Endocytosis Begins inside the Cell

Proximal tubule endocytosis is essential for the production of a protein-free urine. Even with intact glomerular filtration barriers, the proximal tubule must recover up to 2 g per day of low molecular weight (LMW) proteins that escape into the urinary filtrate. Characterization of proteinuric genetic disorders in humans have been key in implicating the principal components of the endocytic machinery, including protein uptake at the cell surface and endosomal recycling. Loss of function mutations in the cell surface receptor megalin result in Donnai–Barrow syndrome, characterized by LMW proteinuria and abnormal neurologic development.¹ Likewise, mutations in megalin’s co-receptor cubilin cause Imerslund–Gräsbeck syndrome marked by LMW proteinuria and megaloblastic anemia.² Even if cell surface receptors are competent for protein uptake, endocytosis is inhibited if the intracellular endolysosomal pathways are perturbed. Reduced activity of inositol polyphosphate-5-phosphatase that regulates intracellular trafficking leads to X-linked Lowe syndrome consisting of oculocerebral deficits and renal Fanconi syndrome.³ Deficient endolysosomal acidification resulting from loss of function of the endosomal chloride channel ClC-5 causes proteinuric Dent disease 1.⁴ In each of these cases, genetic discoveries have illuminated new pathologic processes and guided subsequent mechanistic studies.

Now, in the current issue of JASN, Issler et al. characterize a novel autosomal recessive disorder impinging on endosomal recycling that results in LMW proteinuria and high-frequency sensorineural hearing loss.⁵ Six individuals from a Druze population in Palestine with unexplained proteinuria on dipstick were recruited for genetic analysis after clinical workup, renal ultrasound, and renal biopsy had been unrevealing. Following genetic linkage analysis that pointed to a significant locus on chromosome 11, exome sequencing was conducted that unveiled a novel homozygous variant in Eps15 Homology Domain 1 (EHD1), causative of a missense mutation, p.R398W, in the encoded protein.

EHD1 localizes to the subapical cellular compartment where it has been shown to be involved in endosomal scission that enables recycling endosomes to return to the apical membrane⁶; it has also been implicated in sperm development and early ciliogenesis. However, before the work of Issler et al. EHD1 had not yet been shown to have a role in kidney biology, and here for the first time it is demonstrated to be essential for endocytosis in the proximal tubule. The authors confirm a link between EHD1 dysfunction and proteinuric disease through multiple functional studies with animal models: (1) morpholinos directed against EHD1 homologs in zebrafish reduced LMW dextran uptake in pronephric zebrafish tubules, (2) EHD1^−/− knockout (KO) mice exhibited proteinuria with reduced β2-microglobulin uptake in the tubules, and (3) EHD1^R398W/R398W knock-in mice were comparably proteinuric to the KO model. Interestingly, no defects in the primary cilia of KO mice, R398W knock-in mice, or human patients affected with the EHD1 mutation were appreciable on electron microscopy analysis of the proximal tubules. Notably, the male mice of both KO and knock-in mouse models were infertile, which is consistent with previous data regarding EHD1’s role in sperm development.⁷

In probing the mechanisms of proteinuria resulting from EHD1 dysfunction, Issler et al. discovered that mutant EHD1 was misexpressed in the EHD1^R398W/R398W knock-in mice. Instead of the expected subapical localization, EHD1 was now found in intracellular aggregates. In distal portions of the tubules in the knock-in mice, including the thick ascending limb, these aggregates also appeared as elongated filamentous structures. Issler et al. provide evidence that this misexpression is cell autonomous by finding the same phenotype of intracellular aggregates in LLC-PK1 cells overexpressing EHD1^R398W. Importantly, the authors found that in these structures EHD1 was associated with MICAL-L1 and Pacsin 2, which are two protein factors important for endosomal trafficking and enriched in endosomal recycling membranes, suggesting that the observed structures represent abnormal tubular recycling endosomes.

Structural studies have shown that EHD proteins oligomerize in ring-like assemblies on endosomes.⁸ Crystallographic evidence demonstrates that EHD proteins dimerize via their nucleotide-binding domain to enable the formation of a curved membrane-binding surface, and these dimers in turn oligomerize along the endosomal membrane enabling their scission activity. Issler et al. developed homology models for EHD1 based on published EHD2 and EHD4 structures.^9,10 Modeling of R398 revealed a location for this residue within an alpha-helix at the dimer–dimer interface, where the authors propose it may be involved in a hydrogen bond with E106 of the adjacent dimer. Introduction of the bulky aromatic side chain of tryptophan for the cationic side chain of arginine ablates the hydrogen bond and produces steric hindrance that they hypothesize prevents the dimer–dimer interaction necessary for oligomerization and EHD1 activation. An alternative explanation to this could be that the large side chain of tryptophan destabilizes the alpha-helix itself, leading to unfolding of the local secondary structure and predisposing the mutant EHD1 to aggregation.

In summary, Issler et al. propose a model whereby an autosomal recessive missense mutation in EHD1 perturbs its oligomerization on endosomal membranes and thus prevents its engagement in scission of those endosomes. As a consequence, instead of successfully recycling to the cell membrane to participate in further endocytosis, those endosomes remain intracellular, subsequently aggregating or becoming abnormally elongated. This failure of endosomal recycling and reduction in membrane flux in turn inhibits uptake of LMW proteins at the cell surface via new clathrin-coated vesicles.

In addition to highlighting the kidney-specific functions of EHD1 and introducing this factor as a target for future investigations of proximal tubule transport, the authors’ data highlights several outstanding issues in the understanding of endocytosis. Their immunofluorescence experiments demonstrate an apical localization for cubilin but a more subapical localization for megalin. These localizations belie a simple co-receptor relationship between the two and speak to a more nuanced coordination between megalin and cubilin in protein uptake that extends beyond merely mutually increasing protein binding affinity. Interestingly, no megalin staining was appreciated in the endosomal aggregates that failed to recycle, perhaps pointing to reduced megalin stability at endosomal pH. Further, megalin and cubilin expression at the apical surface remained intact in the disease models, suggesting that despite EHD1 dysfunction, vesicular transport from the Golgi to the cell surface remained intact. That EHD1 dysfunction also had no impact on Arf6 or Rab11 abundance or localization raises questions of how different pools of endosomal membrane are independently regulated. All these observations arising from the work of Issler et al. emphasize the fact that uptake of extracellular proteins requires functional endolysosomal transport, which is a process that begins inside the cell.

Disclosures

The author has nothing to disclose.

Funding

None.