Selectivity of proteinuria in congenital nephrotic syndrome of the Finnish type

Kidney International, Vol. 8 (1975), p. 255—261

Selectivity of proteinuria in congenital nephrotic syndrome of the Finnish type NIIL0-PEKKA HUTTUNEN, ERKKI SAVILAHTI and JUHANI RAPOLA Children's Hospital, University of Helsinki, Helsinki, Finland

responded well to steroid therapy and in whom

Selectivity of proteinuria in congenital nephrotic syndrome of the Finnish type. The selectivity of proteinuria was determined in 14

histologic examination revealed minimal changes

patients with congenital (CNF) and 21 patients with acquired

than in those patients who had glomerulonephritic lesions in their renal tissues. This observation has since been confirmed by other investigators [3—7]. The congenital nephrotic syndrome of Finnish type

nephrotic syndrome by estimating, with quantitative immunological methods, the relative clearance of orosomucoid,

transferrin, IgG and a-2-macroglobulin. Selectivity was generally

as high in patients with CNF, as it was in those with steroidsensitive nephrotic syndrome with minimal changes in renal

(CNF) is known to be resistant to steroid and other immunosuppressive therapy [8]. However, Perheentupa has suggested, on the basis of results obtained

histology. In contrast, selectivity was generally low in patients with

steroid-resistant nephrotic syndrome with glomerulonephritic history. However, in those CNF patients who had particularly severe glomerular lesions, selectivity was low over the molecular weight range of orosomucoid to IgG but notably higher for a-2-

from the immunoelectrophoresis of serum and urinary proteins in CNF patients, that in CNF

macroglobulin. This phenomenon was thought to be a consequence of the heterogeneity of the structural abnormality in the glomerular filter in those CNF patients.

proteinuria is nevertheless highly selective [9]. In the present study we have determined the selectivity of proteinuria in CNF by using a more precise

Sélectivité de Ia protéinurie dans le syndrome nephrotique congén-

ital de type Finnois. La sélectivité de Ia protéinurie a été deter-

technique, and have compared these results with

minée chez 14 malades atteints de syndrome nCphrotique congén-

ital (CNF) et 21 malades atteints de syndrome néphrotique ac-

those obtained for patients with other types of

quis. Les clearances relatives de l'orosomucoIde, de Ia transferrine, de l'IgG et de l'a-2-macroglobuline ont été dCterminées a I'aide de techniques immunologiques quantitatives. La sélectivitC est gCnCr-

childhood nephrotic syndrome.

alement aussi importante chez les malades atteints de CNF que

Methods

chez ceux ayant un syndrome nephrotique sensible aux stéroldes et des modifications histologiques minimes. Au contraire, Ia sClectivité est genéralement faible chez les malades atteints de syndrome nephrotique resistant aux stéroldes avec des lesions histologiques de glomCrulonéphrite. Cependant les malades atteints de CNF qui ont des lesions histologiques particulièrement sévères ont une sClectivité faible pour l'éventail des poids moléculaires compris entre

Patients. This series comprised 14 patients with congenital and 21 patients with acquired nephrotic syndrome treated at Children's Hospital, University of Helsinki, from 1971 to 1973. The patients with ac-

quired disease were divided into two groups ac-

l'orosomucoIde et l'IgG mais une sClectivité élevée pour l'a-2macroglobuline. Ce phénomène est interprété comme Ia consCquence de l'hCterogénéite des anomalies structurales du filtre

cording to their response to steroid therapy. The 35 patients studied were thus grouped as follows:

glomerulaire.

1) CNF patients. This group included 14 con-

Blainey et a! [1] first used the term "selectivity of proteinuria" to indicate that the clearance of proteins through the damaged glomerular filter is inversely proportional to their molecular weight. Estimations

secutive patients with congenital nephrotic syndrome of Finnish type, ages 1 to 20 months. None had been

treated with steroids. One patient had received

tive [2]. Blainey et a! also found that proteinuria was

cyclophosphamide without effect before the time of the present study. 2) Responders. This group consisted of 13 consecutive patients with steroid-sensitive nephrotic syn-

more selective in those nephrotic patients who

drome, ages 3 to 11 yr. Five patients were newly

of protein clearances are based on the observation that the tubular reabsorption of proteins is nonselec-

diagnosed, and eight were patients in relapse. Three

of these latter eight patients had previously been treated with cytostatic agents. At the time of sam-

Received for publication May 2, 1974; and in revised form March 13, 1975. © 1975, by the International Society of Nephrology.

pling, one patient had been treated with steroids for 255

256

Huttunen et a!

three days; none of the other patients received any

orosomucoid, transferrin and a-2-macroglobulin

immunosuppressive therapy during the course of this study.

(Behringwerke AG, Marburg-Lahn, Germany) were

3) Nonresponders. This group consisted of eight consecutive patients, ages 4 to 20 yr, with acquired

laboratory [15]. When the urinary concentration of a-2-macroglobulin was below the detection limit of the electroimmunodiffusion method, urine samples were concentrated 5- to 50-fold by negative pressure dialysis with 8/100 Visking tubing (Union Carbide, Chicago, Illinois, USA). The possible loss of a-2macroglobulin during the concentration procedure

steroid-resistant nephrotic syndrome. For all of them

the onset of the renal disease with nephrotic syndrome had been in childhood. All eight patients had been treated with steroids; four had also been treated with cytostatic agents. None of the patients, however, was receiving any immunosuppressive treatment at the time of sampling. All these patients had massive

proteinuria although five no longer had hypoproteinemia and were edema-free at the time of the study. Renal histology. Renal tissue from all 35 patients

was examined histologically. Biopsy or autopsy specimens were taken from all patients, except three nonresponders, within one year of the present study. In these three nonresponders biopsies had been performed two, three and six years prior to the study,

respectively. The CNF patients had typical

glomerular, tubular and interstitial changes of

used. Anti-IgG serum was a product of our own

was checked in five patients: the content of a-2-

macroglobulin in their urine samples was measured before and after concentration and no detectable loss was observed.

The selectivity of proteinuria was estimated graphically by the method of Blainey et a! [1]. Such selectivity was determined, for each patient, from the regression of log relative clearance of the proteins on

log respective molecular weight. The slope of the graph indicated the degree of selectivity: the steeper

the slope, the higher the selectivity. The degree of selectivity of proteinuria was also determined by the

morphological diagnosis of the patients with acquired nephrotic syndrome was made as recom-

method of Joachim et al [3]. The linearity of the regression slope was determined by the method of least squares, and the degree of selectivity was expressed as an angle derived from the horizontal

mended by Churg, Habib and White [10]. Urine and serum collection. The urine samples were all aliquots of 24-hr collections except in the case of two responders from whom overnight urine

tangent to the slope of the regression line. Sucrose gradient studies. Because the presence, in serum, of immunoreactive fragments of the reference molecules can cause faulty high clearance values, or

various degrees [8]. The severity of the glomerular changes was graded as mild, moderate or severe. The

was collected. Merthiolate was used as the preservative to a final concentration of 1: 10,000. The protein content of urine samples was determined by

the biuret method [11]. Urinary protein content

an abnormal aggregation of the proteins faulty low clearance values, the serum and urine samples of four

selected patients and the serum samples of two

27.8 g/liter in the nonresponders. Serum samples

healthy subjects were analyzed by sucrose gradient ultracentrifugation. A 0.2 ml sample of urine or serum was layered on linear 10 to 25% sucrose gradient made in 0.9%

were taken during or immediately after the collection of urine. Determination of the selectivity of proteinuria. For

NaCl buffered by 0.01 M phosphate, pH 7.2. Centrifugations were carried out in a centrifuge (Beckman Spinco L-50 with a rotor (SW 50) at

varied from 3.7 to 15.5 g/liter in CNF patients, from 1.5 to 53.0 g/liter in the responders, and from 1.8 to

each patient an estimation of the selectivity of proteinuria was made by determining the relative clearance of orosomucoid: mol wt, 44,100; transferrin: mol wt, 90,000; IgG: mol wt, 160,000; and a-2macroglobulin, mol wt, 820,000 [12]. The urine/serum ratio of each protein was established, and the relative clearance of the protein was expressed as a percent of

the clearance of transferrin. The concentration of urinary and serum proteins was measured either by single radial immunodiffu-

50,000 rpm. Centrifugation time was 15 hr when protein molecules the size of IgG or smaller were analyzed. When protein molecules the size of a-2macroglobulin were analyzed, centrifugation time was five hours. The urine samples were analyzed undiluted. In the 15-hr run, the sera of CNF patients were analyzed undiluted or diluted 1:2, that of the responder was analyzed undiluted and those of the

in the samples was too low to be detected by single

control subjects were analyzed diluted 1:4. In the fivehour run, the sera of the CNF patients were analyzed in dilutions from 1:4 to 1:10, that of the responder was analyzed undiluted and those of the control subjects were diluted 1:4. Yeast alcohol dehydrogenase, S20,

radial immunodiffusion. Commercial antisera to

6.72 to 7.71 [161, and horse liver alcohol

sion [13] or by electroimmunodiffusion [14]. The latter method was used when the concentration of proteins

257

•: ; :vttiie.rttxea

rcsa

;n

Proteinuria in congenital nephrotic syndrome

Fig. 1. Morphologic changes in the glomeruli of two CNF patients. A, micrograph showing a slight change consisting of only moderate

mesangial hypercellularity (hematoxylin-eosin, )< 250). B, micrograph showing extensive changes consisting of advanced periglomerular fibrosis and mesangial sclerosis (hematoxylin-eosin, X400).

258

Huttunen

et al

Table 1. Degree of the glomerular changes, relative protein clearances and selectivity angles of the CNF patients

Age months 1

1 1

2 3 4 5 6 9 11

Degree of glomerular changes

Moderate Moderate Mild Mild Mild Mild

Protein clearance, % transferrin

Orosomucoid 119 167 172 172

lgG 6.4 4.3 11.9

125 151

9.4 3.6 32.7 58.9

Mild Moderate Moderate Mild

743

10.0

181

21.8

134

10.2

914

Severe

358

a-2-M acroglobulin 0.05 0.13

Selectivity angle degrees

0.03

70.6 68.9 69.3 72.2

0.11 0.01

75.0

0.13

68.1

67.0

22.3 64.6

0.21 0.07 0.10 0.07 0.09 0.14

14 17

Severe

170

Moderate

279

16.7

0.21

20

Severe

196

64.6

0.15

Mean SD

72.7

69.6 70.0 72.4 68.5 68.5

68.6 2.2

70.1

dehydrogenase, S20, 5.11 [17] (Boeringer, Mannheim, Germany), were used as sedimentation standards. After centrifugations, fractions of five drops were collected. The amount of orosomucoid, transferrin, IgG and a-2-macroglobulin in each fraction was determined by the aforementioned quantitative methods [13, 14]. Statistical analysis. The differences in the angle of selectivity determined for the three groups of patients

were statistically analyzed with the t test.

nearly complete obliteration of glomerular capillaries

(Fig. 1, B). Generally, the glomerular basement

membrane in CNF patients appeared thin except in those glomeruli affected by advanced sclerosis. The

degree of the glomerular changes, the relative clearances of proteins and the selectivity angles determined for CNF patients are presented in Table 1. In

Tables 2 and 3 are presented the morphologic diagnoses, the relative clearances of proteins and the selectivity angles determined for the responders and the nonresponders.

In Fig. 2 are presented the regression slopes for each of the 35 patients. The selectivity slopes plotted for the CNF patients and the responders are, on the

Results

The glomerular histology in CNF varied from patient to patient. Some had only mild changes con-

whole, steep—indicating high selectivity; the corresponding slopes plotted for the nonresponders are

sisting of mesangial hypercellularity without any

fibrosis (Fig. 1, A); others had advanced focal

generally less steep—indicating lower selectivity. The selectivity angles calculated for the CNF patients and

periglomerular fibrosis and mesangial sclerosis with

for the responders were not significantly different.

Table 2. Renal histology, relative protein clearances and selectivity angles of the responders

Age

Renal histology

yr 3 3 3

4 4 4 4 5 7 8

9 9 11

Mean±sn

Minimal change Minimalehange Minimalchange Minimal change Minimal change Minimal change Minimal change Minimal change Minimal change Minimal change Minimal change Minimalchange Minimal change

Protein clearance, % Iransferrin Orosomucoid a-2-Macroglobulin IgG

Selectivity angle degrees

0.09

70.4 69.5

11.0

0.11 0.39

321

8.9

0 02

268 574 229 270 227 236

4.2

0.02 0.06 0.02 0.30 0.02

232 189 517

221

214 279

5.7 9.4

8.7 3.9 5.8 3.3 8,7 5.6 3.4 1.8

0.03 0.08 0.05 0.02

68.0 73.8 73.6 72.6 73.4 67.5 73.4 72.7 70.6 71.6 73.7

71.6±2.2

259

Proteinuria in congenital nephroiic syndrome Table 3. Renal histology, relative clearances of proteins and selectivity angles of the nonresponders

Protein clearance, % transferpm Age

yr 4

II 15 16 16

17 19

20 Mean + sr

Renal histology

.

Orosomucoid

IgG

a-2-Macroglobulin

182

41.2

110

38.6 41.5 27.3

0.32 0.10 6.62 0.23 2.60

Proliferative ON with crescents Pure mesangioproliferative UN MembranoproliferativeGN Focal proliferative UN MembranoproliferativeGN ExtramembranousGN MembranoproliferativeGN MembranoproliferativeGN

147 160 175

10.8

223 176

15.6 30.9

212

13.0

However, the selectivity angles determined for the nonresponders were significantly (P < 0.01) smaller than those calculated for either the CNF patients or for the responders. For the responders, the selectivity slopes (Fig. 2) are rather linear, indicating that there is no great dis-

crepancy between the relative clearances of any reference proteins. However, in some CNF patients and in some nonresponders, the relative clearance of orosomucoid and IgG points to a lower selectivity

than that evidenced by the clearance of a-2macroglobulin (Tables 1 and 3). The discrepancy

between the clearance of a-2-macroglobulin and the

clearance of the smaller proteins, appearing as a marked nonlinearity of the selectivity slope, tends to be greatest in those CNF patients in whom histologic

1000

CNF

1.80

6.90 2.00

47.7 66.9 56.4 59.5

48.7 58.7 59.1 + 8.0

glomerular lesions. Analyzed by sucrose gradient ultracentrifugation were the serum and urine samples of the three CNF patients whose regression slopes were the most nonlinear, the serum and urine samples of a responder who had a rather linear slope, as well as the serum of

two healthy control subjects. The results of the analyses of the samples of two CNF patients, one responder and one control subject are presented in

Fig. 3. Orosomucoid, transferrin, IgG and a-2macroglobulin sedimented at about the same rate in the samples of the CNF patients as in the samples of the responder and the healthy subject. No extra peak or detectable tailing of the peaks was observed for any of the proteins studied. Similar results were ob-

Responders

0

C)

patients.

66.1 68.6

examination had revealed moderate or severe

C)

Fig. 2. Selectivity slopes of individual

Selectivity angle degree

Mol wt

Nonresponders

260

Huttunen et a! Control

Responder mg/IOU ml

ng/lOO ml 100

IOU

F

a a0

IOU

A

./

(5

0 (5 0I-

:( '

50

5

(5

S

10

2

s

I5

Fraction number

2

To

/IOO ml CNF 50 40

G

.

mg/IOU ml

Fig. 3. Sucrose gradient ultracentrifugation analyses of urine and serum samples of two CNF patients and one responder as well as

30

A

,g/lOO ml

I

5 40

1

25

/

0

5

10

.—?,— S

Bottom Fraction number

20

fl I

is

I

I

V

't 1\ I\ ,,,

L—-——

5

II.)

Bottom Fraction number

tamed in the ultracentrifugation analysis of the serum

and urine samples of a third CNF patient and the serum of another control subject. Discussion

The highly selective proteinuria in CNF seems to be an exception to the observation that steroid sen-

sitivity, good prognosis and minimal change

histology are commonly associated with highly selective proteinuria [4, 6, 18]. Morphologically this exception should not be surprising, however, because

the glomerular capillary basement membrane is generally thin in patients with CNF just as it is those

nephrotic patients who have minimal change histology. In those CNF patients who had the most advanced glomerular lesions, however, selectivity was low in the molecular weight range from orosomucoid to IgG but notably high for a-2-macroglobulin. Results that

gave a similar nonlinear selectivity slope were also observed in some patients with glomerulonephritic histology in this series. This difference between the clearance of a-2-macroglobulin and the clearance of smaller proteins has also been observed in nephrotic

syndrome associated with malaria [19], in focal glomerular sclerosis [20, 21] and in renal amyloidosis [211.

IS

3ottom Fraction number

To1

a 8 a

/

A Bottom CNF

aU)

I

mg/IOU ml

.3

Top

oo v

of the serum sample of a healthy control subject. The scale of fractions of the 15-hr

runs is marked with L, that of the 5-hr runs with S. The fractions in which the activity of yeast alcohol dehydrogenase (Y)

and horse liver dehydrogenase (H) was highest are marked with vertical lines. The

content of orosomucoid (0), transferrin (T), IgG (0) and a-2-macroglobulin (A) in serum is indicated by circles and solid lines, those in urine by dotted lines.

The reason for the nonlinearity of the regression slope might lie 1) in the presence of preferentially cleared fragments of some reference proteins in the serum of the patients, 2) in the fact that tubular reabsorption of proteins was nevertheless selective or 3)in

the nature of the damage to the glomerular filter. Ultracentrifugation of the serum and urine samples of the selected CNF patients revealed no detectable amounts of fragmented molecules of either IgG or the other reference proteins. Hence, it seems highly unlikely that the nonlinearity of the regression slope in the CNF patients reflects the faulty clearance of any of the reference proteins. Renal tubular uptake of low-molecular weight proteins has been reported to be somewhat greater than that of albumin [22] but no data are available concerning the possible difference in the tubular reab-

sorption rate between intermediate- and highmolecular weight proteins. The filtered proteins are reabsorped from the renal tubuli by endocytosis [23] and this mechanism, per Se, favors the concept of nonselective absorption. Even in the absence of factual data, it is unlikely that the great observed discrepancy between the clearance of a-2-macroglobulin and that of the smaller proteins can be explained by selective protein reabsorption from renal tubuli. Findings of recent studies suggest that the normal glomerular barrier for proteins is composed of two

261

Proteinuria in congenital nephrotic syndrome

consecutive filters: the glomerular basement membrane acting as a coarse prefilter and the slit pore system between the epithelial foot processes constituting the final barrier [24]. In kidney disease involving serum protein leakage into the urine, the glomerular filter is clearly damaged, but the nature and the site of the damage are still obscure. Some studies on the selectivity of proteinuria have also been concerned with trying to characterize the nature of the underlying glomerular damage responsible for

6. TSAO YC, CHAN WC, GIBSON JB: Persistent proteinuria in children. Arch Dis Child 44:443—453, 1969

7. LINES DR: Selectivity of proteinuria in childhood nephrotic syndrome. Arch Dis Child 44:461—464,

drome. Nephron 9.

glomerular filter can cause a nonlinear regression of log protein clearance on log respective molecular size. Hence, to explain the nonlinearity of the regression slope of some patients with CNF, it can be supposed that in these patients the damage to or the structural abnormality in the glomerular filter, at the level either

Acknowledgments

This investigation was supported by grants from the Sigrid Jusélius Foundation and the Foundation for Pediatric Research, Finland. Prof. Nub Hallman,

Head of the First Department of Pediatrics in Children's Hospital, Helsinki, and Dr. Erkki Hölttä,

11:101—110, 1973

Serum protein turnover in the congential

J:

estimation of total protein in urine. J Lab C/in Med 39:618—623, 1952

12. SCHULTZE HE, HEREMANS JF: Mo/ecu/ar Biology of Human

Proteins. Amsterdam, Elsevier Publishing Company, 1966, vol.

l,pp. 176—181

13. IMMONEN

P: Levels of the serum immunoglobulins yA, yG

and 7M in the malabsorption syndrome in children. Ann PaediatrFenn 13:115—153, 1967 14. LOPEZ M, TSU T, HYSLOP NE JR: Studies of electroim-

munodiffusion:

of the glomerular basement membrane or the slit pore system, is not homogeneous but affects different glomeruli in different manners causing non-Gaussian distribution of "pore sizes" in their glomerular filter.

PERHEENTUPA

nephrotic syndrome. Ann Paediatr Fenn 12:Fasc 4, 1966 10. CHURG J, HABIB R, WHITE RHR: Pathology of the nephrotic syndrome in children. Lancet 1:1299—1302, 1970 II. FOSTER PW, RICK JJ, WOLFSON WQ: Studies in serum proteins: VI. The extension of the standard biuret method to the

the proteinuria [25, 26]. Hardwicke et al [27] have

postulated that heterogeneous damage to the

1969

8. HALLMAN N, N0RIO R, RAPOIA J: Congenital nephrotic syn-

Immunochemical quantitation of proteins in

dilute solutions. Immunochemistry 6:513—526, 1969 15.

SAVILAHTI E: Immunoglobulin-containing cells

in the in-

testinal mucosa and immunoglobulins in the intestinal juice in children. C/in Exp Immunol 11:415-425, 1972 16. SUND H, THEORELL H: Alcohol dehydrogenases, 2, in The Enzymes (2nd ed); edited by BOYER PD, LARDY H, MYRBCK K, New York, Academic Press Inc. 1963, vol.7, p. 25 17. EHRENBERG A, KEITH D: The molecular weight of horse liver alcohol dehydrogenase. Ada Chem Scand 12:465—469, 1958

18. CAMERON JS: The clinical significance of glomerular permeability studies. Proc R Soc Med 59:512—515, 1966 19. S00THILL JF, HENDRICKSE RG: Some immunological studies

of the nephrotic syndrome of Nigerian children. Lancet 2:629—632, 1967

Department of Medical Chemistry, University of Helsinki, engaged with us in valuable discussions.

20. MILLS RJ, RAINE DN, WHITE RHR: Significance of variation in selectivity of proteinuria. Lancet 1:373, 1973

Miss Marjatta Viikari provided technical assistance.

21. BRANCACCIO

D,

RIVOLTA E, GRAZIANI

G,

PIZZOLATO M:

Albumin and a-2-macroglobulin clearance: A new approach-

ing method for studying selectivity of proteinuria in unReprint requests to Dr. Niilo-Pekka Hut tunen, Children's 1-lospital, University of Helsinki, Stenbâckinkatu 11, SF-00290 I-I elsinki 29, Finland.

concentrated urine. Nephron 12:150—156, 1974 22. CORTNEY MA, SAWIN LL, WEISS DD: Renal tubular protein absorption in the rat. J C/in Invest 49:1-4, 1970 23. BOURDEAU JE, CARONE FA: Protein handling by the renal tubule. Nephron 13:22—34, 1974

References

24. SCFINEEBERGER EE: Glomerular permeability to protein

I. BLAINEY JD, BREWER DB, HARDWICKE J, SOOTI-IILL iF: The nephrotic syndrome. Q J Med 29:235—256, 1960 2. HARDWICKE J, SQUIRE JR: The relationship between plasma albumin concentration and protein excretion in patients with

molecules: Its possible structural basis. Nephron 13:7—21, 1974 25. HARDWICKE J, SOOTHILL JF: Glomerular damage in terms of "pore size", in Ciba Foundation Symposium on Renal Biopsy, edited by WOLSTENHOLME GEW, CAMERON MP, London, Churchill, 1961, p. 32

proteinuria. C/in Sci 14:509—530, 1955

26. VERE DW, WALDUCK A: The chemical estimation of renal

3. JOACHIM GR, CAMERON iS, SCHWARTZ M, BECKER EL: Selec-

selective permeability to proteins during steroid-induced

tivity of protein excretion in patients with the nephrotic syn-

remission of the nephrotic syndrome. C/in Sci 30:315—324,

drome. J C/in Invest 43:2332—2346, 1964

4. Hnzci WH, AURICCHIO 5, BENNINGER JL: Clearance spezificher Plasmaproteinfraktionen bei Nierenleiden. K/in Wochenschr43:l 154—1166, 1965 5.

CAMERON JS, WHITE RHR: Selectivity of proteinuria in children with nephrotic syndrome. Lancet 1:463—465, 1965

1966 27. HARDWICKE J, CAMERON JS, HARRISON iF, HULME B, SOOTHILL iF: Proteinuria studied by clearances of individual macromolecules, in Proteins in Norma/ and Pathological Urine, edited by MANUEL Y, REVILLARD JP, BETUEL H, Basel, S.

Karger AG, 1970,

p.

111