PTPN21/Pez Is a Novel and Evolutionarily Conserved Key Regulator of Inflammation In Vivo

Abstract

Drosophila provides a powerful model in which to study inflammation in vivo, and previous studies have revealed many of the key signaling events critical for recruitment of immune cells to tissue damage. In the fly, wounding stimulates the rapid production of hydrogen peroxide (H₂O₂).^1,2 This then acts as an activation signal by triggering a signaling pathway within responding macrophages by directly activating the Src family kinase (SFK) Src42A,³ which in turn phosphorylates the damage receptor Draper. Activated Draper then guides macrophages to the wound through the detection of an as-yet unidentified chemoattractant.^3-5 Similar H₂O₂-activated signaling pathways are also critical for leukocyte recruitment following wounding in larval zebrafish,^6-9 where H₂O₂ activates the SFK Lyn to drive neutrophil chemotaxis. In this study, we combine proteomics, live imaging, and genetics in the fly to identify a novel regulator of inflammation in vivo; the PTP-type phosphatase Pez. Pez is expressed in macrophages and is critical for their efficient migration to wounds. Pez functions within activated macrophages downstream of damage-induced H₂O₂ and operates, via its band 4.1 ezrin, radixin, and moesin (FERM) domain, together with Src42A and Draper to ensure effective inflammatory cell recruitment to wounds. We show that this key role is conserved in vertebrates, because "crispant" zebrafish larvae of the Draper ortholog (MEGF10) or the Pez ortholog (PTPN21) exhibit a failure in leukocyte recruitment to wounds. This study demonstrates evolutionary conservation of inflammatory signaling and identifies MEGF10 and PTPN21 as potential therapeutic targets for the treatment of inflammatory disorders.

Keywords: Draper; Drosophila; Megf10; PTPN21; Pez; inflammation; macrophage; migration; neutrophil; zebrafish.

Figure 1

Pez Is Required for Macrophage Migration to Epithelial Wounds and Functions within the H₂O₂-Src42A-Draper Signaling Pathway (A) Pez locus highlighting mutant alleles. Approximate CB insertion (6.056 kb) site is indicated. Pez² deletion is marked below, adapted from Poernbacher et al. (B) Live imaging of inflammation following laser ablation reveals reduced macrophage recruitment in Pez^CB mutants. Wound margin is denoted by dashed red line. Cell tracks are shown at 1 h. (C) Quantification reveals a significant decrease in macrophage numbers at wounds in the two Pez mutant lines at 40 and 60 min post-injury (n ≥ 10 wounded embryos/genotype; multiple t tests with Holm-Sidak multiple comparisons). (D and E) Cell tracking reveals (D) macrophage speed post-wounding is unaffected in Pez^CB mutants (n ≥ 130 cells from ≥5 embryos/genotype; Mann-Whitney U test), and (E) meandering index is significantly reduced in responding (cells that reach the wound site at any point within 2 h) Pez^CB macrophages (n = 53 responders from ≥5 embryos/genotype; Mann-Whitney U test). (F) Heterozygote (src42A^[E1]/+, draper^Δ5/+, and Pez^CB/+) and transheterozygote (src42A^[E1]/Pez^CB and Pez^CB/+; draper^Δ5/+) mutant embryos at 60 min post-wounding. Wound margin is denoted by dashed red line. (G) Significantly reduced macrophage wound recruitment in transheterozygotes embryos versus Pez^CB/+ (n ≥ 15 wounded embryos/genotype; one-way ANOVA with multiple comparisons). All error bars are mean ± SD. NS, not significant; ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.005, and ^∗∗∗∗p < 0.001. All scale bars represent 20 μm. See also Figure S2 and Videos S1, S2, and S3.

Figure 2

The Role of Pez in Macrophage Wound Recruitment Is Cell Autonomous and Dependent upon the FERM Domain (A) Macrophage-specific expression of Pez-RNAi (TRiP constructs) impairs inflammatory recruitment to wounds (images 1 h post-wounding). Scale bars represent 10 μm. Wound margin is denoted by dashed red line. (B) Pez-RNAi significantly reduces macrophage recruitment to wounds compared to control (n ≥ 21 wounded embryos/genotype; Kruskal-Wallis with Dunn’s multiple comparisons). (C) Pez-sfGFP expression in macrophages (outlined in red) co-expressing either control RNAi or either Pez-RNAi. Scale bars represent 10 μm. (D) Both RNAi lines significantly reduce macrophage Pez-sfGFP intensity levels (n = 18 cells from 6 embryos/genotype; Kruskal-Wallis with Dunn’s multiple comparisons). (E) UAS-Pez expression constructs. FERM domain and PTP domains noted and deletions depicted. For phosphatase dead construct (UAS-Pez^ΔPD), the mutated cysteine is noted. Adapted from Poernbacher et al. (F) Images of wounded Pez² embryos with macrophage-specific expression of indicated Pez constructs, 1 h post-ablation. Scale bar represents 20 μm. Wound margin is marked by dashed red line. (G) Macrophage-specific expression of UAS-Pez, UAS-Pez^ΔPD, and UAS-Pez^ΔPTP (but not Pez^ΔFERM) is sufficient to rescue Pez² wound recruitment defect (n ≥ 13 wounded embryos/genotype; one-way ANOVA with Dunnett’s multiple comparisons to Pez²). (H) Quantification of meandering index reveals specific expression of Pez rescues the inflammatory chemotaxis of Pez² macrophages (n ≥ 42 cells from n ≥ 5 wounded embryos/genotype; unpaired t test). All error bars are mean ± SD. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.005, and ^∗∗∗∗p < 0.001. See also Figure S2 andVideo S4.

Figure 3

Dynamic Pez Puncta Are Stimulated upon Wounding in a Draper-Dependent Manner (A) Diagrams of fluorescently tagged Pez and Draper constructs. For Pez, the FERM and PTP domains are shown. For Draper, the N-terminal extracellular domain is noted, along with the transmembrane domain (TM) and immunoreceptor tyrosine activation motif (ITAM). (B) Pez forms puncta within the cell body and lamellipod. Dynamic lamellipodial puncta flow inward from the cell periphery (denoted by red line). Colored arrows show puncta tracking over 1 min. (C) Dynamic Draper-EGFP puncta (orange arrowheads) induced post-wounding. Red line donates cell periphery; yellow arrow indicates direction of wound. (D) Kymographs of individual Pez-sfGFP and Draper-EGFP puncta (orange arrows) following wounding demonstrate similar dynamics over time. (E) Kymograph of Draper-EGFP punctum reveals colocalization with Pez-mCh following wounding. For all kymographs, the x axes represent distance starting at lamellipod leading edge (174 nm/pixel; 17.4 μm total). The y axes represent time (10 s/pixel; 2.5 min total). (F) Lamellipodial Pez-sfGFP puncta are suppressed in draper^Δ5 and src42A^[E1] mutant macrophages. (G and H) Puncta number (G) and distribution (cell body versus lamellipod; H) significantly altered in mutants (n ≥ 10 cells from ≥5 embryos/genotype; Kruskal-Wallis with Dunn’s multiple comparisons and one-way ANOVA with Tukey’s comparisons, respectively). (I) Images of control, draper^Δ5, and src42A^[E1] mutant macrophages (red outlines) 5 min post-wounding. Direction of wound marked by yellow arrow. (J and K) Analysis of Pez puncta 5 min post-wounding reveals (J) a wound-induced significant increase in puncta number that is dependent on both Draper and Src42A (n ≥ 6 cells from ≥5 embryos/genotype for each condition; Kruskal-Wallis with Dunn’s multiple comparisons) and (K) a wound-induced significant increase in the proportion of puncta residing within the cell body of control cells that is absent in draper and src42A mutants (n ≥ 6 cells from ≥5 embryos/genotype for each condition; one-way ANOVA with Sidak’s multiple comparisons). All error bars are mean ± SD. ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗∗p < 0.001. All scale bars represent 10 μm. (L) Proposed role of Pez in wound-induced Draper clustering. Under basal conditions, Draper’s ITAM domain remains in an inactive state. Following H₂O₂-mediated Src42A activation, phosphorylated Pez is recruited to Draper clusters via its FERM-domain-mediated interaction with Src42a. Acting as an adaptor, Pez coordinates inflammatory Draper signaling via effectors such as Shark, leading to efficient macrophage chemotaxis. See also Figure S2 and Videos S5 and S6.

Figure 4

The Orthologs of Pez (PTPN21) and Draper (MEGF10) Are Required for Leukocyte Recruitment to Wounds in Zebrafish Larvae (A) Representative images of entire control, PTPN21 crispant, and MEGF10 crispant zebrafish 3 dpf larvae expressing either lysc:nls-mScarlet (neutrophil marker) or mpeg1.1:nls-mScarlet and mpeg1.1eGFP (macrophage marker). Scale bars represent 500 μm. (B) Quantification of leukocyte numbers revealed an increase in neutrophils in PTPN21 crispants (n = 10 larvae/genotype; Kruskal-Wallis with Dunn’s multiple comparisons) and a decrease in macrophage in MEGF10 crispants (n ≥ 7 larvae/genotype; one-way ANOVA with Dunnett’s multiple comparisons). (C) For wound studies, zebrafish embryos (one cell stage) were injected with 2 CRISPR guide RNAs (crRNAs) alongside tracrRNA and raised to 3 dpf. Following tailfin transection, fish were stained at 2, 6, and 22 h post-injury (hpi). (D) Images of wounded control larvae and PTPN21 crispants at 2, 6, and 22 hpi time points. Tg(lysC:DsRed) (red) zebrafish co-stained with DAPI (blue) are shown. Quantification zone of 150 μm proximal to the wound margin is marked by the white box across all images. (E) Significantly reduced neutrophils recruited to the wound at 6 hpi in PTPN21 crispant larvae compared to control (n ≥ 18 wounded larvae/genotype for each time point; multiple t test). (F) Images of wounded control larvae and PTPN21 crispants at 2, 6, and 22 hpi time points. Tg(mpeg1.1:nls-mScarlet) (red) zebrafish co-stained with DAPI (blue) are shown. Quantification zone of 150 μm proximal to the wound margin is marked by the white box across all images. (G) Significantly reduced macrophages recruited to the wound at 6 hpi and 22 hpi in PTPN21 crispants (n ≥ 17 wounded larvae/genotype for each time point; multiple t test). (H) Images of wounded control larvae and MEGF10 crispants at 2, 6, and 22 hpi. Tg(lysc:DsRed) (red) zebrafish stained with DAPI (blue) are shown. Quantification zone of 150 μm proximal to the wound margin is marked by the white box across all images. (I) Significantly reduced neutrophils recruited to the wound at all time points in MEGF10 crispant compared to the control (n ≥ 15 wounded larvae/genotype for each time point; multiple t test). (J) Images of wounded control larvae and MEGF10 crispants at 2, 6, and 22 hpi time points. Tg(mpeg1.1:nls-mScarlet) (red) zebrafish co-stained with DAPI (blue) are shown. Quantification zone of 150 μm proximal to the wound margin is marked by the white box across all images. (K) Significantly reduced macrophages recruited to the wound at 6 hpi and 22 hpi in PTPN21 crispant (n ≥ 13 wounded larvae/genotype for each time point; multiple t test). All error bars are mean ± SD. ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗∗p < 0.001. All scale bars represent 100 μm. See also Figure S2.