Analysis of anionic charge of the glomerular basement membrane in aminonucleoside nephrosis

Kidney International, Vol. 35 (1989), pp. 1405—1408

TECHNICAL NOTE

Analysis of anionic charge of the glomerular basement membrane in aminonucleoside nephrosis SHUZO KOBAYASHI, MITSUMASA NAGASE, NisHlo HONDA, KYOKO ADACHI, N0RI0 ICHINOSE, and AKIRA HIsHIDA First Department of Medicine and Department of Chemistry, Hamamatsu University, School of Medicine, Hamamatsu and First Department of Medicine, Teikyo University School of Medicine, Tokyo, Japan

Although there is growing evidence that suggests the importance of the electrostatically charged barrier in the transglomerular permeation of macromolecules, a controversy still exists as to whether or not the anionic charge barrier is deteriorated in nephrotic states [1—5]. The disparity of opinions seems to arise from the paucity of established methods for quantitative evaluation of the anionic charge. In the present study, a method to

copper content in GBM. according to the technique described by Gorsuch [9]. The tissues were heated in 61% nitric acid at 200 to 300°C for 30 minutes. Then, 60% perchloric acid was added. The heating was continued until the solution completely evaporated. When the samples were cooled to room temperature, I N nitric acid (1 ml) and redistilled water (3 ml) were

added and heated at 200 to 300°C for an additional three

morphologically demonstrate the anionic sites in glomerular minutes. Finally, the solution volume was accurately adjusted basement membrane (GBM) was developed, and anionic charge to 10 ml and its copper content was measured with an atomic was quantified. With this method, the change of the charge absorption spectrophotometer (Hitachi-S 18, Tokyo, Japan). barrier was estimated in puromycin aminonucleoside (PAN) The results were expressed as pg/mg GMB protein, which nephrosis in rats. reflects quantitatively the anionic charge of GBM.

To examine whether or not the disruptive procedure of

Methods

In female Wistar rats, weighing approximately 150 g, nephrosis was induced by a single i.v. injection of PAN dissolved in I ml of 0.9% saline at a dose of 10 mg/100 g body weight. In the

pair-fed control rats, the vehicle alone was injected. The

ultrasonication affects the copper content of GBM by removing laminae rarae, copper was measured in GBM ultrasonicated for various lengths of time: 4, 8, 12, 18 and 24 minutes. At the same time, the copper content in the supernatant of the centrifuged

24-hour urinary albumin excretion was measured by the single suspension was likewise examined. The copper content was radial immunodiffusion method [6]. The animals were sacrificed also measured in GBM obtained from the kidney perfused with at days 2, 9, and 42 after the injection to estimate the anionic cuprolinic-blue-free solution. Next, the pH dependency of the charge of GBM. Upon anesthetization with ethyl ether, the left binding of cuprolinic blue to GBM was examined within the kidney was exposed through a midabdominal incision and range of pH 0.7 to 8.2; constant ionic strength was maintained perfused with 25 m of sodium acetate buffer (pH 5.6) for two by using buffer containing 0.3 M MgCI2. GBM was incubated at minutes through a catheter inserted into the abdominal aorta, 37°C for one hour. Also, the binding capacity was examined at and then for 10 minutes with the same buffer containing 0.05% pH 5.6 in the buffer not containing MgCI2. These experiments cuprolinic blue (BDH Chemical, London, UK) and 0.3 M MgCl2 were performed in duplicates or triplicates. For the morphological analysis with an electron microscope, (pH 5.6) at a constant rate of 0.68 mI/mm. Perfused kidneys sodium acetate buffer (25 mM, pH 5.6) was first infused into the were removed and subjected to the harvesting of glomeruli by sequential sieving through graded size mesh screens. Then the left kidney briefly, and then 2% glutaraldehyde in the above glomeruli underwent a four-minute ultrasonication, and the measured buffer containing 0.05% cuprolinic blue and 0.3 M GBM freed from cellular components of glomeruli was obtained MgCl2 was infused. Cortical sections of 1 to 2 mm3 were incised as described by Spiro [7]. The isolated GBM was suspended in and then fixed for two hours in the above fixatives, followed by 12 ml of the above-mentioned buffer. A quarter of the suspen- immersion in 0.5% sodium tungstate for an additional 30 minsion was used for the measurement of protein by the method of utes. After washing overnight in the sodium acetate buffer, the Lowry's eta! [81. The rest was used for the measurement of the tissues were postosmicated, stained en bloc with uranyl acetate, dehydrated in graded ethanols, and embedded in epoxy resin. Ultrathin sections were double-stained briefly as needed for electron microscopic observation (JEOL-l00 CX, Tokyo, Received for publication September 16, 1987 and in revised form December 27, 1988 Japan). Accepted for publication January 12, 1989

© 1989 by the International Society of Nephrology

Statistical analysis was performed by Student's t-test and values were expressed as the mean SE. 1405

Ko/n,vashi Ct a!: Analysis of anionic

4

4%

v.'i

charge / GBM

.1

:4.

-

lA

:

S

•4#t,

.4

:.

4;

1406

Fig. 1. Electron microscopic photographs of isolated GBM stained using cuprolinic blue. Note dye-positive fibrils on both sides of the lamina densa. The sonication time was 4 minutes. x 300,000 C 5,

C

a 10

10

10

C

.

5,

0 5

111 ' 4

8

12

•..

0) E

5

a)

16

20

24

0

0)

Sonication time, minutes

0

2. The influence of the sonication ti,ne on f/u' binding of cuprolinic blue to GBM. Closed and open circles show the copper content in GBM Fig.

9

8

7

6

and the supernatant of the GBM suspension, respectively. Note that the

copper contents of both GBM and supernatant do not significantly change until the sonication time exceeds 12 minutes. ** <0.01, P < 0.05, as compared with the value at 4 minutes.

Results

Rates of urinary albumin excretion were 1.3 0.24 (N = 8), 156 26.5 (N = 8), and 3.8 0.84 (N = 7) mg/day, 2, 9, and 42 days after PAN injection, respectively. In control rats, urinary excretion rate of albumin was less than 1 .2 mg/day throughout the observation period.

5

4

3

2

0

1

pH

3. The influence of pH on the binding of cuprolinic blue under a constant ionic strength. Note that the copper content (binding of the dye to GBM) begins to decrease at pH between 3.0 and 4.0, and at pH 2.0 to 50% of the value at pH 5.6. Fig.

Table 1. Copper (pg/mg GBM protein) bound to GBM in PAN- and saline-treated rats Day after injection

9

2 Saline

PAN

7.4

0.28

(N= 10)

7.7

0.39

Electron microscopic examination revealed the complete removal of cellular components from GBM by a four minute Saline vs. PAN

(N=8)

ultrasonication and the presence of dye-positive fibrils on both sides of the lamina densa (Fig. 1). The standard curve obtained for the determination of copper

SEM.

All data are means

NS

7.9

42 0.43

(N= 13) 3.7

0.33

(N=8)

P < 0.01

7.6

0.59

(N= Ii)

7.3

0.67

(N=7) NS

content using an atomic absorption spectrophotometer was linear within the range of 0.01 to 0.5 ppm. A kidney perfusion 12 minutes did not affect the copper content of GBM (Fig. 2). with dye-free buffer yielded 2.0 0.42 zg (N = 5) of copper per When the sonication time exceeded 18 minutes, however, marked decreases in the copper content of GBM were noted. mg GBM protein in the control rats, while that with cationic dye 13). Sonication for less than Conversely, the copper content in the supernatant of the was as high as 7.9 0.43 sg (N

1407

Kobavashi et at: Analysis of anionic charge of GBM

'-i'- J. p

4 i*&

—":1

-

-

e

.

.1

4. Perfusion staining of GBM with cuprolinic blue, a. Glomerular capillary Fig.

-—-a,-

wall in normal rats. Note the dye-positive fibrils running perpendicular to the laminae densa. Also note that the fibrils are absent beneath the slit membrane. x 165,000 b. Glomerular capillary wall in PAN-induced nephrotic rat (9 days after PAN injection). Note the fibrils are reduced in number and distorted in arrangement. x 165,000

centrifuged glomerular suspension increased with the prolongation of sonication time. At pH between 3.0 and 4.0, the binding rate of cuprolinic blue began to decrease. At pH 2.0, the binding rate was reduced to the 50% of the value at pH 5.6 (Fig. 3). In contrast, the binding of the dye was increased by approximately 50% (13.3 1.82. N = 3) when examined in a buffer solution free of MgCI2. The degree of copper binding to GBM was markedly reduced at day 9 following PAN injection (Table I) and returned to the

control level at day 42, concomitantly with the recovery of albuminuria.

4A). In nephrotic rats (9 days after PAN injection), ultrastructural features obtained by conventional methods were compatible with those described previously, an extensive effacement of the foot processes and vacuolation of the epithelial cells [1—2].

In those animals, the fibrils were reduced in number and distorted in arrangement (Fig. 4B). These changes appeared on

day 2, became most striking on day 9, and were scarcely present at day 42. Discussion

Despite the growing number of attempts to evaluate anionic sites in GBM. there has not been an established method which accurately quantifies an alteration of the anionic charge of the

In control rats, cuprolinic blue staining with the critical electrolyte concentration method [10—Il] demonstrated the anionic sites in GBM as extended fibrils in all layers. These GBM. fibrils were demonstrated most distinctly in the laminae rarae Nevertheless, the change of anionic charge in glomeruli of (Fig. 4A). The fibril was 30 to 40 nm in length and perpendicular PAN nephrosis has been widely examined using various catito the long axis of the GBM. It extended from the anchored onic probes. Some investigators have showed a decrease of portion of the epithelial cells to the lamina densa, with short anionic charge [1, 2, 5], while others [3, 4] have found it arms. The fibrils were not found beneath the epithelial slits (Fig. unaltered. The discrepancy between these data might be attrib-

1408

Kobayashi el a!: Analysis of anionic charge of GBM

uted, in part, to the differences in the probes used. Hitherto,

methrine [18], which they subsequently showed to bind to

carboxyl groups rather than to sulfate groups of proteoglycans in GBM [19]. Therefore, as mentioned earlier, our study still leaves a possibility that carboxyl groups may be also reduced in GBM of PAN nephrosis, thereby contributing, at least in part, fact that smaller or less distinct stains may reflect an even greater to the defects of charge as well as size permselectivity.

quantitative evaluations have been attempted by enumerating the dye-positive sites. These attempts, however, may not truly reflect the changes in the anionic charge, because all the stained sites are counted equally with normally stained sites, despite the

decrease of anionic charge. Farquhar and Kanwar [12] have stated Reprint requests to Dr. Mitsumasa Nagase, First Department of that the binding of cationic ferritin (P1 7.3 to 7.6) was replaced at Medicine, Teilcyo University School of Medicine, il-I Kaga, 2 Chome, lower salt concentrations (0.2 against 0.3 M KCI) in PAN nephrotic Itabashi-Ku, Tokyo 173, Japan. rats excreting maximum urinary protein than in normal control rats. This finding which indicates a reduction, but not a comReferences plete loss of anionic charge of GBM of PAN nephrosis rats, also 1. CAUFIELD JP, FARQUHAR MG: Loss of anionic sites from the

suggests that a mere enumeration of cationic probe positive particles may not truly reflect an alteration of anionic sites. Cuprolinic blue is a cationic probe which was first introduced by Scott [10] to stain the anionic sites in tissue. By applying the critical electrolyte concentration concept, he noticed that the dye, when used in a buffer containing 0.3 M MgCl2, combined preferentially with the sulfate groups rather than with carboxyl groups [11]. He also showed that in this procedure, the dye becomes monovalent; thus, more dyes could bind to the unit substrate [11], thereby resulting in more effective staining. Our present experiment showed that the binding of the dye to GBM decreased markedly at pH of less than 2.0 (Fig. 3). In view of the pK's of the sulfate and carboxyl groups, which are 2.0 and 5.0, respectively, the pH-dependent dissociation curve of duprolinic blue (Fig. 3) also suggests that the dye preferentially reacts with the sulfate groups in GBM. However, the possibility still remains that the dye may, at least in part, react with the carboxyl group, and therefore the decreased binding of the dye

does not necessarily imply an unaltered state of carboxyl groups in GBM of PAN nephrosis [121. Our quantitative study showed a decrease of anionic charge,

particularly the decrease of that endowed by sulfate groups. Fishman and Karnovsky [13] have recently suggested a de-

creased incorporation of "SO4 in the cultured glomerular epithelial cells when exposed to PAN. On the other hand, in PAN nephrosis, increased incorporation of 35S04 into heparan sulfate proteoglycans of glomeruli with no change of "SO4heparan sulfate content in GBM has been reported [14]. Lelongt, Makino and Kanwar [15] have also demonstrated increased de novo synthesis of heparan sulfate proteoglycans into

glomeruli and GBM. These different results concerning the "SO4 incorporation may be explained on the basis of metabolic disOrder at different stages of PAN nephrosis, or of different doses or manners of PAN treatment. Taking into account that 80% of injected PAN is excreted within 24 hours, the majority

of it being within the first eight hours [13], it may also be conceivable that the increased de novo synthesis of sulfated proteoglycans is a compensatory response to the decrease of sulfated proteoglycans induced by increased degradation and/or decreased synthesis occurring at an early stage [14].

It should be also remembered that not only heparan sulfate [16] but also carboxyl groups in GBM possibly play a potential role for charge as well as size barrier [17—19]. In this regard, Bray and Robinson [171 have shown, in an in vitro study, the importance of carboxyl groups in terms of glomerular permselectivity of macromolecules. Indeed Hunsicker, Shearer and Shaffer have presented evidence that charge as well as size selectivity defect is induced by infusion of polycation, hexadi-

glomerular basement membrane in aminonucleoside nephrosis. Lab Invest 39:505—512, 1978

2. CAUFIELD JP: Alterations in distribution of alcian blue-staining

fibrillar anionic sites in the glomerular basement membrane in aminonucleoside nephrosis. Lab Invest 40:503—511, 1979 3. KANWAR YS, JAKUBOWSKI ML: Unaltered anionic sites of glomer-

ular basement membrane in aminonucleoside nephrosis. Kidney mt 25:613—618,

1984

4. KANwAR YS, ROSENZWEIG U, KERJASCHKI DI: Glycosaminogly-

cans of the glomerular basement membrane in normal and nephrotic states. Renal Physiol 4:121—130, 1981 5. MAHAN JD, SIssoN-Ross S. VERNIER RL: Glomerular basement membrane anionic charge site changes early in aminonucleoside nephrosis. Am J Pathol 125:393—401, 1986 6. MANcINI G, CARBONARA AO, HEREMONS JF: Immunochemical quantitation of antigens by single radial immunodiffusion. Immunochemistry 2:235—254, 1965 7. SPIRO RG: Studies on the renal glomerular basement membrane. Preparation and chemical composition. JBio! C/tern 242:1915—1922, 1967 8. LOWRY OH, ROSENBROUGH NJ, FARR Al, RANDALL RJ: Protein measurement with the Folin phenol reagent. J Biol C/tern 193:265— 275, 1951

9. GORSUCH TT: Wet oxidation, in The Destruction of Organic Matter, edited by GORSUCH TT, Oxford, Pergamon Press, 1970, pp. 19—27

10. SCOTT JE: Collagen-proteoglycan interactions. Localization of proteoglycans in tendon by electron microscopy. Biochern J 187: 887—891, 1980

II. SCOTT JE: Affinity, competition and specific interactions in the biochemistry and histochemistry of polyelectrolytes. Biochem Soc Trans 1:787—806, 1973 12. FARQUHAR MG, KANWAR YS: Functional organization of the

glomerulus: Presence of glycosaminoglycans (proteoglycans) in the glomerular basement membrane, in Immune Mechanisms in Rena! Disease, edited by CUMMINGS NB, MICHAEL AF, WILsoN CB, New York, London. Plenum Medical Book Co. 1986, pp. 1—35 13. FISHMAN JA, KARNOVSKY Mi: Effects of the aminonucleoside of puromycin on glomerular epithelial cells in vitro. Am J Pathol 118: 398—407,

1985

14. KLEIN Di, DEHNEL P1, OEGEMA TR, BROWN DM: Alterations in proteoglycan metabolism in the nephrotic syndrome induced by the aminonucleoside of puromycin. Lab Invest 50:543—551, 1984 15. LELONGT B, MAKINO H, KANWAR YS: Status of glomerular proteoglycans in aminonucleoside nephrosis. Kidney mt 31:1299— 1310, 1987 16. KANwAR YS, FARQUHAR MG: Presence of heparan sulfate in the glomerular basement membrane. Proc NatlAcad Sci USA 76:1303— 1307, 1979

17. BRAY 1, ROBINsON GB: Influence of charge on filtration across renal basement membrane films in vitro. Kidney Int 25:527—533, 1984

18. HUNSICKER LO, SHEARER TP, SHAFFER

Si: Acute reversible

proteinuria induced by infusion of the polycation hexadimethrin. Kidney ml 19.

20:7—13, 1981

BERTOLATUS JA, HUNSICKER KG:

Polycation binding to glomerular basement membrane. Effect of biochemical modification. Lab Invest 56:170—179, 1987