Urine glycosaminoglycans in congenital and acquired nephrotic syndrome.

Transcription

1 Kidney nternational, Vol. 4 (1991), pp Urine glycosaminoglycans in congenital and acquired nephrotic syndrome LYDA P. JADREC, GuDo FLLER, and T. MARTN BARRATT nstitute of Child Health, University of London, London, England, United Kingdom Urine glycosaminoglycans in congenital and acquired nephrotic syndrome. To evaluate the specificity of a raised heparan sulphate (H) excretion previously reported in four children with congenital nephrotic syndrome (CN), we measured the urinary excretion of H and chondroitin sulphate (C) in seven children with Finnish-type congenital nephrotic syndrome (CNF), seven with diffuse mesangial sclerosis (DM), nine with focal segmental glomerulosclerosis (FG), 14 with steroid-sensitive nephrotic syndrome of whom eight had a biopsy confirming minimal change histology (N), and 17 controls. The urine H/C ratio in normal children had a median of.36 (observed range.21 to.68) and was independent of age. H/C ratio was significantly greater than controls in CNF (median.8, range.43 to 1.28), DM (median.81, range.49 to 1.13) and FG children (median.66, range.38 to 1.6), but was not in N (median.44, range.28 to.7). There was a positive correlation between the H/C ratio and urine albumin excretion. High H/C ratios are not diagnostic of a particular histological variety of CN. Research on the pathogenesis of proteinuria in renal disease has focused on alterations in the charge and size filtration characteristics of the glomerular basement membrane (GBM) [1 6]. Much of this work has centered around the components of the GBM that are responsible for its charge [7 12]. Heparan sulphate (H), a proteoglycan found predominantly in the lamina rarae of the GBM [11], is the major component of the fixed anionic charge in the GBM [4, 9, 11]. There is also evidence that other anionic groups, such as carboxyl residues linked to GBM glycoproteins, are important [1]. The contribution to the charge barrier of other glycoproteins such as chondroitin sulphate (C) is less clear [7 9]. Vernier et al reported that the concentration of anionic sites in the GBM was decreased in children with congenital nephrotic syndrome (CN) [4]. n their study the cationic probe polyethyleneimine (PE) was used to identify anionic sites, and normal kidney tissue was incubated with heparatinase to demonstrate that H was indeed the major anionic group labelled by PE in the GBM. Vermylen et al [5] found a marked reduction in the GBM content of H in a child with CN histologically characterized as diffuse mesangial sclerosis (DM). They also found an increased excretion of H relative to C in the urine of this case and three other children with CN, two of whom had Received for publication August 6, 199 and in revised form March 15, 1991 Accepted for publication March 19, by the nternational ociety of Nephrology Finnish-type histology (CNF) and the third DM. They suggested that the basic defect was a failure to incorporate H into the macromolecular structure of the GBM with subsequent loss of H into the urine. The purpose of the present study was to investigate H and C excretion in children with various forms of CN together with age-matched controls and children with later onset, acquired nephrotic syndrome. n particular, we wanted to find out whether an increased urine H/C ratio was specific for a particular form of CN. Methods Patients Twenty-three children with early onset N were studied: seven with CNF, seven with DM, nine children with focal glomerulosclerosis (FG), three of whom were under six months of age and all steroid resistant. None were on steroids at the time of study. Fourteen children with acquired steroidsensitive nephrotic syndrome (N) were studied, 1 of whom were receiving steroids at the time of study, at doses ranging from.5 mg/kg alternate day to 2 mg/kg/day. Eight had biopsy evidence confirming minimal change histology. The controls consisted of 17 normal children who either attended the hospital creche or were the children of laboratory staff. Methods Random urine samples were collected with merthiolate 1:1, as preservative, coded, and stored at 7 C until analyzed. The code was not broken until the end of the study. Urinary glycosaminoglycans (GAGs) were isolated with the cationic dye Alcian Blue as described by Whiteman [13]. Alcian Blue in a.5% solution in 5 m sodium acetate buffer ph 5.8 with 5 m magnesium chloride has been shown to precipitate GAGs specifically [13]. The Alcian Blue used was from mpenal Chemical ndustries Ltd. Two milliliters of centrifuged urine was added to 2 ml of the Alcian Blue solution. The resulting GAG precipitate was dissociated with.2 ml 4 M sodium chloride solution and.1 ml methanol. The Alcian Blue was denatured by alkalinizing the mixture with.1 ml.1 M sodium carbonate and.4 ml water, and separated by centrifugation. The GAGs were precipitated from the remaining mixture by the addition of 3 volumes of ethanol and subsequent centrifugation. The supernatant was discarded, the GAGs were dried overnight at room temperature, and dissolved in 2 j.d water. Aliquots of.5 or 1..d were applied to a cellulose acetate strip adjacent to 28

2 Jadresic et al: Urine glycosaminoglycans in N E 1/) C-, U) o.eh Age, years 6 8 Fig. 1. The urinary heparan sulphate! chondroizin sulphate ratio in normal 1 children no significant relationship with age (r =.39, p >.1). Table 1. Urinary GAG excretion in normal controls Age at Controls study (N = 17) years UA/UC H/CR C/CR H/C Median Observed range.4-9, standards containing both H and C at concentrations ranging from.12 mg/m to.75 mg/mi. Electrophoresis was performed in.1 M barium acetate ph 6. for three hours with a current of 7.5 volts/cm. The individual GAGs separated into discrete bands which were visualized by staining with the Alcian Blue reagent. Excess Alcian Blue was removed by washing with a solution containing 5 mm magnesium chloride and 5 mrs sodium acetate. The individual bands were cut out and eluted with ml of a dimethyl suiphoxide reagent, which was prepared by dissolving.51 G anhydrous sodium acetate, 1.27 g MgC2. 6H2 and 1.56 ml glacial acetic acid in 25 dimethyl suiphoxide (igma Chemicals Ltd, spectroscopic grade). The optical density was read at 678 nm. A standard curve was obtained from the standards which were run on the same cellulose acetate strip. The recovery following GAG precipitation and extraction was 54% for H and 62% for C. The within-assay coefficient of variation was 7% for H and 9% for C, and the between-assay reproducibility of the H/C ratio was 13%. The urinary albumin concentration was measured by a single radial immunodiffusion assay [14]. A 2% solution of anti-human albumin antibody (RaHu/Ab Ref 3283, Nordic, Tilburg, Netherlands) in 1.5% agarose gel was used and the concentrations of the samples were calculated by reference to commercial standards (igma Chemicals Ltd). Urine creatinine concentration was measured using a Beckman analyzer utilizing the Jaffé reaction, and results were expressed as the urinary albumin! creatinine ratio (Alb/Cr, ). The Mann-Whitney test and the pearman rank correlation coefficient test were used in the statistical analysis. Results The urinary heparan sulphate/creatinine (H/Cr) ratio (Table 1) in the 17 control children had a median of.8, observed range.3 to.32, and the chondroitin sulphate/creatinine (C/Cr) ratio had a median.23, range.6 to.88. The urinary H/C ratio had a median of.36, range.21 to.68. Both H/Cr and C/Cr decreased with age, but H/C did not (r =.39, P >.1) (Fig. 1). The H/C ratios (Table 2) in the CNF, DM and FG groups were all significantly greater than the control values, but that for N was not (CNF median:.8, range.43 to 1.28, P <.1; DM median:.81, range.49 to 1.13, P <.5; FG median:.66, range.38 to 1.6, P <.1; N median:.44, range.28 to.7, P >.5. There was no difference in the H/C ratio within the N group, between those children with confirmed MCN histology and those without histology (Table 2, Fig. 2). The H/Cr ratio in both CNF and DM was greater than in controls, although it only reached significance level in the CNF group (P <.5). However, there was no significant difference between the C/Cr ratio in the controls and the CNF, DM and FG groups, indicating that the raised H/C ratio is due to an elevation in urine H concentration rather than a decrease in the C concentration (Table 2). The H/C ratio in the CNF, DM, FG and N groups was strongly correlated with the urine Alb/Cr (r =.66, P <.1) (Fig. 3). However, Alb/Cr in the N group was significantly lower than that in the CNF, DM or FG groups (P <.1), but the correlation between H/C and Alb/Cr persisted even when N patients were excluded from the regression analysis (r =.45, P <.5). There was no correlation between H/C and plasma creatmine concentration, nor were there any differences in plasma creatinine concentration between any of the nephrotic subgroups studied. Discussion The observation of Vermylen et al [5] of increased H excretion in the urine of four children with CN was of interest

3 282 Jadresic et a!: Urine glycosaminoglycans in N Table 2. Urinary GAG excretion in CNF, DM, FG and N Age at Age at diagnosis study PCR Pt ex Histology years pmo1/!iter UA/UC H/CR C/CR Congenital nephrotic syndrome Finnish-type M CNF < M CNF < F CNF <.1.8 NA M CNF < NA M CNF < F CNF < M CNF < median < c b range Diffuse mesangial sclerosis 8 M DM M DM M DM M DM M DM < M DM M DM median ,8c c range < Focal glomerulosclerosis 15 M FG M FG M FG M FG M FG F FG F FG F FG M FG median ,lc.7.17 O.66a range Age at Age at diagnosis study PCR Pt ex Histology years p.mo!/liter UA/UC H/CR C/CR H/C H/C teroid dose mg/kg teroid sensitive nephrotic syndrome 24 M MCN alt day 25 M MCN daily 26 M MCN alt day 27 M MCN daily 28 M MCN no steroids 29 M MCN ,7 daily 3 M MCN NA no steroids 31 M MCN no steroids 32 M NA alt day 33 F NA alt day 34 M NA no steroids 35 M NA alt day 36 M NA alt day 37 F NA alt day median c " range Abbreviations are: NA = not available, alt = alternate ap <.1, bp <.1, p <,5, dnot significant when taken in the context of the virtual absence of H in the GBM of one them, suggesting failure of incorporation of H into the GBM. We investigated children with different forms of congenital nephrotic syndrome in order to ascertain whether abnormalities of the H/C ratio were characteristic of a particular type of CN. t also prompted the speculation that urinary H measurement might be of diagnostic value in the assessment of CN. We refined the quantitative assay for H and C in urine used previously [5], introducing separate standard curves in each analysis. The reaction for this was the observation of nonlinearity at high concentrations of GAGs. Moreover, Alcian

4 Jadresic et a!: Urine glycosaminoglycans in N , () ( Ṡ Controls CNF DM FG N Fig. 2. The urinapy heparan suiphate/chondroitin sulphate ratio in controls, CNF, DM, FG and N. Blue precipitates other macromolecules such as Tamm-Horsfall protein, and therefore the calculation of individual GAG concentration from the total GAG measurement and the H/C optical density ratio, as in the previous study, is less accurate than direct measurement. We elected to study the urinary H/C ratio. The use of H/Cr would have been unreliable as the urinary creatinine excretion may be low in children in poor nutritional state. t would have led to artificially elevated values, especially in the CNF group. n addition, we observed that the HC ratio is independent of age. Our results, based on a larger sample of patients, confirm Vermylen et al's findings of an increased H/C ratio in the urine of children with CNF and DM. We also observed a raised H/C ratio in some children with FG. We found that the HC ratio correlated strongly with albuminuria. t might be that increased losses of H from the GBM leads to a greater degree of proteinuria as result of loss of GBM anionic charge. Alternatively, a raised HC ratio might be the non-specific consequence of high levels of proteinuria. t might also be that the loss of H in the urine is related to the extent of structural damage of the glomeruli, which occurs to a varying degree in the different forms of nephrotic syndrome in which we found a raised HC ratio. However, it is not possible on the basis of this study to explain these differences nor their underlying mechanisms. Further studies are needed into degradation and synthesis of H in the GBM in this group of disorders. We confirmed that children with N have a normal urinary H/C ratio. However, they had lower levels of proteinuria. elective proteinuria is unlikely to account for the low urinary H excretion seen in the N group. n CNF there is also selective proteinuria [15], but they had the highest urinary H excretion. Most of the children in the N group were on steroids at the time of study. However, there was no significant difference in the H/C ratios between those children on steroids and those children off steroids. n summary, we have shown that children with CNF, DM and some children with FG, but not children with N, have an increased H content in the urine compared to controls. There was, however, a positive correlation of the H/C ratio with Alb/Cr, suggesting that raised H/C ratios are related to states of heavy albuminuria rather than characteristic of specific histological categories. Acknowledgments Li was supported by a grant from the Kidney Research Aid Fund. We thank the pediatric nephrologists who made samples from their patients available for this study. Alcian Blue used in this study was a gift from mperial Chamical ndustries, Ltd., United Kingdom. Reprint requests to Dr. Lyda Jadresic, Department of Paediatric Nephrology, nstitute of Child Health, 3 Guilford treet, London WCN EH, United Kingdom.

5 284 Jadresic et al: Urine glycosaminogivcans in N ) 1 E U) C-) Alb/Cr, Fig. 3. The relationship between the urinaty albumin/ creatinine and heparan suiphate/creatinine ratio (r =.66, P <.1). ymbols are: ( ) CNF;() DM, () FG; (*) N. References 1. LELONOT B, MAKNO H, KANWAR Y: tatus of glomerular proteoglycans in aminonucleoside nephrosis. Kidney n: 31: , WEENNOii, RENNKE HG: Glomerular permeability and polyanion in adriamycin nephrosis in the rat. Kidney liii 24: , GROGGEL GC, TEVENON J, HOVNGH P, LNKER A, BORDER WA: Changes in heparan sulfate correlate with increased glomerular permeability. Kidney mt 33: , VERNER RL, KLEN Di, sson P, MAHAN JD, OEGEMA TR, BROWN DM: Heparan sulfate-rich anionic sites in the human glomerular basement membrane. N Eng J Med 39:11 19, VERMYLEN C, LEVN M, MOMAN J, BARRATr TM: Glomerular and urinary heparan sulphate in congenital nephrotic syndrome. Pediatr Nephrol 3: , BRENNER BM, HOTETTER TH, HUME MD: Molecular basis of proteinuria of glomerular origin. N Engi J Med 278: , PARTHAARATHY N, PRO RG: Characterization of the glycosaminoglycan component of the renal glomerular basement membrane and its relationship to the peptide portion. Biol Chem 256:57 513, ROENZWEG Li, KANWAR Y: Removal of sulfated (heparan sulfate) or nonsulfated (hyaluronic acid) glycosaminoglycans results in increased permeability of the glomerular basement membrane to '251-bovine serum albumin. Lab nvest 47: , KANWAR Y: Biology of disease. Biophysiology of glomerular filtration and proteinuria. Lab nvest 51:7 21, BERTOLATU JA, HUNCKER LG: Polycation binding to glomerular basement membrane. Effect of biochemical modification. Lab nvest 56:17 179, KANWAR Y, FARQUHAR MG: Anionic sites in glomerular basement membrane: n vivo and in vitro localization to the laminae rarae by cationic probes. Cell Biol. 81: , BLAU EB, HAA ie: Glomerular sialic acid and proteinuria in human renal disease. Lab nvest 28: , WHTEMAN P: The quantitative determination of glycosaminoglycans in urine with Alcian Blue 8GX. Biochem 1 131: , MANCN G, CARBONARA AO, HEREMAN if: mmunochemical quantitation of antigens by single radial immunodiffusion. mt J mmunochem 2: , HUTTUNEN N-P, VEHAKAR M, VUKAR M, LAP L: Proteinuria in congenital nephrotic syndrome of the Finnish type. Gun Nephrol 13: 12 19, 198