What's known on the subject? and What does the study add?
The nucleotides associated with Tamm-Horsfall protein (THP) identified in this study are 2′(3′) isomers, which are uncommon and very little is known about their biochemical pathway and role in interstitial cystitis (IC). The current study provides additional evidence of our earlier finding that THP is abnormal in IC patients. This can adversely affect THP's protective function, and suggests that THP plays a role in the pathogenesis of the disease.
To identify and characterise Tamm-Horsfall protein (THP)-associated metabolites present in patients with interstitial cystitis (IC). Identification of these metabolites would give us insight into the complex interaction of THP with urinary metabolites and its effect on the protective function of THP.
Patients, Subjects and Methods
THP was isolated from the urine specimens of 146 patients with IC and 87 control subjects, and was analysed for total sialic acid (SA) content by 1,2-diamino-4,5-methylenedioxybenzene. 2HCl (DMB)-high performance liquid chromatography (HPLC).
THP-associated metabolites were isolated by Superdex-200 size-exclusion chromatography (SEC) as a second peak (SP2) and the SP2 was further fractionated into six major components by C18 reverse phase-HPLC.
Ion-pair HPLC analysis was performed to identify and quantify THP-associated metabolites.
The metabolite structures were confirmed by reversed-phase HPLC combined with electrospray ionization (ESI), liquid chromatography-tandem mass spectrometry (LC-MS and LC-MS/MS) and by high-resolution ESI-time of flight mass spectrometry (HR-ESI-TOFMS).
The THP-associated metabolites (SP2) were more prevalent in patients with IC than in the controls (40.4% vs 12.6%, P < 0.001) as determined by Superdex-200 SEC.
Superdex-200 SEC showed higher SP2 content in patients with IC than in controls, as determined by area under the peak/100 μg of THP. The mean (sem), for patients was 84.3 (8.1) vs 18.0 (2.4) milli-absorption unit*min, for controls (P < 0.001).
Total SA content of THP was much lower in SP2-positive patients with IC than those who were SP2-negative. The mean (sem) was 41.6 (3.2) vs 92.6 (2.2) nmol/mg THP, respectively (P < 0.001).
Ion-pair HPLC identified SP2 metabolites as nucleotides, namely 5′-CMP, 3′-UMP, 3′-AMP, 3′-GMP, 2′-AMP and 2′-GMP. The total nucleotide content of SP2 was significantly higher in patients with IC than in controls. The mean (sem) was 25.3 (2.9) vs 2.2 (0.2) μg/mg THP, respectively (P < 0.001).
LC-MS and LC-MS/MS confirmed the nucleotides based on retention time on HPLC, and mass to charge ratio (m/z) of the parent ion and the daughter ions. HR-ESI-TOFMS gave further confimation of the nucleotide sturctures with high mass accuracy.
In the THP of patients with IC, there is a direct correlation between reduced SA levels and high prevalence of nucleotides associated with it.
The THP of patients with IC has a much higher content of these nucleotides than control, and these unique nucleotide isomers identified are very consistent in all SP2-positive patients with IC, suggesting biological significance.
This study provides additional evidence that THP is abnormal in patients with IC.
high-resolution ESI-time of flight mass spectrometry
liquid chromatography-tandem mass spectrometry
Pelvic Pain Urgency and Frequency (questionnaire)
labelled Superdex peak 2 (RT of 17.0 min)
Evidence supports the concept that bladder symptoms in interstitial cystitis (IC) are primarily due to a dysfunction in the transitional epithelium of the bladder [1, 2]. The surface umbrella cell has mucus at the outer layer of its apical membrane that is composed of both glycosaminoglycans and glyocoproteins. Scientific evidence has shown that this mucus is responsible for regulating the permeability of the epithelium to small molecules and ions in both rodents and humans [1-4]. When the glycosaminoglycan layer is experimentally injured by protamine sulphate (a highly charged cation), solutes leak through this layer into the bladder wall. Patients with IC are reported to have a dysfunctional, or ‘leaky’, bladder mucous layer . Data supports potassium as a urinary metabolite that can generate bladder symptoms when the epithelium is dysfunctional . Urinary solutes that are cationic could also injure mucus in a similar way to protamine, resulting in a leaky epithelium. A low molecular weight cationic fraction isolated from human urine (referred to as toxic factors) was found to injure cultured urothelial cells . These toxic factors are neutralised by Tamm-Horsfall protein (THP) [7, 8].
THP is the most abundant protein in human urine and is also called uromodulin . THP is ≈30% sugar by weight due to the presence of eight potential N-glycosylation sites that terminate in sialic acid (SA) [10, 11]. Data suggests that negatively charged SA rendering anionic properties to THP, sequesters and neutralizes urinary cationic toxic factors [6, 7]. THP from normal subjects was found to be more protective than THP from patients with IC in preventing urothelial injury from cations . The biological activity of THP to function as a protective molecule was found to reside in the SA content on the N-terminal glycans . In several studies, including a ‘blinded’ multisite study [13, 14], the SA content of THP in patients with IC has been shown to be reduced compared with asymptomatic control subjects. In both studies the THP concentration in urine was equal in both control subjects and patients with IC.
THP in patients with IC was found to have tightly bound compounds in significantly larger quantities compared with the THP of control subjects. The present study was conducted to identify and quantify these protein-associated metabolites in patients with IC and control subjects. Identifying this protein-bound material would give us insight into the complex interaction of THP with urinary metabolites and its effect on the protective function of THP. Additionally it can help us in determining its potential role in IC.
Patients, Subjects and Methods
The study was approved by the University of California San Diego Human Subjects Committee. Control subjects were women who: were recruited for the study; were not urological patients, had no recurrent renal stones, renal failure or renal disease; had scores of 1 or 0 on the Pelvic Pain Urgency and Frequency (PUF) questionnaire ; and had no prior history of bladder infections, IC, overactive bladder, bladder symptoms, dyspareunia, gynaecological pelvic pain, vaginitis, or vulvadynia.
The patients with IC were women with normal urine analyses showing no infection, a minimum of 15 on the PUF questionnaire, and a minimum of 1 year of continuous bladder symptoms consisting of frequency (≥10 voids/24 h), urgency, and bladder generated pain. They met all of National Institute of Diabetes and Digestive and Kidney Diseases clinical criteria for IC (cystoscopy not required) . Patients with recurrent renal stones, renal failure or renal disease were not included.
THP was isolated from urine via the salt precipitation method of Tamm and Horsfall as described previously . The purity of the resulting THP was varified by SDS-PAGE with silver stain. The SA content of THP was measured by 1,2-diamino-4,5-methylenedioxybenzene. 2HCl (DMB)-HPLC . The protein was quantified by Superdex-200 size-exclusion chromatography (Superdex-200 SEC) with ultraviolet (UV) detection at 277 nm . THP was dissolved at 1 mg/mL in 0.02 m sodium phosphate buffer (pH 6.8) containing 4 m urea, and a 100 μL (100 μg) sample was injected onto the Superdex-200 column. Aggregated THP elutes in the void volume (Peak 1) at retention time (RT) of 7.2 min, corresponding to a molecular weight of several million. The THP from many patients with IC produced a broad second peak eluting as an aggregate of small molecules with molecular weights of several thousand. This second peak at a RT of 17.0 min was labelled Superdex peak 2 (SP2).
The SP2 sample was isolated by preparative Superdex-200 SEC of THP. In all, 0.5 mg THP was injected onto the column and the fraction containing SP2 was collected; 4 m urea was removed by dialyzing the sample against milliQ water using a 100–500 Da MWCO Cellulose Ester dialysis membrane (Spectra/Por® Biotech). The dialysate containing THP-associated metabolites was lyophilized and stored dry at –20 °C. To confirm Superdex Peak 1 to be THP, peak 1 was collected, dialyzed as above and analysed by SDS-PAGE with silver stain.
Nucleotides were profiled by ion-pair HPLC using a Dionex Acclaim® 120 C18 column (4.5 × 250 mm) at a flow rate of 0.7 mL/min. Tetrabutylammonium hydrogen sulphate (10 mm) containing 0.03 m potassium dihydrogen phosphate (pH 7.5) was used as an ion-pair reagent with a methanol gradient of 8.3–28.4% in 0–24 min. The nucleotide standards, 2′(3′)-GMP mixed isomers, 2′(3′)-AMP mixed isomers, and 5′-CMP, were purchased from Sigma. The 3′-UMP was obtained from Toronto Research Chemicals, Inc. The dialyzed SP2 sample from 1 mg THP was dissolved in 1 mL milliQ water. A 10% sample was injected onto HPLC. The nucleotides were identified and quantified by comparing elution times and peak areas to those of known standards at UV 254 nm.
To isolate individual nucleotides, the dialyzed SP2 sample was fractionated using reverse-phase (RP)-HPLC with a C18 column similar to above at a flow rate of 1 mL/min. Eluents used were milliQ water (0.01% trifluoroacetic acid [TFA]) with acetonitrile (0.01% TFA) gradient of 0–70% in 0–20 min, followed by 70% constant for 3 min. Data were collected using Dionex Ultimate 3000 workstation with UV 260 nm. The individual peaks P1 through to P6 were collected and pooled from multiple runs. The fractions were dried on speed-vac and stored at –20 °C.
Individual peaks P1 through to P6 were identified by ion-pair HPLC by comparing the RTs with known standards. To confirm identification, each sample was spiked with 0.5 μg of the corresponding standard and re-run. Nucleotides were quantified by comparison with known amounts of standards using Chromeleon® software.
To prepare samples for mass spectrometric analysis, the dialyzed SP2 samples were desalted by RP-HPLC using a Dionex Acclaim® 120 C18 column (4.5 × 250 mm) at a flow rate of 1 mL/min. Eluents used were milliQ water (0.01% TFA) with acetonitrile (0.01% TFA) gradient of 0–30% in 0–20 min followed by 30% constant for 3 min. The UV detector was set at 260 nm. The nucleotides eluted in the range of 5–7 min were collected and dried on speed-vac. The samples were reconstituted in milliQ water (0.01%TFA) and analysed by liquid chromatography-tandem mass spectrometry (LC-MS and LC-MS/MS).
For LC-MS the samples were separated by C18 RP-HPLC using the above-described method on an Agilent 1260 HPLC with UV 260 nm. In all, 20% of the LC flow (0.2 mL/min) was introduced to the ion source and the remaining was diverted to the waste. LC-MS and LC-MS/MS analyses were performed using a Finnigan LCQDECA mass spectrometer with electrospray ionization (ESI) source operated under positive ion mode. The data were collected and analysed using Xcalibur® software.
The same six individual nucleotide peaks (P1–P6) analysed by ion-pair HPLC were analysed by high-resolution ESI-time of flight mass spectrometry (HR-ESI-TOFMS). The Agilent 6230 high-resolution time of flight mass spectrometer was used to perform HR-ESI-TOFMS analysis on all six peaks along with nucleotide standards. Data were collected by flow injection analysis of the sample on the Agilent 6230 HR-ESI-TOFMS using positive ion mode.
All measured values were expressed as the mean (sem). Patient and control groups were compared with the Student's t-test, and the P value was used to determine statistical significance. For SP2-positive patient vs control data and for correlation between SA level and SP2-positive, the P values were obtained by chi-square test.
THP from urine samples of 146 patients with IC and 87 control subjects was analysed by Superdex-200 SEC for protein content (Fig. 1). SDS-PAGE analysis confirmed Superdex Peak 1 to be pure THP, showing single band at ≈85 KDa in comparison with THP isolated by salt precipitation from a control and patient (Fig. 2). Protein-associated metabolites isolated from SP2 were found in 59 patients (40.4%) and 11 controls (12.6%); the difference was significant (P < 0.001). The quantity of SP2 as determined by area under the peak/100 μg THP was much greater in patients with IC than in controls. The peak area mAU*min was calculated by Chromeleon® software using absorbance in milli-absorption unit (mAU) and RT (min). The mean (sem) for patients was 84.3 (8.1) vs 18.0 (2.4) mAU*min for controls (P < 0.001; Fig. 3).
The SA content of THP for SP2-positive patients was much lower than that of SP2-negative patients: mean (sem), 41.6 (3.2) vs 92.6 (2.2) nmol SA/mg THP, respectively (P < 0.001; Fig. 4). The mean (sem) age of patients and controls was similar: 48.7 (1.4) and 45.2 (1.7) years, respectively (P > 0.05). The average age of SP2-positive patients was higher than that of SP2-negative patients, at 52.1 (2.2) vs 46.4 (1.7) years, respectively (P < 0.03; Fig. 5).
The C18RP-HPLC analysis of SP2 indicated that it is composed of six major components, P1 through to P6, which were isolated for further analysis of individual peaks (Fig. 6). The nucleotide constituents of P1–P6 were identified by ion-pair HPLC. The ion-pair HPLC analysis of patients' SP2 sample show a very reproducible profile, with five major peaks corresponding to nucleotides, while control samples show significantly lower intensity of these nucleotides, in addition to other nonspecific peaks in the shorter RTs (Fig. 7).
The nucleotides elute in the order: 5′-CMP, RT 8.4 min; 3′-UMP, and 3′-GMP, RT 22.2 min; 2′-GMP, RT 25.6 min; 3′-AMP, RT 28.6 min; and 2′-AMP, RT 32.5 min (Fig. 8).
Nucleotides were quantified by comparison to known standards using Chromeleon® software. The total nucleotide content of SP2 in patients with IC was significantly higher than in controls. The mean (sem), was 25.3 (2.9) vs 2.2 (0.2) μg/mg THP, respectively (P < 0.001).
The six major components of SP2 (P1–P6) were further analysed by LC-MS and LC-MS/MS to confirm their structures. The data obtained were compared with known nucleotide standards analysed under similar conditions (Table 1). The same samples were also analysed by HR-ESI-TOFMS to confirm the mass with high accuracy (Table 2).
Table 1. The RT (min), parent ion mass to charge ratio (m/z) and daughter ion m/z of the six major SP2 metabolites P1–P6 compared with nucleotide standards by LC-MS and LC-MS/MS using ESI in positive ion mode.
Molecular ion, m/z [M+H]+
Daughter ion, m/z
Table 2. Measured and theoretical exact masses and their δ values in parts per million (ppm) of the six major SP2 metabolites P1–P6 compared with nucleotide standards using HR-ESI-TOFMS in positive ion mode.
The literature reports that THP is abnormal in patients with IC, with reduced glycosylation and lower SA content [13, 17]. The protein is highly anionic, due to SA, and this negative charge is important to its biological activity as a protective urinary macromolecule [8, 12]. Consequently, the THP of patients with IC would be expected to be less protective against urinary cations that could injure the mucus of the urothelium, initiating a cascade of events leading to IC [1, 12]. The present study provides additional evidence of the altered physical properties of this protein in patients with IC.
The THP-associated metabolites were more than four-times greater in patients with IC compared with controls. Of further interest, SP2-positive patients had 55% less SA in their THP vs SP2-negative patients. In all, 78% of SP2-positive patients had SA levels of <60 nmol/mg THP compared with 8% in SP2-negative patients (P < 0.001). The mean (sem) SP2 area/100 μg THP for SP2-positive patients with SA of <60 nmol/mg THP was 113.6 (13.8) vs 18.0 (5.3) mAU*min for SP2-positive patients with SA of >60 nmol/mg THP (P < 0.001). This clearly indicates the inverse relationship of SA content and SP2 metabolites associated with THP. This suggests that an SA content of <60 nmol/mg THP is a good indicator of abnormality in THP.
The HPLC results suggest that most (>80%) of SP2 was composed of mainly 2′ and 3′ nucleotides, which were independently confirmed by mass spectrometric analysis.
That THP from patients with IC had low SA content and a high affinity for nucleotides is particularly interesting, as the 2′ and 3′ isomers are the only ones bound to THP. Not much is known about the biochemical pathway of 2′ and 3′ nucleotides and its role in the disease.
It is reported that kidney produces 2′, 3′-cAMP in response to renal injury by breakdown of mRNA. 2′,3′-cAMP is exported and metabolised to 2′-AMP and 3′-AMP, which are further metabolised to adenosine, a retaliatory metabolite to protect both the injured cells and neighbouring cells [18, 19]. In addition, 2′ and 3′ nucleotides have been reported to be antiproliferative to cultured endothelial cells . However, they have not yet been shown to be cytotoxic to cultured urothelial cells. mRNA breakdown is most likely a universal response to cellular injury and likely to be seen in patients with IC, resulting in higher concentrations of these nucleotides in the urine.
In the present study, THP was shown to avidly bind to nucleotides, suggesting that THP is structurally altered. It is not known if the nucleotides bind more to THP in patients with IC because of the lower SA content and/or if the urinary nucleotide levels are higher in patients with IC. Urinary nucleotide levels have not been reported. Urine factors in patients with IC that are reported consist primarily of inflammatory mediators that probably reflect late changes secondary to the disease, which are essentially reactions to the disease process . Only a few investigators have reported on factors that may be involved in the development of IC, e.g. anti-proliferative factor  or urinary toxic factors . Our interest is in both protective factors (THP) and toxic factors that could injure epithelial mucus .
It was also shown that metabolic poisons increase the release of 2′,3′-cAMP, 3′-AMP, 2′-AMP and adenosine . We have reported the presence of toxic factors in patients with IC, which can also increase production of these nucleotides. The nucleotides, if not a direct epithelial toxic factor, could hinder THP function by binding to it and further reducing its protective biological activity.
THP has been investigated for a potential role in renal stone disease, but its role is not clear. Data has suggested that both urinary THP levels and its SA content are increased in recurrent renal stone formers . This is in contradistinction to another report that suggests that the SA content is the active part of the protein molecule that prevents micro-crystal aggregation, suggesting that stone formation occurs with reduced SA content and/or reduced urinary THP levels . THP probably has more than one role in the urine and in general its high negative charge could chelate cations, micro-crystals, bacterial by-products etc., but there is no evidence to suggest that patients with IC have increased renal stones. There are other factors, e.g. urinary pH and concentration of cations like calcium and/or sodium, which play important roles in stone formation.
In conclusion, the present study indicates the presence of unique nucleotide isomers associated with THP. A predictable correlation is seen in patients with IC, with low THP SA content and high levels of associated nucleotides. This provides additional evidence that the physical properties of THP are altered, resulting in the compromise of its protective function [7, 8, 12]. These data support our earlier findings that THP is abnormal in patients with IC. Lastly, this evidence also supports the overall concept that urine factors may be involved in the development of IC.
Conflict of Interest
The authors have no conflicts of interest or funding sources to disclose.