Urinary Excretion of the Carboxy Terminal Domain of Type IV Collagen Is Associated with Kidney Size and Function in IDDM

Ole Torffvit, MD Jiirgen Wieslander, PhD Lillemor Forsberg, MD Esbjiirn Hederstriim, MD Carl-David Agardh, MD

Departments of Nephrology, internal Medicine, and Diagnostic

Radiology, University of Lund, Lund, Sweden

Reprint requests to: Ole Torffvit, MD, De- partment of internal Medicine, University Hospital, S-221 85 Lund, Sweden. Submitted for publication September 1989, accepted in revised form March 1990.

0 1990 Elsevier Science Publishing Co., Inc. 0691-6632/90/$3.50

ABSTRACT

We evaluated whether urinary excretion of the carboxy terminal do- main (NCl) of Type IV collagen is associated with glomerular filtration rate and kidney size in Type I (insulin-dependent) diabetes mellitus (IDDM). Urinary excretion rate of NCl, glomerular filtration rate (GFR), and kidney size were measured in 16 men with Type I diabetes. Their mean age was 33.3 + 6.1 years with a duration of diabetes of 14.9 + 3.7 years (mean fr SD). The urinary excretion rate of NC1 was higher in the diabetic patients than in 16 healthy control subjects. Urinary excretion of NC1 was associated with both kidney size, parenchymal width, and GFR (I = 0.73, p = 0.001; r = 0.63, p = 0.009; r = 0.53, p = 0.04, respectively). The exact relationship between these factors and basement membrane turnover/synthesis remains to be eluci- dated. (The Journal of Diabetic Complications 4;4:166-169, 1990.)

INTRODUCTION Glomerular filtration rate (GFR) 1-5 and kidney size’-3v6 are increased before development of diabetic nephropathy. GFR is dependent on met- abolic control’ and kidney sizeim5 and reflects hyperfunction4; glomer- ular hyperfiltration has been associated with late diabetic nephropathy in one5 but not in another study.7

The structural hallmark of the glomerulopathy is accumulation of basement membrane-like material* composed of collagenous and non- collagenous proteins. The major collagenous component is Type IV collagen. Two molecules of collagen IV are cross-linked at the carboxy terminal end, and this noncollagenous globular domain is called NCl. Serum concentrations of NC1 have been suggested to reflect synthesis of Type IV collagen in ratsg, and increased urinary excretion of NC1 coexists with high normal albuminuria in insulin-dependent diabetes mellitus (IDDM).” We evaluated whether urinary NC1 excretion is as- sociated with clinical parameters, such as glomerular filtration rate and kidney size, in patients with Type I diabetes.

MATERIALSAND METHODS

Subjects Sixteen men with Type I diabetes that began before age 30, were studied. All had a diabetes duration >7 years, Albustix negative urinalysis, negative urine culture, normal urinary sediment, resting blood pressure 5 150/90 mmHg, serum creatinine levels ~100 pmol/L, and no medication other than insulin. Patient characteristics are shown in Table 1. Urinary excretion rates of NC1 and albumin were compared with the same parameters in 18 age-matched, apparently healthy men who served as control subjects. All participants gave their informed consent, and the study was approved by the local ethics committee.

Analytic Methods Urinary albumin levels were measured with an en- zyme-linked immunosorbent assay previously described.” The method for measuring NC1 has also been described.” The sensitivity was 4 kg/ L, with intra- and interassay variabilities of 11.9% and 39.7%, respec- tively. The day-to-day variation within subjects was 26% for urinary NC1

166

URINARY COLLAGEN AND KIDNEY SIZE

TABLE 1 Patient Characteristics TABLE 2

Age (yr) Diabetes duration (yr) Insulin dose (IUiday) HbAlc (%) Body mass index (kg/m’) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Serum creatinine (pmol/L) Glomerular filtration rate (mllmini1.73m2)

Values are means + SD.

33.3 k 6.1 14.9 f 3.7 47.4 2 13.9

8.0 * 1.2 23.6 i 1.4

126.6 ? 12.5 76.6 t 7.7 82.2 2 9.2

118.6 2 20.1

Patients Controls Variable (n = 16) (n = 16)

Albumin excretion rate (pg/min) 13.0x/:2.6 5.7x/:1.8*

NCI excretion rate (ng/min) 0.96 + 0.32 0.61 f 0.24t Kidney size (cm) 12.0 2 0.9

Kidney parenchymal width (cm) 1.9 k 0.5

Values are means 2 SD. For albumin excretion rate the geo- metric mean x/: tolerance factor was used: * = 2P = 0.007; t = 2P = 0.002.

concentration, and 37% for urinary NC1 excretion rate.” Before analysis of NCl, the urine was concentrated 16.1 + 3.8 times (mean ? SD) by centrifugation at 2,000 g with an Amicon Centriprep CPlO (W.R. Grace, Amicon Division, Danvers, CT); molecular weight cut off level 1,000 Dalton). Urine was obtained by three timed (24 h) collections in the diabetic patients, and at least one col- lection in the control subjects, and was stored in poly- styrene plastic tubes (10 ml/sample) containing 10 p-1 Tween 20 at -20°C until analysis. Analysis for urinary albumin was performed within one month; for NCl, it was performed within 2 years of sampling.

RESULTS

In the diabetic patients, both the urinary excretion rate of NC1 (0.96 2 0.32 ng/min; mean c SD) and albumin (13.0 x/:2.6 Fg/min; geometric mean x/: tolerance factor) were higher than in the control subjects (0.61 4 0.24 ngl min; 2p = 0.002 and 5.7x/:1.8 p,g/min; 2p = 0.007, re-

spectively) (Table 2). Mean sizes of the right and left kidneys were 12.0 2

0.9 cm, and the mean parenchymal width was 1.9 t 0.5

GFR was measured as 5’CrEDTA clearance, corrected for body surface area. Within 3 months of the urine col- lections, glycosylated hemoglobin (HbA,,) levels were measured by ion-exchange chromatography using com- mercially available microcolumns (BioRad, Richmond, CA); normal value ~6%. Serum creatinine concentrations were measured with a kinetic Jaffe reaction. Blood pres- sure was taken with a conventional mercury sphygmo- manometer in the supine position after 5 minutes rest. Korotkoff phase V was considered to denote the diastolic blood pressure (DBP).

r 2.0

E P z 1.5

_I

H

6 t 1.

Q)

Ultrasonography was performed for measurement of kidney size within 6 months of urine collections, using modern real-time equipment (Toshiba 100). Kidney size was measured with the patient in the supine or anterior oblique position as the mean of right and left kidney length from superior to inferior border. Multiple mea- surements were performed, and the longest on each side was noted so that renal size would not be underesti- mated. Approximations were made to 0.5 cm intervals. Kidney parenchymal width was measured by ultrasonog- raphy as the mean between the right and left superior, and middle and inferior width from the point of the pa- pillae to the exterior of the kidney.

Statistics The albumin levels (log 10 transformed) con- formed fairly Well to a normal distribution. These data are therefore presented as geometric mean x/: tolerance fac- tor. The Mann-Whitney two-tailed test was used for com- parison between diabetic patients and controls, and lin- ear regression analyses by least-squares method were used for correlations. Five of 14 urine samples analyzed in five of the diabetic patients, and five in the control subjects, had NC1 levels below the detection limit and were allocated the lowest detectable concentration of 4

IJ.g/L.

0.0 !: , I 1 I I

Q8 1.3 1.8 2.3 2.8 Kidney pamnchymal widthkm)

FIG. 1 Urinary NC1 excretion rate (ng/min) and kidney size (cm) (r = 0.73, p = 0.001) (A). Urinary NC1 excretion rate (ng/min) and kidney parenchymal width (cm) (r = 0.63, p = 0.009) (6).

167

Urinary Excretion of Albumin and NCI in Patients and Controls, and Ultrasonographic Data

0

l

00 0 0

0 a 00

a l l

l l

0

0

l

0 0 aa

*a a 0

l 0

l

TORFFVIT ETA.

0

‘.’ eB

l

l

0 l

0

l l

0.04 , I I 1

70 100 130 160

Glomerular filtration rate (mMnin/1.73m2)

FIG. 2 Urinary NC1 excretion rate (ng/min) and glomerular fil- tration rate (ml/min/1.73m2) (r = 0.53, p = 0.04).

cm (Table 2). GFR was 118.6 + 20.1 ml/min/l .73m2 (Table

1). Urinary excretion rate of NC1 correlated with kidney

size (Figure la; r = 0.73; p = O.OOl), parenchymal width (Figure lb; r = 0.63; p = 0.009), and GFR (Figure 2;

Kidney size (cm)

0 l

j6oi (3 08 l.3 la 23 28

Kidney parenchymal width (cm)

FIG. 3 Glomerular filtration rate (GFR, ml/min/l.73m2) and kid- ney size (cm); (r = 0.54, p = 0.03) (A). GFR (ml/min/l. 73m2) and kidney parenchymal width (cm), (r = 0.55, p 0.03) (6).

r = 0.53; p = 0.04), but the excretion rate of albumin did not (r = -0.07; r = -0.30; r = -0.02, respectively). GFR was related to kidney size (Figure 3a, r = 0.54; p = 0.03) and parenchymal width (Figure 3b, r = 0.55; p = 0.03). No relationship was noted between the urinary excretion rate of NC1 on the one hand and the urinary albumin excretion rate (AER), levels of HbAIc, blood pressure, di- uresis, insulin dose, body mass index, or serum creati- nine, on the other.

DISCUSSION

Basement membrane components in urine were first de- tected in healthy humans” and later in people with kid- ney disease.13 In an earlier study, we showed that diabetic patients have an increased urinary excretion of NC1 when urinary albumin excretion rate is at the upper limit of nor- mal (6.5-30 yg/min),‘O and were able to confirm a pre- vious observation in diabetic mice which showed in- creased urinary concentrations of Type IV collagen.‘4

In the present study, increased urinary excretion of NC1 was associated with size and parenchymal width of the kidneys. The accuracy of using ultrasonography for measurement of kidney length and width was described previousIy’5 and is a reliable and suitable alternative to urography. ” The best estimation of kidney size uses the length alone.” In this study, we also found a better cor- relation between urinary excretion of NC1 and renal length as compared to width. Hypothetically, such an as- sociation might be caused by the diabetic state per se leading to a generally increased synthesis of all kidney material. This cannot be the only explanation, however, as in one study kidney size was unrelated to glomerular basement membrane width.” A first step in development of diabetic nephropathy is thickening of the glomerular basement membrane.* NC1 is part of Type IV collagen, the major collagenous component of the glomerular basement membrane. In early and moderate stages of diabetic nephropathy, Type IV collagen increases in the mesangium and tubular basement membranes.19 Be- cause Type IV collagen occurs in all basement mem- branes, it is unknown whether urinary NC1 excretion rep- resents glomerular or general basement membrane metabolism. However, urinary NC1 represents the latter, and one would expect serum NC1 to be increased, but this was not observed in a previous study of diabetic pa- tients.‘O Another possibility is that urinary NC1 excretion reflects changes in tubular reabsorption and/or secre- tion, but this is presently unknown. Glomerular basement membrane thickening per se is unlikely to be responsible for development of overt proteinuria in diabetic nephro- pathy, in contrast to mesangial and peripheral capillary surface expansion.20 Therefore, it is not surprising that no relationship was observed between urinary excretion rates of NC1 and albumin.

With development of incipient nephropathy, clinically defined by the presence of microalbuminuria (30-200 pg/ min), urinary excretion of NC1 decreases.” This obser- vation is consistent with a previous study which showed reduction of Type IV collagen in the mesangium and glo- merular basement membrane with development of glo- merular sclerosis.”

URINARY COLLAGEN AND KIDNEY SIZE 169

Urinary excretion of NC1 was associated with GFR, which is related to glomerular filtration surface*“** and kidney size.lm3 Increased GFR’,6 and kidney size’ de- crease with strict metabolic control in newly diagnosed diabetes mellitus.’ However, metabolic factors are not the only explanations for increased kidney size, because with longer duration of diabetes the kidney volume does not change when there has been strict metabolic con- tr0L6

6.

7.

In summary, increased urinary excretion of NC1 was associated with GFR, kidney size, and parenchymal width. The exact relationship between these factors and basement membrane turnover/synthesis remains to be elucidated. Further studies are needed to determine if a first step in diabetic nephropathy is increased basement membrane synthesis, and if excretion of NC1 may be an early marker for development of diabetic nephropathy.

8.

9.

10.

11.

ACKNOWLEDGMENTS 12.

This work was supported by Arne Lundberg’s Founda- tion, Gijteborg; the Crafoord Foundation, Lund; the Royal Physiographic Society, Lund; the Medical Faculty, University of Lund; the Nordic Insulinfund, Copenhagen, Denmark; the Swedish Society of Nephrology; the Swed- ish Society of Medicine; and the Swedish Hoechst Dia- betesfund. The authors are grateful to Professor U. Bengtsson for fruitful discussions and to Mrs. A.M. Pers- son for designing the figures.

13.

14.

15.

16.

17. REFERENCES

1.

2.

3.

4.

5.

Mogensen CE, Andersen MJF: Increased kidney size and glomerular filtration rate in untreated juvenile diabetes: Nor- malization by insulin-treatment. Diaberologia 11:221-224, 1975. Puig JG, Antdn FM, Grande C, Pallardo LF, Arnalich F, Gil A, Vdzquez JJ, Garcia AM: Relation of kidney size to kidnev function in early insulin-dependent diabetes. Diabetologia 21:363-367, 1981. Sandahl Christiansen J, Gammelgaard J, Frandsen M, Parv- ing H-H: increased kidney size, glomerular filtration rate and renal plasma flow in short-term insulin-dependent diabetics. Diabetologia 20:451-456, 1981. Brbchner-Mortensen J, Ditzel J: Glomerular filtration rate and extracellular fluid volume in insulin-dependent patients with diabetes mellitus. Kidney Int 21:696-698, 1982. Mogensen CE: Early glomerular hyperfiitration in insulin-

18.

19.

20.

21.

22.

dependent diabetics and late nephropathy. Stand J C/in Lab Invest 46:201-206, 1986. Christensen CK, Sandahl Christiansen J, Schmitz A, Chris- tensen T, Hermansen K, Mogensen CE: Effect of continuous subcutaneous insulin infusion on kidney function and size in IDDM patients: A 2 year controlled study. J Diabefic Com- plications 1:91-95, 1987. Letvang H-H, Jensen S, Brbchner-Mortensen J, Ditzel J: Early glomerular hyperfiltration and the development of late nephropathy in Type 1 (insulin-dependent) diabetes mellitus. Diabetologia 311723-9, 1988. Osterby R: Early phases in the development of diabetic glomerulopathy. Acta MedScand200(suppl574):1-82,1975. Brocks DG, Neubauer HP, Strecker H: Type IV collagen an- tigens in serum of diabetic rats: A marker for basement mem- brane collagen synthesis. Diabetologia 28:928-932, 1985. Torffvit 0, Agardh C-D, Cederholm B, Wieslander J: A new enzyme-linked immunosorbent assay for urine and serum concentrations of the carboxyterminal domain (NCl) of col- lagen IV. Application in type I (insulin-dependent) diabetes. Stand J C/in Lab Invest 49:431-439, 1989. Torffvit 0, Wieslander J.: A simplified enzyme-linked im- munosorbent assay for urinary albumin. Stand J C/in Lab Invest 46545-548, 1986. McPhaul JJ, Dixon FJ: lmmunoreactive basement membrane antigens in normal human urine and serum. J Exp Med 130:1395-1409,1969. Bygren P, Wieslander J, Heinegtird D: Human glomerular basement membrane. Antigenic determinants in human urine. Biochim Bioohvs Acta 553:255-261, 1979. Mellon CD, Sasser’B\ir, Bowman BH: Increased concentra- tions of basement membrane collagen fragment in urine of diabetic mice. Biochem Genet 22(9/10):631-639, 1985. Hederstriim E, Forsberg L: Accuracy of repeated kidney size estimation by ultrasonography and urography in children. Acfa Radio/ Diagnosis 26(5):603-607, 1985. Hedertrijm E, Forsberg L: Kidney size in children assessed by ultrasonography and urography. Acta Radio Diagnosis 26:85-91, 1985. Griffiths GJ, Cartwright G, McLachlan MSF: Estimation of renal size from radiographs: Is the effort worthwhile? C/in Radio/ 26:249-256, 1974. Ellis EN, Steffes MW, Goetz FC, Sutherland DER, Mauer SM: Relationship of renal size to nephropathy in Type 1 (insulin- dependent) diabetes. Diabetologia 28:12-15, 1985. Falk RJ, Scheinman JI, Mauer SM, Michael AF: Polyanti- genie expansion of basement membrane constituents in di- abetic nephropathy. Diabetes 32(suppl 2):34-39, 1983. Mauer SM. Steffes MW. Ellis EN. Sutherland DER. Brown DM, Goetz FC: Structu’ral-functional relationships’ in dia- betic nephropathy. J C/in Invest 74:1143-55, 1984. Ellis EN, Steffes MW, Goetz FC, Sutherland DER, Mauer SM: Glomerular filtration surface in type I diabetes mellitus. Kid- ney Int 29:889-894, 1986. Osterby R, Parving H-H, Nyberg G, Hommel E, Jijrgensen HE, LBkkegaard H, Svalander C: A strong correlation be- tween glomerular filtration rate and filtration surface in di- abetic nephropathy. Diabetologia 31:265-270, 1988..