A congenital neutrophil defect syndrome associated with mutations in VPS45

Abstract

Background: Neutrophils are the predominant phagocytes that provide protection against bacterial and fungal infections. Genetically determined neutrophil disorders confer a predisposition to severe infections and reveal novel mechanisms that control vesicular trafficking, hematopoiesis, and innate immunity.

Methods: We clinically evaluated seven children from five families who had neutropenia, neutrophil dysfunction, bone marrow fibrosis, and nephromegaly. To identify the causative gene, we performed homozygosity mapping using single-nucleotide polymorphism arrays, whole-exome sequencing, immunoblotting, immunofluorescence, electron microscopy, a real-time quantitative polymerase-chain-reaction assay, immunohistochemistry, flow cytometry, fibroblast motility assays, measurements of apoptosis, and zebrafish models. Correction experiments were performed by transfecting mutant fibroblasts with the nonmutated gene.

Results: All seven affected children had homozygous mutations (Thr224Asn or Glu238Lys, depending on the child's ethnic origin) in VPS45, which encodes a protein that regulates membrane trafficking through the endosomal system. The level of VPS45 protein was reduced, as were the VPS45 binding partners rabenosyn-5 and syntaxin-16. The level of β1 integrin was reduced on the surface of VPS45-deficient neutrophils and fibroblasts. VPS45-deficient fibroblasts were characterized by impaired motility and increased apoptosis. A zebrafish model of vps45 deficiency showed a marked paucity of myeloperoxidase-positive cells (i.e., neutrophils). Transfection of patient cells with nonmutated VPS45 corrected the migration defect and decreased apoptosis.

Conclusions: Defective endosomal intracellular protein trafficking due to biallelic mutations in VPS45 underlies a new immunodeficiency syndrome involving impaired neutrophil function. (Funded by the National Human Genome Research Institute and others.).

Figure 1. Clinical Findings

Panel A shows the pedigrees of five affected families. Double lines indicate consanguineous unions. Squares denote male family members, circles female family members, solid symbols affected family members, open symbols unaffected family members, and slashes deceased family members. Arrows point to the children who were studied. Panel B (hematoxylin and eosin) shows extramedullary hematopoiesis in a renal-biopsy specimen from one of the affected children. Megakaryocytes have large, lobulated, hyperchromatic nuclei (arrows) and are the hallmark of extramedullary hematopoietic tissue. Normoblastic red cells are circled. T denotes renal tubule. In Panel C (reticulin stain), a bone marrow–biopsy specimen from one of the children shows prominent, diffuse, intersecting reticulin fibers, representing fibrosis (grade 3 of 4). Bone is shown in the upper left corner of the image. Panel D shows electron-microscopical images of peripheral-blood neutrophils from a child with a mutation in VPS45 and an age-matched control. The micrographs on the left are at lower magnification than those on the right. In the control neutrophils, the number and structure of the mitochondria are normal and there are multiple granules (arrows). In comparison, the patient's neutrophils contain abundant mitochondria (Mito) and far fewer cytoplasmic granules (arrows); also, the Golgi apparatus (G) and rough endoplasmic reticulum (RER) are more developed, and the chromatin (Ch) and cytoplasm are less condensed. N denotes nucleus.

Figure 2. Cellular Phenotypes

In Panel A, immunoblot analysis and graphic representation of densitometry show reduced expression of VPS45 in peripheral blood and fibroblasts and reduced expression of the interacting partners of VPS45 — rabenosyn-5 and syntaxin-16 (STX16) — in fibroblasts from one of the affected patients as compared with a control. Loading was controlled by immunoblotting the same membrane for emerin, β-actin, or α-tubulin. Densitometry was performed with the use of the ImageJ 1.45s Gel Analysis tool (National Institutes of Health). Panel B shows impaired cell motility. Fibroblasts were grown on coverslips coated with fibronectin. The cultures were “wounded” by introducing a scratch (approximately 800 μm in width) across the field of cells; the same field of cells was imaged immediately, after 6 hours, and after 12 hours. The fibroblasts from Patient B-II-9 migrated significantly more slowly toward the wounded area than did control cells (similar results were obtained in fibroblasts from Patient E-II-1). Each experiment was performed in triplicate, and similar results were obtained in four independent experiments. The histogram shows the quantification of the area of the scratch at 0, 6, and 12 hours. Panel C shows flow-cytometric analysis for CD45 and β1 integrin that was performed on peripheral blood in a patient and a control. The patient's blood (with VPS45-mutant neutrophils) has far fewer cells that are positive for β1 integrin. Panel D shows apoptosis in fibroblasts from a patient and a control. Staining for activated caspase 3, an in situ fluorescein isothiocyanate marker and indicator of apoptosis, was increased by a factor of 14 in fibroblasts from the patient as compared with control cells. In Panels A, B, and D, T bars indicate the standard deviation.

Figure 3. Correction of the VPS45 Mutant Phenotype

Panel A shows that in VPS45-deficient fibroblasts, which contain approximately half the normal amount of both VPS45 and its interacting protein, rabenosyn-5, levels of both proteins are restored after transfection of the fibroblasts with a plasmid containing nonmutated VPS45. β-Actin was used as the control protein. Panel B shows cell motility after fibroblasts from a patient were transfected with either empty vector or a vector containing nonmutated VPS45 complimentary DNA (cDNA) and then grown on coverslips coated with fibronectin. The cultures were “wounded” by introducing scratches. The histogram shows the quantification of the area of the scratch at various time points. The VPS45-transfected fibroblasts migrated significantly more rapidly toward the wounded area than did cells transfected with empty plasmid. Panel C shows apoptosis, as assessed by staining for activated caspase 3, in patient fibroblasts transfected with either empty vector or a vector containing non-mutated VPS45 cDNA, as compared with control fibroblasts transfected with empty vector. Transfection with the nonmutated VPS45 vector reduced apoptosis by 50%.

Figure 4. Targeted Knockdown of Vps45 with the Use of Morpholino Oligonucleotides (MOs) That Block Splicing and Translation in Zebrafish Embryos

Panel A is a diagram of morpholino design showing the zebrafish gene vps45 and morpholino antisense strategies to block either the translation of the zebrafish vps45 messenger RNA (ATG-MO) or the splice acceptor site of exon 3 (I2E3-MO). Primers F and R validate exon skipping, including the expected size of amplicons that result from loss of exon 3. Below the diagram is a schematic depiction of the spliced transcript in the I2E3-MO–injected embryos (258 bp) as compared with the uninjected and ATG-MO–injected embryos (359 bp). In Panel B, the top immunoblot shows the results of reverse-transcriptase–polymerase-chain-reaction analysis of vps45 transcript from uninjected and morpholino-injected embryos 5 days after fertilization. The bottom immunoblot shows the results of Western blot analysis of whole-protein lysates from uninjected and morpholino-injected embryos 5 days after fertilization. Lysates from embryos injected with ATG-MO and from those injected with I2E3-MO show marked reduction of vps45 when normalized with β-actin, which was used as a protein-loading control. In Panel C, in situ hybridization of embryos 5 days after fertilization supports functional knockdown of vps45. Shown are representative images of uninjected zebrafish embryos (left), embryos injected with ATG-MO (middle), and embryos injected with I2E3-MO (right). Results of whole-mount in situ hybridization with the use of a digoxigenin-labeled RNA probe against zebrafish myeloperoxidase are shown. The myeloperoxidase detects neutrophils in the caudal hematopoietic tissue (rectangle). In uninjected embryos, the myeloperoxidase signals are readily seen (arrowheads). However, in ATG-MO–injected and I2E3-MO–injected embryos, there is a marked reduction in myeloperoxidase staining, suggestive of a decreased number of mature neutrophils.