When enteropathogenic bacteria breach the intestinal epithelium, they are recognised by epithelial and immune cells that elicit an intestinal regenerative response. However, less is known about whether and how intestinal stem cells (ISCs) directly detect invading pathogenic bacteria and couple this to their proliferation. Here, we show that adult Drosophila midgut ISCs recognise pathogenic bacteria through the peptidoglycan recognition proteins, PGRP-LC and PGRP-LE, and translate this into their proliferation by stimulating Imd-Mkk3-p38 signalling. Moreover, we find that PGRP-LC/LE-Imd-Mkk3-p38 signalling in ISCs regulates p38 activation throughout the midgut epithelium after infection, indicating that ISCs can influence the regenerative microenvironment in a non-cell autonomous manner. Whilst it was previously thought that ISC proliferation in both mammals and flies is driven solely by damage-induced signals after infection, our work reveals that ISCs can directly recognise pathogenic bacteria and mount a strong parallel regenerative response that spreads throughout the midgut epithelium. Increased ISC proliferation after bacterial recognition may also serve as a strategy to repopulate the epithelium with uninfected cells.

PGRP-LC and PGRP-LE promote ISC proliferation and p38 activation in the midgut epithelium after pathogenic bacterial infection

Previously, it was shown that PGRP-LC is required for p38 activation in the midgut epithelium after exposure to peptidoglycan from Ecc15 [33]. Further, it was also shown that p38 is activated in the midgut epithelium after P.e. infection [39]. Thus, we tested whether PGRP-LC was required for p38 activation after P.e. infection. We observed that null PGRP-LC^E12 mutants were not viable as homozygotes on our medium, whereas the putative null PGRP-LC^Δ5 mutants were viable. Thus, we examined p38 activation in PGRP-LC^E12/PGRP-LC^Δ5 transheterozygotes. Using an antibody against phosphorylated human p38 (Thr180/Tyr182) (p-p38) [39], we found that PGRP-LC transheterozygotes showed a decrease in p38 activation following continuous P.e. infection for 18 hours compared to infected control midguts (Fig 1A – 1E). The increased phosphorylated p38 in infected control midguts likely represents both activated p38a and p38b [39]. We then determined whether the intracellular peptidoglycan sensor, PGRP-LE, was necessary for p38 activation in the midgut epithelium after P.e. infection. Indeed, p38 activation in the midgut epithelium also decreased in infected null PGRP-LE¹¹² mutants compared to infected control midguts (Figs 1F and S1A–S1D). These data suggest that peptidoglycan recognition by midgut epithelial cells through PGRP-LC and PGRP-LE promotes p38 activation in the midgut epithelium. Previously, it was shown that p38 activity in ECs is required for ISC proliferation after P.e. infection [39]. Thus, we determined whether PGRP-LC and PGRP-LE were both required for ISC proliferation after P.e. infection. ISC proliferation decreased by 64.4% and 54.8% in PGRP-LC transheterozygotes (PGRP-LC^E12/PGRP-LC^Δ5) and PGRP-LE¹¹² mutants after P.e. infection, respectively, compared to infected control midguts (Fig 1G and 1H). Interestingly, infected PGRP-LC^Δ5 mutants showed only a 23.7% decrease in ISC proliferation compared to infected control midguts (S1E Fig), suggesting that PGRP-LC^Δ5 [45] is a hypomorphic allele of PGRP-LC. Together, these data indicate that PGRP-LC and PGRP-LE are required to promote midgut regeneration after pathogenic bacterial infection, partly by inducing p38 activity in the midgut epithelium.

Pathogenic bacterial recognition by progenitors stimulates ISC proliferation via p38 signalling

Whilst both PGRP-LC and PGRP-LE were necessary for ISC proliferation following pathogenic bacterial infection (Fig 1G and 1H), the specific cell type in which they functioned remained unclear. Since PGRP-LC was previously shown to be required for p38 activation in ECs after exposure to peptidoglycan from Ecc15 and to promote antimicrobial peptide production [33,37,38], we first tested whether PGRP-LC was required in ECs to promote ISC proliferation after pathogenic bacterial infection. Depleting PGRP-LC by RNA interference (RNAi) in ECs with the EC-specific driver mex-GAL4; tubGAL80^ts (Malic enzyme modifier, mex^ts) for 5 or 6 days did not affect ISC proliferation after continuous P.e. infection for 18 hours (S2A and S2C Fig), suggesting that PGRP-LC function in ECs is not essential for promoting midgut regeneration.

Previously, it was shown that PGRP-LC is highly expressed in midgut progenitor cells [46]. Additionally, ISC proliferation in Drosophila is known to increase with age, and both PGRP-LC and PGRP-LE are required for this age-related increase [46]. Thus, we next tested whether PGRP-LC and PGRP-LE were both required in progenitors. We found that both receptors were required in progenitors for ISC proliferation after P.e. infection. Depletion of PGRP-LC in progenitors with the progenitor-specific driver esg-GAL4 tubGAL80^ts (esg^ts) reduced ISC proliferation after P.e. infection by 80.4% and 49.5% with RNAi (1) and RNAi (2), respectively, compared to infected control midguts (Figs 2A and S2B). Similarly, we found that PGRP-LE depletion resulted in a 106.8% decrease in ISC proliferation after P.e. infection compared to infected control midguts (Figs 2H and S2D). Furthermore, the decrease in ISC proliferation due to PGRP-LC and PGRP-LE depletion in progenitors was not due to a decrease in progenitor number as the mean percent progenitors in these midguts was similar to uninfected and infected control midguts (S2I–S2J Fig). Together, these data suggest that DAP-type peptidoglycan recognition through PGRP-LC and PGRP-LE in progenitors stimulates ISC proliferation after P.e. infection.

Since both PGRP-LC and PGRP-LE were required for p38 activation in the midgut epithelium after P.e. infection, we tested whether both PGRPs were required for p38 activation in esg⁺ progenitors. We found increased levels of activated p38 within esg⁺ progenitors in addition to ECs after P.e. infection compared to the basal p-p38 levels in uninfected control midguts (Fig 2B–2E′). However, we found that PGRP-LC depletion with esg^ts reduced p38 activation in progenitors by 105.0% after P.e. infection compared to progenitors in infected control midguts (Fig 2B and 2D–2G′; S1A Appendix). Similarly, PGRP-LE depletion with esg^ts reduced p38 activation in progenitors by 74.2% after P.e. infection compared to progenitors in infected control midguts (Figs 2I and S2E–S2H′; S1B Appendix). Moreover, we surprisingly found that depleting PGRP-LC and PGRP-LE in progenitors using esg^ts reduced p38 activation throughout the midgut epithelium after P.e. infection compared to infected control midguts (Figs 2C–2G′, 2J, and S2E–S2H′). These data suggest that pathogenic bacterial recognition by progenitors through PGRP-LC and PGRP-LE promotes ISC proliferation, possibly through activation of p38 signalling in progenitors and throughout the midgut epithelium after P.e. infection.

We next ascertained whether other gram-negative pathogenic bacteria such as Ecc15 can promote ISC proliferation and activate p38 in the midgut epithelium through PGRP-LC in progenitors. We found that Ecc15 can stimulate ISC proliferation, as previously described [25], as well as p38 activation in the epithelium of control midguts (S2K–S2L Fig). However, depleting PGRP-LC in progenitors inhibited ISC proliferation and p38 activation throughout the midgut epithelium (S2K–S2L Fig). These data indicate that the mechanism by which pathogenic bacterial recognition by progenitors via PGPR-LC and PGRP-LE promotes ISC proliferation is not limited to specific pathogenic bacteria.

Since p38 activation within progenitors and ISC proliferation after P.e. infection required both PGRP-LC and PGRP-LE, we next tested whether p38 signalling was required in progenitors for ISC proliferation after P.e. infection. To do this, we depleted p38 signalling from midgut progenitors by expressing an RNAi that targets both p38a and p38b (p38a+b^RNAi) using esg^ts and infected these flies with P.e.. Depleting p38a and p38b in progenitors resulted in a 73.4% decrease in ISC proliferation after P.e. infection compared to infected control midguts (Fig 3A). This decrease was not caused by reduced progenitor number due to p38a and p38b depletion with esg^ts and P.e. infection as the mean percent esg⁺ progenitors in these midguts was similar to uninfected and P.e.- infected control midguts (S3A Fig). These data suggest that p38 signalling in progenitors is required for ISC proliferation after P.e. infection.

We then examined whether blocking p38 signalling in progenitors affects their levels of activated p38 in progenitors after P.e. infection. Inhibition of p38b resulted in an 87.8% decrease in p38 activation in progenitors after infection compared to progenitors in infected control midguts (Figs 3B and S3B; S1C Appendix). Surprisingly, we also found that depleting p38a and p38b or inhibiting p38b alone in progenitors with esg^ts reduced p38 activation throughout the midgut epithelium after P.e. infection (Fig 3C–3G). These data indicate that in addition to regulating ISC proliferation, p38a and p38b activity in progenitors non-cell autonomously controls p38 activity throughout the midgut epithelium.

DAP-type peptidoglycan alone is insufficient to trigger an ISC regenerative response in healthy midguts

Since peptidoglycan from Ecc15 was shown to activate p38 in the midgut epithelium [33], we tested whether DAP-type peptidoglycan alone could stimulate ISC proliferation in uninfected midguts. We first determined whether midgut epithelial cells recognised DAP-type peptidoglycan after feeding using Dipt(2.2)-lacZ, which reports the expression of the AMP, Diptericin. As expected, DAP-type peptidoglycan or P.e. activated Dipt(2.2)-lacZ in ECs (S3D–S3F′ Fig). This indicates that after adult flies consumed food containing DAP-type peptidoglycan or P.e., midgut ECs, as previously reported [37,38], launched an innate immune response by producing AMPs. Nevertheless, DAP-type peptidoglycan did not trigger ISC proliferation in these flies, while P.e. infection did (Fig 3H), indicating that DAP-type peptidoglycan alone cannot trigger an ISC regenerative response.

We further tested whether exposing the midgut epithelium to DAP-type peptidoglycan from non-virulent P.e. or heat-killed virulent P.e. could stimulate ISC proliferation. Non-virulent P.e. activated the Dipt(2.2)-lacZ reporter but did not stimulate ISC proliferation or epithelial p38 like P.e. (Figs 3I–3J and S3G–S3G′). Thus, DAP-type peptidoglycan released from non-virulent P.e. was sufficient to trigger AMP production but not ISC proliferation. We then fed flies heat-killed P.e. (at 95°C), which surprisingly triggered ISC proliferation (Figs 3K and S3C). These midguts appeared damaged after exposure to heat-killed P.e., suggesting that certain virulence factors may be resistant to 95°C heat treatment. Interestingly, depleting PGRP-LC or PGRP-LE from progenitors decreased ISC proliferation by 81.4% and 82.1%, respectively, after feeding heat-killed P.e. compared to heat-killed P.e.-fed control midguts (Figs 3K and S3C).

Together, these data indicate that DAP-type peptidoglycan alone is not sufficient to stimulate ISC proliferation, and that it possibly requires midgut damage to stimulate ISC proliferation and activate epithelial p38. This is consistent with our experiments feeding damage-causing pathogenic bacteria such as P.e. and Ecc15.

Imd signalling in progenitors promotes ISC proliferation and p38 activation in the midgut epithelium after pathogenic bacterial infection

Previously, it was shown that Imd but not Relish (Rel) is required for p38 activation in the midgut epithelium after exposure to peptidoglycan from Ecc15 [33]. Indeed, like PGRP-LC and PGRP-LE, Imd is also required in midgut progenitors for the age-related increase in ISC proliferation [46]. Thus, Imd may also be involved in progenitors to promote ISC proliferation and p38 activation in the midgut epithelium after infection. To test this, we first examined imd¹ mutants but found that ISC proliferation after infection was similar between imd¹ mutant and control midguts (S4A–S4B Fig). However, we then inhibited Imd specifically in progenitors by either expressing a dominant-negative form of Imd (Imd^D30A) or by RNAi. Inhibiting Imd by either expressing Imd^D30A or imd^RNAi in progenitors reduced ISC proliferation by 55.1% and 83.5%, respectively (Figs 4A and S4C). This decrease was not caused by reduced progenitor number due to Imd^D30A expression with esg^ts and P.e. infection (S4E Fig). Further, we found a 43.9% decrease in p38 activation in progenitors expressing Imd^D30A after infection compared to progenitors in infected control midguts (Fig 4B and 4D–4G′; S2A Appendix). Moreover, we found a decrease in p38 activation in the midgut epithelium in P.e.-infected midguts expressing Imd^D30A or imd^RNAi in progenitors compared to infected control midguts (Figs 4C–4G′ and S4D). These data suggest that Imd is necessary in progenitors for p38 activation in the midgut epithelium and ISC proliferation. Finally, we tested using isogenic control and mutant lines whether Relish was required for ISC proliferation after P.e. infection. Consistent with a prior study using Ecc15 [47], we found that ISC proliferation was similar in Rel^E20 mutant midguts compared to control midguts after P.e. infection (S4F Fig). Together, these data suggest that Imd, but not Relish, is required in progenitors for ISC proliferation and p38 activation in the midgut epithelium.

Mkk3-p38 signalling in progenitors promotes ISC proliferation and p38 activation in the midgut epithelium after pathogenic bacterial infection

Since PGRP-LC/LE-Imd-p38 signalling was required for ISC proliferation after P.e. infection, we next determined whether increasing p38 activity in midgut progenitors is sufficient to stimulate ISC proliferation. To activate p38 signalling within midgut progenitors, we overexpressed the Drosophila Mkk3, Licorne, with esg^ts. Overexpression of Licorne in progenitors caused a significant increase in ISC proliferation after 1 day of expression (Fig 5A). Furthermore, we found that Licorne overexpression in progenitors for 1 day with esg^ts increased p38 activation by 211.1% within progenitors (Figs 5B and S5A–S5B′; S2B Appendix). Interestingly, we also found increased p38 activation in ECs adjacent to Licorne-overexpressing progenitors (S5A–S5B′ Fig), suggesting that p38 activation in progenitors causes p38 activation in ECs. To correlate p38 activation in progenitors with p38 activation in adjacent ECs, we measured the p-p38 fluorescence intensity in progenitors and the adjacent ECs within both control midguts and midguts with progenitors overexpressing Licorne. In midguts with progenitors overexpressing Licorne, we found a high correlation (r = 0.60) between p-p38 levels in progenitors and their adjacent ECs (Fig 5C). This suggests that increases in p38 activation in progenitors correspondingly increases p38 activation in adjacent ECs. In control midguts, we similarly found a high correlation (r = 0.68) between p-p38 levels in progenitors and in adjacent ECs (Fig 5C). This suggests that basal p38 activation in progenitors of unchallenged midguts regulate p38 activation in adjacent ECs. Together, these data show that overexpressing Mkk3 in progenitors can promote ISC proliferation and p38 activation in both progenitors and surrounding ECs.

Since Licorne overexpression stimulated ISC proliferation in the absence of P.e. infection, we next tested whether Licorne was required for ISC proliferation after P.e. infection. We found that depleting licorne with two RNAis (RNAi (1) and RNAi (2)) in progenitors with esg^ts reduced ISC proliferation after P.e. infection (by 73.1% for RNAi (1)) compared to infected control midguts (Figs 5D and S5C). This decrease in ISC proliferation was not due to reduced progenitor number caused by licorne depletion with esg^ts and P.e. infection as the mean percent esg⁺ progenitors in these midguts was similar to uninfected and P.e.- infected control midguts (S5F–S5G Fig). These data suggest that Licorne, likely through p38a and p38b, promotes ISC proliferation after P.e. infection.

We next examined whether Mkk3/Licorne activity in progenitors regulated p38 activation in both the progenitors and throughout the midgut epithelium. We indeed found that depletion of licorne with esg^ts decreased p38 activation in both the progenitors and throughout the midgut epithelium. Depleting licorne in progenitors with two RNAis (RNAis (1) and (2)) decreased p38 activation within progenitors by 87.1% and 102.9%, respectively, after P.e. infection compared to progenitors in infected control midguts (Figs 5E, 5G–5J′ and S5D; S2C Appendix). Further, the same depletion of licorne in progenitors decreased p38 activation throughout the midgut epithelium after infection compared to infected control midguts (Figs 5F–5J′ and S5E).

Together, these data indicate that Mkk3-p38 signalling in progenitors after P.e. infection stimulates ISC proliferation and activates p38 throughout the midgut epithelium.

PGRP-LC/LE-Imd-Mkk3-p38 activity in progenitors activates p38 in the midgut epithelium after pathogenic bacterial infection

So far, we had shown that depleting p38a and p38b, Mkk3, Imd, PGRP-LC or PGRP-LE in midgut progenitors reduced p38 activation after P.e. infection not only within progenitors but also throughout the midgut epithelium. Furthermore, we had found that increasing ISC proliferation through Mkk3/Licorne overexpression led to p38 activation in nearby ECs. Thus, we asked whether it was increased PGRP-LC/LE-Imd-Mkk3-p38 signalling in progenitors that stimulated p38 activation throughout the midgut epithelium or rather increased ISC proliferation. To test this, we blocked ISC proliferation by depleting the Drosophila cdc25 phosphatase homolog, String, in progenitors with esg^ts. This blocked ISC proliferation by 99.4% after P.e. infection compared to infected control midguts (Fig 6A). The decrease in ISC proliferation was not caused by reduced progenitor number due to string depletion with esg^ts and P.e. infection as the mean percent esg⁺ progenitors in these midguts was similar to uninfected and P.e.- infected control midguts (S6A Fig). Furthermore, blocking ISC proliferation did not decrease p38 activation, neither within progenitors (Fig 6B and 6D–6G′; S3A Appendix) nor throughout the epithelium (Fig 6C–6G′). These data suggest that pathogenic bacterial recognition via PGRP-LC/LE-Imd-Mkk3-p38 signalling in progenitors promotes p38 activation in the epithelial regenerative microenvironment.

Recognition of pathogenic bacteria primarily by intestinal stem cells promotes adult midgut regeneration

Because esg⁺ midgut progenitors include ISCs, EBs and EEps, it was not clear in which progenitor cell type(s) the recognition of pathogenic bacteria was important for midgut regeneration. Thus, we tested whether ISCs or EBs sensed pathogenic bacteria. To determine this, we depleted PGRP-LC from ISCs using an ISC-specific driver, esg-GAL4 tubGAL80^ts Su(H)-GAL80 (esg^ts Su(H)-GAL80). This system allowed us to express PGRP-LC^RNAi in ISCs whilst blocking GAL4-mediated expression in Suppressor of Hairless-positive (Su(H)⁺) EBs by GAL80 expression. We found that PGRP-LC depletion from ISCs decreased ISC proliferation by 48.3% after P.e. infection compared to infected control midguts (Fig 7 A). Furthermore, PGRP-LC depletion strongly reduced p38 activation throughout the midgut epithelium after P.e. infection (Fig 7 B–7F). In contrast, we found a milder reduction in ISC proliferation (20.5%) and p38 activation in the midgut epithelium after P.e. infection when PGRP-LC was depleted from EBs using the EB-specific driver, w; Su(H)Gbe-GAL4; tubGAL80^ts UAS-GFP (Su(H)^ts) (Fig 7 G–7H). These data suggest that ISCs are the primary sensors of pathogenic bacteria amongst progenitors after infection.

We similarly asked whether p38a and p38b were required in ISCs or in EBs for ISC proliferation and p38 activation throughout the midgut epithelium. To determine this, we depleted p38a and p38b from ISCs using esg^ts Su(H)-GAL80. We found that p38a and p38b depletion in ISCs strongly reduced their proliferation after P.e. infection by 59.6% compared to infected control midguts (Fig 7 I). Further, p38a and p38b depletion in ISCs strongly reduced p38 activation throughout the midgut epithelium after infection compared to infected control midguts (Fig 7 J). In contrast, depleting p38a and p38b in EBs did not affect ISC proliferation after P.e. infection (S6B Fig).

Together, these data indicate that PGRP-LC-p38 signalling is essential in ISCs to detect pathogenic bacteria after infection and promote their proliferation.

In summary, we have uncovered a mechanism by which pathogenic bacterial infection promotes ISC proliferation in the adult Drosophila midgut. We find that pathogenic bacteria are primarily recognised by ISCs through PGRP-LC and PGRP-LE, activating Imd-Mkk3-p38 signalling within ISCs. This p38 activity stimulates ISC proliferation but also stimulates p38 activation throughout the midgut epithelium—a component of the regenerative microenvironment (Fig 8).

Fly stocks

The following fly stocks were used: w¹¹¹⁸, yw, w; esg-GAL4; tubGAL80^ts UAS-GFP (esg^ts), esg-GAL4 UAS-2X EYFP; Su(H)Gbe-GAL80, tubGAL80^ts (esg^ts Su(H)-GAL80), w; Su(H)Gbe-GAL4; tubGAL80^ts UAS-GFP (Su(H)^ts) and mex-GAL4; tubGAL80^ts (mex^ts). UAS-imd^RNAi (101834KK) [53], UAS-licorne^RNAi (1) (106822KK) [39], w; UAS-p38a+b^RNAi (1) (34238GD) [39], w; UAS-p38a+b^RNAi (2) (52277GD) [39], w; UAS-PGRP-LC^RNAi (1) (51968GD), UAS-PGRP-LC^RNAi (2) (101636KK) [46,54], UAS-PGRP-LE^RNAi (108199KK) [46] and KK control line (60100KK) from the Vienna Drosophila Stock Center (VDRC). PGRP^Δ5, Dipt2.2-lacZ; Drs-GFP, yw, w; UAS-imd^D30A/MRKS, UAS-mCherry^RNAi, UAS-licorne^RNAi (2) (JF01433) and UAS-string^RNAi (HMS00146) from the Bloomington Drosophila Stock Center (BDSC). yw; UAS-licorne [39] from the Zürich Fly ORFeome project (FlyORF) and UAS-p38b^antisense [55] from the Kyoto Stock Center. PGRP-LC^E12 and PGRP-LE¹¹² are from Bruno Lemaitre (École Polytechnique Fédérale de Lausanne, EPFL) and w; imd¹, isoDrosDel Rel^E20 and isoDrosDel w¹¹¹⁸ are from Mark Hanson (University of Exeter).

Drosophila melanogaster housing and husbandry

Drosophila melanogaster were housed at either 18°C or 25°C (65% humidity) under a light-dark cycle and fed a Drosophila medium containing water, agar, glucose, dry yeast, wheat flour, 10% Nipagin and propionic acid.

Drosophila genetics

Flies raised at 18°C were shifted to 29°C to induce GAL4-mediated UAS-transgene expression. Experiments were performed using 5–10 day old, adult, female Drosophila melanogaster. Typically, between 15 and 20 flies were used per experiment. Flies were first selected based on genotype, then randomly chosen for use in experiments. Genotypes for all experiments can be found in S1 and S2 Tables.

Pathogenic bacterial culture and preparation for oral infection

Pseudomonas entomophila (P.e.) was grown by inoculating 3 mL of Luria Broth (LB) with 100 µg/mL rifampicin with a single colony of P.e. grown on a LB-rifampicin plate. The 3 mL culture was grown for 24 hours at 30°C at 130 RPM. 500 µL of this culture was diluted into 50 mL of fresh LB with rifampicin, which was then grown for 18 hours at 30°C at 130 RPM. The bacteria were then centrifuged at 3000g for 30 min. The bacterial pellet was then resuspended in 2 mL of 5% sucrose and diluted in 5% sucrose to CFU (colony forming unit) = 7.0 × 10⁹ bacteria/mL. Non-virulent P.e. ΔgacA were prepared similarly but resuspended to CFU = 4.1 × 10¹⁰ bacteria/mL. For feeding of heat-killed P.e., P.e. were prepared as above and then 1 mL of resuspended P.e. (CFU = 7.0 × 10⁹ bacteria/mL) was heated at 95°C for 30 min. P.e. viability after heat treatment was determined by streaking out lysates onto a LB-agar plate containing 100 µg/mL rifampicin. P.e. ΔgacA was provided by Bruno Lemaitre (École Polytechnique Fédérale de Lausanne, EPFL).

Erwinia carotovora carotovora 15 (Ecc15) was grown by inoculating 3 mL of Luria Broth (LB) with 100 µg/mL rifampicin with a single colony of Ecc15 grown on a LB-rifampicin plate. The 3 mL culture was grown for 24 hours at 30°C at 130 RPM. 2.5 mL of this culture was diluted into 250 mL of fresh LB with rifampicin, which was then grown for 18 hours at 30°C at 130 RPM. The bacteria were then centrifuged at 3000g for 30 min. The bacterial pellet was then resuspended in 1 mL of 5% sucrose and diluted in 5% sucrose to CFU = 3.8 × 10¹¹ bacteria/mL.

Oral infection of Drosophila with bacteria and feeding of heat-killed bacteria

Flies were fed food supplemented with 500 µL of 5% sucrose or P.e., P.e. ΔgacA or Ecc15 resuspended in 5% sucrose. This was done by adding 500 µL of each to the top layer of the food within a vial (25 mm wide) and mixing it into the food to a depth of approximately 3.7 mm. Heat-killed P.e. was fed similarly. Flies were continuously exposed to P.e., heat-killed P.e., P.e. ΔgacA or Ecc15 for 18 hours or for 12 hours in Fig 2I–2J and S1B Appendix.

Peptidoglycan feeding

Flies were exposed for 24 hours with a 45x26 mm rectangular Whatman filter paper soaked with either 250 µL of 5% sucrose or 5 mg/mL DAP-type peptidoglycan (PGN) suspended in 5% sucrose. PGN was an ultrapure preparation of DAP-type peptidoglycan isolated from Escherichia coli K12 (InvivoGen). 5 mg of PGN was added to 1 mL LAL reagent water and sonicated six times using a Bioruptor UCD-200 (Diagenode) at 320 W for 30 seconds in an ice bath, followed by 30 seconds of rest. Sucrose was then added to make a solution of 5 mg/mL PGN in 5% sucrose.

Histology

Adult Drosophila midguts were dissected in cold PBS (phosphate-buffered saline) and then fixed in 8% formaldehyde containing phosSTOP (Roche) for 1 hour. After fixation, midguts were washed at least three times in PBS, 0.1% Triton X-100 for 10–15 min. Midguts were then blocked for at least 30 min in blocking solution containing PBS, 0.1% Triton X-100, 1% bovine serum albumin (BSA) and 2% normal goat serum (NGS). After blocking, midguts were incubated overnight with primary antibodies in blocking solution at 4°C. The following primary antibodies were used: rabbit polyclonal anti-phospho Ser 10 histone 3 (Millipore; 06-570; 1:1000), mouse monoclonal anti-phospho Ser 10 histone 3 (Sigma-Aldrich; 3H10; 05-806; 1:1000), rabbit polyclonal anti-phosphorylated p38 (Thr180/Tyr182) (Cell Signaling; 9211; 1:200) and chicken polyclonal anti-ß-galactosidase (abcam; ab9361; 1:1000). Midguts were then washed three times as above in PBS containing 0.1% Triton X-100, stained in PBS, 0.3% Triton X-100, 0.1% BSA with Alexa Fluor-conjugated secondary antibodies (Life Technologies; 1:1000) and Hoechst 33258 (Life Technologies) and mounted in Vectashield (Vector Laboratories). Specificity of the antibody against phosphorylated p38 (Thr180/Tyr182) was previously tested in Patel and colleagues, 2019 [39].

Microscopy and image processing

Samples were analysed using a Leica DMI6000 epifluorescence microscope and a Leica SP5-II confocal laser scanning microscope. Images were processed using ImageJ (NIH) and Adobe Photoshop. Confocal images are presented as maximum intensity projections or as single slices. Each z-stack for each midgut within an experiment was acquired with the same laser intensity and gain. Regions R4a and R4b were imaged from randomly chosen midguts within a sample. For single slice images to which the Fire LUT (Look-Up Table) was applied, the range (minimum and maximum pixel) was set from 0 to 255 in ImageJ prior to applying the LUT.

The number of mitoses per midgut was determined by counting the number of phospho-Ser 10 histone 3-positive cells from whole adult female midguts. The mean number of mitoses per midgut and 95% confidence interval are presented for each genotype and treatment.

Image quantitation

Phosphorylated p38 in progenitors: the Modular Image Analysis (MIA) plugin in Fiji (ImageJ) was used to trace around and individually track esg > GFP⁺ progenitors across a series of 10 consecutive slices from 0.25 or 0.5 μm z-stacks of the progenitor layer to generate a series of linked regions of interest (ROIs). These linked ROIs were applied onto corresponding slices of the p-p38 channel to measure the mean raw integrated density of each progenitor across the 10 slices. The mean raw integrated density of each progenitor was divided by the mean projected area of each progenitor to obtain the normalised p-p38 fluorescence intensity per progenitor. To exclude GFP⁺ early ECs from our analysis, we used a macro that we developed to identify cells with a nuclear size comparable to those of esg > GFP⁺ progenitors found in homeostatic midguts.

Phosphorylated p38 in the midgut epithelium: sum intensity projections of 10 slices from 0.5 µm z-stacks of the progenitor layer from regions R4a and R4b were generated for the p-p38 channel using ImageJ. The freehand or polygon selection tool was used to trace around the epithelium to measure the raw integrated density and epithelial area. The raw integrated density of p-p38 was then divided by the area of selected epithelium to obtain the normalised fluorescence intensity of p-p38 in the whole epithelium. These intensities were then further normalised by subtracting the background (raw integrated density/area) determined from a region of the image without tissue to obtain the final normalised p-p38 fluorescence intensity for the whole epithelium.

Phosphorylated-p38 measurements between progenitors and adjacent ECs: a slice from a 0.25 µm z-stack containing the majority of progenitors within the field was selected from the esg > GFP channel using ImageJ. A line (thickness = 5) was drawn between a progenitor and all adjacent ECs, starting within the progenitor. The middle point of the line was placed between these cells. The grey value along the line was determined from the corresponding slice from the p-p38 channel. The first half of the grey values were attributed to the progenitor, with the latter half to the adjacent EC. The grey values were then averaged. This process was repeated for all progenitors in the field. Average grey values for progenitors were graphed against those for adjacent ECs, and a Pearson’s correlation coefficient was determined using GraphPad Prism.

% esg-positive cells: maximum intensity projections of 10 slices from 0.25, 0.5 or 1 µm z-stacks of the progenitor layer from regions R4a and R4b were generated for the esg > GFP and DNA (Hoechst) channels in ImageJ. The total number of esg > GFP⁺ cells with small nuclei and the total number of nuclei were counted. The number of esg > GFP⁺ cells with small nuclei was then divided by the total number of nuclei to obtain % esg-positive cells.

Quantitative RT-PCR

RNA was isolated from 24 midguts using the PicoPure RNA extraction kit and DNase treated during RNA isolation using the Qiagen RNase-free DNase Set. The RNA concentration was determined using a NanoDrop ND-1000 spectrophotometer. 200−250 ng of RNA was used for cDNA synthesis using the QuantiTect Reverse Transcription Kit or the Superscript IV First-Strand Synthesis System. qPCR was performed using the QuantiNova SYBR Green PCR Kit and Applied Biosystems QuantStudio 5 Real-Time qPCR System. Croquemort (Crq) was used as a reference. Each assay was performed in triplicate on two independent biological replicates. Fold changes were determined using the comparative Ct (ΔΔCt) method. A standard curve was performed to calculate primer efficiencies, which were between 90% and 110% for all primers. The following primers were used: PGRP-LE_F 5′-ACAGAAGCCCATGGACGAGC-3′; PGRP-LE_R 5′-ATCCCCGCGACTCAATGTGG-3′; PGRP-LC_F 5′-GCCGAAGCGGAGGATTATACGG-3′; PGRP-LC_R 5′-TGGTTAAGGTCGCGCAGGTG-3′; Crq_F 5′-CAGAGCTCTCCTCCGAATTG-3′ and Crq_R 5′-ATGCCGGTGATGAGAAAGAC-3′.

Statistical analysis

Statistical analyses were conducted using GraphPad Prism 10. A two-sided Mann–Whitney test was applied to determine statistical significance. The significance level is denoted by an * for p ≤ 0.05, ** for p ≤ 0.01, *** p ≤ 0.001, and **** for p ≤ 0.0001, and ns, for not significant, p > 0.05.

S1 Fig. PGRP-LC and PGRP-LE promote for ISC proliferation and p38 activation throughout the midgut epithelium after pathogenic bacterial infection.

PGRP-LE is required for epithelial p38 activation after P.e. infection (A–D). Epithelial p-p38 (red) levels increased in infected control midguts (B) compared to uninfected control midguts (A). Epithelial p-p38 levels decreased in infected PGRP-LE¹¹² midguts (D) compared to infected control midguts (B). Epithelial p-p38 levels were similar between uninfected control midguts (A) and uninfected PGRP-LE¹¹² midguts (C). PGRP-LC is required for ISC proliferation after P.e. infection (E). The mean mitoses per midgut increased in infected control midguts (n = 34 pooled from two experiments) compared to uninfected control midguts (n = 36 pooled from two experiments) (Mann–Whitney; p < 0.0001). The mean mitoses per midgut decreased in infected PGRP-LC^Δ5 midguts (n = 34 pooled from two experiments) compared to infected control midguts (Mann–Whitney; p = 0.0032) (E). Data in A–D, acquired from R4a-b region. In E, mean (red line) and 95% confidence intervals (blue bars) are shown. DNA is in blue. Scale bars are 20 µm. Numerical data for this figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003951.s001

(TIF)

S2 Fig. PGRP-LC in progenitors but not in enterocytes stimulates ISC proliferation and p38 activation after pathogenic bacterial infection.

PGRP-LC is required in progenitors and not enterocytes (ECs) for ISC proliferation after P.e. infection (A–B). The mean mitoses per midgut increased in infected control midguts (n = 34 pooled from two experiments) compared to uninfected control midguts (n = 39 pooled from two experiments) (Mann–Whitney; p < 0.0001). The mean mitoses per midgut increased in infected midguts expressing PGRP-LC^RNAi (1) in ECs (n = 34 pooled from two experiments) compared to infected control midguts (Mann–Whitney; p = 0.0156) (A). The mean mitoses per midgut increased in infected control midguts (n = 23 from one experiment) compared to uninfected control midguts (n = 24 from one experiment) (Mann–Whitney; p < 0.0001). The mean mitoses per midgut decreased in infected midguts expressing PGRP-LC^RNAi (2) in progenitors (n = 19 from one experiment) compared to infected control midguts (Mann–Whitney; p = 0.0002) (B). Quantitative real-time PCR showed a decrease in both PGRP-LC and PGRP-LE transcripts in whole midguts after expressing PGRP-LC^RNAi (1) and PGRP-LE^RNAi in ECs (n = two experiments for both) (C–D). PGRP-LE in progenitors promotes p38 activation in progenitors and throughout the midgut epithelium after P.e. infection (E–H). p-p38 (E, F: red; E′, F′: Fire LUT) increased in progenitors (F: GFP; F–F′: green-dashed outline) and throughout the midgut epithelium after P.e. infection (F–F′) compared to uninfected control midguts (E–E′). p-p38 (H: red; H′; Fire LUT) decreased in progenitors and throughout the midgut epithelium in midguts expressing PGRP-LE^RNAi with esg^ts (H–H′) compared to infected control midguts (F–F′). Depleting PGRP-LC or PGRP-LE from progenitors does not affect their number after P.e. infection (I–J). The percent esg⁺ cells was similar between infected and uninfected control midguts and between infected control midguts and infected midguts expressing PGRP-LC^RNAi (1) or PGRP-LE^RNAi (n = 6 from one of two experiments) (I–J). PGRP-LC signalling in progenitors promotes ISC proliferation and epithelial p38 activation after Ecc15 infection (K–L). The mean mitoses per midgut increased in Ecc15-infected control midguts (n = 18 from one experiment) compared to uninfected control midguts (n = 21 from one experiment) (Mann–Whitney; p < 0.0001) (K). The mean mitoses per midgut decreased in infected midguts expressing PGRP-LC^RNAi (1)′ in progenitors (n = 19 from one experiment) compared to infect′ed control midguts (Mann–Whitney; p = 0.0128) (K). Epithelial p-p38 ′fluorescence intensity increased in Ecc15-infected control midguts ′′(n = 9 from one experiment) compared to uninfected control midguts (n = 9 from one experiment). Epithelial p-p38 fluorescence intensity decreased in infected midguts expressing PGRP-LC^RNAi (1) in progenitors′ (n = 9 from one experiment) compared to infected control midguts (L). Data in E–J and L, acquired from R4a-b region. In A-B and K, mean (red line) and 95% confidence intervals (blue bars) are shown. In I–J and L, standard deviation (blue bars) is shown. DNA is in blue. Scale bars are 20 µm. Numerical data for this figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003951.s002

(TIF)

S3 Fig. DAP-type peptidoglycan and non-virulent P.e. elicit an innate immune response.

Inhibiting p38a and p38b in progenitors does not affect their number after P.e. infection (A–B). The percent esg⁺ cells was similar between infected and uninfected control midguts and between infected control midguts and infected midguts expressing p38a+b^RNAi (1) (n = 3–5 from one experiment) or p38b^antisense (n = 5 from one experiment) (A, B). PGRP-LE signalling in progenitors promotes ISC proliferation after exposure to heat-killed P.e. (C). The mean mitoses per midgut increased in control midguts exposed to heat-killed P.e. (n = 17 from one experiment) compared to control midguts (n = 19 from one experiment) (Mann–Whitney; p < 0.0001). The mean mitoses per midgut decreased in midguts expressing PGRP-LE^RNAi in progenitors exposed to heat-killed P.e. (n = 18 from one experiment) compared to control midguts exposed to heat-killed P.e. (Mann–Whitney; p = 0.0001) (C). Diptericin expression increases in the midgut epithelium after exposure to DAP-type peptidoglycan or gram-negative pathogenic bacteria (D–G). Diptericin expression reporter (Dipt2.2-lacZ) (D–G′; B-gal, red) was activated in midgut enterocytes exposed to DAP-type peptidoglycan, P.e. or P.e. ΔgacA. Data in A, B, acquired from R4a-b region. Data in D–G′, acquired from R5. In C, mean (red line) and 95% confidence intervals (blue bars) are shown. In A–B, standard deviation (blue bars) is shown. DNA is in blue. Scale bars are 20 µm. Numerical data for this figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003951.s003

(TIF)

S4 Fig. Imd signalling in progenitors stimulates ISC proliferation and p38 activation after pathogenic bacterial infection.

ISC proliferation is not consistently affected in imd mutant midguts after P.e. infection (A–B). The mean mitoses per midgut increased in infected control midguts (n = 50 pooled from three experiments) compared to uninfected control midguts (n = 47 pooled from three experiments) (Mann–Whitney; p < 0.0001). The mean mitoses per midgut in infected imd¹ midguts (n = 49 pooled from three experiments) were similar to infected control midguts (Mann–Whitney; p = 0.1049) (A). The mean mitoses per midgut for each genotype and treatment per experiment (n = 3) described in A (B). Imd activity in progenitors promotes ISC proliferation and p38 activation throughout the midgut epithelium after P.e. infection (C–D). The mean mitoses per midgut increased in infected control midguts (n = 19 from one of two experiments) compared to uninfected control midguts (n = 18 from one of two experiments) (Mann–Whitney; p < 0.0001) (C). The mean mitoses per midgut decreased in infected midguts expressing imd^RNAi in progenitors (n = 19 from one of two experiments) compared to infected control midguts (Mann–Whitney; p < 0.0001) (C). Epithelial p-p38 fluorescence intensity increased in infected control midguts (n = 6 from one experiment) compared to uninfected control midguts (n = 6 from one experiment). Epithelial p-p38 fluorescence intensity decreased in infected midguts expressing imd^RNAi in progenitors (n = 6 midguts from one experiment) compared to infected control midguts (D). Inhibiting Imd signalling in progenitors does not decrease their number after P.e. infection (E). The percent esg⁺ cells mildly increased between infected control midguts (n = 6 from one of two experiments) and uninfected control midguts (n = 6 from one of two experiments). The percent esg⁺ cells further increased in infected midguts expressing imd^D30A in progenitors (n = 6 from one of two experiments) compared to infected control midguts (E). Relish is not required for ISC proliferation after P.e. infection (F). The mean mitoses per midgut increased in infected control midguts (n = 17 from one experiment) compared to uninfected control midguts (n = 14 from one experiment) (Mann–Whitney; p < 0.0001). The mean mitoses per midgut in infected Rel^E20 midguts (n = 17 from one experiment) was similar to infected control midguts (Mann–Whitney; p = 0.3890) (F). Data in D–E, acquired from R4a-b region. In A, C and F, mean (red line) and 95% confidence intervals (blue bars) are shown. In B and D–E, standard deviation (blue bars) is shown. Numerical data for this figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003951.s004

(TIF)

S5 Fig. Mkk3/Licorne is required and sufficient to promote ISC proliferation and p38 activation after pathogenic bacterial infection.

Licorne overexpression in progenitors increased p38 activation in progenitors and in adjacent ECs (A–B). p-p38 (A, B: red; A′, B′: Fire LUT) increased in midgut progenitors (A, B: GFP; A′, B′, green-dashed outline) and in nearby epithelial cells after overexpressing licorne with esg^ts compared to control midguts (A–B′) (n = one of two experiments). Licorne activity in progenitors promotes ISC proliferation after P.e. infection (C). The mean number of mitoses per midgut increased in infected control midguts (n = 21 from one experiment) compared to uninfected control midguts (n = 19 from one experiment) (Mann–Whitney; p = 0.0169). The mean number of mitoses per midgut decreased in infected midguts expressing licorne^RNAi (2) in progenitors (n = 15 from one experiment) compared to infected control midguts (Mann–Whitney; p = 0.0465) (C). Licorne is required in progenitors for p38 activation in progenitors and throughout the midgut epithelium after P.e. infection (D–E). The progenitor p-p38 fluorescence intensity increased in infected control midguts (n = 5 from one experiment) compared to progenitors in uninfected control midguts (n = 5 from one experiment). The progenitor p-p38 fluorescence intensity decreased in infected midguts expressing licorne^RNAi (2) with esg^ts (n = 5 from one experiment) compared to progenitors in infected control midguts (D). The epithelial p-p38 fluorescence intensity per midgut from those analysed in D are shown in E. The epithelial p-p38 fluorescence intensity increased in infected control midguts compared to uninfected control midguts. The epithelial p-p38 fluorescence intensity decreased in infected midguts expressing licorne^RNAi (2) in progenitors compared to infected control midguts (E). Depleting licorne from progenitors does not affect their number after P.e. infection (F-G). The percent esg⁺ cells was similar between infected and uninfected control midguts and between infected control midguts and infected midguts expressing licorne^RNAi (1) (n = 6 from one of two experiments) and licorne^RNAi (2) (n = 5 from one experiment) (F–G). Data in A–B′ and D–G, acquired from R4a-b region. In C, mean (red line) and 95% confidence intervals (blue bars) are shown. In D–G, standard deviation (blue bars) is shown. DNA is in blue. Scale bars are 20 µm. Numerical data for this figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003951.s005

(TIF)

S6 Fig. p38 activity in enteroblasts does not promote ISC proliferation after pathogenic bacterial infection.

Depleting string from progenitors does not affect their number after P.e. infection (A). The percent esg⁺ cells was similar between infected and uninfected control midguts and between infected control midguts and infected midguts expressing string^RNAi in progenitors (n = 7 from one of two experiments) (A). p38 activity is not required in enteroblasts (EBs) for ISC proliferation after P.e. infection (B). The mean number of mitoses per midgut increased in infected control midguts (n = 36 pooled from two experiments) compared to uninfected control midguts (n = 35 pooled from two experiments) (Mann–Whitney; p < 0.0001). The mean number of mitoses per midgut in infected midguts expressing p38a+b^RNAi (1) in EBs (n = 35 pooled from two experiments) was similar to infected control midguts (Mann–Whitney; p = 0.3462) (B). Data in A, acquired from R4a-b region. In B, mean (red line) and 95% confidence intervals (blue bars) are shown. In A, standard deviation (blue bars) is shown. Numerical data for this figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003951.s006

(TIF)

S1 Appendix. PGRP-LC and PGRP-LE promote p38 activation in progenitors after pathogenic bacterial infection.

The distribution (histogram) of the normalised p-p38 fluorescence intensity per progenitor is shown for uninfected control midguts (n = 175), P.e.-infected control midguts (n = 142), uninfected midguts expressing PGRP-LC^RNAi (1) in progenitors (n = 233) and P.e.-infected midguts expressing PGRP-LC^RNAi (1) in progenitors (n = 153). The mean p-p38 level decreased by 105.0% in progenitors of infected midguts expressing PGRP-LC^RNAi (1) with esg^ts compared to progenitors in infected control midguts (Mann–Whitney; p < 0.0001). Basal p-p38 in progenitors of uninfected midguts expressing PGRP-LC^RNAi (1) with esg^ts was similar to progenitors in uninfected control midguts (A). The distribution (histogram) of the normalised p-p38 fluorescence intensity per progenitor is shown for uninfected control midguts (n = 108), P.e.-infected control midguts (n = 202), uninfected midguts expressing PGRP-LE^RNAi in progenitors (n = 117) and P.e.-infected midguts expressing PGRP-LE^RNAi in progenitors (n = 178). The mean p-p38 level decreased by 74.2% in progenitors of infected midguts expressing PGRP-LE^RNAi with esg^ts compared to progenitors in infected control midguts (Mann–Whitney; p < 0.0001). Basal p-p38 in progenitors of uninfected midguts expressing PGRP-LE^RNAi with esg^ts was similar to progenitors in uninfected control midguts (B). The distribution (histogram) of p-p38 fluorescence intensity per progenitor is shown for uninfected control midguts (n = 143), P.e.-infected control midguts (n = 143), uninfected midguts expressing p38b^antisense in progenitors (n = 136) and P.e.-infected midguts expressing p38b^antisense in progenitors (n = 142). The mean p-p38 level decreased by 87.8% in progenitors of infected midguts expressing p38b^antisense with esg^ts compared to progenitors in infected control midguts (Mann–Whitney; p < 0.0001). Basal p-p38 in progenitors of uninfected midguts expressing p38b^antisense with esg^ts was similar to progenitors in uninfected control midguts (C). Data acquired from R4a-b. Histogram x-axis bin width is 50 (A–C). Numerical data for this figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003951.s007

(TIF)

S2 Appendix. Imd and Mkk3/Licorne promote p38 activation in progenitors after pathogenic bacterial infection.

The distribution (histogram) of the normalised p-p38 fluorescence intensity per progenitor is shown for uninfected control midguts (n = 181), P.e.-infected control midguts (n = 187), uninfected midguts expressing imd^D30A in progenitors (n = 164) and P.e.-infected midguts expressing imd^D30A in progenitors (n = 173) (A). The mean p-p38 level decreased by 43.9% in progenitors of infected midguts expressing imd^D30A with esg^ts compared to progenitors in infected control midguts (Mann–Whitney; p < 0.0001). Basal p-p38 in progenitors of uninfected midguts expressing imd^D30A with esg^ts was similar to progenitors in uninfected control midguts (A). The distribution (histogram) of the normalised p-p38 fluorescence intensity per progenitor is shown for control midguts (n = 140) and midguts overexpressing licorne in progenitors (n = 184) (B). The mean p-p38 level increased by 211.1% in progenitors of midguts overexpressing licorne with esg^ts compared to progenitors in control midguts (Mann–Whitney; p < 0.0001) (B). The distribution (histogram) of the normalised p-p38 fluorescence intensity per progenitor is shown for uninfected control midguts (n = 174), P.e.-infected control midguts (n = 225), uninfected midguts expressing licorne^RNAi (1) in progenitors (n = 212) and P.e.-infected midguts expressing licorne^RNAi (1) in progenitors (n = 191). The mean p-p38 level decreased by 87.1% in progenitors of infected midguts expressing licorne^RNAi (1) with esg^ts compared to progenitors in infected control midguts (Mann–Whitney; p < 0.0001). Basal p-p38 in progenitors of uninfected midguts expressing licorne^RNAi (1) with esg^ts was similar to progenitors in uninfected control midguts (C). Data acquired from R4a-b. Histogram x-axis bin width is 50 (A), 200 (B) or 100 (C). Numerical data for this figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003951.s008

(TIF)

S3 Appendix. Blocking ISC proliferation does not affect p38 activation in progenitors after pathogenic bacterial infection.

The distribution (histogram) of the normalised p-p38 fluorescence intensity per progenitor is shown for uninfected control midguts (n = 121), P.e.-infected control midguts (n = 127), uninfected midguts expressing string^RNAi in progenitors (n = 143) and P.e.-infected midguts expressing string^RNAi in progenitors (n = 166) (A). The mean p-p38 level increased by 205.2% in progenitors of infected midguts expressing string^RNAi with esg^ts compared to progenitors in infected control midguts (Mann–Whitney; p < 0.0001). Basal p-p38 in progenitors of uninfected midguts expressing string^RNAi with esg^ts was similar to progenitors in uninfected control midguts (A). Data acquired from R4a-b. Histogram x-axis bin width is 100 (A). Numerical data for this figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003951.s009

(TIF)