SEARCH

SEARCH BY CITATION

Keywords:

  • Tamm-Horsfall protein;
  • interstitial cystitis;
  • glycosylation

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS, SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

OBJECTIVE

To confirm abnormal glycosylation of Tamm-Horsfall protein (THP) in patients with interstitial cystitis (IC).

PATIENTS, SUBJECTS AND METHODS

The sialic acid content of THP, a critical component of its biological activity, is reduced in patients with IC. N-glycan shows reduced levels of high molecular weight tri- and tetra-antennary sialylated oligosaccharides. These results are supported by quantitative monosaccharide analysis of neutral and amino sugars in patients vs control subjects. THP was isolated from urine samples of 23 patients with IC and 24 control subjects by salt precipitation. The sialic acid contents were measured using 1,2-diamino-4,5-methylene dioxybenzene-high performance liquid chromatography analysis. For N-glycan profiling, purified THP was treated with peptide:N-glycosidase F to release N-glycans. The purified N-glycans were labelled with 2-aminobenzamide and were profiled by high-pH anion exchange chromatography (HPAEC) with fluorescence detection. The neutral and amino sugars were determined by HPAEC with pulsed amperometric detection.

RESULTS

The total sialic acid in patients was half of that in controls. There was a pattern of reduced level of high molecular weight sialylated oligosaccharide in 17 of 23 patients vs four of 24 controls. The total neutral and amino sugars showed a ≈30% reduction in patients. The mean (sem) for the controls was 133.79 (6.51) vs 94.76 (6.67) nmol/200 µg of THP for patients (P < 0.001).

CONCLUSIONS

THP in patients with IC has reduced sialylation and overall glycosylation, and by inference, THP has a role in the pathophysiology of IC.


Abbreviations
IC

interstitial cystitis

THP

Tamm-Horsfall protein

GAG

glycosaminoglycan

DMB

1,2-diamino- 4,5-methylene dioxybenzene

PNGaseF

peptide:N-glycosidase F

PGC

porous graphitized carbon

2-AB

2-aminobenzamide

HPAEC-PAD

high-pH anion-exchange chromatography with pulsed amperometric detection

RT

retention time.

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS, SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Interstitial cystitis (IC)/painful bladder syndrome is a chronic bladder disorder characterized by the symptoms of urinary frequency, urgency and pelvic pain. Several theories have been proposed for its pathogenesis; one for which there is substantial evidence is that there is a functional deficit in the protective function of the bladder mucous layer in individuals with IC [1–3]. Bladder surface mucus is composed in part of glycosaminoglycans (GAGs), which are highly sulphated with anionic properties. Because of its hydrophilic nature, it forms a tightly bound water layer on the urothelium, acting as a protective barrier against toxic metabolites in the urine [4]. In most patients with IC who are tested, there is a mucosal permeability defect that allows urinary solutes such as potassium to penetrate the bladder mucosa and provoke symptoms by depolarizing nerves and muscle [1,3,5,6]. Although there is significant evidence from experimental data that this permeability defect occurs in IC, the underlying cause is unknown.

Another mystery has been the role of Tamm-Horsfall protein (THP) in the urinary tract. THP is the most abundant protein in normal human urine and is synthesized in the thick ascending limb of the loop of Henle in the kidney and is known as uromodulin when isolated from urine during pregnancy [7]; its physiological role in the urinary tract is yet to be determined. THP is 25–30% carbohydrate by weight with eight potential N-glycosylation sites, which are glycosylated with complex sialylated di-, tri- and tetra-antennary chains with small amounts of oligomannose type (Man5-Man8) structures [8,9]. It acts as a natural inhibitor of microbial infection of the urinary tract and the bladder [10] and has immunomodulating properties that are carbohydrate-regulated. THP has also been investigated for its role in the formation of calcium oxalate stones [11]. Its ubiquity in vertebrate species [12] suggests that it serves an important purpose in the urinary tract, but specific, conclusive evidence of that purpose is unknown.

The role of THP in preventing urinary constituents from inflicting damage on the urothelial surface has been shown in earlier studies [13–15]. Normal urine contains metabolic by-products, low molecular weight cationic compounds that are cytotoxic to cultured urothelial cells and that we refer to as toxic factors [14,15]. In animal models, the interaction between toxic factors and the mucous layer of the bladder causes loss of epithelial barrier function and allows the urinary solute potassium to move into the bladder interstitium and cause contractions of bladder smooth muscle in a manner proposed to also occur in IC [15].

In earlier studies we noted that anionic THP sequesters and neutralizes cationic urinary toxic factors [14,15], and THP from healthy individuals is significantly more protective against the known cytotoxic effects of the cation protamine sulphate [1,6,13] than THP from patients with IC. The protective ability appears to depend on the sialic acid present in the protein [15]. The THP concentration in urine was found to be identical in healthy controls and patients with IC [16]. However, the sialic acid content of THP was found to be reduced in people with IC, supporting the notion that THP functions as a protective macromolecule in urine, preventing toxic factors from injuring the mucosa [16].

The aim of the present study was to corroborate the sialic acid findings by comparing fingerprint profiling of THP N-glycans of patients with IC and control subjects, identifying differences in the sialic acid-containing high molecular weight oligosaccharides. In addition, the total monosaccharide content of the protein was determined for both groups, as a reduction in these sugars in patients would also help to confirm the importance of the protein’s glycosylation in its role as a protective urinary macromolecule.

PATIENTS, SUBJECTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS, SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Urine samples totalling ≥500 mL (from several voids) were collected from female patients diagnosed with IC based on all National Institute of Diabetes and Digestive and Kidney Diseases clinical criteria, excluding cystoscopy. In addition, patients must have had continuous symptoms for ≥12 months, a 24-h voiding frequency of ≥12, a Pelvic Pain and Urgency/Frequency score of ≥15 and analogue scales (0–10) scores for pain and urgency of ≥5 [17]. Patients included new and treatment-naive individuals, and those undergoing therapy.

Urine (500 mL) was collected from female control subjects who had no history of any urinary bladder problems, e.g. infection, irritative voiding symptoms or incontinence; no history of pelvic pain or vaginal problems such as vaginitis, vulvodynia or dyspareunia; and who scored 0 on the Pelvic Pain and Urgency/Frequency Patient Symptom Scale [17]. The study procedures were approved by the institutional review board and written informed consent was obtained.

THP was isolated from urine samples using the salt precipitation method of Tamm and Horsfall. In all, 500 mL of fresh urine from each participant was used. Briefly, urine was first centrifuged for 2 min at 3200g to remove crystalloid and other debris. It was next incubated in the presence of 0.58 m NaCl overnight at 4°C, followed by centrifugation at 3200g for 30 min. The THP pellet was rinsed twice with 0.58 m NaCl and re-suspended in purified (milliQ) water. This protein was then desalted by ultrafiltration using the Amicon Ultra 30 kDa molecular weight threshold filter (Millipore, Billerica, MA, USA). Desalted protein was lyophilized and stored dry at −70°C. The purity of the resulting THP was verified by SDS-PAGE. The quantity of protein used in all assays was determined by weighing lyophilized THP on a balance (accurate to 0.01 mg).

Sialic acids were analysed using 1,2-diamino-4,5-methylene dioxybenzene (DMB)-HPLC analysis. The lyophilized THP samples (0.5–1.0 mg) were used to prepare a stock solution of 1 mg/mL in milliQ water. For hydrolysis 100-µg samples were heated to 80°C in 2 m acetic acid for 3 h. The released sialic acids were collected by ultrafiltration through a 10 000 molecular weight threshold filter and derivatized with DMB (Sigma Chemical Co., St Louis, MO, USA) as described by Hara et al.[18]. The fluorescent, DMB-derivatized sialic acids were analysed on the UltiMate 3000 system by reverse phase-HPLC using Acclaim 120 C18 column (Dionex Corp., Sunnyvale, CA, USA) at a flow rate of 0.9 mL/min. Samples were eluted isocratically with acetonitrile (9%) and methanol (7%) in milliQ water for 30 min. The excitation and emission wavelengths were 373 and 448 nm, respectively. The DMB-derivatized sialic acids were identified and quantified by comparing elution times and peak areas to known standards that were similarly treated.

To prepare N-glycans, 100 µg of purified THP was treated with peptide:N-glycosidase F (PNGaseF; New England BioLabs, Ipswich, MA, USA) in 50 mm sodium phosphate buffer, pH 7.5, at 37°C for 24 h. The digest was applied to a Sep-Pak® C18 cartridge (Waters Corp., Milford, MA, USA) equilibrated in water; the protein and most of the detergent bound to the resin. The run-through and water washes containing N-glycans were applied to a porous graphitized carbon (PGC) cartridge. N-glycans bound to PGC in aqueous conditions, and salts and detergents were not retained. N-glycans were eluted from PGC with 30% acetonitrile containing 0.1% trifluoroacetic acid and were vacuum-dried and stored at −20°C.

The purified glycans were labelled with 2-aminobenzamide (2-AB) as described by Bigge et al.[19]. Briefly, samples were dissolved in 10 µL of a solution of 0.44 m 2-AB in 35% acetic acid in dimethylsulphoxide containing 1 m sodium cyanoborohydride. The samples were incubated at 65°C for 2.5 h. The 2-AB-labelled glycans were purified using GlycoClean S cartridge (Glyko, San Leandro, CA, USA) following the manufacturer’s glycan cleaning protocol.

For the desialylation experiment the sample was treated with PNGase F and neuraminidase (BioLabs) in a double-digest and the released N-glycans were purified and labelled with 2-AB as above.

The 2-AB-labelled N-glycans were profiled using the CarboPac PA1 (4 × 250 mm, Dionex) anion exchange column, with a guard column (4 × 50 mm), at a flow rate of 1 mL/min. Glycans were separated in 100 mm sodium hydroxide with a sodium acetate gradient of 0–250 mm in 0–75 min. The data were collected using the UltiMate 3000 HPLC system with RF 2000 fluorescence detector with λex set at 330 nm, and λem at 420 nm with gain 1, 0.5-s response, and medium sensitivity.

For the monosaccharide analysis of neutral and amino sugars, the purified THP was hydrolysed in 2 m trifluoroacetic acid at 100°C for 4 h to release monosaccharides. After drying the hydrolysate, samples were dissolved in water and analysed with high-pH anion-exchange chromatography with pulsed amperometric detection (HPAEC-PAD). Neutral and amino sugars were separated on a CarboPac PA 10 column (4 × 250 mm) with AminoTrap guard column (4 × 50 mm) using an isocratic gradient of 18 mm sodium hydroxide at a flow rate of 1 mL/min. Data were collected using the Dionex DX-600 HPLC system. Identification and quantification of all sugars were obtained by comparison with the standards.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS, SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

The total sialic acid contents were determined by DMB-HPLC analysis of THP for 24 controls and 23 patients with IC. The mean (sem) sialic acid content for controls and patients was 86.57 (4.70) vs 48.42 (4.01) nmol/mg THP, respectively (P < 0.001). The sialic acid level in the patients was almost half that in controls (Fig. 1).

image

Figure 1. Sialic acid content of THP from the urine of 24 controls and 23 patients with IC. The horizontal bars represent the mean. No control subjects had ≤55 nmol of sialic acid, vs 65% of patients. As a diagnostic test these data indicate 65% sensitivity and 100% specificity. The sialic acid content was significantly lower in patients with IC (P < 0.001).

Download figure to PowerPoint

The 2-AB profile of THP N-glycans was obtained by HPAEC with fluorescence detection for the same set of controls and patients as above. The peaks were assigned by referring to standard glycoproteins. Ribonuclease B (from bovine pancreas; Sigma) was used for assigning neutral high-mannose oligosaccharides, while fetuin (from fetal calf serum; Sigma) was used for assigning various sialylated oligosaccharides (Fig. 2).

image

Figure 2. 2-AB profiling of N-glycans: ribonuclease B (5 µg) was used for assigning neutral N-glycans, which shows as a cluster of high mannose peaks of Man5-Man9 in the RT of 21–27 min. Fetuin (10 µg) was used to assign various sialylated N-glycans, which appear as four major clusters of peaks based on sialic acid residues corresponding to monosialyl (RT 34–36 min), disialyl (RT 44–47 min), trisialyl (RT 54–56 min) and tetrasialyl (RT 70–71 min).

Download figure to PowerPoint

To further confirm the assignment of highly sialylated peaks, THP samples from controls were profiled with sialidase treatment. In the sialidase-treated control THP profile, highly sialylated peaks were absent, while in the patient profile, all highly sialylated peaks were of lower intensity (Fig. 3).

image

Figure 3. 2-AB N-glycan profiling obtained from 30 µg of THP. 1. Control, 2. Desialylated control, 3. IC patient. Profile 1 of the control sample shows a major neutral peak at 23.1 min and a major sialylated peak at 61.4 min. Profile 2 of desialylated control shows complete absence of sialylated peaks and the presence of new neutral peaks at 20.1, 26.2, and 33.5 min. Profile 3 of the patient sample shows a reduction in the intensity of the major sialylated peaks.

Download figure to PowerPoint

The ratio of the area under the major peak of neutral oligosaccharide (retention time, RT, 23.1 min) to the area under the major peak of sialylated (RT 61.4 min) oligosaccharide was calculated for the 24 controls and 23 patients. The mean (sem) for controls was 1.94 (0.10) vs 2.40 (0.14) for patients (P < 0.01; Fig. 4). The higher ratio in patients was consistent with the DMB-HPLC results showing lower sialylation in patients than in controls.

image

Figure 4. A scatter plot of the ratio of the area under the major peak of neutral oligosaccharide to the area under the major peak of sialylated oligosaccharide for 24 controls and 23 patients with IC. The horizontal bars represent the mean. There was a pattern of reduced high molecular weight sialylated oligosaccharide in 17 of 23 patients vs four of 24 control subjects (P < 0.001, chi-squared test).

Download figure to PowerPoint

To evaluate the total sugar content of THP, neutral and amino sugars were quantified by the HPAEC-PAD method for the same 24 controls and 23 patients. The mean (sem) for controls was 133.79 (6.51) vs 94.76 (6.67) nmol/200 µg THP for patients (P < 0.001; Fig. 5).

image

Figure 5. A scatter plot of HPAEC-PAD analysis of neutral and amino sugars of control subjects and patients. The horizontal bars represent the mean; 8% of the controls vs 57% of patients had sugars of <100 nmol/200 µg THP.

Download figure to PowerPoint

The monosaccharide analysis showed the presence of fucose, N-acetylgalactosamine, N-acetylglucosamine, galactose and mannose. In patients the amounts of all individual sugars were lower than in controls, with an ≈30% reduction in total monosaccharide content (Table 1). The relative ratio of monosaccharides in controls for fucose to N-acetylgalactosamine, N-acetylglucosamine, galactose, mannose and sialic acid was 1:1.04:6.53:4.90:4.51:2.33, and the respective values in patients were 1:0.99:5.28:4.07:3.59:1.52.

Table 1.  The mean monosaccharide content of THP of patients and asymptomatic controls
MonosaccharideMean, nmol/200 µg THP
ControlIC
  1. The neutral and amino sugars of THP for the 24 controls and 23 patients were determined by HPAEC-PAD and sialic acid was determined by DMB-HPLC analysis. Patients with IC had significantly lower monosaccharide levels (P < 0.001).

Fucose  7.44  6.35
N-acetylgalactosamine  7.74  6.31
N-acetylglucosamine 48.61 33.51
Galactose 36.46 25.82
Mannose 33.59 22.82
Sialic acid 17.31  9.68
Total151.15104.49

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS, SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Several studies have shown that the toxic factors in urine cause a functional deficit in the GAG-rich mucous layer that insulates the transitional epithelium of the bladder [1–4,15,20]. In vitro studies in urothelial cell models showed that cationic toxic urinary solutes of low molecular weight (500–1000 Da) have significant cytotoxic activity [14,15]. In animal models, exposure of the bladder mucous layer to toxic factors and protamine results in the loss of epithelial barrier function [1,5,6,15]. The loss of the epithelial barrier allows the urinary solute potassium to move into the bladder interstitium and provoke sensory nerves and contractions of bladder smooth muscle [1,3,5,6,15]. These findings illustrate the mechanism we have proposed for the initiation of IC/painful bladder syndrome, whereby a dysfunctional bladder mucosa results in the production of frequency, urgency, pain, incontinence, and subsequent bladder tissue damage.

The possibility of such a cascade of events leading to disease raises the question of what mechanism is normally in place to counteract toxic factors in urine. Several in vitro studies found that THP from normal subjects considerably reduces the toxicity of protamine sulphate, a cationic substance used as a positive control in those studies [13–15]. The highly charged quaternary amines contained in protamine sulphate are known to destroy the protective mucosal barrier [1,2,6,13]. THP from normal subjects is significantly more protective against cations than is THP from patients with IC, but when desialylated its protective effect is significantly reduced. These findings suggest that anionic charge on THP could bind to the cationic toxic factors via electrostatic interaction and is responsible for its biological activity.

Although it has been shown that the THP concentration is about equal in normal subjects and patients with IC, the sialic acid content is significantly reduced in patients. Sialic acid is present as a terminal sugar of N-linked glycans and is responsible for the protective biological activity of THP [16]. Mass spectrometry methods of THP N-glycans showed a preponderance of high molecular weight chains of >3800 Da as tri- and tetra-antennary structures containing three or four sialic acid residues [8,9].

The present glycan profiling showed reduced sialylation in patients, as indicated by reduced amounts of tri- and tetrasialylated oligosaccharides. The results of sialic acid analysis and glycan profiling indicate that there should be less total glycosylation in patients, which was supported by the monosaccharide analysis of neutral and amino sugars. This showed a significantly lower total glycan content in patients than in control subjects. These data confirm the mass spectroscopy observations [16] of the pattern of reduced levels of high molecular weight sialylated oligosaccharides, substantiating that patients with IC have abnormal glycosylation, with a lower sialic acid content than in asymptomatic, healthy control subjects.

In the present study we improved the purity of THP from raw urine by pre-centrifuging the urine at 3200g for 2 min before the salt precipitation step; this resulted in the removal of contaminating crystalloids and increased the sensitivity and specificity of using THP as a diagnostic test. If it is assumed that ‘normal’ THP contains ≥65 nmol sialic acid/µg of THP then only three of 23 (13%) patients were normal while 22 of 24 (92%) control subjects were normal. This yields a sensitivity of 87% with a specificity of 92% in diagnosing IC/painful bladder syndrome.

One limitation of the present study is that it was conducted at one centre; currently, a multicentre trial is underway.

In conclusion, taken together, the present data have major implications for explaining the function of THP and for understanding the causes of and tailoring new treatments for IC. These are the first substantial data to propose a major function for THP in the urinary tract, i.e. as a protective urinary macromolecule that functions by biologically inactivating cations that have the potential to injure the mucous layer. Because such protective activity would prevent the subsequent induction of IC-associated symptoms and pathology, defective THP could be the primary factor in the cause of IC. Also, if a defective protein initiates the IC cascade, then genetics is implicated in IC.

Historically, IC has been perceived as a syndrome, named and diagnosed in terms of its signs and symptoms (e.g. ‘Hunner’s ulcer’ and ‘painful bladder syndrome’). If the current data are validated by additional studies, then a new model for understanding IC as a disease will emerge. This model for IC would be based on its underlying pathophysiology, perhaps warranting the name ‘THP deficiency disease’, ‘lower urinary dysfunctional epithelial disease’, or something similar.

ACKNOWLEDGEMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS, SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

We thank Sayaka Imagawa at UCSD Glycotechnology core facility for performing monosaccharide analysis.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS, SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES