, versus constitution of auxin into fly media here), the period of auxin exposure, and/or potential fly strain-specific effects. As an additional example where activating the AGES machinery was detrimental in our hands, we observed shortening of lifespan with auxin feeding in da-GAL4, AGES flies but not in UAS-InR^DN flies (Figure 5C-D). These findings could be explained by a detrimental effect of GAL4 activity itself, consistent with previous reports (Dravecz et al., 2022); or, alternatively, activation of the TIR1 enzyme itself may exert additional effects that shorten lifespan. In either case, our findings support the need for appropriate controls (e.g. driver-alone and responder-alone) in future work using the AGES system.

As a proof of principle for AGES as a viable alternative system for ageing research, our study also attempted to recapitulate a well-known pro-longevity phenotype frequently observed upon reduction of IIS via expression of InR^DN (Ikeya et al., 2009; Slack et al., 2011; Bolukbasi et al., 2017). Unfortunately, our recapitulation here remained limited, with reduced resistance to environmental stressors in flies treated with auxin regardless of genotype (Figure 6), and a shortening of overall lifespan when da-GAL4, AGES flies were fed auxin (Figure 5). These results make it difficult to interpret whether InR^DN expression affected lifespan in this context, and serve as a cautionary tale for using the AGES system in lifespan studies, at least when using the da-GAL4 driver. In summary, our results suggest that researchers using flies to study ageing should broadly consider the potential trade-offs between metabolic effects associated with auxin, the bioavailability of auxin at tissues of interest with their selected dose, and potential differences in induction attainable across sexes and AGES transgene insertion sites, when deciding whether AGES may be best suited for their given investigation.

Despite the limitations of the AGES system we describe above, our findings also leave room for the system to be a useful tool in certain domains of ageing research. For example, while our studies here have used a single GAL4 driver that expresses throughout fly tissues, it would be informative to evaluate whether AGES can recapitulate other lifespan and healthspan phenotypes upon reduced IIS with tissue-specific drivers that may not exert the same magnitude of systemic effects. In addition, our ubiquitous driver system was still able to recapitulate the reduced fecundity phenotype frequently described with lowered IIS (Figure 5B) without any change in fecundity in control genotypes, consistent with previous findings describing auxin’s lack of impact on oogenesis (Fleck et al., 2024). Taken together, these results suggest that AGES may be well suited for phenotypes related to egg development and fecundity, despite its impact on some other phenotypes.

Finally, we should also acknowledge several limitations within our study. Here we prioritised the use of female flies due to our observation that males show greater leakiness in transgene expression and lower levels of transgene induction upon auxin feeding (Figure 1). Our studies therefore miss potential sex-dependent differences in lifespan assays and metabolic indices, in contrast to previous work which demonstrated sex-specific feeding behaviours in contexts of auxin exposure (Fleck et al., 2024). As an additional limitation, whilst we used KCl supplementation in control conditions to control for the effects of potassium derived from the K-NAA (auxin) compound, we did not fully control for potential confounding factors introduced by dietary chloride on fly physiology. We have also focused on a single ubiquitous driver (da-GAL4) and a single pro-longevity intervention (InR^DN expression) that, while considered a classical pro-longevity intervention, also shows genotype-dependent effects on its modulation of longevity (Ibrahim et al., 2026). Lastly, whilst we utilised a genetic background strain (w^Dah), diet, and chemical-inducing feeding regimen that are frequently used in ageing studies (Ziehm, Piper and Thornton, 2013; Piper and Partridge, 2016), we cannot rule out strain-specific and diet-dependent differences from our results that may differ from other experimental procedures that utilise the AGES system.

Overall, our findings demonstrate that AGES remains a pertinent tool to modulate gene expression, even in the aged fly. However, our results also serve as a cautionary tale for the application of AGES when studying aspects of healthspan and longevity, primarily due to confounding effects of auxin and/or activation of the AGES machinery on metabolism. This caution aligns with previously described alterations in feeding behaviour upon auxin supplementation (Fleck et al., 2024). Here, we suggest that future studies using the AGES system would benefit from the inclusion of rigorous control genotypes (e.g. auxin-treated driver-alone and responder-alone controls). As an attractive future approach that may avoid some of these confounding effects, proposed alternate variants of the TIR1 receptor have the potential for efficacy at significantly lower concentrations of alternative auxins (McClure et al., 2022; Zhang et al., 2022). These systems may therefore exert reduced off-target effects on metabolism, thus providing additional opportunities for the use of AGES in future ageing studies within Drosophila.

Supplementary material

Supplementary Table 1.

Data availability

Source data underlying all figures are available at https://doi.org/10.17632/rw4mvv8mk2.1.

Acknowledgements

We thank Tracy Lamont, Chloe Buchanan, and the rest of the Glasgow Academic Support Unit (ASU) team for their tireless and professional work to prepare the fly media needed to carry out these experiments to a high standard. We further thank Jonathan Booth for experimental support/advice and the Sanz Montero lab for equipment sharing throughout this project. We also gratefully acknowledge Colin McClure (Queens University Belfast) for sharing Drosophila AGES stocks and advice on the AGES system, and Tony Southall (Imperial College London) for insightful comments and suggestions on this manuscript. This work was supported by the Wellcome Trust [301529/Z/23/Z]. AD was supported by a UKRI Future Leaders Fellowship [MR/Y019660/1].

Additional information

Author Contributions

T.M. and N.W. conceptualised the study; T.M., K.G., A.J.D. and N.W. designed the methodology; T.M., K.G., A.J.D. and N.W. conducted the investigation; T.M. and N.W. conducted formal analysis; T.M. wrote the original draft of the manuscript; T.M., K.G., A.J.D. and N.W. wrote, reviewed, and edited the manuscript; T.M. and N.W. visualised the study; N.W. supervised the study; N.W. acquired funding.

Open Access Statement

For the purpose of open access, the authors have applied a Creative Commons Attribution (CC BY) licence to any Author Accepted Manuscript version arising from this submission.

Funding

Wellcome Trust (WT)

https://doi.org/10.35802/301529

• Nathan Wooding

UK Research and Innovation (UKRI) (MR/Y019660/1)

• Adam J Dobson