dOCRL maintains immune cell quiescence by regulating endosomal traffic

Abstract

Lowe Syndrome is a developmental disorder characterized by eye, kidney, and neurological pathologies, and is caused by mutations in the phosphatidylinositol-5-phosphatase OCRL. OCRL plays diverse roles in endocytic and endolysosomal trafficking, cytokinesis, and ciliogenesis, but it is unclear which of these cellular functions underlie specific patient symptoms. Here, we show that mutation of Drosophila OCRL causes cell-autonomous activation of hemocytes, which are macrophage-like cells of the innate immune system. Among many cell biological defects that we identified in docrl mutant hemocytes, we pinpointed the cause of innate immune cell activation to reduced Rab11-dependent recycling traffic and concomitantly increased Rab7-dependent late endosome traffic. Loss of docrl amplifies multiple immune-relevant signals, including Toll, Jun kinase, and STAT, and leads to Rab11-sensitive mis-sorting and excessive secretion of the Toll ligand Spåtzle. Thus, docrl regulation of endosomal traffic maintains hemocytes in a poised, but quiescent state, suggesting mechanisms by which endosomal misregulation of signaling may contribute to symptoms of Lowe syndrome.

Fig 1. dOCRL is required in hemocytes to restrict hemocyte abundance.

(A-B) Absolute quantification of hemocytes in hemolymph extracted from wandering third instar larvae. (A) docrl mutants exhibit increased circulating hemocytes. (B) Re-expression of dOCRL in hemocytes is sufficient to suppress hemocyte number. (C) 2D projection of confocal images showing hemocyte distribution in larvae (left). Red arrows indicate salivary glands (a common site of expression for GAL4 drivers); white arrows indicate hematopoietic pockets. Scale bar is 200 μm. Quantification of circulating (released upon opening of the larval cuticle) and sessile (retained in the larval carcass) hemocytes in control and docrl^Δ3 larvae (right). Control genotype is w; CD8-GFP/+; he-GAL4/+. (D) docrl mutant larvae exhibit melanotic masses, which are rescued by docrl gene duplication Dup(1;3)DC402. (E) Frequency of visible melanotic masses. (F) Epifluorescence and reflected light images of a docrl^Δ3 larva showing GFP-labeled hemocytes clustered around a melanotic mass (arrow). Scale bar is 500 μm. (right) 2D projection of confocal image of hemocytes surrounding a melanotic mass. Scale bar is 50 μm. Associated with S1 and S2 Figs.

Fig 2. Aberrant hemocyte differentiation and activation in docrl mutants.

(A) Re-expression of dOCRL-GFP rescue transgene in hemocytes (he-GAL4) and also in muscle (mef2-GAL4), but not lymph gland (dome-GAL4 and ser-GAL4), nephrocytes (dot-GAL4), or fat body (lsp2-GAL4) rescues hemocyte number in docrl^Δ3 larvae. Data are presented as mean +/- SEM. For dome-GAL4 (marked with ~) only very small larvae were recovered, which may have exhibited lower hemocyte number due to developmental stage. (B) 2D projection of confocal images of lymph glands. P1 (magenta) labels differentiating hemocytes and phospho-H3 (green) labels dividing cells. (C) 2D projection of confocal image of hemocytes stained for the lamellocyte marker L1 (magenta). Scale bar is 20 μm. (D) Quantification of L1 staining. Greater numbers of docrl^Δ3 hemocytes are L1-positive (left). This defect is partially rescued by re-expression of dOCRL-GFP in hemocytes (right). (E) Most docrl^Δ3 L1-positive cells do not exhibit normal lamellocyte morphology. Individual values in A represent independent samples of hemocytes collected from 2–4 larvae. Sample N in panels D-E represent number of cells counted. Associated with S3 Fig.

Fig 3. OCRL regulates PIP₂ homeostasis in diverse endosomal compartments.

(A) Endogenously tagged dOCRL (magenta) co-localizes strongly with Clc-GFP (green), single channel confocal slices shown below in gray. (B) Pearson correlations between endogenously tagged dOCRL and trafficking compartments in live primary hemocytes. dOCRL localizes only moderately with other endosomal compartments. (C-D) docrl^Δ3 hemocytes exhibit increased PH-PLC-Cherry marker of PIP₂ (magenta) in live primary hemocytes in all membrane compartments examined (green). (E) The phosphatase activity of dOCRL is necessary but not sufficient to restrict hemocyte number. Scale bars in A,C are 10 μm. Data are presented as mean +/- SEM. Individual values in B,D represent single cells. Associated with S4 Fig.

Fig 4. Loss of docrl causes increased actin assembly in hemocytes.

(A) docrl^Δ3 hemocytes exhibit greater F-actin assembly (phalloidin, green) and a spikier morphology than control cells. (B-C) Quantification of F-actin intensity (B) and ‘spikiness’ (C). Both phenotypes are rescued by hemocyte-autonomous re-expression of dOCRL-GFP. (D) Infection of larvae with M. luteus promotes assembly of F-actin in hemocytes. (E) Sparse re-expression in hemocytes with HH-LT-GAL4 reduces actin assembly in GFP+ rescue hemocytes, but not GFP- hemocytes that do not express the rescue construct. (F) Expression of phosphatase alone moderately rescues actin accumulation (phalloidin staining, gray), though actin accumulation at ruffles (arrowhead) and intracellular puncta (arrows) are still observed. Data are presented as mean +/- SEM. Sample N in panels B-F represent individual cells pooled equally from three independent samples of 2–4 larvae. Scale bars are 10 μm.

Fig 5. Aberrant endosomal compartment structure in docrl mutants.

(A-C) Loss of docrl disrupts the structure of multiple endosomal compartments, including fragmentation and perinuclear accumulation of Clc (A) and Rab5 (B), and expansion of Rab7 (C). Images show single confocal slices at the plasma membrane and an internal Z plane. Scale bar is 10 μm. (D) Quantification of endosomal compartment levels in docrl^Δ3 mutants. Rab7 levels are increased ~3 fold, while Clc and Rab11 levels are slightly increased. N in panel D represent individual cells equally pooled from three independent sets of larvae. (E) Hemocytes from injured wild-type larvae exhibit defective Rab7-positive endosome structure. Scale bar is 5 μm. All panels are representative single confocal slices from live primary hemocytes. Associated with S5 Fig.

Fig 6. Aberrant endosomal compartment function in docrl mutants.

(A) docrl mutant hemocytes exhibit defective scavenger receptor mediated endocytosis of mBSA (magenta). (B) docrl mutant hemocytes are competent for phagocytosis. Images show 2D maximum intensity projection of confocal stacks, showing myr-mRFP-expressing hemocytes incubated with Alexa-488 labeled E. coli. Lower panels show ZX projections taken at the plane marked by arrows in the respective projections. Z sections show internalized E. coli. (Middle) Quantification of mean number of E. coli particles. (Left) Fraction of particles completely engulfed by hemocytes. Data are presented as mean +/- SEM. (C) The structure and abundance of Lysotracker-positive (acidified) compartments is altered in docrl mutants, and rescued cell-autonomously by re-expression of docrl. Image shows 2D maximum intensity projection of confocal stacks. (D) docrl mutants exhibit defective accumulation of autophagic compartments, marked by GFP-mCherry-Atg8a (GC::Atg8A). Quantification shows Pearson’s Correlation Coefficient per cell +/- s.e.m. for GFP and mCherry (left) and mean fluorescence +/- s.e.m. for each fluorophore/cell (right). All panels are shown with the same brightness and contrast. Individual values in graphs represent single cells. Scale bars are 10 μm. Associated with S6 Fig.

Fig 7. Endosomal sorting defects underlie hemocyte activation in docrl mutants.

(A) Relative hemocyte number in larvae with hemocyte-specific manipulation of the indicated Rab GTPases. Inhibiting the internalization step of endocytosis has no effect on hemocyte number, while manipulation of endosomal sorting recapitulates the docrl mutant phenotype. Statistical significance noted reflects pairwise tests with wild type control. (B) The phosphatase activity of dOCRL is sufficient to rescue scavenger receptor endocytosis of mBSA. (C) Manipulation of endosomal sorting via Rab11 and Rab7 rescue docrl-induced hemocyte overabundance. Data sets were each normalized and statistical significance noted relative to docrl mutant controls taken at the same time. Data are presented as mean +/- SEM. (D) Expression of Rab7^DN, but not Rab11^CA, rescues docrl-induced actin (left) and Lysotracker (right) accumulation. Individual values in A,C are independent samples pooled from 2–4 larvae. Multiple independent experiments are shown together for clarity, all statistical comparisons reflect independent controls. Data points in B and D represent single cells pooled equivalently from three independent larval collections. Statistical comparisons shown in B and D are to the relevant control group (green). Associated with S7 Fig.

Fig 8. Loss of docrl activates multiple immune-relevant signaling pathways.

(A) Jak/STAT reporter 10X STAT-GFP is induced in docrl mutant muscle tissue. (B) Loss of docrl induces phospho-JNK (green) but not 10X STAT reporter (magenta) in isolated hemocytes (quantified in adjacent graphs). (C) qPCR shows that Toll target Drs is not upregulated in docrl^Δ3 larvae as compared to sterile or M. luteus wounded animals. (D) Spz-GFP accumulation (gray) is mildly, but significantly increased in docrl^Δ3 mutant hemocytes. Images show 2D confocal projections of fixed hemocytes. (E) Secreted Spz-GFP is increased in docrl^Δ3 hemolymph, measured by western blot of cell-free hemolymph (see blot in S8D Fig). (F) Hemocyte-specific expression of Rab11^CA causes strong Spz-GFP accumulation in docrl^Δ3 mutant hemocytes (quantified in G). Images show 2D confocal projections of fixed hemocytes, top and bottom panels are two example cells of each genotype to show the range of typical Spz-GFP intracellular distribution; Rab11 samples are contrast-adjusted relative to controls. (H) Hemocyte-specific expression of constitutively active Rab11 rescues the level of Spz-GFP secreted into the hemolymph, by western blot. Statistical significance noted is relative to all other data sets. Scale bars in B, D, F are 10 μm. Data are presented as mean +/- SEM. Individual values in B, D, G are cells pooled equivalently from three larval collections. N in panel C, E, H are independent samples collected from 4 pooled larvae each. Associated with S8 Fig.