Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
CLIN.CHEM.40/9, 1739-1743 (1994) #{149} Molecular Pathology Tamm-Horsfall Glycoprotein: Role in Inhibition and Promotion of Renal Calcium Oxalate Stone Formation Studied with Fourier-Transform Infrared Spectroscopy Rainer Kn#{246}rle,”5 Peter Schnierle,2 Alexander Koch,1 Niels-Peter Buchholz,3 Franz Hering,4 Theodor Ackermann,’ and Georg Rutishauser3 Tamm-Horsfall glycoprotein (THP) from healthy probands inhibits the precipitation of calcium oxalate, whereas THP from individuals who repeatedly develop calcium oxalate stones has no effect or even promotes precipitation. Using Fourier-transform infrared spectroscopy, we found a structural differentiation between these functionally different THP5: a decisive difference in sialic acid content. Quantitative analysis for sialic acid showed the same results. THP from healthy probands had a high sialic acid content (51 ± 9 glkg), whereas THP from recurrent stone formers had a decreased sialic acid content (21 ± 4 g/kg). This explains the dual role of THP in the precipitation of calcium oxalate and the formation of renal stones and shows the importance of glycosylationin the function of this glycoprotein. Indexing Terms: glycated proteins/sieiic acid/kidney stones/uiine/ uromodulin In individuals who develop renal stones and in healthy persons, urine is a supersaturated solution of calcium oxalate, the most common component of kidney stones (1). The urinary concentration of calcium oxalate is at least four times higher than its solubility (2), suggesting that crystal formation occurs in both healthy persons and stone formers (3). Consequently, supersaturation cannot be the sole explanation for the formation of calcium oxalate stones. The main theories concerning stone formation in the urinary tract are the matrix theory and the theory of inhibitors (or, better, lack of inhibitors). Based on results from analyses of many stones, the matrix theory (4) proposes that urinary macromolecules promote stone formation by binding inorganic ions and acting as a framework for the deposition of stone salts. Such macromolecules should be present only in the urine of stone formers or should at least be present in excess in such persons. The inhibitor theory (5) explains the fact that healthy persons do not develop clinically significant stones by postulating the presence of inhibitors in urine that prevent nucleation, growth, and aggregation of crystals in the urinary tract. Accordingly and con‘Department of Physical Chemistry, University of Freiburg, Albertstr. 23a, 79104 Freiburg, Germany. 2Iituth of Inorganic Chemistry, University of Basel, Spitalstr. 51, 4056 Base!, Switzerland. 3Clinic of Urology, Department of Surgery, University Hospital Kantonsspital Base!, Spitalstr. 21, 4031 Basel, Switzerland. Urological Clinic, Kantonsspital Baden, 5404 Baden, Switzerland. 5Author for correspondence. Fax +41/761/2036189. Received February 17, 1994; accepted May 31, 1994. Hans Seiler,2 stone formation in patients with urolithiasis would be indicative of a lack or deficiency of these inhibitors. In the inhibitor theory, pathogenesis of stone formation is regarded as the result of an imbalance between supersaturation with calcium oxalate and inhibitory activity (6). Several studies have identified naturally cccurring inhibitory substances (7), of which there are usually considered to be at least two groups (8): (i) the so-called chelating inhibitors, e.g., magnesium ions, which reduce free ion activity and therefore reduce urinary supersaturation, and (ii) the mostly macromolecular polyanionic inhibitors, which do not influence supersaturation but bind to crystal surfaces, block growing sites, and modifSr attractive or repulsive forces between crystals so that growth and especially crystal aggregation is inhibited (9). All renal calcium oxalate stones contain a macromolecular organic matrix (10), and most include TarnmHorsfall glycoprotein (THP) (11). The most abundant protein in human urine and present in the kidneys of all placental mammals, THP is -30% carbohydrate by weight. The physicochemical and biological properties of THP have been studied extensively, but its physiological function has remained obscure (12). According to its cDNA sequence, THP exhibits eight possible N-glycosylation sites, of which five usually are utilized (13). Proteolytic digests of THP reveal that every carbohydrate moiety shows a different glycosylation pattern (14), and the site heterogeneity is substantial. The resulting glycoforms of THP produced (15) have different physical and biochemical properties, which may lead to functional diversity. Early studies on the chemical composition of THP showed -4% (by weight) glycosylation with sialic acid (16). Under physiological conditions, sialic acid is negatively charged, causing a low p1 value for the glycoprotein (17). Because of its presence in renal stones, THP has been proposed to play a role in renal stone formation. Some groups have described a promoting influence of THP on the precipitation of calcium oxalate, according to the matrix theory (18, 19), whereas others have shown that it has inhibitory properties and probably acts as an aggregation inhibitor (20, 21). Our earlier studies with the oxalate tolerance test (22) showed that THP from healthy probands had an inhibitory effect on the precipitation of calcium oxalate, whereas THP from recurrent calcium oxalate stone formers had no effect or often even a promoting one (23). In the present study we used Fourier-transform infrared (FTIR) spectroscopy and resolution enhancement by versely, CLINICAL CHEMISTRY, Vol. 40, No. 9, 1994 1739 Fourier seif-deconvolution to elucidate possible structural differences between the two functionally different THPs revealed by the oxalate tolerance test. The spectroscopic methods were complemented with biochemical approaches, e.g., polyelectrolyte titration or the thiobarbituric acid assay for sialic acids. MaterIals and Methods Reagents and Chemicals Neuraininidase acetylneuraminic (from Clostridium perfringens), Nacid (sialic acid), deuterium oxide, thiobarbituric acid, and fetuin were purchased from Sigma Chemie, Deisenhofen, Germany. Poly-(1,l,-dimethyl-3,5diinethylene)-piperidinium chloride, potassium polyvinylsulfate, and Toluidine Blue 0 were supplied from Aldrich Chemie, Steinheim, Germany. All other reagents were purchased from Merck1 Darmstadt, Germany. THP from 12 healthy probands and 13 recurrent stone formers was isolated by a modified method of Tanun and Horsfall (24). Urine samples (24-h) were precipitated three times with 0.58 mol/L NaCl and the precipitates dialyzed for 72 h. The final products were lyophilized. Healthy probands were chosen randomly. Of both sexes and various ages, they had no known stone episode so far and no other disease at that time. Stone formers were from the Kantonsspital Base! and Kantonsspital Baden, patients classified as recurrent calcium oxalate stone formers; they were sex- and age-mixed and had no urinary tract infections. We determined the concentration of THP in solution spectrophotometrically at 277 nm, using an absorptivity of 10.8 cm’ for 10 gIL (25). Asialo-THP was obtained from the healthy probands’ THP by two independent methods, acid hydrolysis and enzymatic digestion (26). Acid hydrolysis was carried out in 50 mmol/L H2S04 at 80#{176}C for 1 h. After adjusting the pH of the hydrolysate to -5.5, we extensively dialyzed the glycoprotein against water and lyophilized the product. For enzymatic cleavage, THP was incubated at 37#{176}C for 24 h with neuraminidase, 50 kUIL, in 50 mmol/L sodium acetate/acetic acid buffer, pH 5.5, containing 9 g/L NaCl, 1 g/L CaC12, and 0.2 g/L NaN3. Neuraminidase from C. perfringens is specific for the cleavage of terminal a-2,3-, a-2,6-, and a-2,8-linked sialic acid. After digestion, the glycoprotein was recovered by extensive dialysis against water and then lyophilized. FTIR Investigations of solutions of proteins in the infrared region are usually carried out in deuterium oxide (D2O) because of the strong absorbance of water (H20) in this spectral region. For H/D exchange we lyophilized the dialyzed protein, dissolved it three times in deuterium oxide, and lyophilized again. The solutions were measured in a 50-pm pathlength CaF2 cell. Spectra were recorded with a Bruker IFS-113v FTIR spectrometer equipped with a mercury-cadmium-telluride detector (Bruker Analytische Messtechnik, Karlsruhe, Germany). The spectrometer was controlled by a CS-43 computer and OPUS-software (also from Bruker). The re- 1740 CLINICALCHEMISTRY,Vol.40, No. 9, 1994 sults of 1024 scans were combined to produce a spectrum of 2 cm’ resolution between wavenumbers of 1850 and 1300 cm’. THP Charge The charge of THP was measured by polyelectrolyte titration (27). A known amount of protein (50-150 g) was preincubated in a solution of poly-(1,1,-dimethyl3,5-dimethylene-piperidinium chloride). After adding the indicator Toluidine Blue 0, we titrated the excess polycation with potassium polyvinylsulfate at 23#{176}C with continuous stirring. The titration was monitored by continuous measurement of absorbance at 520 nm. To improve the accuracy of determination, we titrated dilution series of known THP content. The number of charged residues per protein molecule was calculated from the differences of the added and retitrated amounts of polycation preincubated with the protein sample. Sialic Acid Sialic acid was determined in THP from 10 healthy probands and recurrent stone formers by the thiobarbituric acid assay. After acid hydrolysis and subsequent dialysis the diffusate was lyophilized. The amount of sialic acid was determined colorimetrically according to Aminoff (28), a sensitive method specific for free sialic acid. In the reaction sequence, periodate oxidation and coupling of the prechromogen with thiobarbituric acid yields a red chromophore with an absorbance maximum at 549 nm (29). The amount of sialic acid can easily be calculated from the absorptivity at 549 nm. This assay was tested with fetuin, a glycoprotein containing 61 g/kg sialic acid (30). Acid hydrolysis revealed a sialic acid content of 58 ± 3 g/kg, corresponding to a relative error of the determination of 5%. Results Spectra of globular proteins measured as solutions in D20 usually show two intense broad bands reflecting vibrations of the anude group: These are usually described as amide-I’ and amide-Il’ band. The conformationally sensitive amide-I’ band has an absorbance maximum between 1620 and 1660 cm’ and mainly results from a C=O stretching mode. The amide-Il’ band (15001400 cm1), which is not sensitive to the type of secondary structure, is approximated to a N-D vibration. In addition to the ainide group of the peptide backbone, amino acid side chains absorb in the infrared (31). The carboxylate groups of aspartic acid and glutamic acid result in bands at 1583 and 1564 cm1, tyrosine at 1514 and 1616 cm1. The infrared spectrum of glycoproteins results from vibrations both of the core protein and the carbohydrate moiety. Contributions of the carbohydrate to the spectrum in the spectral region between 1850 and 1300 cm’ are assigned to vibrations of amide groups (in N-acetylhexosamines), carboxylate groups (in uronic acids), and the C-H and C-O bond (32). FTIR spectra of THP from healthy probands and recurrent stone formers show two absorbance maxima, at 1638 and 1454 cm (Fig. 1, top). These bands consist of C C L 1800 1700 1600 1400 1500 Wovenumber, cm’ C ‘1 C 1800 1700 1600 1500 Wovenumber. Differences between the spectra of THP from healthy probands and that from recurrent stone formers are greatest in the intensity of the band at 1608 cm’, which could be assigned to the asymmetric stretching vibration of the carboxylate group of sialic acid (32). All healthy probands showed a clearly distinct band at 1608 cm’, but the band was nearly absent in 12 of 13 spectra of THP from recurrent stone formers; the remaining contributions in this region result from vibrations of tyrosine side chains. Spectra of asialo-THP, whether obtained after chemical or enzymatic cleavage of sialic acid, were identical. Comparison with the spectra of untreated THP (Fig. 2) shows the disappearance of the carboxylate vibration at 1608 cm’, a clear indication that this band results from vibrations of sialic acid. Moreover, the spectra of asialo-THP resemble those of THP from recurrent stone formers by having no spectral feature at 1608 cm’ (Fig. 2). The differences of the relative intensities at 1608 cm’ most probably result from different degrees of glycosylation with sialic acid. FTIR spectroscopic investigations of fetuin, another glycoprotein containing sialic acid comparable to the amount in THP, and its asialo derivative confirmed these results (Fig. 3). The deconvoluted spectra of fetuin 1400 cm’ Fig. 1. Top: Infrared spectra of solutionsof THP in 020; bottom: deconvolutedspectraof the solutions. (A) healthy proband; (B) recurrent stone former. Deconvolutionwas carried out by usinga half-width athalf-height of20 cm mentfactor of2. anda resolution enhance- a number of overlapping components that can be identified by application of Fourier self-deconvolution (33). This procedure decreases the width of the bands, leading to an increased separation and thus better visualization of the overlapping component bands comprising the composite band contour. Deconvolution by half-width at half-height of 20 cm1 and a resolution enhancement factor of 2 resolved bands at 1676, 1662, 1652, and 1636 cm’ (Fig. 1, bottom), which are assigned to vibrations of the peptide backbone and correspond to the secondary structure of the core protein (34). The bands at 1636 and 1676 cm’ can be assigned to (3-sheets, whereas the 1652 cm’ band arises from both a-helix and random coils. The band at 1662 cm’ is characteristic for 13-turns. The spectra of both of the functionally different forms of THP are identical in this spectral region, indicating that the secondary structure of the core protein is unaffected in the functional differentiation. From the relative areas, the 1636 cm’ band appears to be the most prominent, indicating a high content of 13-sheets. 1750 1700 1650 Wovenumber. 1600 1550 cm’ -j 1750 1700 1650 Wovenumber, 1600 1550 cm’ Fig. 2. Deconvolutedspectra of THP (A) and asialo-THP (B) from a healthyproband (top) and from an oxalate stoneformer (bottom). CLINICALCHEMISTRY,Vol.40, No.9, 1994 1741 I., 0 A 1750 1700 1650 Wovenumber. 1600 1550 cm’ Fig. 3. Deconvoluted spectra of fetuin (A) and asialofetuin (B). show a clearly distinct band at 1607 cm’ that is lacking in the spectrum of asialofetuin. The band at 1454 cm1 mainly results from the amide-Il’ band. This region also contains contributions of the carbohydrate moiety. The complex structure of the sialylated, sulfated, and N-acetylgalactosamine-contaming oligosaccharide chains of THP leads to several absorbances between 1700 and 1350 cm’, making a complete assignment of all frequencies in this spectral region difficult. The results obtained by FTIR spectroscopy are supported by biochemical methods. The sialic acid content of THP was significantly higher (P <0.001) in 10 healthy persons (51 ± 9 g/kg, range 34-64 g/kg) than in 10 recurrent renal stone formers (21 ± 4 g/kg, range 13-29 g/kg). Healthy probands reveal a distinctly greater content of sialic acid in THP than do recurrent stone formers, in good agreement with earlier investigations of THP (35). Polyelectrolyte titration also shows a loss of titratable negatively charged groups at THP from recurrent stone formers (15 ± 6, range 7-20; in contrast to THP from healthy persons showing 40 ± 10, range 34-50; P < 0.05). These result from sialic acids as well as from sulfated carbohydrates and the amino acid side chains. Discussion Application of FTIR spectroscopy allowed us to structurally differentiate between the functionally different THPs from healthy probands and from recurrent calcium oxalate stone formers. Inhibitory THP from healthy persons was more highly glycosylated with sialic acid than the promoting THP from recurrent stone formers. These results confirm the concepts that (a) glycosylation is important for the function of THP (36) and (b) a lack of sialic acid is the first step in the conversion of mucosubstances to mineralizable matrix (37), which may lead to renal stone formation. Our results show that terminal sialic acid is essential for the inhibitory function of THP from healthy persons. Knowledge of these structural and functional differences of THP makes it possible to combine inhibitor 1742 CLINICALCHEMISTRY,Vol.40, No. 9, 1994 theory and matrix theory of renal stone formation with regard to this glycoprotein. As shown by our study and earlier results (38), THP from healthy persons is a polyanionic macromolecule. According to the proposed mechanism by which urinary macromolecules inhibit crystal growth and aggregation (9), such a molecule could bind to the surfaces of growing calcium oxalate crystals, block the growing sites, and modifr attractive or repulsive forces between the crystals, thereby impeding or preventing aggregation of the crystals. Consequently (39), no large crystal clusters would be formed to obstruct the renal tubules and any microcrystals formed due to supersaturation could easily be washed out from the urinary tract. THP lacking sialic acid (i.e., with fewer negatively charged groups), as is found in patients with calcium oxalate urolithiasis, would no longer be capable of binding effectively to these calcium oxalate crystals. In contrast, THP from stone formers presents an uncharged surface that might, according to matrix theory, act as an additional surface for heterogeneous nucleation and thus provide a framework for the deposition of stoneforming salts. This supports the idea (21) that stone formers are no longer fully protected against the formation of large crystal aggregates, which can be deposited in the urinary tract. Although kidney stone formation is a complex multifactorial disease, the loss of this protection mechanism of THP through an increase in the asialo-THP glycoforms in urine seems to represent a major factor in renal stone formation. One possible explanation for the lack of sialic acid in the carbohydrate moiety of THP from recurrent stone formers may be the inhibition or a defect of an enzyme involved in the processing of the oligosaccharide chains. Inhibition of processing of glycoproteins has been reported in heavy alcohol consumption, leading to an increase in asialotransferrin glycoforms (40). Several inherited diseases are associated with abnormalities in the genes for glycosidases or glycosyltransferases (41). A deficiency of N-acetylglucosamine transferase II gives rise to congenital dyserythropoietic anemia type 11(42). Other examples of genetic diseases involving defective oligosaccharide biosynthesis are I-cell disease and pseudo-Hurler polydystrophy, in which a deficiency of phospho-N-acetylglucosaminyl transferase activity is the primary cause (43). In this regard, investigations of THP from one family by circular dichroism showed an inherited molecular abnormality of the protein in children of recurrent stone formers (44), suggesting that the recurrent formation of renal stones may be an inherited disease caused by a deficiency of glycosidases or glycosyltransferases. In conclusion, FTIR spectroscopy allowed a structural differentiation between THP from healthy probands and that from recurrent calcium oxalate stone formers. It also demonstrated the importance of glycosylation for the functional activity of this glycoprotein, showing that the functional differences between the two THPs result from different degrees of glycosylation with sialic acid, with the secondary structure of the core protein being unaltered. Depending on the glycosylation, THP acts as an inhibitor or promoter of calcium oxalate precipitation and thus of the renal stone formation process. Deconvolution of the infrared spectra was carried out with RAMOP, a software written by D. J. Moffat, National Research Council of Canada The technical assistance of Ursula Friedrich is gratefully acknowledged. This work was supported by the Deutsche Forschungsgemeinschaft through SFB 60. References 1. Coe FL, Parks JH, Asplin JR. The pathogenesis and treatment of kidney stones [Reviewl. N Engi J Med 1992;16:1141-52. 2. Coe FL, Parks JH. Physical chemistry of calcium stone disease. In: Coe FL, Parks JH, eds. Nephrolithiasis: pathogenesis and treatment, 2nd ed. Chicago: Year Book Medical, 1988:38-58. 3. Goidwasser B, Weinerth JL, Carson CC. Calcium stone disease: an overview [Review]. J Urol 1986;135:1-9. 4. Boyce WH, King JS Jr. Present concepts concerning the origin of matrix and stones. Ann Rev NY Acad Sci 1963;104:563-78. 5. Howard JE, Thomas WC, Smith LH, Barker LM, Wadkuns CL. A urinary peptide with extraordinary powers against biological “calcification” (deposition of hydroxyapatite crystals). Trans Assoc Am Physicians 1966;79:137-45. 6. Robertson WG, Peacock M, Marshall RW, Marshall DH, Nordin BEL. Saturation-inhibition index as a measure of the risk of calcium oxalate stone formation in the urinary tract. N Engi J Med 1976;294:249-52. 7 Ryall RL, Hannett RM, Marshall VR. The effect of urine, pyrophosphate, citrate, magnesium and glycosaminoglycans on the growth and aggregation of calcium oxalate crystals in vitro. Cliii Chim Acta 1981;112:349-56. 8. Robertson WG, Scurr DS. Modifiers of calcium oxa]ate crystallization found in urine. I. Studies with a continuous crystallizer using an artificial urine. J Urol 1986;135:1322-6. 9. Robertson WG, Scurr DS. Modifiers of calcium oxalate growth found in urine. II. Studies on their mode of action in an artificial urine. J Urol 1986;136:128-31. 10. Khan SR, Hackett RL. Role of organic matrix in urinary stone formation: an ultrastructural study of crystal matrix interface of calcium oxalate monohydrate stones. J Urol 1993;150:239-45. 11. Fraij BM. Separation and identification of urinary proteins and stone-matrix proteins by mini-slab sodium dodecyl sulfatepolyacrylamide gel electrophoresis. Clin Chem 1989;35:658-62. 12. Kumar S, Muchmore S. Tamm-Horsfall protein-uromodulin (1950-1990) [Review]. Kidney mt 1990;37:1395-401. 13. Williams J, Marshall RD, van Halbeek H, Vliegenthart JFG. Structural analysis of the carbohydrate moieties of human Tamm- Horsfall glycoprotem. Carbohydr Res 1984;134:141-55. 14. Afonso AMM, Charlwood PA, Marshall RD. Isolation and from digests of human TammRes 1981;89:309-19. 15. HArdK, van Zadeihoff G, Moonen P, Kamerling JP, Vliegenthart JFG. The Asn-linked carbohydrate chains of human TammHorsfall glycoprotein of one male. Eur J Biochem 1992;209:895characterisation of glycopeptides Horafall glycoprotein. Carbohydr 915. 16. Fletcher AP, Neuberger A, Ratcliffe WA. Tamm Horsfall glycoprotein. The chemical composition. Biochem J 1970;120:41724. 17. Rambausek M, Dulawa J, Jann K, Ritz E. Tamm Horsfall glycoprotein in diabetes mellitus: abnormal chemical composition and colloid stabffity. Eur J Cliii Invest 1988;18:237-42. 18. Yoshioka T, Koide T, Utsunomiya M, It.atani H, Oka T, Sonoda T. Possible role oxalate crystallization. 19. Rose GA, Sulainian calcium oxalate crystal Urol 1982;127:522-4. 20. Grover PK Ryall of Tamm-Horsfall glycoprotein in calcium Br J Urol 1989;64:463-7. S. Tamm-Horsfall mucoproteins formation in urine: quantitative RL, Marshall promote studies. J VR. Does Tamm-Horsfall mucoprotein inhibit or promote calcium oxalate crystallization in human urine? Clin Chim Acta 1990;190:223-38. 21. Hess B, Nakagawa Y, Coo FL. Inhibition of calcium oxalate monohydrate crystal aggregation by urine proteins. Am J Physiol 1989;257:F’99-106. 22. Brielmann T, Hering F, Seiler H, Rutishauser G. The oxalate tolerance value: a whole urine method to discriminate between calcium oxalate stone formers and others. Urol Res 1985;13:291-5. 23. Schnierle P, Sialm F, Seller H, Hering F, Rutishauser G. Investigations on macromolecular precipitation inhibitors of calcium oxalate. Urol Res 1992;20:7-11. 24. Tamm I, Horsfall FL. Characterization and separation of an inhibitor of viral hemagglutination present in urine. Proc Soc Exp Biol Med 1950;74:108-14. 25. Hamlin LM, Fish WW. Physical properties of Tamm-Horsfall glycoprotein and its glycopolypeptide. mt J Pept Protein Res 1977;10:270-6. 26. Neuberger A, Ratdliffe WA. The acid and enzymic hydrolysis of 0-acetylated sialic acid residues from rabbit Tamm-Horsfall glycoprotein. Biochem J 1972;129:683-93. 27. Horn D, Heuck CC. Charge determination of proteins with polyelectrolyte titration. J Biol Chem 1983;258:1665-70. 28. Ainunoff D. Methods for the quantitative estimation of N-acetylneurauninic acid and their application to hydrolysates of sialomucoids. Biochem J 1961;81:384-92. 29 Schauer R. Chemistry, metabolism and biological functions of sialic acids [Review]. Adv Carbohydr Chem Biochem 1982;40:131234. 30. Spiro RG. Studies on fetuin, a glycoprotein of fetal serum. I. Isolation, chemical composition and physicochemical properties. J Biol Chem 1960;235:2860-9. 31. Chirgadze YN, Fedorov OV, Trushina NP. Estimation of amino acid residue side-chain absorption in the infrared spectra of protein solutions in heavy water. Biopolymers 1975;14:679-94. 32. Tipson RS, Parker FS. Physical methods for structural analysis: infrared spectroscopy. In: Pigman W, Horton D, Wander JD, eds. The carbohydrates: chemistry and biochemistry. New York: Academic Press, 1980:1394-436. 33. Kauppinen JK, Moffatt DJ, Mantsch HH, Cameron DG. Fourier self-deconvolution: a method for resolving intrinsically overlapped bands. App! Spectrosc 1981;35:271-6. 34. Byler DM, Susi H. Examination of the secondary structure of proteins by deconvolved FTIR spectra. Biopolymers 1986;25:46987. 35 Malek RS, Boyce WH. Observations on the ultrastructure and genesis of urinary calculi. J Urol 1977;117:336-41. 36. Hession C, Decker JM, Sherblom AP, Kumar S, Yue RJ, Mattaliano RJ, et al. Uromodulin (Taunm-Horsfall glycoprotein): a renal ligand for lymphokines. Science 1987;237:1479-84. 37. van Aswegen CH, van der Merve CA, du Plessis DJ. Sialic acid concentration in the urine of men with and without renal stones. Urol Res 1990;18:29-33. 38. Scurr DS, Robertson WG. Modifiers of calcium oxalate crystallization found in urine. III. Studies on the role of TanimHorsfall mucoprotein and of ionic strength. J Urol 1986;136:505-7. 39. Edyvane KA, Hibberd CM, Harnett RM, Marshall VR, Ryall RL. Macromolecules inhibit calcium oxalate crystal growth and aggregation in whole human urine. Clin Chim Acta 1987;167:32938. 40. Storey EL, Anderson GJ, Mack U, Powell LW, Halliday JW. Desialylated transferrin as a serological marker of chronic excessive alcohol ingestion. Lancet 1987;i:1292-4. 41. Rademacher TW, Parekh RB, Dwek RA. Glycobio!or. Ann Rev Biochem 1988;57:785-838. 42. Fukuda MN, Dell A, Scartezzin P. Primary defect of congenital dyserythropoietic anemia type II. Failure in glycosylation of erythrocyte lactosarnunoglycan proteins caused by lowered N-acetylglucosaxninyltransferase II. J Biol Chem 1987;262:7 195206. 43. Kornfeld S. Trafficking of lysosomal enzymes in normal and disease states [Review]. J Clin Invest 1986;77:1-6. 44. Hess B, Nakagawa Y, Parks JH, Coo FL. Molecular abnormality of Tamm-Horsfall glycoprotein in calcium oxa]ate nephrolithiasis. Am J Physiol 1991;260:F569-78. CLINICAL CHEMISTRY, Vol. 40, No. 9, 1994 1743