The kidney releases a nonpolymerizing form of uromodulin in the urine and circulation that retains the external hydrophobic patch domain

Abstract

Uromodulin [Tamm-Horsfall protein (THP)] is a glycoprotein uniquely produced in the kidney. It is released by cells of the thick ascending limbs apically in the urine and basolaterally in the renal interstitium and systemic circulation. Processing of mature urinary THP, which polymerizes into supramolecular filaments, requires cleavage of an external hydrophobic patch (EHP) at the COOH-terminus. However, THP in the circulation is not polymerized, and it remains unclear if nonaggregated forms of THP exist natively in the urine. We propose that an alternative processing path, which retains the EHP domain, can lead to a nonpolymerizing form of THP. We generated an antibody that specifically recognizes THP with retained EHP (THP + EHP) and established its presence in the urine in a nonpolymerized native state. Proteomic characterization of urinary THP + EHP revealed its COOH-terminus ending at F617. In the human kidney, THP + EHP was detected in thick ascending limb cells and less strongly in the renal parenchyma. Using immunoprecipitation followed by proteomic sequencing and immunoblot analysis, we then demonstrated that serum THP has also retained EHP. In a small cohort of patients at risk for acute kidney injury, admission urinary THP + EHP was significantly lower in patients who subsequently developed acute kidney injury during hospitalization. Our findings uncover novel insights into uromodulin biology by establishing the presence of an alternative path for cellular processing, which could explain the release of nonpolymerizing THP in the circulation. Larger studies are needed to establish the utility of urinary THP + EHP as a sensitive biomarker of kidney health and susceptibility to injury.NEW & NOTEWORTHY In this work, we discovered and characterized a novel form of uromodulin that does not polymerize because it retains an external hydrophobic patch at the COOH-terminus. These findings establish an alternative form of cellular processing of this protein and elucidate new aspects of its biology. We also provide evidence suggesting that measuring urinary nonpolymerizing uromodulin could be a promising assay to assess the risk of acute kidney injury.

Keywords: Tamm-Horsfall protein; kidney injury; thick ascending limb; uromodulin.

Graphical abstract

Figure 1.

Annotated structure of uromodulin [Tamm-Horsfall protein (THP)] and its structural domains. A: schematic diagram of the THP structure and its functional domains. The top shows the initial 640-amino acid THP precursor, and the bottom shows its processed mature urinary form. SP, signal peptide; I, II, III, and IV, EGF domains; D8C, cysteine-rich domain; ZP, zona pellucida domain; GPI, glycosylphosphatidylinositol anchor at 614; CCS, consensus cleavage site at 587; IHP, internal hydrophobic patch (430–436 and 456–462); EHP, external hydrophobic patch (598–607). B: amino acid sequence of full-length THP with color coded domain annotation. The following are indicated in the sequence: red, signal peptide; green, EFG-like domains; blue, D8C disulfide bond domain; purple, ZP-N and ZP-C domains; lavender, EHP; light green, IHP; yellow, CCS; cyan, GPI anchor attachment site; magenta, sites of N-linked glycosylation.

Figure 2.

Nonpolymerizing uromodulin [Tamm-Horsfall protein (THP)] with retained external hydrophobic patch (EHP) is produced in the kidney. A: schematic representation of the THP precursor with annotation of its functional domains, focusing on the COOH-terminus (enlarged). A polyclonal antibody against the indicated sequence encompassing the EHP was developed. B, right: denaturing immunoblot where THP with retained EHP (THP + EHP) was detected by anti-EHP in the urine (U). THP + EHP has a higher molecular weight than mature urinary THP; the latter was detected by reprobing with an antibody (anti-THP) that was raised against mature urinary THP that lacks the EHP (B, left). No EHP signal was detected in mature urinary polymerizing THP (pTHP) that was purified based on its aggregative properties. C: native separate immunoblots revealed that THP + EHP is a dimer in its native form (arrow head, right), whereas mature urinary THP without EHP polymerizes into high-molecular-weight multimers (arrows, left). D: human urine samples were treated with a mixture of deglycosylating enzymes for 0, 1, and 16 h and then resolved by SDS-PAGE followed by immunoblot analysis. Blots were probed with anti-THP or anti-EHP. E: human urine samples were resolved by SDS-PAGE followed by immunoblot analysis in the presence or absence of EHP blocking peptide. CCS, consensus cleavage site; D8C, cysteine-rich domain; GPI, glycosylphosphatidylinositol anchor.

Figure 3.

Proteomic characterization of urinary uromodulin [Tamm-Horsfall protein (THP)] with the external hydrophobic patch (THP + EHP). A: polymerizing THP (pTHP) was removed from the urine by two consecutive salt precipitations. The supernatants were resolved by SDS-PAGE followed by Western blot analysis and probing for THP or EHP. Start, starting urine material; ppt I, urine supernatant after the first salt precipitation, drawn immediately (left lane) or 3 days after storage at 4°C (right lane); ppt II, urine supernatant after the second salt precipitation; pTHP, pTHP that was precipitated. B: bottom-up mass spectrometry characterization on immunoprecipitated THP + EHP was carried out with trypsin and chymotrypsin, showing ∼100% coverage- excluding signal peptide (gray highlight indicates covered sequences) and recovered peptide details are shown for the COOH-terminus. The enlarged box area focuses on the end coverage of the COOH-terminal region, with the sequence ending at F617 in two chymotryptic peptides. C: enlarged peptide sequence of the COOH-terminus, based on the results shown in B. Green and red arrows show sites of potential chymotrypsin and trypsin cleavage, respectively. The data searches included de novo peptide identifications and semispecific protease cleavage, which allows for one terminus of an identified peptide to be nonprotease specific to detect the COOH-terminal end of THP + EHP with the highest confidence level. CCS, consensus cleavage site; GPI, glycosylphosphatidylinositol anchor.

Figure 4.

Localization of uromodulin [Tamm-Horsfall protein (THP)] with the external hydrophobic patch (THP + EHP) in the human kidney. A: immunofluorescence confocal microscopy of human kidney sections stained to detect THP and THP + EHP (referred to as EHP in the figure). EHP staining appeared not only in thick ascending limbs (TALs) but diffusely in other tubules as well, but not in glomeruli. NC, negative control for EHP without primary antibody. Scale bar = 200 µm. B–E: representative high-magnification images from a separate experiment are shown. Scale bars = 100 µm. The separate channels for THP and EHP from the image shown in B are shown in C and D, respectively. The right panels in B–D represent a magnification of the area outlined by the box in the left panel in B. Nonpolymerizing THP + EHP colocalizes with mature THP in TAL tubules, where it is highly expressed at the apical domain (white arrows). However, a less intense THP + EHP signal can also be detected in other segments such as proximal tubules (PTs; blue arrowhead). E: negative control without primary EHP antibody (NC). F: quantitation of the THP + EHP signal per tubule type from three separate fields. CD, collecting duct; DCT, distal convoluted tubule; Glom, glomeruli. ****P < 0.0001 compared to TAL fluorescence.

Figure 5.

Immunoblot analysis of immunoprecipitated serum uromodulin [Tamm-Horsfall protein (THP)] demonstrates the presence of the external hydrophobic patch (EHP). Immunoprecipitation (IP) of serum THP was performed followed by Western blot analysis with anti-THP (raised against mature protein, left) or with anti-EHP (right) and compared with urine (U) and purified polymerizing THP (pTHP). Serum THP and THP + EHP have comparable molecular weights to each other (*) and to urinary THP + EHP (cyan arrow). In the urine, anti-THP recognizes total THP, the majority of which is composed of pTHP (red arrow for the urine band and green arrow head for pTHP) and, to a lesser extent, nonpolymerizing THP + EHP of higher molecular weight (cyan arrow). Anti-EHP recognized serum THP and only urinary THP + EHP (cyan arrow) and did not recognize pTHP (green arrow head), as expected. These findings support the presence of a retained EHP in the serum.

Figure 6.

Proteomic characterization of serum uromodulin [Tamm-Horsfall protein (THP)] confirms the presence of the external hydrophobic patch (EHP). A: bottom-up mass spectrometry (MS) characterization on immunoprecipitated serum THP was carried out with trypsin and chymotrypsin (gray highlight indicates covered sequences and recovered peptide details are shown). The COOH-terminal covered region ends at R606. B: results from a separate experiment are shown focusing on the end coverage of the COOH-terminal region, which ends at K607 (end of the EHP domain). C: MS/MS spectrum view showing fragment matches and relative ion intensities of COOH-terminal peptides from A, left, and B, right. D: enlarged peptide sequence of the COOH-terminus, based on A and B, highlighting the recovered COOH-terminal ends of serum THP and urinary THP + EHP. CCS, consensus cleavage site; GPI, glycosylphosphatidylinositol anchor.

Figure 7.

Detection of uromodulin [Tamm-Horsfall protein (THP)] with the external hydrophobic patch (THP + EHP) in the urine of patients at risk for acute kidney injury (AKI). A, left: design of the sandwich ELISA for the measurement of THP + EHP. Epitope mapping was performed for a commercially available monoclonal antibody (Ab) [anti-THP (m)] and a synthetic hybrid peptide combining the recognizable sequence in mature THP along with the EHP sequence was generated for calibration. A representative standard curve using the hybrid peptide is shown in A, right. B: levels of nonpolymerizing (THP + EHP) and total THP in the urine of a cohort of 20 patients with cirrhosis was measured on admission. Total and nonpolymerizing THP did not have a direct significant correlation (left). After adjustment to total urinary proteins (there was no statistical difference in total urinary protein between the two groups), only THP + EHP, but not total THP, was lower in patients who subsequently developed AKI (middle and right).