Dietary restriction improves intestinal cellular fitness to enhance gut barrier function and lifespan in D. melanogaster

Abstract

Loss of gut integrity is linked to various human diseases including inflammatory bowel disease. However, the mechanisms that lead to loss of barrier function remain poorly understood. Using D. melanogaster, we demonstrate that dietary restriction (DR) slows the age-related decline in intestinal integrity by enhancing enterocyte cellular fitness through up-regulation of dMyc in the intestinal epithelium. Reduction of dMyc in enterocytes induced cell death, which leads to increased gut permeability and reduced lifespan upon DR. Genetic mosaic and epistasis analyses suggest that cell competition, whereby neighboring cells eliminate unfit cells by apoptosis, mediates cell death in enterocytes with reduced levels of dMyc. We observed that enterocyte apoptosis was necessary for the increased gut permeability and shortened lifespan upon loss of dMyc. Furthermore, moderate activation of dMyc in the post-mitotic enteroblasts and enterocytes was sufficient to extend health-span on rich nutrient diets. We propose that dMyc acts as a barometer of enterocyte cell fitness impacting intestinal barrier function in response to changes in diet and age.

Author summary

Dietary restriction (DR) is a robust environmental method to slow aging and age-related diseases in diverse organisms. Age-related disruption of gut integrity has been observed in both mammals and fruit flies and is a determinant of lifespan. In Drosophila, DR is able to slow the age-related decline in gut integrity. Although commensal dysbiosis has been proposed as a leading cause of gut barrier dysfunction, antibiotic treatment does not prevent the age-related increase in gut permeability. We identify that an intrinsic mechanism regulates gut barrier function through regulation of enterocyte apoptosis by ‘cell competition’. We show DR up-regulates dMyc expression in the gut which enhances enterocyte cellular fitness, prevents the age-related decline in gut integrity, and contributes to DR-induced lifespan extension. Conversely, on a rich diet, inhibition of dMyc in the enterocytes leads to cell death that enhances gut permeability and leads to systemic inflammation and shortened lifespan.

Introduction

The intestinal epithelium forms a permeable barrier that segregates the internal and external environments, allowing for the absorption of nutrients, but at the same time keeping toxic substances and pathogens from entering the body. The intestine also manages the interaction between the host and the gut microbiome. Increasing gut permeability is one of the risk factors for developing inflammatory bowel disease (IBD, including ulcerative colitis and Crohn's disease) [1] and contributes to systemic immune activation, which promotes the progression of chronic inflammation [2], a known risk factor for aging and some age-related diseases [3,4].

Several mechanisms have been postulated to influence intestinal permeability. These include changes in the microbiota, luminal secretion of mucins and anti-microbial peptides (AMPs), and tight junction proteins [2,5]. However, the role of intestinal cell turnover in modulating intestinal permeability remains underexplored. Removal of intestinal cells by apoptosis, a type of programmed cell death, is an active process that is used to eliminate unwanted cells which are then replaced by dividing intestinal stem cells. Increased intestinal apoptosis and gut barrier dysfunction have been linked to multiple diseases, including necrotizing enterocolitis (NEC), IBD, intestinal cancer, and HIV infection [6–8]. However, the causal link between intestinal apoptosis and gut barrier function remains to be established.

Disruption of intestinal homeostasis is one of the hallmarks of aging in both vertebrate and invertebrate species [9–11]. There are several similarities between the mammalian and Drosophila gut architecture, making flies an attractive model to study gut barrier disorders [12–15]. The intestinal epithelium in Drosophila is comprised of intestinal stem cells (ISC), enteroblasts (EB), enterocytes (EC), and enteroendocrine cells (EE). Loss of intestinal barrier function leads to the increased systemic production of AMPs, which are regulated by Toll and Immune deficiency (IMD) innate immune pathways [16,17]. Loss of intestinal barrier function is also associated with increased mortality in aging flies [18,19] and is likely a consequence of age-associated inflammation.

Dietary restriction (DR), which is the reduction of specific nutrients without causing malnutrition, is a robust environmental intervention that slows aging and age-related diseases in a diverse set of species including yeast, worms, fruit flies and rodents [20–23]. In D. melanogaster, DR imposed by reduction of yeast in the diet not only extends lifespan [24,25] but is also able to slow the age-related decline in gut integrity [19,26]. Similarly in the mouse gut, calorie restriction (CR) is known to alter epithelial structure and function, including villi length, crypt depth and cell turnover [27–29]. Thus, studying the mechanisms by which DR improves gut integrity has great significance for understanding aging and certain intestinal disorders where nutrition is a risk factor.

We demonstrate that the rate of enterocyte apoptosis in the rich nutrient conditions can be attenuated by DR in Drosophila melanogaster. We hypothesize that nutrient-dependent as well as age-related increase in apoptosis in enterocytes holds the key to understanding phenomena like gut inflammation, gut permeability, and aging. We show that a rich diet reduces intestinal cellular fitness due to the reduction of dMyc expression and enhances cell competition-mediated cell death. Cell competition is defined as short-range elimination of unfit cells (loser cells) by apoptosis when confronted by fitter neighboring cells (winner cells), which has been shown to occur in the fly intestine [30,31]. In the Drosophila larval wing disc, relative dMyc expression levels have been shown to determine loser and winner cells. Wing disc cells with lower dMyc expression levels compared to neighboring cells become loser cells and are eliminated by apoptosis [32]. Importantly, this Myc-dependent cell competition mechanism is conserved in mammalian embryos [33,34]. Myc-high naive cells remove Myc-low differentiating cells to maintain the purity of the pluripotent cell pool [35]. However, cell competition and the role of dMyc in intestinal post-mitotic cells has not been described before. We have identified a critical role for diet-dependent modulation of dMyc in regulating the age-related cellular fitness in ECs. Furthermore, we show that dMyc-dependent regulation of intestinal cell death is crucial for intestinal barrier function and organismal survival. Our findings highlight the importance of understanding mechanisms that balance intestinal apoptosis with repair upon aging and dietary modulation, which are likely to play a significant role in age-related diseases and various intestinal disorders.

Results

Heterozygous dMyc mutant flies display a lifespan extension [36]. Furthermore, reduced expression of Myc also extends lifespan and slows the onset of age-related pathologies like osteoporosis, cardiac fibrosis, and immunosenescence in mice, suggesting conserved effects of Myc on aging [37]. While these findings have shown that reduced Myc levels are beneficial, a recent report demonstrated that DR increases dMyc protein abundance and boosts the innate immune response [38]. Hence, dMyc may have different effects in specific tissues under different nutrient conditions. We first examined if modulation of dMyc in different tissues is necessary for nutrient-dependent lifespan changes. We imposed DR and ad libitum (AL) conditions on adult D. melanogaster using diets that differed only in the yeast content (0.5% and 5% yeast extract in the media for DR and AL diets respectively) [39–41]. Surprisingly, EB/EC-specific knockdown of dMyc during the adult stage using a drug (RU486)-inducible Gal4 driver (5966-GS Gal4; referred to as 5966-GS) reduced the maximal DR-mediated lifespan extension. EB/EC-specific dMyc knockdown reduced lifespan by 32% in flies on DR compared to control flies not administered RU486. However, EB/EC-specific dMyc knockdown in flies on AL resulted in a slight lifespan reduction (19%) (Figs 1A and 4D, S2 and S3 Tables). Two independent RNAi strains confirmed dMyc’s effect on lifespan (Figs 1A, S1A and S1B). We used the Cox proportional-hazards model to statistically determine the gene diet-interaction in influencing lifespan upon inhibition of dMyc. The gene-diet interaction terms was highly significant (p < 0.0001) (S4 Table). Similar results were also observed upon knockdown with an EC-specific driver, Np-1-Gal4 (S1C Fig and S2 Table). To achieve adult-specific dMyc knockdown and avoid developmental defects by the loss of dMyc, we used the temperature sensitive Gal4 suppressor, Gal80^ts in combination with NP-1- Gal4 [42]. We also observed that dMyc knockdown using Act5C-GS, which targets multiple tissues including the intestine also prevented the maximal lifespan extension by DR (S1D and S1E Fig and S2 Table). Compared to dMyc knockdown using a 5966-GS driver, DR-dependent lifespan extension was not altered when dMyc expression was inhibited in both the ISCs and EBs (using 5961-GS). Similarly, dMyc knockdown in the fat body (using S₁106-GS) failed to affect DR-dependent lifespan extension (Fig 1B–1D, S2 and S3 Tables). These results suggest temporal importance for the effect of dMyc expression on lifespan. Although reduction of dMyc in the whole body from the developmental stage extends lifespan, an inhibition of dMyc in the ECs in the adult stage significantly diminishes the benefit of DR on lifespan.

Fig 1. Intestinal dMyc modulates lifespan in a diet-dependent manner.

(A-C) Kaplan-Meier survival analysis of tissue-specific knockdown of dMyc upon AL and DR. (A) Enteroblasts and enterocytes (5966-GS>dMyc RNAi), (B) Intestinal stem cells and enteroblasts (5961-GS>dMyc RNAi), (C) Fat bodies (S₁106-GS>dMyc RNAi). (D) Median lifespan calculated from A-C are shown. Statistical analysis of the survival curves and the number of flies are provided in S2, S3 and S4 Tables. See also S1 Fig.

Fig 4. Antibiotic treatment partially rescues the decrease in lifespan but not enterocyte cell death due to loss of dMyc.

(A) Diptericin mRNA expression in dissected fat bodies of 21 day old 5966-GS>dMyc RNAi flies with/without antibiotic treatment. (B, C and D) puc mRNA (B), hid mRNA (C) and upd3 mRNA (D) expressions in dissected guts of 21 day old 5966-GS>dMyc RNAi flies with/without antibiotic treatment. (E) Mitotic ISCs quantification in 21 day old 5966-GS>dMyc RNAi flies with/without antibiotic treatment. Error bars indicate SEM of 13 guts. (F) Smurf gut permeability assay in 20 day old 5966-GS>dMyc RNAi flies with/without antibiotic treatment. Control (–RU486) without antibiotic (AL: n = 185), 5966-GS>dMyc RNAi (+RU486) without antibiotic (AL: n = 222, DR: n = 199), Control (–RU486) with antibiotic (AL: n = 233), 5966-GS>dMyc RNAi (+RU486) with antibiotic (AL: n = 173, DR: n = 238) Error bars indicate SEM of 12 different vials. (G) Kaplan-Meier survival analysis of 5966-GS>dMyc RNAi flies with/without antibiotic treatment upon AL and DR. (A-D) Error bars indicate SD from 3 independent biological replicates. ‘AB’ represents antibiotic. Statistical analysis of the survival curves and the number of flies are provided in S2 and S3 Tables.

Gut barrier dysfunction has been associated with increased mortality in flies [18,19]. Therefore, we examined gut integrity in EB/EC-specific dMyc knockdown flies using the Smurf assay. In this assay, flies are fed a blue food dye that normally does not cross the intestinal barrier. However, flies that display a loss of gut integrity turn blue and are thus termed, ‘Smurfs’ [19,43]. Consistent with a previous report [19], the number of Smurf flies increased with age. Control populations (without RU486) reared under DR conditions exhibited reduced percentage of Smurfs compared to AL conditions (Fig 2A). DR diet also delayed the age-related induction of the antimicrobial peptide, Diptericin, in the fat body and gut of w1118 flies (S2B and S2C Fig) which suggests a reduction in systemic inflammation. EB/EC-specific dMyc knockdown, on the other hand, resulted in a significant increase in the percentage of Smurfs upon DR (Fig 2A). Diptericin expression was increased in flies with EB/EC-specific dMyc knockdown at a middle age on AL condition, but showed increased expression at a later age upon DR (Figs 2B and S2D). These results suggest that dMyc in ECs is necessary to maintain the gut barrier function.

Fig 2. dMyc regulates gut barrier function.

(A) Smurf gut permeability assay in 5966-GS>dMyc RNAi 28 and 35-day old flies. d28:–RU486 (AL: n = 335, DR: n = 422), +RU486 (AL: n = 308, DR: n = 380). d35:–RU486 (AL: n = 84, DR: n = 237), +RU486 (AL: n = 77, DR: n = 151). (B) Age-dependent changes in mRNA expression of Diptericin (Dpt) in dissected fat bodies of 5966-GS>dMyc RNAi flies. Young, middle and old represent day 7, 21 and 35 of adulthood, respectively. mRNA expression for flies at day 0 was set to 1. ‘nc’ represents samples that were not collected. (C) upd3 mRNA expression in dissected guts from 21 day old 5966-GS>dMyc RNAi flies. (D) Mitotic ISCs quantification in 21 day old 5966-GS>dMyc RNAi flies. Error bars indicate SEM of 38 guts. (** p < 0.01 by t-test). (B and C) Error bars indicate SD from three independent biological replicates. (** p < 0.01, * p < 0.05 by t-test). See also S2 Fig.

A dynamic equilibrium exists between damaged enterocytes and ISC proliferation to maintain intestinal homeostasis [44]. Damaged enterocytes undergo apoptosis and release the interleukin-6-like cytokine, upd3, to enhance ISC proliferation and initiate intestinal repair [44]. Thus, we examined whether dMyc knockdown in the ECs alters intestinal homeostasis by modulating secretion of upd3 from ECs, to induce proliferation of ISCs. We observed increased expression of upd3 in 21-day old dMyc knockdown flies under both AL and DR conditions (Fig 2C). EB/EC-specific dMyc knockdown upon DR also resulted in a significant increase in ISC proliferation as measured by the mitotic cell proliferation marker phospho-histone H3, in 21-day old flies (Fig 2D). However, no significant change in ISC proliferation on AL diet was observed (Fig 2D), presumably because a threshold for activating ISC proliferation in the AL conditions is higher than that of DR. These data suggest that dMyc in ECs influences diet-dependent changes in intestinal permeability as well as ISC proliferation, thus impacting function and homeostasis of the intestinal epithelium.

Next, we investigated whether the increase in ISC proliferation upon loss of dMyc using the 5966-GS driver is a result of increased cell death in the intestine. We performed acridine orange fluorescence staining and the TUNEL assay to assess apoptosis [45]. Consistent with the low appearance of Smurf flies upon DR (Fig 2A), control flies reared on DR showed fewer numbers of apoptotic cells in the gut compared to flies on AL (Figs 3A, 3A’ and S3A). Knockdown of dMyc in the gut using 5966-GS significantly increased both the number of TUNEL and acridine orange positive cells upon DR (Fig 3A, 3A’ and S3A). Similar to the ISC proliferation data (Fig 2E), we did not observe a significant difference in the number of apoptotic cells in EB/EC-specific dMyc knockdown flies on the AL diet, despite these flies expressing high levels of upd3 (Fig 2C). In Drosophila, activated JNK signaling induces apoptotic cell death through induction of the pro-apoptotic gene hid [46,47]. dMyc knockdown on both diets resulted in an upregulation of the JNK targets, hid and puckered when examined in 21-day old flies (Figs 3B and S3B). We also confirmed the activation of JNK signaling upon EB/EC-specific dMyc knockdown using the puc^E69 reporter strain (puc-lacZ) [48] at day 21 of age (S3C and S3C’ Fig). When we inhibited JNK signaling in the EB/EC-specific dMyc knockdown background, we found it failed to rescue the reduction of lifespan. (S3D Fig and S2 Table). We also tested the inhibition of JNK signaling alone using 5966-GS. JNK inhibition in EBs/ECs strongly reduced lifespan on both AL and DR conditions (S3E Fig and S2 Table). These results argue that several downstream pathways are likely to be important to explain the dMyc knockdown phenotypes and that JNK is one of them but not sufficient for rescuing the lifespan. Additionally, JNK signaling in the EBs/ECs is required to optimally enhance fly survival. Together, these data suggest that dMyc plays a crucial role in modulating the diet-dependent changes in intestinal cell death with age.

Fig 3. dMyc knockdown in the EBs/ECs causes apoptosis.

(A) TUNEL assay using dissected guts from 28 day old 5966-GS>dMyc RNAi flies. Representative image (n = 25). Scale bar indicates 40 μm. (A’) Quantification of TUNEL positive cells from 25 images. (*** p < 0.001 by t-test). (B) hid mRNA expression in dissected guts of 21 day old 5966-GS>dMyc RNAi flies. (C) Smurf gut permeability assay in 14 day old 5966-GS>UAS-reaper flies.–RU486 (AL: n = 367, DR: n = 377), +RU486 (AL: n = 184, DR: n = 383). Error bars indicate SEM of 17 different vials. (** p < 0.01 by t-test). (D) Smurf gut permeability assay in 5966-GS, dMyc RNAi; + flies and 5966-GS, dMyc RNAi; UAS-p35 flies at 14 days of age. 5966-GS, dMyc RNAi; + flies–RU486 (AL: n = 218, DR: n = 226), +RU486 (AL: n = 233, DR: n = 218), 5966-GS, dMyc RNAi; UAS-p35 +RU486 (AL: n = 260, DR: n = 264). Error bars indicate SD of 11 different vials. (* p < 0.05 by t-test). (E) Mitotic ISCs quantification in 14 day old 5966-GS, dMyc RNAi; + and 5966-GS, dMyc RNAi; UAS-p35 flies. Error bars indicate SEM of 22 guts. (** p < 0.01, * p < 0.05 by t-test). (F) Kaplan-Meier survival analysis of 5966-GS, dMyc RNAi; + flies and 5966-GS, dMyc RNAi; UAS-p35 flies upon AL and DR. Statistical analysis of the survival curves, number of flies are provided in S2 and S3 Tables. (B) Error bars indicate SD of three independent biological replicates. (** p < 0.01, * p < 0.05 by t-test). See also S3 and S4 Figs.

To verify whether induction of cell death in the EBs/ECs is sufficient to increase the number of flies with intestinal permeability, we ectopically induced apoptosis in EBs/ECs and measured gut permeability. Overexpressing the pro-apoptotic gene reaper in the enterocytes using 5966-GS increased the number of Smurf flies (Fig 3C). Next, we investigated whether gut dysfunction caused by loss of dMyc is due to increased apoptosis. We induced p35 (a universal apoptosis inhibitor) in the dMyc-RNAi background. p35 overexpression significantly reduced the number of Smurf flies (Fig 3D) and ISC proliferation (Fig 3E) in both AL and DR conditions. Furthermore, p35 overexpression was able to partially rescue the lifespan reduction seen in 5966-GS-specific dMyc knockdown flies under both nutrient conditions (Fig 3F, S2 and S3 Tables). We also observed similar results by inhibition of dronc (an initiator caspase-9 ortholog) in the dMyc-RNAi background (S4A, S4B and S4C Fig). These data support the notion that enterocyte cell death is necessary to cause intestinal permeability and reduce lifespan upon inhibition of dMyc.

Intestinal apoptosis has been linked with changes in microbiome composition in D. melanogaster. Inhibition of the intestinal homeobox gene caudal leads to overexpression of AMPs. This overexpression results in gut epithelial cell apoptosis, which is mediated by the microbiome [49]. Thus, we examined whether the gut microbiota contributes to dMyc mediated cell death in the ECs. We reared EB/EC-specific dMyc knockdown flies on antibiotic diets after eclosion, to address the role of intestinal bacteria in modulating cell death. We observed that antibiotic treatment is sufficient to reduce fat body-specific expression of Diptericin in flies at 21 days on both AL and DR diets (Fig 4A). Importantly, antibiotic treatment did not rescue the up-regulation of apoptosis indicator genes (puc, hid and upd3) in the gut (Fig 4B–4D), or the ISC hyper-proliferation in EB/EC-specific dMyc knockdown flies (Fig 4E). Furthermore, antibiotics failed to rescue gut integrity in EB/EC-specific dMyc knockdown flies (Fig 4F). However, antibiotic treatment was sufficient to partially extend the lifespan in dMyc knockdown flies on both DR and AL diets (Fig 4G). These results suggest that increased apoptosis in dMyc knockdown flies is not dependent on the influence of gut bacteria; however, the reduction of bacterial load in the gut can diminish mortality in these conditions. These data are consistent with the idea that EC cell death compromises gut barrier function and thus exposes the internal tissues to infiltration by bacteria or bacterial antigens, resulting in systemic inflammation and increased mortality.

To further investigate the mechanisms by which dMyc regulates EC fate, we quantified dMyc mRNA expression in the gut under AL and DR conditions. Flies fed on AL diet showed an age-dependent reduction in the expression of dMyc mRNA in the intestine but not the fat body (Figs 5A and S5A). Notably, this age-dependent decrease in dMyc expression was attenuated upon DR (Fig 5A). This age-dependent downregulation of dMyc mRNA cannot be explained by the number of ECs, because DR flies showed higher dMyc mRNA at old age after normalizing by expression level of Pdm1, an EC-specific marker [50] (S5B Fig). We also observed that dMyc expression in the posterior midgut was maintained upon DR until old age in dMyc:GFP-tagged flies [36] (Fig 5B and 5B’). Age-related reduction of dMyc expression is consistent with the increased apoptosis observed in the flies in AL conditions compared to the DR diet (Fig 3A). As dMyc-deficient cells have been shown to be removed from the larval imaginal disc by cell competition [51], we hypothesized that the same phenomenon occurred in dMyc-deficient enterocytes. To investigate this hypothesis, we created genetic mosaic dMyc knockdown EBs/ECs in the adult intestine. We utilized the CoinFLP-Gal4 system [52], which contains transcriptional STOP cassette in between canonical FRT sites and FRT3 sites. Act5C-Gal4 is expressed when recombination occurs between canonical FRT sites but not in the FRT3 sites. We induced these two types of recombination events only in the post-mitotic intestinal cells, the EBs and ECs, by temporal activation of UAS-FLP for 24 hours under the control of 5966-GS, in young flies upon DR diet. Thus, we named this the 5966-GS: Coin-Flip-out system (S5C and S5D Fig). Since this system carries the UAS-EGFP transgene, flip-out EBs/ECs are GFP positive, which allows one to monitor the turnover of post-mitotic intestinal cells. We found that the area of dMyc knockdown flip-out cells was significantly reduced at 7 days after flip-out event (AFO) compared to that at 48 hours AFO upon DR (Fig 5C and 5C’). In contrast, we did not observe a significant difference in the area of the GFP-positive cell between 48 hours and 7 days AFO in the WT flip-out cells (Fig 5C and 5C’). Elimination of dMyc knockdown flip-out cells is not DR-specific, as we also observed similar results on AL conditions (S5E and S5E’ Fig), suggesting that importance of dMyc on EC health. Next, we asked whether loss of dMyc flip-out cells is caused by cell death using SYTOX orange nucleic acid staining. SYTOX orange was incorporated into the nucleus of GFP-positive dMyc flip-out cells at 48 hours AFO upon DR while WT flip-out cells did not show the staining with SYTOX (Fig 5D). Furthermore, inhibition of apoptosis by overexpressing Drosophila inhibitor of apoptosis 1 (DIAP1) in dMyc flip-out cells was sufficient to inhibit elimination of dMyc knockdown cells (Fig 5C and 5C’). Thus, the loss of dMyc in the intestine leads to a reduction in cellular fitness and eliminates cells by apoptosis. We also found that wild-type ISCs respond to the loss of dMyc by enhancing ISC proliferation, which was quantified using a phospho-histone H3 antibody (Fig 5E and 5E’).

Fig 5. dMyc mediates improved intestinal cellular fitness upon dietary restriction.

(A) dMyc mRNA expression in dissected guts from w1118 upon AL and DR was measured with age. mRNA expression from flies at day 0 was set to 1. (B) Immunostaining of dMyc-GFP in dissected guts. dMyc:GFP flies were fed AL and DR diets for 7 days (Young) and 35 days (Old). Representative image (n = 10–13). Scale bar indicates 40 μm. (B’) Quantification of dMyc-GFP-positive cells from 10–13 images. (** p < 0.01 by t-test). (C) GFP-positive flip-out EBs/ECs were observed at 48 hours (Top panels) and 7 days after flip-out (AFO) (Bottom panels) in the posterior midgut upon DR. (Left) WT flip-out EBs/ECs. (Center) dMyc RNAi flip-out EBs/ECs. (Right) dMyc RNAi, UAS-DIAP1 flip-out EBs/ECs. Scale bar indicates 40 μm. (C’) Quantification of the size of GFP positive cells. (*** p < 0.001 by t-test). (D) SYTOX orange staining in the WT flip-out EBs/ECs (Top) and dMyc RNAi EBs/ECs (Bottom) in the posterior midgut upon DR at 48 hours AFO. (Left) merged image. (Middle) GFP-positive flip-out EBs/ECs. (Right) SYTOX staining. White arrow head indicates SYTOX positive cell. Nuclei were stained with Hoechst 33342. Scale bar indicates 20 μm. (E) Immunostaining of mitotic ISCs using dissected guts from dMyc RNAi flip-out flies upon DR at 48 hours AFO. (Top-left) merged image. (Top-right) GFP-positive dMyc RNAi EBs/ECs with pH3 staining. Bottom panels are magnified images of the yellow square in the top-left image. Scale bars of top and bottom panels indicate 40 and 20 μm, respectively. (E’) Mitotic ISCs quantification in the gut from WT flip-out and dMyc RNAi flip-out flies (n = 15). See also S5 Fig.

Finally, we examined whether EB/EC-specific overexpression of dMyc is sufficient to increase fly survival. Indeed, overexpressing dMyc in EBs/ECs was sufficient to extend lifespan on AL diets, but it reduced lifespan on the DR diet (S6A Fig). In order to overcome this detrimental effect, we activated dMyc expression in the middle of life by feeding RU486 from 21 days of age. Activation of dMyc in EBs/ECs later in life was able to delay the onset of death on AL diets, but it slightly reduced maximum lifespan on AL. There was a lack of a significant effect on the lifespan under DR conditions (Figs 6A and S6B). In order to get a modest activation of dMyc in the EB/ECs, we fed RU486 to flies only 2 days a week during the entirety of adult life. We found that intermittent dMyc overexpression in the EBs/ECs extended health-span on AL and did not show any detrimental effect on fly survival compared to control flies (Figs 6B and S6C). These data support our notion that slightly enhanced dMyc levels in the ECs extend lifespan, however, excess levels of dMyc appears to have a detrimental effect on flies.

Fig 6. The modest activation of dMyc in EBs/ECs extends health-span on the rich nutrient diets.

(A and B) Kaplan-Meier survival analysis of enteroblasts and enterocytes specific dMyc overexpression (5966-GS>UAS-dMyc) upon AL and DR. (A) RU486 was administrated from day 21 of age. (B) RU486 was administrated every Monday and Tuesday during the adult stage. Statistical analysis of the survival curves and the number of flies are provided in S2 Table. See also S6 Fig.

Discussion

The intestine is subject to continuous cellular turnover. To maintain gut homeostasis, both proliferative and regenerative capacities have to be regulated. Previous studies have demonstrated that upon exposure to acute stresses there is an increase in EC apoptosis which induces ISC proliferation to enhance repair of the intestine [53]. We hypothesize that enterocytes undergo cell death in response to cellular stresses imposed by overnutrition and age, which are then repaired by the elimination of damaged cells and their eventual replacement by differentiated ISCs. Clonal analysis revealed that gut turnover in protein-poor conditions is slower than turnover in protein-rich conditions in Drosophila [54]. Consistently, we found that ISC proliferation is significantly decreased upon DR while apoptosis is reduced (Figs 2D & 3A). These results suggest that gut turnover in DR conditions is slower than AL conditions due to the reduced nutrient stress of DR. A number of labs have previously shown an age-related increase in ISC proliferation [10,55]. Our data suggest that the increases in ISC proliferation are a compensatory mechanism for the age-related increase in enterocyte apoptosis.

The gastrointestinal tract forms an excellent ecological niche for a wide variety of commensal microbes that live in proximity with the mucosal epithelial barrier [56]. Age-related intestinal barrier dysfunction is linked to microbiota dysbiosis [18]. In our experiments, antibiotic treatment partially rescued the lifespan but did not rescue the intestinal cell death-associated phenotypes in EB/EC-specific dMyc knockdown flies (Fig 4). This suggests that dMyc knockdown-mediated apoptosis and cytokine expression are not due to microbiota changes.

The commensal microbial ecosystem contributes to the maintenance of the overall intestinal barrier architecture by regulating the organization of epithelial tight junctions lining mucosal surfaces [56]. Tight junctions are complexes that seal adjacent epithelial cells and prevent the trafficking of elements across the gut epithelial barrier [57]. Intestinal barrier permeability has been shown to increase with age in mammals [11,58,59], perhaps due to differential expression of tight junction proteins (such as occludins and claudins) [5]. In Drosophila, epithelial cell integrity is regulated by an apical protein complex composed of septate junctions, which are the Drosophila analog to tight and adherence junctions [60]. Recent reports have described that gut barrier dysfunction is tightly associated with a reduction of cell junction components [18,26]. Therefore, we investigated whether dMyc regulates the expression of the septate junction protein, Discs large (Dlg). Although we observed that EC/EB-specific dMyc knockdown alters epithelial cell integrity, we were still able to detect Dlg expression in 5966-GS-specific dMyc knockdown flies on both diets (S7 Fig). These data suggest that dMyc does not maintain the gut barrier function through modulation of cell junction proteins.

We show that dMyc expression declines in an age-dependent manner, especially in the gut, in rich-nutrient conditions (Fig 5A and 5B). Furthermore, we show that loss of intestinal dMyc is detrimental and leads to increased cell death and intestinal permeability. Thus, dMyc may act as a barometer of fitness in adult enterocytes. Previous studies have shown that reduction of Myc in the whole body has beneficial effects on lifespan in mice and flies [36,37], but here we show that inhibition of dMyc in the ECs during the adult stage reduces lifespan. In Drosophila, Myc has diverse tissue-specific effects [61–63]. One possibility to explain these contradictory results is that developmental or tissue-specific knockdown of dMyc may have beneficial effects in adult life, while knockdown of dMyc in the gut is detrimental. Consistent with this notion, a previous study has shown that DR-dependent increases in dMyc abundance improves immune response and resistance against pathogenic bacteria infection in adult flies [38]. Hence, the role of dMyc in different tissues may be altered by dietary composition and age.

A recent study revealed that cell competition contributes to healthy aging and tissue homeostasis by eliminating unfit cells, especially during development [64]. Our study suggests that enterocytes in the adult aging intestine maintain homeostasis through cell competition. Our study provides a major role for dMyc-mediated cell competition in the adult intestine upon dietary shifts which influences intestinal permeability and lifespan (Fig 7). Loss of dMyc with age or under high nutrient stress could alter the balance of cellular fitness and induce cell death by cell competition to eliminate unfit cells. On the other hand, reduced gut turnover upon DR may compromise the replacement of dying cells upon dMyc knockdown, leading to gut barrier dysfunction and abrogation of lifespan extension. Cell competition in the intestinal stem cell compartment in the fly has been shown to be mediated by the JNK and JAK-STAT pathways [30]. We find induction of JNK and the JAK-STAT ligand Upd3 in animals with intestine-specific knockdown of dMyc (Figs 2 and S3). However, the inhibition JNK signaling fails to rescue the lifespan reduction observed in dMyc knockdown flies. (S3D Fig). Multiple mechanisms are likely at play to regulate the cell death upon inhibition of dMyc. It is possible that the innate immunity system modulates cell competition in the gut because Toll-related receptors (TRRs)/NFkB signaling contributes to loser cell elimination during Drosophila development [65]. Further analysis is needed to unveil the role of the innate immunity pathway on cell competition during aging and dietary shifts.

Fig 7. The role of dMyc in the fly gut under different nutrient conditions.

Schematic diagram showing how intestinal dMyc regulates gut homeostasis and lifespan under different nutrient conditions. dMyc depletion induces EC cell death by cell competition due to loss of cellular fitness and its increases gut permeability. AL conditions enhance cellular damage and reduce levels of dMyc leading to loss of cellular fitness and causing EC cell death, increased ISC proliferation as a consequence of Upd3 secretion and increased gut permeability which shortens lifespan. These changes are reversed by DR which upregulates dMyc levels to improve intestinal homeostasis in the gut.

Loss of intestinal homeostasis is associated with many diseases, including IBD, autoimmune diseases, chronic inflammation, cancer, obesity, and diabetes [66,67]. The intestine is a highly metabolically active tissue which is exposed to the environment and has to adapt to various dietary changes as well as the microbial environment. Thus, the rate of damage accumulating in the aging gut is likely to be significantly higher than in other tissues. Our study demonstrates a critical role for Myc in diet- and age-induced changes in gut homeostasis. We hypothesize that Myc and pathways regulating cell competition are possible targets for therapeutic interventions against a range of age-related and inflammatory diseases.

Materials and methods

Fly culture, stocks and lifespan analysis

Flies were reared on standard laboratory diet (Caltech food recipe; 8.6% Cornmeal, 1.6% Yeast, 5% Sucrose, 0.46% Agar, 1% Acid mix) [41,68]. Emerged adults were transferred within 3–5 days to yeast extract diet (8.6% Cornmeal 5% Sucrose, 0.46% Agar, 1% Acid mix, and variable concentrations of yeast extract). The AL diet contains 5% yeast extract while the DR diet has 0.5% yeast extract. For Gene-Switch Gal4 drivers, RU486 was dissolved in 95% ethanol and was used at a final concentration of 100 μM (the media is then referred to as '+RU486'). The control AL or DR diet contained the same volume of 95% ethanol and is referred to as '–RU486'. Lifespan analysis was followed as described previously [41]. For antibiotic treatment, 50 μg/ml each of kanamycin sulfate, ampicillin sodium, tetracycline hydrochloride, and erythromycin were mixed with autoclaved diets. Survival curves were created using the product-limit method of Kaplan and Meier. The log-rank (Mantel-Cox) test was used to evaluate differences between survivals and determine P values. We used the Prism software package (GraphPad Software) to carry out the statistical analysis and to determine lifespan values. We analyzed the significance of the interaction between two variables in several of the survival outcomes and determine P values using Cox proportional hazards analysis implemented in the R package 'survival'. The following strains were obtained from the Bloomington stock center: UAS-dMyc RNAi ^TRiP-1 (25784), UAS-dMyc RNAi ^TRiP-2 (36123), UAS-dMyc (9674), dMyc-GFP.FPTB (38633), UAS-dronc RNAi (32963) and CoinFLP-Gal4; UAS-EGFP (58751). UAS-dMyc RNAi (v2947) was obtained from the Vienna Drosophila RNAi Center.

qRT-PCR

Total RNA was extracted from 12 female guts, 8 female fat bodies (fly abdomen) or 5 female whole flies using Quick-RNA MiniPrep Kit (Zymo Research). cDNA was synthesized using QuantiTect Reverse Transcription Kit (QIAGEN). 1 μg of total RNA was used per sample. qPCR reaction was performed in duplicate on each of 3 independent biological replicates using SensiFAST SYBR No-ROX Kit (BIOLINE). Error bars indicate SD. Samples were normalized with an endogenous control, ribosomal protein 49 (rp49). The primer sets used for qPCR are summarized in S1 Table.

Immunohistochemistry

Flies were dissected in PEM (100 mM Pipes, 2mM EGTA and 1 mM MgSO₄). Dissected guts were fixed with 4% formaldehyde in PEM for 45 minutes. Samples were washed for 10 minutes three times with PEM then incubated with 1% NP40/PEM for 30 minutes. Samples were washed for 10 minutes three times with TBS-TB (20 mM Tris-HCl, 130 mM NaCl, 1 mM EDTA, 0.1% Triton X-100 and 0.2% BSA) and blocking was performed with 5% goat serum in TBS-TB for 2 hours at room temperature. Samples were incubated with primary antibody overnight at 4°C, were then washed for 10 minutes three times with TBS-TB, and incubated with secondary antibody for 2 hours at room temperature. Nuclei were stained using DAPI. Samples were mounted with Mowiol mounting buffer and analyzed by confocal microscope (Zeiss: LSM780) and fluorescence microscope (KEYENCE: BZ-X710). The following antibodies were used in this study: anti-rabbit GFP (Life technologies: 1/500), anti-rabbit phospho-histone H3 (Millipore: 1/1,000), anti-rabbit β-galactosidase (MP: 1/500), anti-mouse Dlg (DSHB: 1/50), anti-rabbit Alexa fluor 488 (Life technologies: 1/500), anti-mouse Alexa fluor 488 (Life technologies: 1/500) and anti-rabbit Alexa fluor 555 (Life technologies: 1/500).

TUNEL assay

Apoptotic cells were detected using the ApopTag Red In Situ Apoptosis Detection Kit (Millipore: S7165). Flies were dissected in PEM. Dissected guts were fixed with 4% formaldehyde in PEM for 45 minutes. Then we followed the manufacture’s protocol. Nuclei were stained using DAPI. Samples were mounted with Mowiol mounting buffer and analyzed by the fluorescence microscope (KEYENCE: BZ-X710).

Acridine orange staining

Dissected guts were incubated with acridine orange (Sigma: 5 μg/ml) and Hoechst 33342 (Life technologies: 10 μg/ml) in PBS for 5 minutes at room temperature. Samples were rinsed with PBS twice, then mounted with PBS and immediately analyzed by microscope (Olympus: BX51).

SYTOX orange nucleic acid staining

Dissected guts were incubated with SYTOX Orange Nucleic Acid Stain (Invitrogen: 1 μM) and Hoechst 33342 (Invitrogen: 10 μg/ml) in PEM for 10 minutes at room temperature. Samples were rinsed with PEM twice, then mounted with PEM and immediately analyzed by microscope (KEYENCE: BZ-X710).

Smurf gut permeability assay

Smurf assay was adapted as described [19]. Female flies were fed either AL or DR diets before the assay. Flies were placed in an empty vial containing a piece of 2.0 cm x 4.0 cm filter paper. 350 μl of blue dye solution, 2.5% blue dye (FD&C #1) in 5% sucrose, was used to wet the paper as feeding medium. Flies were maintained with feeding paper for 24 hours at 25°C.

Supporting information

S1 Fig. dMyc expression in ECs has an important role for lifespan extension upon dietary restriction.

(A and B) Kaplan Meier survival analysis of 5966-GS>dMyc RNAi flies upon AL and DR using the other two different RNAi strains. (C) Kaplan Meier survival analysis of EC-specific dMyc knockdown (using Np-1-Gal4, tub-Gal80^ts) upon AL and DR at 29°C. (D) Kaplan Meier survival analysis of ubiquitous knockdown of dMyc (using Act5C-GS-Gal4) upon AL and DR. (E) Median lifespan was calculated from Kaplan Meier survival analysis of ubiquitous knockdown of dMyc (using Act5C-GS-Gal4) under 5 different yeast extract conditions. Statistical analysis of the survival curves and the number of flies are provided in S2 Table.

(TIF)

S2 Fig. Dietary restriction slows down the incidence of systemic and local infection.

(A) Experimental timeline for 5966-GS>dMyc RNAi flies. (B and C) Diptericin (Dpt) mRNA expression in dissected fat bodies (B) and guts (C) from w1118 upon AL and DR was measured with age. mRNA expression from flies at day 0 was set to 1. (D) Age-dependent changes in mRNA expression of Diptericin in dissected guts in 5966-GS>dMyc RNAi flies. Young, middle and old represent day 7, 21 and 35 of ages, respectively. mRNA expression for flies at day 0 was set to 1. ‘nc’ represents samples that were not collected. (B-D) Error bars indicate SD from 3 independent biological replicates.

(TIF)

S3 Fig. JNK inhibition failed to rescue dMyc knockdown mediated phenotypes.

(A) Acridine orange staining using dissected guts from 30 day old 5966-GS>dMyc RNAi flies. Error bars indicate SEM of 12 guts. (*** p < 0.001 by t-test). (B) puc mRNA expression in dissected guts of 21 day old 5966-GS>dMyc RNAi flies. (C) Lac Z staining of dissected guts from 21 day old 5966-GS, dMyc RNAi; puc^E69 (puc-lac Z) flies. Representative image (n = 11). Scale bar indicates 50 μm. (C’) Quantification of puc-lacZ positive cells from 11 images. (** p < 0.01, * p < 0.05 by t-test). (D) Kaplan-Meier survival analysis of 5966-GS, dMyc RNAi; + flies and 5966-GS, dMyc RNAi; UAS-Bsk^DN flies upon AL and DR. (E) Kaplan Meier survival analysis of EBs/ECs-specific JNK inhibition (5966-GS> UAS-Bsk^DN) upon AL and DR. Statistical analysis of the survival curves and number of flies are provided in S2 Table.

(TIF)

S4 Fig. Inhibition of apoptosis rescues dMyc knockdown mediated gut dysfunction.

(A) Smurf gut permeability assay in 5966-GS, dMyc RNAi; + flies and 5966-GS, dMyc RNAi; dronc RNAi 20 days old flies. 5966-GS, dMyc RNAi; + flies–RU486 (AL: n = 61, DR: n = 51), +RU486 (AL: n = 40, DR: n = 70), 5966-GS, dMyc RNAi; dronc RNAi +RU486 (AL: n = 84, DR: n = 77). Error bars indicate SD of 4 different vials. (* p < 0.05 by t-test). (B) Mitotic ISCs quantification in 14 days old 5966-GS, dMyc RNAi; + and 5966-GS, dMyc RNAi; dronc RNAi flies. Error bars indicate SEM of 10 guts. (*** p < 0.001, * p < 0.05 by t-test). (C) Kaplan Meier survival analysis of 5966-GS, dMyc RNAi; + and 5966-GS, dMyc RNAi; dronc RNAi flies upon AL and DR. Statistical analysis of the survival curves and number of flies are provided in S2 Table.

(TIF)

S5 Fig. Dietary restriction up-regulates dMyc expression in the gut but not other tissues.

(A) dMyc mRNA expression in dissected fat bodies from w1118 upon AL and DR was measured with age. mRNA expression from flies at day 0 was set to 1. (B) dMyc mRNA expression in dissected guts from w1118 upon AL and DR was measured with age. dMyc mRNA expression was normalized by Pdm1 mRNA expression. mRNA expression from flies at day 0 was set to 1. (C) Schematic diagram for the 5966-GS: Coin-Flip-out system. CoinFLP-Gal4 system (Bosch et al., 2015) is utilized to induce dMyc RNAi mosaic cells in the post mitotic intestinal cells, EBs and ECs, as 5966-GS is allowed to express UAS-FLP in EBs and ECs during RU486 administration. Then, flies were maintained without RU486 and were dissected at 48 hours and 7 days after flip-out event (AFO). Flies were cultured at 18°C from 3^rd instar larvae in order to reduce a leaky expression of Gal4. (D) Schematic diagram for the timeline of 5966-GS: Coin-Flip-out system. (E) GFP-positive flip-out EBs/ECs were observed at 48 hours (Top panels) and 7 days after flip-out (AFO) (Bottom panels) in the posterior midgut upon AL. (Left) WT flip-out EBs/ECs. (Right) dMyc RNAi flip-out EBs/ECs. Scale bar indicates 40 μm. (E’) Quantification of the size of GFP positive cells. (** p < 0.05 by t-test). (A and B) Error bars indicate SD of 3 independent biological replicates.

(TIF)

S6 Fig. The modest activation of dMyc in EBs/ECs extends health-span on the rich nutrient diets.

(A-C) Kaplan-Meier survival analysis of enteroblasts and enterocytes specific dMyc overexpression (5966-GS>UAS-dMyc) upon AL and DR. (A) RU486 was administrated from a day of sorting. (B) RU486 was administrated from day 21 of age. (C) RU486 was administrated every Monday and Tuesday during the adult stage. Statistical analysis of the survival curves and the number of flies are provided in S2 Table.

(TIF)

S7 Fig. EB/EC-specific dMyc knockdown alters gut integrity, but not the expression of septate junction protein.

(A) Immunostaining of Discs large (Dlg) using dissected guts from 28 day old 5966-GS>dMyc RNAi flies. Representative image (n = 10). Scale bar indicates 20 μm.

(TIF)

S1 Table. Primer sets used for qRT-PCR.

(DOCX)

S2 Table. Statistical analysis of the survival curves.

(DOCX)

S3 Table. Summary of the independent repeats of the lifespan analysis.

(DOCX)

S4 Table. The statistical analysis for EB/EC or ISC/EB-specific dMyc knockdown lifespan.

(DOCX)

Data Availability

All relevant data are within the paper and its Supporting Information files.

Funding Statement

This work was funded by grants from the American Federation of Aging Research and Hillblom foundations (PK, SDK), and NIH (R01AG038688 & RO1AG045835) (PK). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.