Biochemical
Abnormalities of the Human Diabetic Basement Membrane Jean Canivet,
Anne Cruz, and Hyacinthe
Since the biochemical composition of the diabetic glomerular basement membrane is still a controversial area, a study was carried out using kidneys from seven diabetic and seven nondiabetic subjects. In diabetic membranes, the glycine. hydroxylysine, glucose, galactose, and hydroxylysine-linked disaccharide unit content was increased together with a decrease in the half-cystine and sialic acid content. These findings support the view of a biochemical alteration in the human diabetic glomerular basement membrane.
of the capillary basement T HICKENING membrane is the characteristic morphological feature of diabetic microangiopathy and has been extensively studied in most tissues.’ The biochemical composition has been investigated in glomerular basement membrane from human diabetic kidneys.2m5Results have been conflicting. Beisswenger and Spiro* have reported that analysis of eight individual diabetics showed a distinct chemical alteration characterized by a significant decrease in lysine with an equivalent increase in hydroxylysine and hydroxylysinelinked disaccharide units (HLDU). However, these abnormalities have not been observed in other studies, whereas a significant decrease in the half-cystine content has been reported.3-5 Therefore, the aim of this study was a reevaluation of the half-cystine, lysine, hydroxylysine and HLDU content in the human diabetic glomerular basement membrane isolated from individual patients and with no pooling of the material obtained from each subject.
From the Department of Endocrinology and Metabolism, Hopital Saint Louis, Universitk Paris, France. Receivedfor publication October 24, 1978. Supported in part by the Delegation Getterale a la Recherche Scientifique et Technique (D.G.R.S.T.), Grant 74.7.0588. Address reprint requests to Jean Canivet, Department of Endocrinology and Metabolism, Hopital Saint Louis, 2 Place du Docteur Alfred Fournier, 75475 Paris Cedex IO, France. o 1979 by Grune & Stratton, Inc. 0026-0495/79/2812~04%01.00/0
1206
Glomerular
Moreau-Lalande
MATERIALS
AND
METHODS
Materials Human kidneys were obtained at autopsy within 6 hr following death from 7 diabetic and 7 nondiabetic subjects, and were immediately stored at -3OOC. Diabetic patients were 5 men and 2 women, ages ranging from 39 to 77 yr (mean 50.6 + 5.0) previously treated with diet and insulin or oral hypoglycemic drugs; selection was based solely on the duration of diabetes over 10 yr (range l&3 1 yr, mean 19.6 f 3.5); in all of them, morphological examination of the kidneys showed various degrees of diabetic glomerulosclerosis. Nondiabetic subjects were 6 men and 1 woman, ages ranging from 27 to 78 yr (mean 58.6 * 5.0); these were patients previously admitted to hospital who had died from diseases unrelated to any kidney disorder and with no renal impairment on pathologic examination; none of them had a history of diabetes and fasting blood glucose levels had been within the normal range.
Methods Preparation of the glomerular basement membrane. This was undertaken on both kidneys of each subject and was run until the final lyophilization without pooling the material obtained from several subjects. It was carried out by the method reported by Spiro,6 using a Branson Sonifier model B12 for the disruption of the glomeruli. Some steps were slightly modified as follows. Complete fragmentation of the glomeruli conducted under the phase microscope control required a total sonic disruption time of 6 min (instead of 4-5 min) with 6 bursts of 1 min and a cooling period allowed between bursts until temperature came down to 4°C. The sonicated material was filtrated again through a 200-mesh screen for the elimination of any resistant hyalinized glomeruli and tubular fragments that could have been present, and was thereafter centrifuged. For each washing, following centrifugation, suspension of the sticking pellet of the membrane was performed by a 5-set sonication burst so that the pellet could be more easily suspended in distilled water. Evaluation of the purity of the preparations. This was performed on the 14 preparations. Deoxyribonucleic acid (DNA) analyses were carried out by the method of Burton’ modified by Spiro.6 Uranic acid content was estimated by the method of Dische.’ Total lipid content was determined according to Spiro.6 For electron microscopy, basement membrane samples were fixed in cold glutaraldehyde and embedded in Epon. Analysis of amino acids. Each sample from the 7 normal and 7 diabetic subjects was hydrolized in 6N hydrochloric acid at 105°C for 24 hr at a concentration of 1 mg membrane/ml HCl in sealed tubes; aliquots were evaporated and analysis was performed on each sample in a Jeolco
Metabolism, Vol. 26, No. 12 Khxember), 1979
DIABETIC GLOMERULAR
BASEMENT
MEMBRANE
1207
Calculations. All determinations were made cate. Statistical analyses were made with Student’s
Analyzer, model JLC 5 AH. For the determination of half-cystine, samples of the membrane were first oxidized with pcrformic acid by the method of Schram et aL9 and then hydrolysis was carried out. For the determination of methionine, hydrolysis was performed in an atmosphere of nitrogen in sealed tubes. Analysis of sugar components. This was carried out on separate samples from 4 normal and 4 diabetic subjects among the selected patients. For the determination of glucose, galactose, mannose, and glucosamine, each sample (1 mg) of the membrane was submitted to methanolysis (0.5 ml methanolic- 1.5M hydrochloric acid/ 1 mg membrane dry weight) for 24 hr at 85OC in sealed glass tubes and was thereafter treated with a trimethylsilylating agent; the determination was achieved by the gas chromatographic system of Bhatti et al.;” sugars were evaluated as trimethylsilyl ethers. For the fucose, the membrane was hydrolized in 0.1N sulfuric acid (1 ml/2 mg of membrane dry weight) at 100°C for 8 hr in sealed glass tubes and the determination was carried out using the Dische-Shettles cysteine reaction.” Sialic acid was evaluated according to the method of Svennerholm.” Analysis of HLDU and lysine. This was carried out on separate samples from the 7 normal and 7 diabetic subjects. Each sample was submitted to alkaline hydrolysis (0.1 ml 2N NaOH/l mg membrane dry weight) for 24 hr at 105°C in sealed glass tubes; aliquots of hydrolysate were deposited on 3 mm Whatmann paper and were submitted to high-voltage electrophoresis separation (4500 V for 2 hr in pyridineacetic acid-water 1:10:89, pH 3.6 buffer) according to Moczar and Moczar” in a Gilson electrophoretor, model D;
RESULTS
Isolated glomeruli as seen under phase microscope appeared to be free from tubular elements (less than 0.5%). Under the electron microscope, the membranes appeared as amorphous structures, with no cellular contaminants. Average dry weight of the lyophylized membrane was 44 + 2.3 mg/ 100 g kidney cortex from normals, and 71 .O + 2.6 mg/ 100 g kidney cortex from diabetics. DNA content was less than 0.1% and lipid content was less than 1% of the membrane (dry weight). Results of amino acid analyses from each subject are shown in Table 1 and the average values are summarized in Table 2. In diabetics, as compared with nondiabetic subjects, an increased content in glycine and hydroxylysine and a decreased content in half-cystine were observed (Table 2). The mean values for neutral sugar components are shown in Table 3. An increased content in galactose and glucose was observed in diabetics, with a decreased content in sialic acid. Analyses showed a significantly increased content in HLDU in diabetics as compared with nondiabetics (p < 0.01, Table 4); with the elec-
after drying the paper, revelation with ninhydrine-cadmium and elution with methanol, HLDU and lysine were measured using a Jobin-Yvon spectrophotometer model Monospac at 500 nm. Table 1. Amino Acid Composition
of Glomerular
Basement
(Residues/1000
Membrane
in Seven Diabetics and Seven Nondiabetics
Amino Acid Residues)
NondiabeticSubjects Amano Acld
Lysine
1
2
24.4
21.6
3
19.2
4
5
24.0
24.2
in duplit test.
DiabeticSubjects 6
19.3
7
19.9
6
20.3
9
15.1
10
11
12
13
14
17.3
25.4
20.1
19.6
19.5
H&dine
16.2
10.5
14.7
15.5
15.3
16.2
14.9
17.3
12.3
15.3
17.2
13.2
16.4
14.0
Arginine
48.7
34.3
33.6
50.4
42.0
43.7
41.2
49.4
29.2
37.9
45.2
40.1
42.6
44.5
Hydrowylysine
21.5
19.6
21.8
23.9
22.3
23.3
27.3
29.5
27.7
27.6
30.6
27.9
30.7
27.0
Hydroxyproline
64.3
94.4
96.1
67.9
93.0
69.2
96.2
97.1
122.9
119.9
84.2
111.6
106.0
86.3
Aspartic acid
65.3
70.3
73.1
67.9
73.7
67.6
63.3
67.6
66.4
63.6
67.0
63.6
63.0
63.5
Threonine
36.2
42.1
41.8
34.5
37.8
35.0
32.7
37.3
36.6
35.1
36.3
33.6
34.1
37.6
Serme
45.8
52.8
52.9
45.8
46.8
47.1
45.1
47.0
49.3
45.5
46.4
44.4
43.8
47.6
Glutamic acid
96.6
96.2
97.0
95.1
96.3
97.0
94.4
99.6
95.3
92.2
97.7
94.6
96.0
101.1
Proline
78.4
67.5
66.2
66.6
76.6
54.6
76.6
71.5
61.5
72.4
60.0
74 2
75.4
65.7
Glycme
205.8
175.4
176.8
207.8
196.1
201.0
202.0
208.2
204.9
214.5
199.1
209.4
220.6 53.6
216.2
Alanine
58.5
58.9
57.9
56.3
56.4
60.2
54.2
59.2
50.1
45.6
60.0
57 1
51.9
Half-cystine
24.7
26.2
25.7
19.8
24.2
20.6
24.0
21.1
16.8
16.4
20.8
16.0
20.4
16.7
Valine
41.0
44.4
42.2
32.6
39.2
40.3
55.0
42.2
36.2
36.2
44.4
36 0
38.8
37.4
Methionine
10.6
15.7
14.4
22.5
15.8
12.1
11.9
10.8
13.5
11.6
10.8
10.7
10.4
13.4
lsoleucine
30.9
33.6
32.5
31.3
31.7
33.3
29.3
31.7
32.5
33.7
31.6
29.6
30.0
32.6
Leucine
62.6
65.5
64.7
67.3
65.1
75.7
69.2
61.6
62.0
66.0
61.6
60.4
59.2
67.5
Tyrosine
17.5
19.6
19.6
18.0
16.5
17.3
16.5
17.1
17.0
16.9
16.1
160
16.8
19.4
Phenylalanine
32.6
31.7
31.5
30.0
31.4
30.6
30.0
32.2
29.7
29.0
34.6
30 9
33.5
32.4
CANIVET, CRUZ, AND MOREAU-LALANDE
1208
Table 2. Amino Acid Composition of Glomerular
Table 4. HLDU and Lysine Content in Nondiabetic and
Basement Membrane: Average Values (Residues/1000
Diabetic Subjects (mg/lOO mg Membrane Dry Weight)
Amino Acid Residues, Mean + SD) AminoAcids
Nondiabetic Subjects
Subjects
HLDU
Lysine
Nondiabetic
DiabeticSubjects
P
1
3.68
2.59
Lysine
21.8
f 0.90
19.8 + 1.20
NS
2
3.90
2.74
Histidine
14.8 + 0.58
15.1 + 0.75
NS
3
2.83
1.74
Arginine
41.9
f 2.55
41.3
k 2.45
NS
4
3.55
2.52
Hydroxylysine
22.8
% 0.96
28.8
k 0.60
Abnormalities of the Human Diabetic Basement Membrane Jean Canivet,
Anne Cruz, and Hyacinthe
Since the biochemical composition of the diabetic glomerular basement membrane is still a controversial area, a study was carried out using kidneys from seven diabetic and seven nondiabetic subjects. In diabetic membranes, the glycine. hydroxylysine, glucose, galactose, and hydroxylysine-linked disaccharide unit content was increased together with a decrease in the half-cystine and sialic acid content. These findings support the view of a biochemical alteration in the human diabetic glomerular basement membrane.
of the capillary basement T HICKENING membrane is the characteristic morphological feature of diabetic microangiopathy and has been extensively studied in most tissues.’ The biochemical composition has been investigated in glomerular basement membrane from human diabetic kidneys.2m5Results have been conflicting. Beisswenger and Spiro* have reported that analysis of eight individual diabetics showed a distinct chemical alteration characterized by a significant decrease in lysine with an equivalent increase in hydroxylysine and hydroxylysinelinked disaccharide units (HLDU). However, these abnormalities have not been observed in other studies, whereas a significant decrease in the half-cystine content has been reported.3-5 Therefore, the aim of this study was a reevaluation of the half-cystine, lysine, hydroxylysine and HLDU content in the human diabetic glomerular basement membrane isolated from individual patients and with no pooling of the material obtained from each subject.
From the Department of Endocrinology and Metabolism, Hopital Saint Louis, Universitk Paris, France. Receivedfor publication October 24, 1978. Supported in part by the Delegation Getterale a la Recherche Scientifique et Technique (D.G.R.S.T.), Grant 74.7.0588. Address reprint requests to Jean Canivet, Department of Endocrinology and Metabolism, Hopital Saint Louis, 2 Place du Docteur Alfred Fournier, 75475 Paris Cedex IO, France. o 1979 by Grune & Stratton, Inc. 0026-0495/79/2812~04%01.00/0
1206
Glomerular
Moreau-Lalande
MATERIALS
AND
METHODS
Materials Human kidneys were obtained at autopsy within 6 hr following death from 7 diabetic and 7 nondiabetic subjects, and were immediately stored at -3OOC. Diabetic patients were 5 men and 2 women, ages ranging from 39 to 77 yr (mean 50.6 + 5.0) previously treated with diet and insulin or oral hypoglycemic drugs; selection was based solely on the duration of diabetes over 10 yr (range l&3 1 yr, mean 19.6 f 3.5); in all of them, morphological examination of the kidneys showed various degrees of diabetic glomerulosclerosis. Nondiabetic subjects were 6 men and 1 woman, ages ranging from 27 to 78 yr (mean 58.6 * 5.0); these were patients previously admitted to hospital who had died from diseases unrelated to any kidney disorder and with no renal impairment on pathologic examination; none of them had a history of diabetes and fasting blood glucose levels had been within the normal range.
Methods Preparation of the glomerular basement membrane. This was undertaken on both kidneys of each subject and was run until the final lyophilization without pooling the material obtained from several subjects. It was carried out by the method reported by Spiro,6 using a Branson Sonifier model B12 for the disruption of the glomeruli. Some steps were slightly modified as follows. Complete fragmentation of the glomeruli conducted under the phase microscope control required a total sonic disruption time of 6 min (instead of 4-5 min) with 6 bursts of 1 min and a cooling period allowed between bursts until temperature came down to 4°C. The sonicated material was filtrated again through a 200-mesh screen for the elimination of any resistant hyalinized glomeruli and tubular fragments that could have been present, and was thereafter centrifuged. For each washing, following centrifugation, suspension of the sticking pellet of the membrane was performed by a 5-set sonication burst so that the pellet could be more easily suspended in distilled water. Evaluation of the purity of the preparations. This was performed on the 14 preparations. Deoxyribonucleic acid (DNA) analyses were carried out by the method of Burton’ modified by Spiro.6 Uranic acid content was estimated by the method of Dische.’ Total lipid content was determined according to Spiro.6 For electron microscopy, basement membrane samples were fixed in cold glutaraldehyde and embedded in Epon. Analysis of amino acids. Each sample from the 7 normal and 7 diabetic subjects was hydrolized in 6N hydrochloric acid at 105°C for 24 hr at a concentration of 1 mg membrane/ml HCl in sealed tubes; aliquots were evaporated and analysis was performed on each sample in a Jeolco
Metabolism, Vol. 26, No. 12 Khxember), 1979
DIABETIC GLOMERULAR
BASEMENT
MEMBRANE
1207
Calculations. All determinations were made cate. Statistical analyses were made with Student’s
Analyzer, model JLC 5 AH. For the determination of half-cystine, samples of the membrane were first oxidized with pcrformic acid by the method of Schram et aL9 and then hydrolysis was carried out. For the determination of methionine, hydrolysis was performed in an atmosphere of nitrogen in sealed tubes. Analysis of sugar components. This was carried out on separate samples from 4 normal and 4 diabetic subjects among the selected patients. For the determination of glucose, galactose, mannose, and glucosamine, each sample (1 mg) of the membrane was submitted to methanolysis (0.5 ml methanolic- 1.5M hydrochloric acid/ 1 mg membrane dry weight) for 24 hr at 85OC in sealed glass tubes and was thereafter treated with a trimethylsilylating agent; the determination was achieved by the gas chromatographic system of Bhatti et al.;” sugars were evaluated as trimethylsilyl ethers. For the fucose, the membrane was hydrolized in 0.1N sulfuric acid (1 ml/2 mg of membrane dry weight) at 100°C for 8 hr in sealed glass tubes and the determination was carried out using the Dische-Shettles cysteine reaction.” Sialic acid was evaluated according to the method of Svennerholm.” Analysis of HLDU and lysine. This was carried out on separate samples from the 7 normal and 7 diabetic subjects. Each sample was submitted to alkaline hydrolysis (0.1 ml 2N NaOH/l mg membrane dry weight) for 24 hr at 105°C in sealed glass tubes; aliquots of hydrolysate were deposited on 3 mm Whatmann paper and were submitted to high-voltage electrophoresis separation (4500 V for 2 hr in pyridineacetic acid-water 1:10:89, pH 3.6 buffer) according to Moczar and Moczar” in a Gilson electrophoretor, model D;
RESULTS
Isolated glomeruli as seen under phase microscope appeared to be free from tubular elements (less than 0.5%). Under the electron microscope, the membranes appeared as amorphous structures, with no cellular contaminants. Average dry weight of the lyophylized membrane was 44 + 2.3 mg/ 100 g kidney cortex from normals, and 71 .O + 2.6 mg/ 100 g kidney cortex from diabetics. DNA content was less than 0.1% and lipid content was less than 1% of the membrane (dry weight). Results of amino acid analyses from each subject are shown in Table 1 and the average values are summarized in Table 2. In diabetics, as compared with nondiabetic subjects, an increased content in glycine and hydroxylysine and a decreased content in half-cystine were observed (Table 2). The mean values for neutral sugar components are shown in Table 3. An increased content in galactose and glucose was observed in diabetics, with a decreased content in sialic acid. Analyses showed a significantly increased content in HLDU in diabetics as compared with nondiabetics (p < 0.01, Table 4); with the elec-
after drying the paper, revelation with ninhydrine-cadmium and elution with methanol, HLDU and lysine were measured using a Jobin-Yvon spectrophotometer model Monospac at 500 nm. Table 1. Amino Acid Composition
of Glomerular
Basement
(Residues/1000
Membrane
in Seven Diabetics and Seven Nondiabetics
Amino Acid Residues)
NondiabeticSubjects Amano Acld
Lysine
1
2
24.4
21.6
3
19.2
4
5
24.0
24.2
in duplit test.
DiabeticSubjects 6
19.3
7
19.9
6
20.3
9
15.1
10
11
12
13
14
17.3
25.4
20.1
19.6
19.5
H&dine
16.2
10.5
14.7
15.5
15.3
16.2
14.9
17.3
12.3
15.3
17.2
13.2
16.4
14.0
Arginine
48.7
34.3
33.6
50.4
42.0
43.7
41.2
49.4
29.2
37.9
45.2
40.1
42.6
44.5
Hydrowylysine
21.5
19.6
21.8
23.9
22.3
23.3
27.3
29.5
27.7
27.6
30.6
27.9
30.7
27.0
Hydroxyproline
64.3
94.4
96.1
67.9
93.0
69.2
96.2
97.1
122.9
119.9
84.2
111.6
106.0
86.3
Aspartic acid
65.3
70.3
73.1
67.9
73.7
67.6
63.3
67.6
66.4
63.6
67.0
63.6
63.0
63.5
Threonine
36.2
42.1
41.8
34.5
37.8
35.0
32.7
37.3
36.6
35.1
36.3
33.6
34.1
37.6
Serme
45.8
52.8
52.9
45.8
46.8
47.1
45.1
47.0
49.3
45.5
46.4
44.4
43.8
47.6
Glutamic acid
96.6
96.2
97.0
95.1
96.3
97.0
94.4
99.6
95.3
92.2
97.7
94.6
96.0
101.1
Proline
78.4
67.5
66.2
66.6
76.6
54.6
76.6
71.5
61.5
72.4
60.0
74 2
75.4
65.7
Glycme
205.8
175.4
176.8
207.8
196.1
201.0
202.0
208.2
204.9
214.5
199.1
209.4
220.6 53.6
216.2
Alanine
58.5
58.9
57.9
56.3
56.4
60.2
54.2
59.2
50.1
45.6
60.0
57 1
51.9
Half-cystine
24.7
26.2
25.7
19.8
24.2
20.6
24.0
21.1
16.8
16.4
20.8
16.0
20.4
16.7
Valine
41.0
44.4
42.2
32.6
39.2
40.3
55.0
42.2
36.2
36.2
44.4
36 0
38.8
37.4
Methionine
10.6
15.7
14.4
22.5
15.8
12.1
11.9
10.8
13.5
11.6
10.8
10.7
10.4
13.4
lsoleucine
30.9
33.6
32.5
31.3
31.7
33.3
29.3
31.7
32.5
33.7
31.6
29.6
30.0
32.6
Leucine
62.6
65.5
64.7
67.3
65.1
75.7
69.2
61.6
62.0
66.0
61.6
60.4
59.2
67.5
Tyrosine
17.5
19.6
19.6
18.0
16.5
17.3
16.5
17.1
17.0
16.9
16.1
160
16.8
19.4
Phenylalanine
32.6
31.7
31.5
30.0
31.4
30.6
30.0
32.2
29.7
29.0
34.6
30 9
33.5
32.4
CANIVET, CRUZ, AND MOREAU-LALANDE
1208
Table 2. Amino Acid Composition of Glomerular
Table 4. HLDU and Lysine Content in Nondiabetic and
Basement Membrane: Average Values (Residues/1000
Diabetic Subjects (mg/lOO mg Membrane Dry Weight)
Amino Acid Residues, Mean + SD) AminoAcids
Nondiabetic Subjects
Subjects
HLDU
Lysine
Nondiabetic
DiabeticSubjects
P
1
3.68
2.59
Lysine
21.8
f 0.90
19.8 + 1.20
NS
2
3.90
2.74
Histidine
14.8 + 0.58
15.1 + 0.75
NS
3
2.83
1.74
Arginine
41.9
f 2.55
41.3
k 2.45
NS
4
3.55
2.52
Hydroxylysine
22.8
% 0.96
28.8
k 0.60
