The phosphate exporter xpr1b is required for differentiation of tissue-resident macrophages

Abstract

Phosphate concentration is tightly regulated at the cellular and organismal levels. The first metazoan phosphate exporter, XPR1, was recently identified, but its in vivo function remains unknown. In a genetic screen, we identified a mutation in a zebrafish ortholog of human XPR1, xpr1b. xpr1b mutants lack microglia, the specialized macrophages that reside in the brain, and also displayed an osteopetrotic phenotype characteristic of defects in osteoclast function. Transgenic expression studies indicated that xpr1b acts autonomously in developing macrophages. xpr1b mutants display no gross developmental defects that may arise from phosphate imbalance. We constructed a targeted mutation of xpr1a, a duplicate of xpr1b in the zebrafish genome, to determine whether Xpr1a and Xpr1b have redundant functions. Single mutants for xpr1a were viable, and double mutants for xpr1b;xpr1a were similar to xpr1b single mutants. Our genetic analysis reveals a specific role for the phosphate exporter Xpr1 in the differentiation of tissue macrophages.

Figure 1. xpr1b is required autonomously in macrophages for microglia development

(A) 5 dpf zebrafish larvae were stained with the microglial vital marker neutral red (Herbomel et al 2001). Left panel shows the dorsal view of xpr1b heterozygous larva with microglia (arrow) in the optic tectum; (B) Whole-mount in situ hybridization at 5 dpf reveals the expression of the microglial marker apoe in xpr1b heterozygous but not in xpr1b mutants. (C) Lateral view of neutral red stained heterozygous and homozygous mutants shows normal whole-animal morphologies at 5 dpf. (D) Schematic representation of the st87 locus. (E) Chromatogram depicts the lesion (arrow) in the coding sequence of xpr1b. (F) Graphical representation of the domain structure of Xpr1 and the location of the st87 lesion (asterick). (G) Diagram represents the transgenic constructs driving the expression of mcherry reporter or wildtype xpr1b under the control of macrophage (mpeg1), neurons (huC) or neutrophils (lyz) tissue promoters. Below are representative images of neutral red stained xpr1b mutant fish expressing mcherry (left) or xpr1b (right) under the macrophage lineage specific promoter. Microglia are present only when xpr1b is expressed under the mpeg1 promoter. (H) Graph shows the percentage of mutants rescued by different tissue specific promoters. The number of embryos analysed for each condition, mcherry (C) or xpr1b, are shown below. Scale bar = 50 μm. See also Figure S1.

Figure 2. xpr1b is required for brain colonization by microglial progenitors

(A and B) Representative z-projections of mpeg1:EGFP labelled cells at 2–4 dpf in the head (A) or 4 dpf in the trunk (B), in heterozygous larvae (top panel) or xpr1b mutants. (C, D and E) Graphs show the average number of mpeg1:EGFP positive cells in the highlighted region of fish schematics: CNS (eye, midbrain and hindbrain) (C), head periphery (D) and trunk (E). xpr1b mutants show a reduction in the average number of macrophages in the CNS region starting at 2dpf. Error bars show s.e.m. Statistical significance was determined by paired two-tailed Student’s test *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001 (2 dpf, mutants n =7, siblings n = 16; 3 dpf, mutants n = 6, siblings n = 16; 4d pf, mutants n = 4, siblings n = 9). (F) Whole mount in situ hybridization at 2 and 3 dpf, shows mfap4+ macrophages in the brain of wildtype heterozygous (top panel) larva but not in xpr1b mutants (lower panel). (G) Whole mount in situ hybridization at 3dpf, shows apoe+ microglia in the brain of wildtype heterozygous (top panel) larva but not in xpr1b mutants (lower panel). (H) Serial single-plane images of stable transgenic fish expressing the macrophage marker mpeg1:EGFP and the vasculature marker kdrl:mCHERRYCAAX. Images show extensive penetration of macrophages into the brain of wildtype heterozygous fish (top panels) and only a few in xpr1b mutants (lower panels) (arrowheads indicate macrophages). (I) Z-stack images of stable transgenic fish expressing the microglia marker apoe:EGFP. Images show microglia cells with processes in wildtype fish at 5 and 10 dpf (top panels) whereas no microglia cells are present in xpr1b mutants, even at 10 dpf (right panel). Scale bar = 50 μm. See also Figure S2–S3 and Movies S1–S4

Figure 3. xpr1b is required for differentiation of microglia and Langerhans cells

(A) Representative Z-stack images of the head region of stable mpeg1:EGFP transgenic fish at 3 and 4 dpf. In wildtype heterozygous fish macrophages display an array of morphologies from rounded (arrows) to multiple process bearing cells (arrowhead). In contrast, xpr1b mutants display a more restricted array of morphologies. (B and C) Graphs show the percentage of mpeg1:EGFP positive cells that contain 0, 1, 2, 3 or 4 or more processes at 3 dpf (B) or 4 dpf (C) (3 dpf siblings n = 6; mutants n = 5. 4 dpf siblings n = 4; mutants n = 4). (D and E) Representative Z-stacks of the superficial layers of the trunk of wildtype heterozygous and xpr1b mutant fish at 10 (D) and 21 dpf (E). Insets show representative images of Langerhans cells. (F) Graph shows the average number of Langerhans cells in the field of view in the area of the trunk highlighted in the schematics. Error bars show s.e.m. *P ≤ 0.05, using paired two-tailed Student’s t test (10 dpf siblings n = 5; mutants n = 6. 21dpf siblings n = 7; mutants n = 6). Scale bar = 50 μm. See also Figure S4.

Figure 4. xpr1b is required for normal skeletal development

(A) μCT-derived volumetric reconstructions of the neural (top) and hemal (bottom) 14^th vertebral arches from a representative wild type (n=6), heterozygous (n=7) and mutant (n=6) fish. xpr1b mutant fish have excess bone on the inner surface of the arch compared to wild type and heterozygous fish, reducing the overall aperture of the canal. Scale bar = 100 μm. (B–G) Graphs showing quantified values for the vertebral arch defect depicted in (A). xpr1b mutant fish exhibit significant reductions in canal aperture (B, E) with a concomitant increase in the total amount of bone as measured by bone volume (BV; C, F) and surface area (SA; D, G). (H) Volumetric reconstructions of the 14^th vertebral bodies from a representative wild type, heterozygous and mutant fish. xpr1b mutant vertebral bodies appear larger than their wild type and heterozygous siblings. Scale bar = 100 μm. (I, J) Graphs depicting the significant increase in the total amount of bone present in mutant vertebrae as measured by BV (I) and SA (J). Error bars show s.e.m. Statistical significance was determined by unpaired, two-tailed t tests. * P < 0.05, ** P < 0.005, *** P < 5×10⁻⁴. (K) Lateral view of wildtype (left panel) and xpr1b mutant (right panel) adult fish. Inset shows the presence of an ectopic vessel in xpr1b mutants (right panel, arrowheads) that is absent in wildtype fish. See also Figure S5.

Figure 5. xpr1a encodes a functional phosphate exporter but is not required for microglia development or embryonic development

(A) Graphic representation of the domain structure of Xpr1b and Xpr1a showing their shared identity. (B) Graph representing the percentage of rescued xpr1b mutants when either xpr1b, control mcherry (C) or xpr1a are expressed in the macrophage lineage. (C) Representative dorsal image of a neutral red stained 5 dpf xpr1b mutant larva rescued by expression of xpr1a in the macrophage lineage. (D) Top panel shows diagram depicting the TALE nuclease targeted nucleotide sequence and the corresponding amino acid sequence of xpr1a. Bottom panels show the two alleles generated that have frameshift mutations in xpr1a coding sequence resulting in premature stop codons. (E) Dorsal views of neutral red stained 5 dpf xpr1a mutant (left) and xpr1b;xpr1a double mutant (right), showing that xpr1a mutants have a normal number of microglia cells. (F) Lateral views of neutral red stained 5dpf xpr1a mutant (left) and xpr1b;xpr1a double mutant (right), show grossly normal development at 5 dpf. Scale bar = 50 μm.