Uromodulin in Kidney Injury: An Instigator, Bystander, or Protector?

Abstract

Uromodulin, also known as Tamm-Horsfall protein, is a glycoprotein expressed exclusively by renal tubular cells lining the thick ascending limb of the loop of Henle. Although the physiologic functions of this protein remain elusive, significant progress has been made over the last decade that highlights the importance of uromodulin in the pathophysiology of various diseases, such as medullary cystic kidney disease, urinary tract infections, and nephrolithiasis. Meanwhile, there is a renewed interest in the role of uromodulin in kidney injury, both acute and chronic. In this article, we review the existing evidence that supports a role for uromodulin in acute kidney injury (AKI), chronic kidney disease (CKD), and renal inflammation. Contrary to the conventional view of uromodulin as an instigator in kidney injury, new data from uromodulin knockout mice reveal a protective role for this protein in AKI, possibly through down-regulating interstitial inflammation. In CKD, uromodulin excretion, when adjusted for kidney function, is increased; the significance of this remains unclear. Although it has been suggested that uromodulin exacerbates progressive kidney injury, we propose that the elevation in uromodulin secretion is instead reactive to injury, and reflects an increase of uromodulin in the renal parenchyma where it slows the injury process.

Background

Tamm-Horsfall protein was discovered by Igor Tamm and Frank Horsfall in 1950 when they precipitated a protein from the urine that inhibited hemagglutination of viruses ^1–2. This glycoprotein is exclusively expressed in the kidney, specifically by cells lining the thick ascending limb of the nephron ^3–4. In 1985, Muchmore and Decker isolated a protein named uromodulin from the urine of pregnant women that had immunosuppressive effects on T cells in vitro ⁵. Shortly thereafter, Pennica et al demonstrated by amino acid sequencing that the uromodulin and Tamm-Horsfall protein are in fact identical ⁶. Recent discoveries have underscored the importance of uromodulin (encoded by the UMOD gene) as a regulatory protein in health ^7–8 and in various conditions, such as medullary cystic kidney disease ⁹, glomerulocystic kidney disease ¹⁰, urinary tract infections ^11–12, nephrolithiasis ¹³, and acute kidney injury ^14–15. More recently, polymorphisms in the UMOD gene have been strongly linked to chronic kidney disease ¹⁶, further raising the interest in the role of this protein in progressive kidney injury ¹⁷. There are previous comprehensive reviews that discussed broadly the biology and role of uromodulin in various diseases ^18–21. In the present review, we focus on the potential role of uromodulin in kidney injury, both acute and chronic, in light of more recent in vivo work based on uromodulin knockout and transgenic mice. We will also discuss how the measurement of urinary uromodulin can be optimized for use as a biomarker for kidney disease.

Case Vignette

Over the course of 13 years, a 52 year white man with type 2 diabetes mellitus, hypertension, atrial fibrillation and hyperlipidemia underwent serial measurements of his kidney function and urine albumin levels, as well as two 24 hour urine collections (Table 1). At the beginning of the follow up period, he had normal kidney function and normoalbuminuria. Urine collection at that time showed 105.2 mg of protein excreted per 24 hours. The patient was maintained throughout this period on multiple medications, including an angiotensin-converting enzyme inhibitor, a diuretic, and a beta-blocker. His diabetes management was challenging, requiring combination therapy with insulin, metformin, and glipizide. His Hemoglobin A1C trend is also shown in table 1. In the last five years, he developed progressive microalbuminuria, which stabilized at an albumin-creatinine ratio of 48.7 mg/g during year 13. A repeat 24 hour-urine collection 11 years after the initial collection showed 238.0 mg protein excreted. His kidney function remained preserved.

Table 1.

Laboratory values of the case vignette

Year	SCr (mg/dl)	eGFR (ml/min/1.73m2)*	HbA1c (%)	ACR (mg/g)	24-h Protein Excretion (mg)
1	0.9	93.5	-	-	105.2
2	0.9	93.1	9.9	0	-
3	0.9	92.8	8.8	9.8	-
4	1.0	81.9	9.7	-	-
5	0.9	92.1	10.8	15.8	-
6	0.9	91.8	10.7	16.0	-
7	0.9	91.5	9.2	15.4	-
8	0.9	91.2	10.3	24.0	-
9	0.9	90.9	12.0	29.9	-
10	0.9	90.6	10.8	43.0	-
11	0.9	90.3	9.6	96.1	-
12	0.8	103.1	8.7	69.9	238.0
13	0.9	89.7	8.6	48.7	-

Note: Conversion factors for units: SCr in mg/dl to μmol/L, x88.4; eGFR in mL/min/1.73m² to mL/s/1.73m², x0.01667.

Abbreviations: ACR, albumin-creatinine ratio; SCr, Serum Creatinine; eGFR, estimated glomerular filtration rate; HbA1c, hemoglobin A1c.

^*eGFR calculated using the 4-variable MDRD [Modification of Diet in Renal Disease] Study equation.

Although uromodulin was not measured in this patient, this protein accounts for the majority of the urinary protein in healthy individuals. Therefore, we expect that the first 24 hour urine collection consisted predominantly of uromodulin. Can the baseline levels of uromodulin predict the susceptibility of patients to acute or chronic kidney injury? As we discuss below, diabetic nephropathy at an early stage can increase the urinary excretion of uromodulin. This may be reflected in the subsequent increase in the 24 hour urinary excretion of protein, which may not be solely accounted for by the development of microalbuminuria. What is the significance of increased uromodulin on the pathogenesis of kidney disease? Does it predict deterioration of kidney function?

Pathogenesis

Uromodulin synthesis and secretion

Uromodulin is an 80–90 kDa protein ^{5–6, 18, 22} expressed solely in the thick ascending limb (TAL), ^3–4 with no production in the macula densa cells ²³. It contains several epidermal growth factor-like domains and a zona pellucida domain, and is heavily glycosylated (30% of molecular weight) ^{18, 22}. Within the TAL, uromodulin is predominantly apically targeted, a process facilitated by the addition of a glycosylphosphatidylinositol (GPI) anchor, an apical targeting signal that is acquired in the endoplasmic reticulum ^24–25. Protease cleavage releases uromodulin from the GPI anchor to be secreted in the urine ²⁶. Interestingly, as has been demonstrated in independent studies, there is also a lesser yet significant basolateral release of uromodulin ^{3, 27}. For example, using immuno-electron microscopy of the rat kidney, Bachmann et al showed that the ratio of apical to basolateral uromodulin was about 2:1 ³. In addition, uromodulin is detected in the serum of healthy individuals at a concentration of 30–540 ng/ml ^28–30. The significance and mechanism of this basolateral release are not entirely understood. Additionally, little is known about uromodulin expression and release at the ultra-structural level in stressed conditions. More recent evidence suggests that basolateral release, as opposed to apical secretion, may even be significantly increased during kidney injury ¹⁴.

The urinary uromodulin excreted in 24 hours by humans ranges from 9 to 66 mg in most reports ^29–33, although several other studies reported as much as 70 mg to 113 mg of uromodulin in 24-hour urine ^{28, 34}. While some degree of variability is to be expected due to the differences in nephron mass between participants, an additional reason for the discrepancies is the use of different methodologies in assaying and reporting urinary uromodulin and the small sample size within each study. Although the rate of uromodulin excretion may not be related to kidney function in healthy individuals ³³, the relationship of uromodulin excretion to kidney function becomes more apparent in established kidney diseases. Unfortunately, the lack of large-scale studies that define a standardized normal rate of excretion of uromodulin is a major hindrance to correlating its urinary level to kidney health and disease.

Factors that control uromodulin expression, secretion, and excretion have not been well explored. It is known that the transcription factor hepatic nuclear factor 1 β beta (HNF1B) positively regulates the expression of uromodulin ³⁵. HNF1B binds to the promoter elements of the UMOD gene, and inactivation of HNF1B in vivo is associated with decreased UMOD transcription ³⁵. Interestingly, mutations in HNF1B are associated with features of familial juvenile hyperuricemic nephropathy ³⁶, possibly through a uromodulin link. Mice deficient in COX-2 (prostaglandin-endoperoxide synthase 2) also had profoundly lower uromodulin than wild-type counterpart ³⁷, but it is unclear whether the effect of COX-2 deletion is specific to uromodulin or if it affects the TAL globally. Limited studies in humans and animals have described experimental manipulations or pharmacologic agents that affect uromodulin expression or secretion. Some studies suggested that uromodulin excretion depends on urinary volume ³⁸, although this could not be confirmed by others ³⁹. Factors that could increase uromodulin expression or excretion are dietary salt alone or combined with furosemide ^40–41 and high protein diet ³⁹. Factors that could decrease uromodulin expression or excretion are angiotensin-converting enzyme inhibitors ⁴², possibly colchicine ^43–44 and urinary tract obstruction ⁴⁵. It is important to note that the UMOD mutations causing medulary cystic kidney disease/familial juvenile hyperuricemic nephropathy are associated with a significantly decreased secretion of normal uromodulin ⁴⁶, which is explained by the fact that these mutations hinder the trafficking of uromodulin to the apical membrane of the TAL ^{27, 47–48}.

Uromodulin and kidney injury: a historical perspective

Initial studies localized uromodulin to renal casts seen in acute kidney injury ⁴⁹, and subsequently in almost all urinary casts ^50–51. This led to the notion that the aggregating properties of uromodulin promote cast formation and obstruction in AKI ^{18, 52–53}. After localization of uromodulin to the thick ascending limb ^3–4, predominantly in an apical membrane domain, the findings of “abnormal deposits” of uromodulin in the interstitium during tubulo-interstitial injury suggested a pathologic role of this protein in interstitial inflammation ^{49, 54–56}. The potential deleterious effect of uromodulin was further supported by the fact that intravenous administration of uromodulin in rodents was highly immunogenic ⁵⁷. Therefore, it was speculated that interstitial uromodulin deposits could potentially cause an immunologic response that promulgates interstitial inflammation. However, these conclusions did not consider the possibility that the presence of uromodulin in a damaged area may be coincidental or reactive. In fact, more conclusive evidence was recently provided with the use of uromodulin knockout mice indicating that this protein in effect may be protective in acute kidney injury ¹⁵. However, a cardinal unresolved issue which perpetuates the confusion surrounding the role of uromodulin in kidney injury is the relationship of uromodulin to inflammation.

Uromodulin and inflammation

When uromodulin was rediscovered by Muchmore and Decker, it was described as having immunosuppressive effects on T cells in vitro ⁵. Uromodulin was also found to bind renal cytokines and lymphokines (interleukin 1 and tumor necrosis factor α) ⁵⁸, thereby inhibiting inflammation. However, a number of subsequent studies suggested a pro-inflammatory role of uromodulin, specifically in activating neutrophils ^59–61, monocytes ^62–63 and dendritic cells ⁶⁴. More recently, Saemann and colleagues demonstrated that uromodulin activates myeloid dendritic cells via Toll-like receptor 4 (TLR4), triggering them to reach a fully mature phenotype ⁶⁴. Bone marrow-derived macrophages are also activated by uromodulin. Even though intravenous administration of uromodulin is highly immunogenic and produces cytokines ^{57, 64}, several clinical and experimental observations indicate that uromodulin deposits in the kidney do not have a crucial role in the inflammatory process. For example, there was no correlation between uromodulin accumulation in the interstitium and tubular damage on kidney biopsies ⁶⁵. In addition, an absent or low inflammatory response to extratubular uromodulin deposits has also been reported in kidney allografts with interstitial disease ⁶⁶. Although in vitro experiments suggest that uromodulin antibody formation may be crucial for engaging an inflammatory response ⁶⁷, these antibodies do not necessarily form in response to extratubular uromodulin in vivo ⁶⁸. In fact, it remains unclear whether the uromodulin found in the interstitium is derived from urine extravasation or from basolateral tubular secretion. We recently showed that ischemic injury shifts uromodulin from the apical towards the basolateral domain ¹⁴, suggesting that interstitial uromodulin emanates from increased basolateral release. This was associated with down-regulation of inflammatory signaling in contiguous proximal tubules ¹⁴.

How can we then reconcile the pro-inflammatory properties of uromodulin on myeloid cells in vitro with the anti-inflammatory properties reported in epithelial cells during kidney injury in vivo? We believe there are two different scenarios. First, the opposing effects of uromodulin on different cell types (epithelial vs. myeloid) may cooperatively be geared towards preventing unchecked and persistent inflammation after AKI. For instance, by stimulating renal dendritic cells ^{64, 69}, uromodulin may promote a prompt inflammatory response to clean up cellular debris immediately after the onset of kidney injury ⁷⁰. Simultaneously, uromodulin may also down-regulate inflammatory signaling in epithelial cells ¹⁴. Both effects could serve to minimize the inflammatory burst after injury and/or signal the beginning of repair. The second scenario could be that basolaterally released uromodulin may not have the same pro-inflammatory properties on immune cells as the urinary uromodulin. For example, basolateral uromodulin may not undergo the same type and/or extent of glycosylation as compared to apical uromodulin ⁷¹, which could affect the glycoprotein’s function. This possibility is consistent with the finding that the pro-inflammatory effect of uromodulin on neutrophils may be mediated by a single class of carbohydrate-specific receptors ⁷², although other receptors for uromodulin on neutrophils could also exist ⁷³. Furthermore, Pfistershammer et al recently demonstrated the presence of additional receptors for uromodulin (scavenger receptors, not typically carbohydrate specific) that do not mediate any pro-inflammatory signaling ⁷⁴. We hypothesize that alteration of uromodulin glycosylation by basolateral release limits the interaction of uromodulin to specific receptors present on renal epithelial cells, and these receptors mediate a predominant anti-inflammatory effect.

Table 2 summarizes the possible outcome of various interactions of uromodulin with various cell types within the kidney or urinary tract.

Table 2.

Possible outcome of interactions of uromodulin with cell types within the kidney or urinary tract

Cell Type	Possible Site(s) of Interaction	Outcome	References
Epithelium	Epithelial basolateral surface	Anti-inflammatory	14–15
T Lymphocyte	Urinary space	Anti-inflammatory	5
Neutrophil	Urinary space; interstitium	Pro-inflammatory	59–61
Macrophage	Urinary space; interstitium	Pro-inflammatory	62–64
Dendritic Cell	Interstitium	Differentiation; migration to lymph node	64 69

Recent Advances

Uromodulin and cast formation in ischemic AKI

Uromodulin has been implicated in the pathogenesis of cast formation during ischemic AKI, but the data supporting this premise is either circumstantial ⁷⁵ (casts staining positive for uromodulin) or based on in vitro or ex vivo aggregating properties of the glycoprotein ^{49, 53}. More conclusive evidence emerged recently in an ischemia reperfusion model of AKI in uromodulin knockout mice ¹⁵. Cast formation was not only unaffected by the absence of uromodulin, but in fact was markedly increased in uromodulin knockout mice, correlating with more severe injury ¹⁵. Therefore, it appears that uromodulin is not essential for tubular cast formation in AKI. Nevertheless, this conclusion does not apply to myeloma kidney, for which Sanders and colleagues have elegantly shown that uromodulin binds specifically to Bence Jones proteins to form myeloma casts ^{44, 76}.

Uromodulin expression and secretion after AKI

Only few published studies have investigated the change of uromodulin expression during AKI. Safirstein and colleagues described a decrease in uromodulin messenger RNA in rats after ischemia-reperfusion injury produced by 50 minutes of clamping ⁷⁷. Using 30 minutes’ clamping, Yoshida et al also observed a decrease in uromodulin gene expression using microarray analysis ⁷⁸. To our knowledge, there are no conclusive data on the changes in uromodulin protein expression in experimental AKI ¹⁵. We recently showed that ischemia partially shifted uromodulin from the default apical pathway to the basolateral side ¹⁴. This could theoretically influence its pattern of expression and excretion in AKI.

Several studies have examined uromodulin excretion during clinical AKI in the setting of cardiac surgery ⁷⁹, intensive care unit ⁸⁰ and kidney transplant ⁸¹. Although these are small studies, they suggest that the excretion of this protein is decreased during AKI. Further proteomic studies are needed to determine whether a decrease in urinary uromodulin would be a suitable biomarker for AKI ⁸². Could the baseline rate of urinary uromodulin excretion be predictive of susceptibility to AKI? A study by Romero et al suggested that a higher uromodulin level in patients receiving a liver transplant was associated with a lower risk of developing AKI ³⁴. A high excretion rate in patients without pre-existing kidney disease could be suggestive of a larger nephron mass, thereby more renal reserve. Alternatively, if uromodulin does have anti-inflammatory protective properties in kidney injury ^14–15, a higher urinary level could be protective.

Role of uromodulin in AKI: a case for a protective tubular cross-talk

Renal ischemia-reperfusion in uromodulin knockout mice is significantly worse than in wild type animals. This is associated with increased inflammatory response and increased neutrophil infiltration in the outer-medulla ¹⁵. The phenotype of injury is predominantly that of tubular necrosis affecting the S3 segments but not the uromodulin-deficient TAL ^14–15. The proximal S3 segments have an increased expression of the pro-inflammatory protein TLR4 and a neutrophil chemoattractant, the chemokine CXCL2. Neutralization of CXCL2 is protective, and partially rescues the S3 segments, suggesting that a TLR4-CXCL2 pro-inflammatory pathway may be important in the pathophysiology of injury in the uromodulin knockout mouse ¹⁴.

That uromodulin produced in TAL affects the susceptibility of neighboring proximal tubules to injury strongly suggests a uromodulin-dependent TAL-S3 tubular crosstalk, a phenomenon previously proposed by Safirstein ^{77, 83}. A shift of uromodulin from the apical to the basolateral surface and the presence of this protein in the interstitium during AKI could suggest that uromodulin itself is the principal mediator of such crosstalk. The presence and up-regulation of a putative uromodulin receptor on proximal tubule also supports this notion ¹⁴. However, we cannot rule out a more complex mechanism that involves other intermediary molecules from TAL or uromodulin interactions with other cells in the interstitium. Figure 1 is a conceptual model that summarizes potential mechanisms for this uromodulin-dependent TAL-S3 protective tubular cross-talk during AKI. We also cannot rule out that a pre-existing TAL tubular dysfunction in uromodulin knockout mice may have a contributory role in the increased susceptibility to AKI by affecting auto-regulatory mechanisms within the nephron such as tubulo-glomerular feedback. For example, in a different model of uromodulin knockout mice, Bachmann et al showed increased expression of distal nephron transporters and deceased expression of important juxtaglomerular effectors such as COX-2 and renin ⁷. However, these mice did not have a difference in any solute excretion compared to wild type ⁷. More recently, Mutig et al also showed that uromodulin could play a permissive role for the proper functioning sodium/potassium/chloride (Na⁺/K⁺/2Cl⁻) cotransporter ⁸⁴. Indeed, further characterization of the physiological phenotype of uromodulin knockout mice will clarify the importance of these factors during AKI.

Figure 1. Conceptual model of the uromodulin-dependent protective cross-talk in Acute Kidney Injury (AKI).

This figure represents a cross-section of the outer medulla, where thick ascending limb (TAL), and S3 segments of the proximal tubules are contiguous. During AKI, S3 segments are predominantly injured, and express several pro-inflammatory factors. Path 1 illustrates a direct role of basolateral released uromodulin in down-regulating inflammatory signaling in neighboring S3 segments. Path 2 shows the possibility of a secondary paracrine mediator, released by the TAL in a uromodulin-dependent process with the net effect of protecting neighboring tubule. Path 3 shows a possible “hybrid” mechanism, whereby the protective effect of uromodulin depends on its interaction with an interstitial cell (IC), which in turn regulates or inhibits inflammatory signaling in S3 segments.

The translational relevance of the protective role of uromodulin in AKI remains to be seen. Further investigations are needed to determine whether stimulating uromodulin expression in clinical AKI is protective and to clarify the role of measuring urinary uromodulin in that context.

Uromodulin expression and excretion in chronic kidney disease

The rationale for measuring urinary uromodulin would be to either provide a prognostic tool or a biomarker of disease. There are a multitude of small studies in humans that have measured the rate of urinary uromodulin excretion in chronic disease states (comprehensively reviewed in²⁰). Interpreting these studies is limited by small sample size. Uromodulin excretion could be reflective of at least two factors: the functioning nephron mass present and the regulation of expression imparted by injury ³³. Indeed, uromodulin excretion does decline with reduced GFR ^{30, 33}. However, to better correlate excretion with biologic function, uromodulin excretion should be adjusted to the remaining functioning nephron mass by normalizing the 24-hour uromodulin excretion to glomerular filtration rate, similar to what was shown by Thornely et al ³³. These investigators demonstrated that the adjusted 24-hour uromodulin excretion (reported as μg/ml of creatinine clearance) had a higher range in patients with CKD ³³. Furthermore, they showed in kidney biopsies that intact tubules in patients with CKD are associated with a significantly higher excretion rate. Therefore, the absolute uromodulin excretion declines with the reduction in total nephron mass seen with CKD, but the amount of uromodulin secreted by each single functioning nephron unit is increased. Consequently, in patients with early CKD and a preserved GFR, we expect uromodulin excretion per 24 hours to be increased. Indeed, this has been clearly demonstrated in early diabetes without GFR impairment as discussed below. This could also explain the findings by Kottgen et al that increased levels of uromodulin precede the onset of CKD ⁸⁵. We advocate the use of 24 hour uromodulin excretion normalized to GFR as the best index for measurement. Potential uses and interpretation of this “uromodulin index” are outlined in figure 2.

Figure 2. Measurement of urinary uromodulin in the form of an index.

The interpretation of this index depends on the clinical context: no kidney disease, at risk for kidney disease, and established disease. In patients with normal kidney function and no risk factors for kidney disease, a low index could suggest susceptibility to acute kidney injury (AKI) (reference 34). In patients with established risk factors for chronic kidney disease (CKD), such as diabetes mellitus, a high index could be reactive to subtle injury, and thereby predict the development of CKD (reference 85). In patients with established CKD, a high index could signal ongoing injury and further progression. However, in very advanced disease, a low index could imply depletion of tubular reserves and progression to end-stage. Abbreviations: GFR, glomerular filtration rate.

Uromodulin in diabetic kidney disease

The relationship of uromodulin with CKD is probably best studied in diabetic nephropathy. Experiments in rats show that uromodulin excretion is persistently increased after inducing diabetes using streptozotocin ^{38, 86}, and this occurs independently of renal mass ³⁸. Uromodulin excretion in patients with diabetes mirrors these experimental findings. Patients with diabetes and early kidney disease but without a significantly reduced GFR have an increase in 24 hour uromodulin excretion ^87–89. This could correlate with the hyper-filtration stage of diabetes ⁸⁸. However, the effect of urine flow on uromodulin excretion rate remains unclear ^{38–39, 90} and uromodulin excretion frequently does not correlate with microalbuminuria ^{87, 89}, suggesting an independent effect of diabetes on distal tubular function. With more advanced disease, 24 hour excretion decreases ^91–92, which could be a reflection of a decrease in functioning renal mass. Baseline uromodulin measurements could potentially be useful in predicting worse future outcomes ⁹³. However, successive longitudinal measurements which could be more informative have yet to be performed.

Uromodulin function in chronic kidney disease

The recent association of uromodulin variants with chronic kidney disease ^{16, 85, 94}, in addition to the link of UMOD mutations with familial juvenile hyperuricemic nephropathy ⁹, revitalized the field of uromodulin research ¹⁷, but a causative link between uromodulin and CKD is still missing. Uromodulin production per functioning nephron unit increases in chronic kidney disease ³³. It appears that this increase in urinary excretion is also associated with an increase in basolateral release of uromodulin evidenced by an increase in the level of serum uromodulin in CKD, as previously shown by Thornelly ³³ and more recently by Prajczer et al ⁹⁵. Interestingly, when serum uromodulin does not increase, it is associated with a worse GFR and a higher tubular atrophy score ⁹⁵. The significance of increased basolateral/interstitial and serum uromodulin is unclear. Although the increased uromodulin in the interstitium could stimulate a pro-inflammatory response and aggravate injury ^95–96, we propose that this increase is in effect reactive and counter-inflammatory^14–15. Figure 3 depicts the various proposed pathways whereby uromodulin exerts a modulatory role in chronic kidney disease.

Figure 3. Possible mechanisms whereby uromodulin can modulate chronic injury.

(A) a possible role of apical uromodulin in maintaining the integrity of the thick ascending limb (TAL) by forming a protective layer at the apex of TAL. Uromodulin may also regulate the function of some ion channels and transporters (references 7–8, 84). If injury affects the normal sorting or function of uromodulin at the apical membrane, a breach of TAL integrity and significant electrolyte dysregulation can occur. Panel B illustrates a pro-inflammatory role of uromodulin. Chronic injury can lead to increased basolateral/interstitial uromodulin (1). In this location, uromodulin can stimulate inflammatory signaling (2) through an immunologic response, thereby aggravating injury (3) and promoting subsequent fibrosis. Panel C shows the possibility that uromodulin has a protective role in chronic injury. A persistent inflammatory response during chronic injury (1) can stimulate the release of basolateral/interstitial uromodulin (2). This reactive increase of uromodulin in the interstitium serves to counter the inflammatory response and limit its progression (3). Supportive data for this model are derived from acute injury experiments on uromodulin knockout mice.

The association between single-nucleotide polyporphism (SNP) common variants in the UMOD gene and the risk of CKD using genome-wide association studies (GWAS) has generated significant enthusiasm. Two SNPs (the T allele of reference SNP identification number 12917707 [rs12917707] and the C allele of rs4293393) within the promoter region of the UMOD gene are associated with a decreased risk of CKD and a lower level of urinary uromodulin excretion ^{16, 85, 97}. Do these associations give insight into the role of uromodulin in the pathogenesis of CKD? One explanation could be that if uromodulin has a pathogenic role in CKD, a lower urinary level associated with these variants may be protective. However, a major problem underlying this interpretation is that UMOD SNPs account for an extremely small portion of the variance in kidney function ¹⁶, arguing against a major pathogenic role of the above described loci. Unfortunately, this is a common pitfall of GWAS ⁹⁸, and a few SNPs may unevenly associate with undiscovered rare variants, structural mutations, or other interactions that could be more directly linked to disease pathogenesis ^98–99. Therefore, an alternative explanation would be that the “protective” UMOD promoter variants may be associated with an undiscovered interaction (affecting uromodulin or its regulators) that is responsible for the phenotypic “resistance to CKD”. Consequently, uromodulin excretion may be reactive to such phenotype, rather than related directly to the SNP variant itself. More detailed identification of UMOD mutations could help bridge the gap in knowledge between SNP genotype and disease phenotype.

Can we learn about uromodulin function in CKD from the mutations that have been linked to familial juvenile hyperuricemic nephropathy/medulary cystic kidney disease? Indeed, these mutations of uromodulin are associated with progressive kidney disease, and share common features such as accumulation of misfolded uromodulin in the endoplasmic reticulum, and decrease in secretion of the normal protein ^{10, 46–48}. However, a cardinal question that remains unanswered is whether the described progressive injury is related to a gain of abnormal function of the mutant protein, to a loss of function of the normal protein, to the stress in the endoplasmic reticulum from the accumulation of the misfolded mutant protein, or a combination of these factors. Also, is the modulatory effect of uromodulin mediated by its activity on the apical side or on the de novo basolateral-interstitial pathway? Studies of chronic injury models in the uromodulin knockout mice will clarify whether a uromodulin loss of function is a key factor in the progression of kidney disease. Further characterization of mice transgenic for uromodulin mutants ⁴⁸ will also be crucial to understanding the functions of this protein in chronic kidney disease.

Summary

In returning to the case vignette, this patient has several risk factors for kidney disease. If we assume that uromodulin is the predominant protein excreted in his urine, then his uromodulin index would have increased from 1.1 mg/unit eGFR (105.2 mg/93.5 ml/min/1.73m²) in year 1 to 2.3mg/unit eGFR (238.0 mg/103.1ml/min/1.73m²) in year 12, suggesting a high likelihood of kidney disease progression. This patient should undergo intensive therapy to control his CKD risk factors. This case illustrates how the future incorporation of uromodulin measurements in the care of patients with CKD could potentially influence management and help in providing a better prognosis.

In conclusion, a renewed interest in the understanding of the role of uromodulin in kidney injury has been witnessed over the last few years. This has been enhanced by the availability of knockout and transgenic mice, which provide invaluable tools for in vivo functional studies. It appears that uromodulin has a protective role in acute kidney injury, but the link between uromodulin and chronic kidney disease is still not evident. Indeed, further studies are needed to clarify the role of uromodulin in kidney injury and inflammation. In addition, this review underscores the need for large-scale human studies to validate the use of urinary uromodulin as a tool to investigate prognosis and management options of kidney disease.