Drosophila activins adapt gut size to food intake and promote regenerative growth

Abstract

Rapidly renewable tissues adapt different strategies to cope with environmental insults. While tissue repair is associated with increased intestinal stem cell (ISC) proliferation and accelerated tissue turnover rates, reduced calorie intake triggers a homeostasis-breaking process causing adaptive resizing of the gut. Here we show that activins are key drivers of both adaptive and regenerative growth. Activin-β (Actβ) is produced by stem and progenitor cells in response to intestinal infections and stimulates ISC proliferation and turnover rates to promote tissue repair. Dawdle (Daw), a divergent Drosophila activin, signals through its receptor, Baboon, in progenitor cells to promote their maturation into enterocytes (ECs). Daw is dynamically regulated during starvation-refeeding cycles, where it couples nutrient intake with progenitor maturation and adaptive resizing of the gut. Our results highlight an activin-dependent mechanism coupling nutrient intake with progenitor-to-EC maturation to promote adaptive resizing of the gut and further establish activins as key regulators of adult tissue plasticity.

Fig. 1. The activin signaling pathway control turnover of intestinal cells.

a–c Confocal image of a posterior (a, b″) or whole (c) midgut from a babo-Gal4>UAS-GFP reporter line stained for GFP (green, a-c), Armadillo (Arm) and Prospero (Pros) (Red, a, a″) or Dl (Red, b, b″), and DNA (blue, a, b) reveal enriched expression in delta-positive ISCs and other diploid cells (EBs) (b), not marked by nuclear Pros (a). This experiment was repeated independently 3 times with similar results. d–g Activin signaling maintains homeostatic gut size. Quantifications of midgut lengths after escargot-Gal4 (esg>)-mediated knockout of babo (e–f) (n = 25, 28 biologically independent guts) or smox (g) (n = 19, 24 biologically independent guts) in stem and progenitor cells using CRISPR/Cas9. h Schematic diagram of the esg^ReDDM system. Membrane tethered CD8::GFP and nuclear localized H2B::RFP is specifically co-expressed in stem and progenitor cells with the esg> driver. As progenitors differentiate into mature cells, loss of esg-Gal4 expression terminates further production of GFP and RFP. Differential stability of the fluorophores results in rapid degradation of GFP whereas RFP remains in differentiated cells for an extended duration of time. For all ReDDM experiments, RFP-positive/GFP-negative polyploid cells were scored as newly produced ECs (i–o′) Activin signaling maintains homeostatic cell turnover. Representative confocal images of dissected posterior midguts from control flies (i, m) and flies expressing babo^RNAi (j), smox^RNAi (k), and smox^SDVD (n) in stem and progenitor cells using esg^ReDDM tracing for 10 days and quantification of cell turnover in each condition (l, o) (n = 28, 27, 20 and n = 13, 16 biologically independent guts). Expression of babo^RNAi and smox^RNAi reduces the number of new mature cells (RFP⁺-only) whereas expression a constitutively active smox construct (smox^SDVD) increases the number relative to control. Expression of smox^SDVD also expands the pool of stem and progenitor cells (GFP⁺, RFP⁺) relative to controls (o′) (n = 13, 17 biologically independent guts). Significance was tested with two-tailed unpaired t-tests (f, g, o, o′) and Kruskal Wallis (l) with post-hoc multiple comparison analysis. Data are presented as mean values ±SD. Source data are provided as a Source Data file.

Fig. 2. Babo isoforms A and C regulate different steps of ISC-to-EC differentiation.

a RT-qPCR analysis on dissected control midguts showing differential expression levels of the babo-A, babo-B and babo-C isoforms (n = 6, 5, 6). b–e′ Quantification of tissue turnover in dissected posterior midguts from control flies and flies expressing babo-A^RNAi, babo-B^RNAi and babo-C^RNAi in stem and progenitor cells using esg^ReDDM tracing (pool of two experiments, n = 44, 47, 41, 45). f Schematic of the Cfs^ts system used to distinguish midgut cell populations. Stem and progenitor cells are collectively visualized by expressing nuclear localized UAS-his2b::CFP with esg-Gal4, EBs are marked by nuclear GBE-Su(H)-nlsGFP, while Ubi-his2av::mRFP causes all cells to express nuclear RFP. Thus, ISC and EBs are both marked by CFP, but only EBs are GFP-positive, whereas mature ECs and EECs are both marked by RFP but distinguished by nuclear size. g–j″) Babo-A and Babo-C isoforms control distinct stages of ISC-EC differentiation. Representative confocal images of dissected posterior midguts from control flies (h, h″) and flies expressing babo-A^RNAi (i, i″) or babo-C^RNAi (j, j″) in stem and progenitor cells using Cfs^ts and quantification of cell type numbers in each condition (g, g″′) (pool of two experiments, n = 28, 33, 30, 38). k–n Babo-C is required in EBs to promote the EB-to-EC transition. Representative confocal images of dissected posterior midguts from control flies (k) and flies expressing babo-A^RNAi (l) and babo-C^RNAi (m) in EBs using Su(H)^ts > UAS-GFP and quantification of EB numbers in each condition (n) (pool of two experiments, n = 40, 40, 39). o–q Representative confocal images of dissected posterior midguts from control flies (o, o′) and flies expressing babo-A^RNAi (p, p′) in ISCs using Cfs^ts coupled to Su(H)-Gal80 and quantification of EBs (GFP-positive; q) (n = 19, 17). Significance was tested with a two-tailed Mann-Whitney test (q), one-way ANOVA (g, g′, g″) or Kruskal Wallis (e, e′, g″′, n) with post-hoc multiple comparison analysis. Data are presented as mean values ±SD. Source data are provided as a Source Data file.

Fig. 3. EB-derived Actβ supports regenerative growth and tissue repair.

a, b Quantification of PH3+ cells in midguts dissected from Ecc15 infected control flies and flies with ISC/EB-specific RNAi- (a; n = 22, 15, 15, 15) or CRISPR/Cas9-mediated (b; n = 20, 18) babo knockdown. c Mated flies with ISC/EB-specific knockdown of babo display reduced survival to oral Pe infection (n = 98, 93). d, e Quantification of PH3+ cells in midguts dissected from control flies and flies expressing babo-ARNAi, babo-BRNAi or babo-CRNAi in ISCs/EBs 16 h post Ecc15 (d; n = 18, 20, 19, 20, 19, 20, 20) or Pe (babo-ARNAi; e; n = 15, 16, 17) infection. f–h′ Representative confocal images of dissected posterior midguts from control flies (f, f″) and flies expressing babo-ARNAi (g, g″) in ISCs/EBs 16 hours post Ecc15 infection, and quantification of cell populations (h, h′n = 14, 13). i RT-qPCR analysis on dissected midguts 6- and 16 hours post Pe infection (n = 5, 5, 5). j–m″ Representative confocal images of dissected posterior midguts harboring actb-Gal4>UAS-GFP and Su(H)-lacZ (labeling EBs) stained for GFP (Green), lacZ (red), Pros and Dl (white), and DNA (blue) 16 hours post Ecc15 (k, k″) or Pe (l, l″) infection, and quantification of EECs, ISCs, and EBs positive for GFP signal (m, m″; n = 7, 10, 12). n–q Quantification of PH3+ cells in midguts dissected from control flies and flies expressing actβRNAi in EBs (n; pool of two experiments, n = 23, 45, 45) or actβRNAi (o, q; n = 22, 21, 22 and n = 20, 19, 20) or Cas9 + actβgRNA (p; n = 20, 19) in ISC+EBs (esgts) 16 hours post Ecc15 (n–p) or Pe (q) infection. r–u′ Representative confocal images of dissected posterior midguts from control flies (r) and flies expressing babo-ARNAi (s) and babo-CRNAi (t) in ISCs/EBs 48 hours (16H on Ecc15 + 32H recovery on normal food) post Ecc15 infection and quantification of cell turnover using esgReDDM tracing for 10 days (u, u′; n = 21, 25, 19). Significance was tested with two-tailed unpaired t-tests (b, h, h′,p), one-way ANOVA (a, d, e, i, n, u′) or Kruskal Wallis (m, m′, m″, o, q, u) with post-hoc multiple comparison analysis and a Mantel-cox Log-rank test (c). Data are presented as mean values ±SD. Source data are provided as a Source Data file.

Fig. 4. Daw couples calorie intake with EB differentiation to promote adaptive growth.

a, b Representative images (a) and quantification of gut length (b; n = 10, 9, 10, 9) of dissected midguts of fed, 24H starved, 48H starved and 24H refed controls. c, c″′ Quantification of the total number of cells in the posterior midgut of fed, starved, and refed animals using Cfsts (n = 7, 9, 6). d Representative image of dissected midguts of control (d) 48H starved (d′) esg-lacZ animals labeled with TUNEL (in red) to visualize apoptotic cells. d″ Quantification of the proportion of TUNEL+ that are ISC/EBs (lacZ+) (n = 9,9). e RT-qPCR analysis on dissected control midguts in fed, 24H or 48H starved, and 48 H starved + refed flies (n = 6, 6, 5, 6). f–i Representative confocal images of dissected posterior midguts from control flies (f, f′) or flies with ISC/EB-specific knockdown of babo-C (g, g′) starved for 48 h (f, g) and refed for 24 hours (f′, g′). h–i Quantification of EB numbers (h) and midgut length (i) (n = 15, 15, 12, 17) in f, g′. j–l Quantification of cell turnover in dissected posterior midguts from fed, starved, or starved + 96H refed control flies (j) or flies with ISC/EB-specific knockdown of babo-CRNAi (k) using esgReDDM tracing for 10 days (l) (n = 18, 20, 20). m–p Quantification of ISC + EB numbers (m–o; n = 21, 23) and midgut lengths (p; n = 30, 30, 27) in dissected posterior midguts from 48H starved + 24H refed control flies (m) or flies with EC-specific daw knockdown (n). q, r Mated flies with ISC/EB-specific knockdown of Babo-C (q; n = 145, 147, 148) or EC-specific knockdown of daw (r; n = 150, 148, 149) show reduced resistance to successive cycles of starvation/refeeding. s Model of how activins regulate different steps of ISC-to-EC maturation in response to distinct environmental cues. Significance was tested with a two-tailed Mann–Whitney (d″, o), one-way ANOVA (b, c, c′, c″, c″′, e, l, p) or two-way ANOVA (h, i) with post-hoc multiple comparison analysis and Mantel-cox Log-rank tests (q, r). Data are presented as mean values ±SD. Source data are provided as a Source Data file.