A reductionist approach to studying renal claudins provides insights into tubular permeability properties

Abstract

The tight junction (TJ) is a specialized region of cell-to-cell contact located at the most apical aspect of the junctional complex of epithelial or endothelial cells. Forming a continuous belt-like structure, TJs play an important role in establishing cellular barriers to the external environment and selectively regulating the paracellular permeability of solutes, electrolytes, and water. TJs are composed of intracellular and transmembrane proteins. The claudins are a transmembrane family of TJ proteins, interacting both within the same cell and between adjacent cells. Claudins form either pores or barriers across the paracellular space, and their unique composition regulates the permeability of the shunt. In this issue of Acta Physiologica, Pouyiourou et al.¹ shed new light on the role of claudins expressed in the kidney.

The kidney is critical for electrolyte and water balance as it amends their urinary excretion to maintain electrolyte composition, osmolality, and blood pressure. Transport along the renal tubule depends on both the paracellular and transcellular movement of electrolytes. In the proximal tubule, paracellular reabsorption of cations occurs via CLDN2 and CLND12,² while anions permeate this segment via CDLN10a.³ In the thick ascending limb, the paracellular shunt is cation-selective, with CLDN10b appearing to form pores preferentially permeable to monovalent cations, while pores formed by CLDN16 and CLDN19 are preferentially permeable to divalent cations. Notably, CLDN14 is highly regulated by the calcium-sensing receptor, and when plasma calcium levels increase, CLDN14 expression increases markedly contributing to a paracellular barrier to this segment.⁴ The distal nephron is a tighter epithelium than the proximal tubule and thick ascending limb. However, it exhibits some anion-selective permeability, potentially mediated by CLDN4 and CLDN8, although conflicting evidence exists regarding their contribution. CLDN3 and CLDN7 are expressed in the distal nephron and likely contribute to the barrier properties of these epithelia.

Overexpression of individual claudins in cell culture has produced variable effects on ion permeability. These variations are likely due to the cell line used, and a result of interactions with endogenous claudins or changes in the expression of endogenous claudins.⁵ Consistent with this, overexpression of CLDN4 in the OK proximal tubule model increases transepithelial resistance but also upregulates the expression of endogenous CLDN1, CLDN6, and CLDN9.⁶ As discussed previously,⁵ when overexpressed in LLC-PK1 cells, CLDN16 increases sodium permeability with only moderate effects on magnesium permeability. However, humans with pathogenic mutations in CLDN16 display hypomagnesemia with hypercalciuria and nephrocalcinosis due to severe wasting of divalent cations (calcium and magnesium). Furthermore, animal models with targeted deletion of specific claudins can display alterations in the composition of other claudins in the TJ. For instance, knockout of Cldn10a caused redistribution of CLDN2 to the CLDN10a-devoid TJ in the proximal tubule,³ while the deletion of Cldn10b, resulted in increased CLDN16 expression and reorganization in the thick ascending limb TJ.^{7, 8}

Given these challenges, it has been difficult to ascribe a pore- or barrier-forming function to a specific claudin, as changes in the expression of or interactions with other claudins may account for the observed alterations in permeability. To address the role of specific claudins in the absence of the confounding effects of other claudins, Otani et al. generated an MDCK cell model that is nearly devoid of claudins.⁹ They named this line quinKO as it has CLDN1-4 and CLDN7 ablated from it. Importantly, CLDN12 and CLDN16 are still expressed; however, they do not appear to form TJ strands. In this issue of Acta Physiologica, Pouyiourou et al.¹ employed the quinKO cell line to examine the permeability properties of the proposed pore-forming claudins expressed in the kidney.

The authors employed a detailed series of dilution potential and bionic dilution potential measurements to dissect the specific permeability characteristics of most renal claudins. The detailed permeability studies of this extensive investigation are too many to review here. Instead, we sought to highlight select key results that inform renal physiology. The overexpression of CLDN2 in quinKO cells revealed that this claudin confers cation selectivity. Consistent with its role in mediating paracellular sodium and calcium transport in the proximal tubule, the permeability of both these ions was sixfold greater than that of chloride. Interestingly, magnesium permeability in CLDN2-expressing quinKO cells was half that of calcium, which helps explain why significantly more calcium than magnesium reabsorption occurs from the proximal tubule. The authors also examined the permeability properties of another highly expressed claudin in the proximal tubule, namely CLDN10a. Consistent with the data from the Cldn10a knockout mouse, this claudin when overexpressed in quinKO cells, conferred anion permeability, with the permeability of chloride over sodium >4. The glomerular filtrate contains significant concentrations of both bicarbonate and phosphate. The authors thus also examined the paracellular permeability properties of these anions. Here they found that the permeability of both was markedly less than that for chloride. This is consistent with the physiological observations that there is significant paracellular chloride reabsorption from the proximal tubule and that bicarbonate and phosphate are reabsorbed by transcellular processes. Importantly, having greatly reduced phosphate and bicarbonate permeability would prevent diffusive flux in both directions, thereby enabling the tight regulation of the reabsorption of these ions via well-delineated transcellular pathways.

CLDN16 and CLDN19 are critical for calcium and magnesium reabsorption in the thick ascending limb, and CLDN19 is required for the insertion of CLDN16 into the TJ.¹⁰ Using the quinKO model, this was clearly recapitulated, and the permeability characteristics were addressed. Confocal imaging showed CLDN16 localization to the junctional region, but the signal was discontinuous when colocalized with the TJ protein occludin. In contrast, when CLDN19 was co-expressed, it allowed further recruitment of CLDN16 to the TJ region. Detailed freeze-fracture electron microscopy revealed that CLDN16 did not form continuous strands when expressed alone, whereas CLDN19 alone formed a meshwork . This meshwork became more compact when the two claudins were co-expressed. Consistent with a failure to form TJ strands, overexpression of CLDN16 alone produced a leaky epithelium with low transepithelial resistance and high paracellular flux. CLDN19 overexpression alone formed a tight epithelium with high resistance and barrier properties, while co-expression of CLDN16 increased monovalent, and to a greater extent, divalent cation permeability. Importantly, in the CLDN16 and CLDN19 complex, the calcium and magnesium permeabilities were several times higher than those for sodium. This helps explain the physiological observation that CLDN16 and CLDN19 primarily mediate divalent cation reabsorption in the thick ascending limb and provides the first cell model where the permeability to calcium and magnesium conferred by these claudins can be studied in the absence of confounding effects from other claudins as discussed above. Furthermore, CLDN10b, which is also expressed in the thick ascending limb, when expressed in quinKO cells, conferred much higher permeability to sodium than to calcium and magnesium, in line with an important role in mediating paracellular sodium flux in the thick ascending limb.

CLDN4 and CLDN8 are predominantly expressed in the distal nephron. When expressed by itself, CLDN4 dramatically increased the transepithelial resistance and reduced paracellular flux, consistent with CLDN4 being a barrier-forming claudin. In contrast, when CLDN8 was expressed alone, the epithelial layer displayed very low transepithelial resistance and high paracellular flux. Using immunofluorescence microscopy and freeze-fracture electron microscopy, CLDN8 was observed to be largely absent from the junctional region when expressed alone and failed to form continuous TJ strands. Importantly, co-expression with CLDN4 allowed the integration of CLDN8 in TJ strands. CLDN4 showed mild anion selectivity at an acidic pH when expressed by itself, but not with CLDN8. CLDN4 has previously been suggested to form a paracellular chloride channel in the collecting duct. However, absolute chloride permeability was very low when expressed in quinKO cells, suggesting that CLDN4 may not support a high flux of chloride across the segment. Overall, this work is consistent with CLDN4, together with CLDN8, creating a barrier along the distal nephron.

In conclusion, Pouyiourou et al.¹ confirm key permeability characteristics of renal claudins and provide new important insights into their functions that aid our understanding of epithelial transport along the renal tubule.

R. Todd Alexander: Writing – review and editing; writing – original draft. Henrik Dimke: Writing – review and editing; writing – original draft.

The laboratory of R. Todd Alexander is supported by grants from the Canadian Institutes of Health Research, the Natural Sciences and Engineering Research Council of Canada, and the Women and Children’s health research Institute. The laboratory of Henrik Dimke is supported by grants from the Independent Research Fund Denmark, the Carlsberg Foundation, and the Novo Nordisk Foundation, including a distinguished investigator award.