A patient with multisystem dysfunction carries a truncation mutation in human SLC12A2, the gene encoding the Na-K-2Cl cotransporter, NKCC1

Abstract

This study describes a 13-yr-old girl with orthostatic intolerance, respiratory weakness, multiple endocrine abnormalities, pancreatic insufficiency, and multiorgan failure involving the gut and bladder. Exome sequencing revealed a de novo, loss-of-function allele in SLC12A2, the gene encoding the Na-K-2Cl cotransporter-1. The 11-bp deletion in exon 22 results in frameshift (p.Val1026Phefs*2) and truncation of the carboxy-terminal tail of the cotransporter. Preliminary studies in heterologous expression systems demonstrate that the mutation leads to a nonfunctional transporter, which is expressed and trafficked to the plasma membrane alongside wild-type NKCC1. The truncated protein, visible at higher molecular sizes, indicates either enhanced dimerization or misfolded aggregate. No significant dominant-negative effect was observed. K⁺ transport experiments performed in fibroblasts from the patient showed reduced total and NKCC1-mediated K⁺ influx. The absence of a bumetanide effect on K⁺ influx in patient fibroblasts only under hypertonic conditions suggests a deficit in NKCC1 regulation. We propose that disruption in NKCC1 function might affect sensory afferents and/or smooth muscle cells, as their functions depend on NKCC1 creating a Cl^- gradient across the plasma membrane. This Cl^- gradient allows the γ-aminobutyric acid (GABA) receptor or other Cl^- channels to depolarize the membrane affecting processes such as neurotransmission or cell contraction. Under this hypothesis, disrupted sensory and smooth muscle function in a diverse set of tissues could explain the patient's phenotype.

Keywords: autonomic bladder dysfunction; chronic pain; orthostatic hypotension; small intestinal dysmotility.

Figure 1.

The NKCC1-DFX mutant is nonfunctional. (A) Sequence of a portion of SLC12A2 exon 22 in patient shows wild-type allele (top) and 11-bp deletion in mutant allele (bottom). Translation results in a Phe residue inserted after Asp1025 followed by two stop codons (p.Val1026Phefs*2). (B) Schematic representation of the topology of human NKCC1 and position of the truncation. (C) Oocytes injected with wild-type but not mutant NKCC1 cRNA exhibit bumetanide-sensitive K⁺ influx under isotonic and hypertonic conditions. Small bumetanide-sensitive component observed with mutant is due to native amphibian NKCC1. (D) Western blot analysis of oocyte lysates from wild-type and mutant NKCC1 complementary RNA (cRNA) injected oocytes. (E) Absence of significant dominant-negative effect of mutant NKCC1 on wild-type NKCC1 in Xenopus laevis oocytes. The curve represents a dose–response of K⁺ influx versus amount of wild-type NKCC1 cRNA injected. The bars represent K⁺ influx of oocytes injected with equivalent amount of mutant cotransporter cRNA. The bars represent means ± S.E.M. (n = 20–25 oocytes). (***) P < 0.001; n.s., nonsignificant; P > 0.05, one-way ANOVA followed by Tukey's multiple comparison posttests, WT, wild-type.

Figure 2.

The behavior of the NKCC1-DFX mutant in human cells. (A) Characterization of major K⁺ influx pathways in fibroblasts from patient. Flux was measured in control (cnt) conditions or the absence of inhibitors, in the presence of 1 mM ouabain (ou) to inhibit the Na⁺/K⁺ pump, in the presence of 20 µM bumetanide (bu) to inhibit NKCC1, and in the presence of both inhibitors (ou + bu). (***) P < 0.001; (*) P < 0.05, one-way ANOVA followed by Tukey's multiple comparison posttests. (B) Similar experiment performed under isosmotic (310 mOsM) and hyperosmotic (385 mOsM) conditions in fibroblasts from UDP_2780 patient compared with control human fibroblasts. Bars represent means ± S.E.M (n = 3 dishes). (***) P < 0.001, (**) P < 0.01, (*) P < 0.05, n.s., nonsignificant; P > 0.05, one-way ANOVA followed by Tukey's multiple comparison posttests. (C) Fluorescence images of HeLa cells transfected with tdTomato-NKCC1 (red) and EGFP-NKCC1-DFX (green). Images captured at three consecutive Z-sections with confocal microscopy reveals colocalization at the plasma membrane. Scale bar, 10 µm.

Figure 3.

Additional SLC12A2 frameshift and stop codon variants in human population. The Ensembl genome browser database (species/gene/transcript/exons) color lists a large number of variants for each human gene, and we identified 13 unique variations leading to NKCC1 truncation. The majority of these mutations are also reported in the ExAC database (http://exac.broadinstitute.org/). These variants are labeled 1–13, and their position in the protein is indicated by arrows. The location of the NKCC1-DFX mutation is shown by the red arrow. The 12 transmembrane domains are designated by alternate black and gray boxes crossing the plasma membrane (blue). The location of the exons is indicated above the protein with size of introns proportionally drawn as vertical lines. DNA and amino acid sequence of the 13 variants are listed under the (1) Ensembl database and/or (2) ExAC database.

Figure 4.

Two models are described to account for how NKCC1 function leads to multiorgan failure. (A) Each organ sends afferent fibers to the spinal cord to modulate autonomic function. These neurons have their cell bodies in dorsal root ganglia. Parallel arrows indicate spinal cord ascending and descending pathways. (B) All major organs have smooth muscle cells where NKCC1 accumulates Cl⁻ above thermodynamic potential equilibrium. Opening of Cl⁻ channels depolarizes the membrane triggering the opening of voltage-sensitive Ca²⁺ channels, Ca²⁺ entry, and contraction. NKCC1 dysfunction leads to collapse of Cl⁻ gradients in both sensory (DRG) neurons and smooth muscle cells, thereby contributing to the multiorgan pathology.