insulin-like growth factor-i and insulin-like growth factor-binding proteins in the nephrotic...

4
Pediatr Nephrol (1996) 10:355-358 IPNA 1996 Pediatric 51ephrolooy Review article Insulin-like growth factor-I and insulin-like growth factor-binding proteins in the nephrotic syndrome Stella M. Feld and Raimund Hirschberg Division of Nephrologyand Hypertension, Harbor-UCLA Medical Center and UCLA,Torrance,California,USA Received July 7, 1995; receivedin revisedform November6, 1995; acceptedNovember27, 1995 Abstract. Similar to findings in the nephrotic syndrome in humans, rats with the doxorubicin-induced nephrotic syn- drome (which resembles minimal change disease) have reduced serum levels of insulin-like growth factor-I (IGF-I). This is mainly caused by glomerular ultrafiltration of IGF-I-containing binding protein complexes, primarily of a molecular weight of approximately 50 kilodaltons, and urinary losses of the peptide. Despite urinary excretion of IGF-binding protein (IGFBP)-2, serum levels are increased more than twofold in the nephrotic syndrome compared with controls, because of increased synthesis of this binding protein by the liver. In contrast, the liver synthesis of IGFBP-3, the predominant binding protein in normal serum, is unchanged in the nephrotic syndrome. However, binding and serum levels of IGFBP-3 are reduced in ne- phrotic rat serum, apparently due to proteolytic degradation of IGFBP-3. The glomerular ultrafiltration of IGF-I, which leads to biologically significant IGF-I concentrations of about 1.35 nM in proximal tubule fluid, may have meta- bolic consequences, such as increased tubular phosphate absorption. Hypothetically, tubule fluid IGF-I may also contribute to progressive tubulointerstitial fibrosis which is sometimes present in protractive nephrotic glomer- ulopathies. The profound changes in the IGF-I/IGFBP system in the nephrotic syndrome may also contribute to systemic metabolic abnormalities and growth failure. Key words: Nephrotic syndrome - Insulin-like growth factor-I - Insulin-like growth factor binding proteins-2, -3, and -4 - Insulin-like growth factor binding protein-3 protease Correspondence to: R. Hirschberg,Box 404, Division of Nephrology, Harbor-UCLA Medical Center, 1000 West Carson Street, Torrance, CA 90509, USA Introduction Insulin-like growth factor-I (IGF-I) is a peptide growth factor [molecular weight approximately 7,500 kilodaltons (kDa)] which is present at rather high concentrations in the serum of adults and many research animals. Serum levels of IGF-I are low at birth but rapidly rise thereafter. In serum, ---99% of IGF-I is bound in specific binding protein complexes. The majority (80%) of circulating IGF-I is present in a 150-kDa protein complex that contains IGF- binding protein-3 (IGFBP-3) as the ligand binding moiety, and an acid-labile subunit (approximately 85 kDa). Most of the remaining IGF-I is present in a smaller protein complex of 45-50 kDa, which contains IGFBP-1, -2, -4, or some IGFBP-3 [1-4]. Less than 1% of the serum IGF-I is present in the unbound form [1]. The binding of IGF-I in the large or small protein complex is believed to contribute to the regulation of the bioactivity and tissue availability of IGF-I. The 150-kDa complex is limited to the intravascular space, whereas the 50-kDa complex is able to reach extravascular tissue re- ceptor sites [5, 6]. Cells and tissues that release IGF-I usually also release IGFBPs which appear to modify the autocrine/paracrine actions of the peptide. IGF-I has mi- togenic, metabolic, growth-promoting, and renal hemody- namic actions [7, 8]. IGF-I is synthesized in many organs, including the kidney, but most of the circulating IGF-I and IGFBPs are released from the liver [9, 10]. IGF-I acts through specific receptors present in the membrane of many cells and tis- sues. In the kidney, IGF-I receptors are present on the three residential glomerular cell types and on the basolateral as well as apical membrane of proximal tubules [8, 11]. IGF-I synthesis occurs in the glomerulus and in the distal nephron [8]. Due to its small molecular weight, free serum IGF-I undergoes glomerular ultrafiltration, may bind to apical tubule receptors, may be absorbed or degraded, or may undergo urinary excretion. Several investigators have demonstrated that in normal subjects small amounts of IGF-I are found in urine (approximately 50-200 pg/mg

Upload: stella-m-feld

Post on 10-Jul-2016

221 views

Category:

Documents


5 download

TRANSCRIPT

Page 1: Insulin-like growth factor-I and insulin-like growth factor-binding proteins in the nephrotic syndrome

Pediatr Nephrol (1996) 10:355-358 �9 IPNA 1996 Pediatric

51ephrolooy

Review article

Insulin-like growth factor-I and insulin-like growth factor-binding proteins in the nephrotic syndrome

Stella M. Feld and Raimund Hirschberg

Division of Nephrology and Hypertension, Harbor-UCLA Medical Center and UCLA, Torrance, California, USA

Received July 7, 1995; received in revised form November 6, 1995; accepted November 27, 1995

Abstract. Similar to findings in the nephrotic syndrome in humans, rats with the doxorubicin-induced nephrotic syn- drome (which resembles minimal change disease) have reduced serum levels of insulin-like growth factor-I (IGF-I). This is mainly caused by glomerular ultrafiltration of IGF-I-containing binding protein complexes, primarily of a molecular weight of approximately 50 kilodaltons, and urinary losses of the peptide. Despite urinary excretion of IGF-binding protein (IGFBP)-2, serum levels are increased more than twofold in the nephrotic syndrome compared with controls, because of increased synthesis of this binding protein by the liver. In contrast, the liver synthesis of IGFBP-3, the predominant binding protein in normal serum, is unchanged in the nephrotic syndrome. However, binding and serum levels of IGFBP-3 are reduced in ne- phrotic rat serum, apparently due to proteolytic degradation of IGFBP-3. The glomerular ultrafiltration of IGF-I, which leads to biologically significant IGF-I concentrations of about 1.35 nM in proximal tubule fluid, may have meta- bolic consequences, such as increased tubular phosphate absorption. Hypothetically, tubule fluid IGF-I may also contribute to progressive tubulointerstitial fibrosis which is sometimes present in protractive nephrotic glomer- ulopathies. The profound changes in the IGF-I/IGFBP system in the nephrotic syndrome may also contribute to systemic metabolic abnormalities and growth failure.

Key words: Nephrotic syndrome - Insulin-like growth factor-I - Insulin-like growth factor binding proteins-2, -3, and -4 - Insulin-like growth factor binding protein-3 protease

Correspondence to: R. Hirschberg, Box 404, Division of Nephrology, Harbor-UCLA Medical Center, 1000 West Carson Street, Torrance, CA 90509, USA

Introduction

Insulin-like growth factor-I (IGF-I) is a peptide growth factor [molecular weight approximately 7,500 kilodaltons (kDa)] which is present at rather high concentrations in the serum of adults and many research animals. Serum levels of IGF-I are low at birth but rapidly rise thereafter. In serum, ---99% of IGF-I is bound in specific binding protein complexes. The majority (80%) of circulating IGF-I is present in a 150-kDa protein complex that contains IGF- binding protein-3 (IGFBP-3) as the ligand binding moiety, and an acid-labile subunit (approximately 85 kDa). Most of the remaining IGF-I is present in a smaller protein complex of 45-50 kDa, which contains IGFBP-1, -2, -4, or some IGFBP-3 [1-4]. Less than 1% of the serum IGF-I is present in the unbound form [1].

The binding of IGF-I in the large or small protein complex is believed to contribute to the regulation of the bioactivity and tissue availability of IGF-I. The 150-kDa complex is limited to the intravascular space, whereas the 50-kDa complex is able to reach extravascular tissue re- ceptor sites [5, 6]. Cells and tissues that release IGF-I usually also release IGFBPs which appear to modify the autocrine/paracrine actions of the peptide. IGF-I has mi- togenic, metabolic, growth-promoting, and renal hemody- namic actions [7, 8].

IGF-I is synthesized in many organs, including the kidney, but most of the circulating IGF-I and IGFBPs are released from the liver [9, 10]. IGF-I acts through specific receptors present in the membrane of many cells and tis- sues. In the kidney, IGF-I receptors are present on the three residential glomerular cell types and on the basolateral as well as apical membrane of proximal tubules [8, 11]. IGF-I synthesis occurs in the glomerulus and in the distal nephron [8].

Due to its small molecular weight, free serum IGF-I undergoes glomerular ultrafiltration, may bind to apical tubule receptors, may be absorbed or degraded, or may undergo urinary excretion. Several investigators have demonstrated that in normal subjects small amounts of IGF-I are found in urine (approximately 50-200 pg/mg

Page 2: Insulin-like growth factor-I and insulin-like growth factor-binding proteins in the nephrotic syndrome

356

NS IGF-I mRNA ~7.8kb

)GF-I mRNA ~1.9kb

IGF-t mRNA -0.8-1.2kb

5

4

. 3 - ,-f

2-

Kidney IGFI mRNA

(~7.8kb)

1 "3-

o N NS

Fig. 1. Renal steady-state insulin-like growth factor-I (IGF-1) mRNA levels. The left panel depicts an example of autoradiograms after Northern analysis of Kidney RNA from a normal (N) and a nephrotic (NS) rat. The bar graph shows densitometric quantifications after correction for [3-actin mRNA levels (n = 6). Control values are arbitrarily set at "1". kb, kilobases;* p <0.05

creatinine) [12-15]. IGFBPs have consistently been de- monstrated in urine collected from normal subjects [15, 16]. Western ligand and immunoblot assays have identified IGFBP-3 and -2 in normal urine [15]. Ligand blotting also demonstrated a doublet of bands at 28/29 kDa [15] which may contain IGFBP-1 [16]. IGFBP-3 levels in urine from normal adults have been estimated at about 15-30 ng/mg creatinine [15]. This estimate may roughly correspond to a urinary excretion of IGFBP-3 of < 50 gg/24 h ( < 2 nmol/ 24 h). The urinary excretion of IGFBP-2 in normal subjects has been estimated at about 4 rig/rag creatinine [16].

IGF-I and IGFBPs in the experimental nephrotic syndrome

Studies were performed in this laboratory in pair-fed rats at approximately 20 days after a single intravenous injection of doxorubicin, which results in a stable nephrotic syn- drome with selective proteinuria [-----70 kDa by urine so- dium dodecyl sulfate polyacylamide gel electrophoresis (SDS-PAGE)]. In these animals, renal and nephron hemo- dynamics are essentially normal indicating that renal failure is absent [17]. Compared with controls serum IGF-I levels are reduced by one-third in nephrotic rats [18]. This may result, at least in part, from urinary excretion of IGF-I in nephrotic rats, which occurs at a rate of 129_+ 10 pg/min [18].

In contrast to normal animals, in nephrotic rats there are measurable and biologically significant amounts of IGF-I present in glomerular ultra.filtrate. The IGF-I levels in early proximal tubule fluid collected by nephron micropuncture from nephrotic rats are, on average, 1.35 nM, whereas they are below the detection limit in most normal rats [17]. This finding suggests that the increased glomerular ultrafiltra- tion and urinary excretion of IGF-I in the nephrotic syn- drome occurs in association with IGFBP-complexes. In-

deed, Western ligand blot analysis of nephrotic rat proximal tubule fluid demonstrates the presence of IGFBP-2, but not 1GFBP-3, suggesting that it is the 50-kDa complex that primarily undergoes glomerular ultrafiltration rather than the 150-kDa complex [8, 17]. We recently confirmed that IGFBP-2 is present in proximal tubule fluid in nephrotic rats by Western immunoblot (unpublished observation). Bladder urine contains IGFBP-2, but also some IGFBP-3, as shown by Western ligand and immunoblot analysis [18]. Our finding that IGFBP-3 is found in urine but not in proximal tubule fluid may have several possible explana- tions: (1) urinary IGFBP-3 represents distal tubule secretion of the binding protein; (2) small amounts of the 150-kDa complex undergo glomerular ultrafiltration, but the levels of IGFBP-3 in proximal tubule fluid are below the detec- tion limit of the assays; (3) the urinary IGFBP-3 results from ultrafiltration of 50-kDa protein complexes which contain IGFBP-3 in only small quantities, and IGFBP-3 is detectable in bladder urine but not in (less-concentrated) proximal tubule fluid.

We next examined whether the IGF-I synthesis in liver and kidney is reduced, which may contribute to the ob- served reduction in serum IGF-I levelsl Steady-state IGF-I mRNA levels in nephrotic rat livers were similar to con- trols. Although no definite proof can be provided, this finding suggests that the synthesis of IGF-I in liver is not reduced in the nephrotic syndrome [18]. In nephrotic rat kidney the steady-state IGF-I mRNA levels are slightly reduced compared with normal rats (Fig. 1), and a lower rate of renal IGF-I synthesis may contribute to the reduced serum levels in the nephrotic syndrome. Whether the synthesis of IGF-I in the nephrotic syndrome is also re- duced in other organs is presently unknown.

The amount of intact IGFBP-3 in serum that binds IGFs in the nephrotic syndrome appears to be reduced [18]. In normal rat serum, Western ligand blotting demonstrates binding of IGF-I predominantly to IGFBP-3, and to a far lesser extent to IGFBP-2, -4, and possibly -! (Fig. 2) [18]. In nephrotic rat serum, binding of IGF-I to IGFBP-3 and -4 is reduced, but binding to IGFBP-2 is increased, although the net total serum IGF-I binding capacity in Western li- gand blots is reduced [18]. IGFBP-2 is truly increased in nephrotic serum, as demonstrated by Western im- munoblotting [18]. This rise in serum IGFBP-2 levels is caused, at least in part, by an increase in the liver synthesis of this binding protein, as indicated by a four- to fivefold increase in the hepatic steady-state levels of IGFBP-2 mRNA and by increased levels of IGFBP-2 in liver lysates, as demonstrated by immunoprecipitation and subsequent Western analysis [18].

In contrast, IGFBP-3 mRNA levels are similar in ne- phrotic and control liver, suggesting that the hepatic synthesis of IGFBP-3 is unchanged [18]. In addition to losses through urinary excretion in the nephrotic syndrome, serum IGFBP-3 levels also appear to be reduced by pro- teolytic degradation. First, Western immunobIotting using a specific anti-rat IGFBP-3 antibody demonstrates con- sistently an additional band at approximately 27 kDa in nephrotic rats which is not present in control rat sera and which does not bind IGF-I in Western ligand blots [18]. Second, nephrotic but not normal rat urine contains strong

Page 3: Insulin-like growth factor-I and insulin-like growth factor-binding proteins in the nephrotic syndrome

357

2 IGFBP-3 2 IGFBP-2 ,, -, 2 IGFBP-.4 / / /Y

-o -o / / / y -~ ~_ 2 / / / / 2

/ / . ,7 / / / /

N NS N NS N NS

Fig. 2. Binding of 125iodine-IGF-I to IGF-binding proteins ( I G F B P ) in vitro. Sera from normal (N) and nephrotic (NS) rats (n = 6 each) were analyzed by Western ligand blot. Bands were semiquantified by densitometry. Values in normal sera were arbitrarily set at "1". ~p < 0.05

Serum U r i n e ..... Discussion i i J

N NS NS NS

Fig. 3. Native gel analysis of IGFBP protein complexes. Sera (2.5 gl) or nephrotic rat urine (5 gl) were incubated with 0.3 gCi of 125I-IGF-I in 20 gl TRIS-buffered saline, pH 7.4, for 12 h at 4 ~ C. Samples were taken up in 25 p,1 of 2x sample buffer void of sodium dodecyl sulfate (SDS) and reducing agent and were not denatured. After electropho- resis in SDS-free polyacrylamide gels (10%) using a SDS-free TRIS- glycine buffer system, the dried gel was autoradiographed. Because of the small amounts of total urinary protein excretion in normal rats, this analysis was not performed in normal rat urine

proteolytic activity towards IGFBP-3 that can be demon- strated even at > 1,000-fold dilution of pooled nephrotic urine [18].

In summary, the availability of intact IGFBP-3 to bind IGF-I in serum is reduced and binding of IGF-I to IGFBP-2 and its immunoreactive levels are increased in the ne- phrotic syndrome. Compared with normal serum, this may result in a shift of serum IGF-I binding from the 150-kDa to the 50-kDa protein complex which can access extravascular tissue binding sites.

To test this hypothesis, we performed native gel elec- trophoresis analysis of normal and nephrotic rat serum and urine (Fig. 3). In this technique IGFBP complexes are in- cubated with 125iodine-IGF-I in vitro and are then electro- phoresed without prior denaturing. With this technique, one major band can be visualized in normal rat serum (Fig. 3). In nephrotic rat serum, this band appears reduced and a second band is visualized (Fig. 3). Nephrotic urine contains two bands, one of which co-migrates with the lower band found in nephrotic rat serum (Fig. 3). Thus, in vitro IGF-I binding to serum IGFBP complexes differs markedly be- tween normal and nephrotic rat serum. Unfortunately, this method does not allow for an estimate of the molecular weights of the protein complexes. In native gels protein mobility depends much on the natural charge of proteins, whereas in denaturing SDS-PAGE all proteins become heavily charged and the mobility depends primarily on the molecular mass of proteins.

The literature contains only a few studies of IGF-I/IGFBPs in the experimental or human nephrotic syndrome. Thabet et al. [19] also demonstrated a decrease in serum IGF-I and IGFBP-3, as well as an increase in IGFBP-2, by Western ligand blot in rats with puromycin aminonucleoside ne- phrosis [19]. In contrast to the findings from this laboratory, IGFBP-4 was increased in sera from these latter rats. However, normal rats that were pair fed with the nephrotic rats did not demonstrate this rise in serum IGFBP-4 [19]. Furthermore, Thabet et al. [19] described a decrease in hepatic IGF-I mRNA levels in their nephrotic rat model which we cannot confirm in the doxorubicin-induced ne- phrosis. Both sets of data, however, are compatible with the hypothesis that binding of IGF-I in nephrotic serum is shifted from the large to the small molecular weight protein complex.

Garin et al. examined serum and urine from nephrotic children [20]. Serum IGF-I levels were reduced by 50%. Most of the urine IGF-I was excreted as a molecular form of 45 kDa, some smaller amounts were found at an ap- parent molecular weight of 150 kDa [20]. The presence of the large molecular weight complex in urine of these pa- tients suggests the presence of non-selective proteinuria that allows ultrafiltration of large molecular weight pro- teins. In contrast, our doxorubicin nephrosis model some- what resembles minimal change disease with selective proteinuria (--<70 kDa).

The glomerular ultrafiltration of IGF-I leads to the ex- posure of apical tubule receptors to supraphysiological le- vels of IGF-I. Circumstantial evidence and in vitro ex- periments suggest that this may contribute to increased phosphate retention [21, 22]. Furthermore, tubule fluid IGF-I may increase the expression of extracellular matrix genes and possibly contribute to the interstitial fibrosis that is sometimes observed in chronic nephrotic glomer- ulopathies. However, this hypothesis will need experi- mental confirmation. We recently demonstrated that IGF-I raises procollagen (zl(I) and oO(IV) mRNA levels and the excretion of collagen type I and IV in cultured glomerular mesangial cells [23] and increases the secretion of these two collagen proteins in cultured mouse proximal tubule cells (unpublished observation).

Page 4: Insulin-like growth factor-I and insulin-like growth factor-binding proteins in the nephrotic syndrome

358

Acknowledgements. Work performed in this laboratory and cited or described in this paper was supported by a grant from the American Heart Association, National Center, Dallas, Tex., USA (90-1218). Stella Feld is a recipient of a NIH-National Research Service Award. Janine LaPage, BS, and Ilyse Vanderah, MS, skillfully assisted in parts of this work. Supportive advice, the donation of cDNA probes or of antibodies by Charles Roberts, PhD, Shunichi Shimasaki, PhD, and Nicholas Ling, PhD, is greatly appreciated.

References

1. Frystyk J, Skjaerbaek C, Dinesen B, Orskov H (1994) Free insulin- like growth factor (IGF-I and IGF-II) in human serum. FEBS Lett 348: 185-191

2. Hintz R, Liu F (1977) Demonstration of specific plasma binding sites for somatomedin. J Clin Endocrinol Metab 45:988-995

3. Baxter R, Martin J (1989) Structure of the Mr 140,000 growth hormone-dependent insulin-like growth factor binding protein complex: determination by reconstitution and affinity-labeling. Proc Natl Acad Sci USA 86:6898-6902

4. Holly J (1993) Insulin-like growth factor binding proteins in diabetic and non-diabetic states. In: Flyvbjerg AS, Orskov H, Alberti K (eds) Growth hormone and insulin-like growth factor I in human and experimental diabetes. Wiley, Chichester, UK, pp 47-76

5. Hodgkinson S, Davis S, Burleigh B, Henderson H, Gluckman P (1987) Metabolic clearance rate of insulin-like growth factor-I in fed and starved sheep. J Endocrinol 115:,233-240

6. Blat C, Villaudy J, Binoux M (1994) In rr proteolysis of serum insulin-like growth factor (IGF) binding protein-3 results in in- creased availability of IGF to target' cells. J Clin Invest 93: 2286-2290

7. Hirschberg R, Kopple J, Blantz R, Tucker B (1991) Effects of recombinant human insulin-like growth factor I on glomerular dynamics in 'the rat. J Clin Invest 87: 1200= 1206

8. Hirschberg R, Kopple J (1995) IGF-I and renal function. Diabetes Rev 3:177-195

9. D'Ercole A, Underwood L (1987) Estimation Of :tissue con- centrations of somatomedin C/insulin-like groWth factor I. Meth- ods Enzymol 146i 227-233 ' .

10. Roberts C, Laskry ' S, Lowe W, Seaman W, LeRoith D (1987) Molecular ct0nit~g of rat insulin-like growth factor I com- plementary deoxyribonucleic acids: differential messenger ribo-

23.

nucleic acid processing and regulation by growth hormone in extrahepatic tissues. Mol Endocrinol 1:243-248

11. Chin E, Zhou J, Bondy C (1992) Anatomical relationships in the pattern of insulin-like growth factor (IGF)-I, IGF binding protein- 1 and IGF-1 receptor gene expression in the rat kidney. En- docrinology 130:3237-3245

12. Yokoya S, Suwa S, Maesaka H, Tanaka T (1988) Immunoreactive somatomedin C/insulin-like growth factor I in urine from normal subjects, pituitary dwarfs, and acromegalics. Pediatr Res 23: 151-154

13. Hizuka N, Takano K, Tanaka I, Asakawa K, Miyakawa M, Hor- ikawa R, Shizume K (I987) Demonstration of insulin-like growth factor I in human urine. J Clin Endocrinol Metab 64: 1309-1312

14. Quattrin T, Albini C, Cara J, Vandlen R, Mills B, MacGillivray M (1987) Quantitation of urinary somatomedin-C and growth hor- mone in preterm and fullterm infants and normal children. J Clin Endocrinol Metab 66:792-797

15. Gargosky S, Hasegawa T, Tapaniainen P, MacGillivray M, Hase- gawa Y, Rosenfeld R (1993) Urinary insulin-like growth factors (IGF) and IGF-binding proteins in normal subjects, growth hor- mone deficiency, and renal disease. J Clin Endocrinol Metab 76: 1631-1637

16. Zumkeller W, Hall K (1990) Immunoreactive insulin-like growth factor II in urine. Acta Endocrinol (Copenh) 123:499-503

17. Hirschberg R (1993) IGF-I is ultrafiltered into the urinary space and may exert biological effects in proximal tubules in the ne- phrotic syndrome. J Am Soc Nephrol 4:771

18. Hirschberg R, Kaysen G (1995) Insulin-like growth factor I and its binding proteins in the experimental nephrotic syndrome. En- docrinology 136: 1565-1571

19. Thabet M, Challa AS, Chan W, Liu F, Hintz R, Chan J (1994) Insulin-like growth factor and growth hormone in the nephrotic syndrome. Am J Physiol 266:E102-E106

20. Garin E, Grant M, Janet H, Silverstein M (1989) Insulinlike growth factors in patients with active nephrotic syndrome. Am J Dis Child 143:865-,867

21. Quigley R, Baum M (1991) Effects of growth hormone and in- sulin-like growth factor I on renal convoluted tubule transport. J Clin Invest 88:368-374

22. Hirschberg R, Ding H, Warmer C (1995) Effects of insulin-like growth factor I on phosphate transport in cultured proximal tubule cells. J Lab Clin Med 125, 428-434 Feld S, Hirschberg R, Artishevsky A, Nast C, Adler S (1995) Insulin-like growth factor i induces mesangial proliferation and increases mRNA and secretion of collagen. Kidney Int 48:45-51

L i t e r a t u r e a b s t r a c t

Epidemiol Infect (1995) 114:441-450

Verotoxinogenic Citrobacterfreundii associated with severe gastroenteritis and cases of haemolytic uraemic syndrome in a nursery school: green butter as the infection source

H. Tschiipe, R. Prager, W. Streckel, A. Fruth, E. Tietze, and G. Biihme

A summer outbreak of severe gastroenteritis followed by haemolytie uraemic syndrome (HUS) and thrombotic thrombocytopenic purpura in a nursery school and kindergarten is described. Sandwiches prepared with green butter made with contaminated parsley were the likely vehicle of infection. The parsley originated from an organic garden in

which manure of pig origin was used instead of artificial fertilizers. Clonally identical verotoxinogenic Critrobacterfreundii were found as causative agents of HUS and gastroenteritis and were also detected on the parsley.