Socs36E Controls Niche Competition by Repressing MAPK Signaling in the Drosophila Testis

Abstract

The Drosophila testis is a well-established system for studying stem cell self-renewal and competition. In this tissue, the niche supports two stem cell populations, germ line stem cells (GSCs), which give rise to sperm, and somatic stem cells called cyst stem cells (CySCs), which support GSCs and their descendants. It has been established that CySCs compete with each other and with GSCs for niche access, and mutations have been identified that confer increased competitiveness to CySCs, resulting in the mutant stem cell and its descendants outcompeting wild type resident stem cells. Socs36E, which encodes a negative feedback inhibitor of the JAK/STAT pathway, was the first identified regulator of niche competition. The competitive behavior of Socs36E mutant CySCs was attributed to increased JAK/STAT signaling. Here we show that competitive behavior of Socs36E mutant CySCs is due in large part to unbridled Mitogen-Activated Protein Kinase (MAPK) signaling. In Socs36E mutant clones, MAPK activity is elevated. Furthermore, we find that clonal upregulation of MAPK in CySCs leads to their outcompetition of wild type CySCs and of GSCs, recapitulating the Socs36E mutant phenotype. Indeed, when MAPK activity is removed from Socs36E mutant clones, they lose their competitiveness but maintain self-renewal, presumably due to increased JAK/STAT signaling in these cells. Consistently, loss of JAK/STAT activity in Socs36E mutant clones severely impairs their self-renewal. Thus, our results enable the genetic separation of two essential processes that occur in stem cells. While some niche signals specify the intrinsic property of self-renewal, which is absolutely required in all stem cells for niche residence, additional signals control the ability of stem cells to compete with their neighbors. Socs36E is node through which these processes are linked, demonstrating that negative feedback inhibition integrates multiple aspects of stem cell behavior.

Fig 1. MAPK signaling is elevated in Socs36E mutant clones.

A) Model of the Drosophila JAK/STAT pathway. The ligand Unpaired (Upd, magenta) is produced by hub cells and binds to and activates the receptor Domeless (Dome, gray) on the surface of CySCs. This results in activation of the JAK Hopscotch (Hop) (blue), leading to tyrosine phosphorylation (maroon circles) of Dome. The phosphorylated cytoplasmic domain of the receptor serves as a docking site for a Stat92E dimer (green). Stat92E is phosphorylated leading to the generation of an active Stat92E dimer that translocates to the nucleus, binds to a consensus TTCNNNGAA site, and alters gene expression. Socs36E is one of the best-characterized Stat92E target genes and encodes a negative regulator of JAK/Receptor activity (orange). B) Model of the MAPK pathway. The EGF ligand Spitz (Spi, magenta) is produced by germ line cells. Spi activates the EGF receptor (Egfr, dark gray) on the surface of CySCs, which triggers the canonical MAPK pathway. Activation of Egfr leads to the recruitment of Son of Sevenless (Sos, dark blue), a guanine exchange factor for Ras85D (Ras) that converts Ras from an inactive GDP-loaded form (light gray) to an active GTP-loaded form (yellow). Ras activates Rolled (Rl, blue), a Drosophila MAPK, which inhibits transcriptional repressors (orange) Yan and Capicua (Cic) and activates Pointed (Pnt), a transcriptional activator (green). C-E) MAPK activity (dpERK, red) in control clones, Egfr mutant clones and Socs36E mutant clones. Clones are negatively-marked and are identified by the lack of GFP. C) dpERK can be detected in a control CySC clone (C,C’, arrow), which lacks GFP (green), at similar levels to a neighboring wild type CySC (C,C’, arrowhead). These data indicate that MAPK activity is normally present in CySCs. MAPK activity (dpERK) is decreased in an Egfr mutant CySC (D,D’, arrow), compared to neighboring wild type CySCs (D,D’, arrowheads). E) dpERK is elevated in Socs36E mutant CySCs (E,E’ arrows) compared to a neighboring wild type CySC (E,E’, arrowhead). F) Quantification of fluorescence intensity in clones, expressed as a ratio of intensity within a CySC compared to that of its immediate unmarked wild type neighbor CySC. Asterisks indicate statistical significance ** P<0.01 and *** P<0.001. Tj is blue in C-E. The hub is outlined by a dotted line. Scale bar = 5 μM.

Fig 2. MAPK gain-of-function in CySCs causes them to out-compete GSCs at the niche.

A) A normal complement of Vasa-positive GSCs and Zfh1-positive CySCs at the niche and Eya-positive differentiating cyst cells is observed in a control a Tj-Gal4 (Tj>+) testis. B) GSCs are displaced by CySCs over-expressing dominant-active form of MAPK (Rl^SEM). C) GSCs are almost entirely out-competed by CySCs over-expressing a dominant-active Ras (Ras^V12). In this image, only one GSC remains in contact with the niche and there are numerous CySCs immediately contacting hub cells. We note that the position of the hub frequently shifts slightly basally in Tj>Ras^V12 over-expressing testes. D) A testis with control MARCM clones at 14 dpci, identified by the expression of GFP within the clone. Note that some CySCs and their differentiating progeny are clonally marked with GFP and there is a normal complement of GSCs at the niche. E) A testis with UAS-Ras^V12 MARCM clones at 14 dpci. The entire somatic lineage is clonally marked, indicating that Ras^V12-expressing CySCs out-competed wild type CySCs. There are two GSCs in contact with the niche, indicating that Ras^V12-expressing CySCs also out-competed most GSCs in this testis. Germ line differentiation is also perturbed, presumably because somatic cells with high Ras activity cannot adequately support gonial cells [44]. F) Graph displaying the number of GSCs in control (Tj-Gal4), Tj>Rl^SEM, Tj>λTop, and Tj>Ras^V12 testes. GSC number is moderately but significantly reduced in Tj>Rl^SEM and Tj>λTop and substantially reduced in Tj>Ras^V12 testes. G) Graph displaying the number of GSCs in testes containing control, Rl^SEM-expressing and Ras^V12 -expressing CySC clones. Note the significant drop in GSCs when CySC clones expressing UAS-Rl^SEM and UAS-Ras^V12 are present. **P< 0.01, ***P<0.001. Vasa is red in all panels. Zfh1 is green and Fas3 and Eya are blue in A-C. MARCM clones, which express GFP, are green and Tj is blue in D and E. The hub is outlined by a dotted line. Scale bar = 5 μM.

Fig 3. MAPK signaling regulates CySC numbers.

A) A representative testis from an Egfr/+ animal contains a rosette of GSCs in contact with the niche surrounded by a row of Zfh1-positive CySCs. The arrow highlights a differentiating cyst cell that has started to express Eya. B) A testis from an Egfr^ts animal that was upshifted to the restrictive temperature of 29°C for 10 days. Note the accumulation of many small Vasa-positive early germ cells and a reduction in the number of Zfh1-positive cells. Arrows mark somatic cells close to the niche that have turned on expression of the differentiation marker Eya. C) Markers of the somatic lineage in a control Tj-Gal4 (Tj>+) testis. Tj (blue) is expressed at low levels in the hub and at higher levels in CySCs and early cyst cells. Zfh1-positive CySCs (green) are observed near the niche and Eya (red) begins to be expressed in differentiating cyst cells. Arrow marks a differentiating cyst cell that is several rows away from the niche and that has upregulated expression of Eya. D) Tj>MAPK RNAi testis. Note the reduction in Zfh1-positive cells and upregulation of Eya in somatic cells close to the niche, which is not observed in controls (compare to arrow in C). E) A normal complement of Vasa-positive GSCs and large differentiating spermatogonia in a control Tj-Gal4 (Tj>+) testis. F) In a Tj>MAPK RNAi testis, there is a block in germ cell differentiation, an accumulation of many small early germ cells and a total lack of differentiating spermatogonia, similar to Egfr^ts testes at the restrictive temperature (see B). G,H) Graphs showing the number of CySCs, defined as Zfh1-positive, Eya-negative cells, in Egfr/+, Egfr^ts, control (Tj-Gal4), Tj>MAPK RNAi, Tj>λTop, and Tj>Rl^SEM testes. (G) Note a significant drop in CySCs in Egfr^ts testes compared to Egfr/+. H) Note significant decrease in CySCs in Tj>MAPK RNAi testes compared to control and a significant increase in the number of CySCs in Tj>λTop and Tj>Rl^SEM testes (see S1 Table for “n” values). Vasa is red in A,B,E,F. Zfh1 is green in all panels. Fas3 and Eya are blue in A,B,E,F and red in C,D. The hub is outlined by a dotted line. Scale bar = 5 μM.

Fig 4. MAPK is required autonomously for CySC self-renewal.

In A-D, clones are marked by the absence of GFP and in E, MARCM clones are marked by the presence of GFP. A) Testis with a control CySC clone (arrow) at 2 dpci. B) Testis with a control CySC clone at 7 dpci (B,B’, white arrow). Yellow arrow denotes a labeled differentiating cyst cell. C) Testis with an Egfr mutant clone (arrow) at 2 dpci, indicating that clones of this genotype can be induced. D) At 7 dpci, Egfr mutant CySC clones cannot be recovered. However, the differentiating offspring of these mutant clones are observed (yellow arrow). E) At 2 dpci, an Egfr mutant CySC (arrow) has already begun to differentiate as it has low levels of Zfh1 and has started to express Eya. F) Graph showing clone recovery rates for negatively-marked clones (mutant GSCs and mutant CySCs) at 2 and 7 dpci. See Table 1 for “n” values. Loss of Egfr or Ras does not affect the maintenance of GSCs. By contrast, CySC clones lacking either gene have reduced clone recovery rates at 2 dpci and are not maintained at 7 dpci. Vasa is red and Tj is blue in A-D. Zfh1 is red and Eya is blue in E. The hub is outlined by a dotted line. Scale bar = 5 μM.

Fig 5. Socs36E regulates CySC competition via MAPK.

All clones were generated by the MARCM technique and are positively-marked by the expression of GFP. A) UAS-Ras^N17 CySC clones are not recovered at 7 dpci but their differentiating offspring are observed (arrow). B) By contrast, Socs36E, UAS-Ras^N17 CySCs are readily recovered at 7 dpci (arrow). C, D) A Sos mutant CySC is recovered at 2 dpci (C, arrow) but not 7 dpci (D). E) Sos mutant CySCs are rescued when an activated form of Ras (Ras^V12), which functions downstream of Sos, is also expressed in the clones (E,E’, arrows). F) Sos, Socs36E double mutant CySCs are also recovered at 7 dpci (arrow). G) Sos, Socs36E double mutant clones have robust staining for Stat92E (arrow). H) Socs36E mutant clones have colonized the niche at 14 dpci. I) Sos, Socs36E double mutant CySC clones are recovered at 14 dpci (arrow), indicating clone persistence. However, unlike Socs36E single mutant clones, Sos, Socs36E double mutant clones do not colonize the niche, indicating that MAPK signaling regulates the competitiveness of Socs36E mutant clones. J,K) Graphs showing CySC clone recovery rates at 2 (blue bars), 7 (red bars) and 14 (green bars in K) dpci. Whereas CySCs expressing a dominant-negative Ras (UAS-Ras^N17) are not recovered at 7 dpci, Socs36E, UAS-Ras^N17 CySCs are robustly recovered (J). In K, control CySC clones have reduced recovery rates over time, consistent with stochastic loss and replacement [18]. By contrast, Socs36E mutant CySCs are not lost over time, indicating their competitive advantage. While Sos single mutant clones are not recovered at 7 and 14 dpci, Sos, Socs36E double mutant clones are readily recovered and in a similar fashion to control clones. Vasa is red in all panels. Tj is blue in A-F, H,I. Stat92E is blue in G. The hub is outlined by a dotted line. Scale bar = 5 μM.

Fig 6. Socs36E mutant CySCs can self-renew without JAK/STAT signaling.

All clones were generated by the MARCM technique and are positively-marked by the expression of GFP. A,B) CySC clones that express a dominant-negative form of Dome (Dome^Δcyt) (A), which inhibits JAK/STAT signaling, or that are depleted for Stat92E (UAS-Stat RNAi) (B), are not recovered at 7 dpci. However, their differentiating cyst descendants are observed at this time point (A’,B’, arrow). C,D) By contrast, when Socs36E is also removed from Dome^Δcyt (C, arrows) or from UAS-StatRNAi (D, arrows) CySC clones, these clones are now maintained as stem cells. E) Socs36E clones have elevated levels of Stat92E (E,E’, arrow) compared with wild type CySCs (E,E’, arrowhead). Note that GSCs have higher levels of stabilized Stat92E protein than wild type CySCs. F) Some Socs36E, Dome^Δcyt CySCs have residual levels of stabilized Stat92E (F,F’, arrowhead) and some have undetectable levels of stabilized Stat92E (F,F’, arrow). G) Socs36E, UAS-Stat RNAi CySCs have undetectable Stat92E levels (G’, arrow) whereas wild type CySCs in the same testis have moderate to high levels of Stat92E (G’, arrowheads). H) Graph of CySC clone recovery rates at 2 (blue bars) and 7 (red bars) dpci. CySCs expressing Dome^Δcyt or UAS-Stat RNAi are not recovered at 7 dpci. By contrast, Socs36E, UAS-Dome^Δcyt CySCs are recovered at robust levels at 7 dpci. Socs36E, UAS-StatRNAi CySCs are maintained at moderate levels at 7 dpci. Vasa is red in A-G. Tj is blue in A-D and Stat92E is blue in E-G. The hub is outlined by a dotted line. Scale bar = 5 μM.

Fig 7. Model for the role of Socs36E in CySCs.

Upd is produced by hub cells (green) and activates Stat92E (pSTAT) in CySCs (blue). JAK/STAT signaling upregulates the expression of Socs36E, which negatively regulates both JAK/STAT and MAPK signaling. pSTAT in CySCs primarily controls self-renewal (bigger arrow) but may also impact competition (smaller arrow). Spi is produced by GSCs/germ cells (red) and results in the activation of MAPK (pMAPK), which primarily regulates competitive behavior (bigger arrow) of CySCs but may also impact self-renewal (smaller arrow). In the absence of Socs36E, both JAK/STAT and MAPK pathways have elevated activity, resulting in increased self-renewal and increased competition.