Nestin as a Marker in the Classification of Adrenocortical Tumors
Authors
A. Lachenmayer1, U. D. Lichtenauer2, T. Cox3, M. Schott3, L. K. Malendowicz4, P. E. Goretzki5, K. Cupisti1, W. A. Scherbaum3, S. R. Bornstein6, H. S. Willenberg3
Affiliations
Affiliation addresses are listed at the end of the article
Key words
· fetal adrenal gland
· nestin
· tumor
· adrenal adenoma
@ adrenocortical carcinoma
Abstract
&
Expression of the intermediate filament, nestin, was long believed to be restricted to neuroecto- dermal stem cells. However, nestin expression has recently been detected in several tumors. Since adrenocortical carcinoma, a tumor entity still very difficult to classify, may gain the ability to aberrantly express neuroectodermal proteins including chromogranin A and synaptophysin, we asked the question whether nestin might also be detected in adrenocortical carcinomas, and if so, whether it might serve as a tool for clinical pathology. Therefore, we studied the expression of nestin in normal adrenal glands, adrenocorti-
cal adenomas, and adrenocortical cancers using specific immunohistochemistry and semi-quan- titative reverse transcriptase-polymerase chain reaction. Immunostaining was nestin-positive in 1 out of 9 normal adrenal glands (11%), 2 out of 20 adrenocortical adenomas (10%), and 13 out of 16 adrenocortical carcinomas (81%). Expression of nestin mRNA could be detected in all micro- dissected tissues, independently of their grade of dedifferentiation. We conclude that our findings provide further evidence that nestin, as a marker, is not restricted to neuronal stem cells and nestin expression is worth to be studied in adrenocorti- cal tumors.
received 27.03.2008 accepted 08.01.2009
Bibliography DOI 10.1055/s-0029-1202788 Published online: March 17, 2009 Horm Metab Res 2009; 41: 397-401 @ Georg Thieme Verlag KG Stuttgart . New York ISSN 0018-5043
Correspondence H. S. Willenberg, MD Department of Endocrinology Diabetes and Rheumatology University Hospital Düsseldorf Moorenstraße 5 40225 Düsseldorf Germany
Tel .: +49/211/811 78 10
Fax: +49/211/811 78 60
Holger.Willenberg@ uni-duesseldorf.de
Introduction & The fact that the mesoderm-derived adrenocor- tical cells may exhibit neuroendocrine properties is a process observed frequently, but not under- stood. However, expression of neuroendocrine markers such as synaptophysin and chromo- granin A by steroid-producing cells in adrenocor- tical carcinomas [1-5] and primary pigmented nodular adrenal disease in Carney’s complex [6,7] is interpreted as a step towards dedifferen- tiation. Together with other criteria such as mitotic index, sizes of nuclei, and invasion of ves- sels, expression of neuroendocrine markers is used to distinguish between benign and malig- nant adrenal neoplasia [8]. However, it is often not possible to make a safe diagnosis based on histological studies on the adrenal tumor. There- fore, the issue of more research for new and more specific markers was brought up in the 2002 con- sensus conference held by the National Institutes of Health, USA [9].
Intermediate filaments are a major component of the cellular cytoskeleton, and their cell-specific expression in normal tissues and differential
expression in tumors has been of paramount value in tumor diagnosis [10,11]. Nestin is a recently identified intermediate filament protein that belongs to the sixth class of intermediate filaments [12]. Previous studies have demon- strated that nestin is abundantly expressed in neuroectodermal stem cells, and is expressed by various cells in the central nervous system [13,14] and in skeletal muscle progenitor cells, but downregulated as cellular differentiation proceeds [13-16]. In addition, predominant nes- tin expression has been demonstrated in tumors thought to arise from immature cells such as primitive neuroectodermal tumors, medullo- blastomas, pediatric rhabdomyosarcomas, and gastrointestinal stromal tumors [17-23].
These findings suggest that detection of nestin in cells may be a useful molecular tool for identify- ing neuroectodermal stem cells and characteriz- ing tumors originating from immature cells such as stem or progenitor cells. In this study, we examined nestin expression in normal human adult and fetal adrenal glands, adrenocortical adenomas, and adrenocortical carcinomas (ACCs) in order to elucidate the potency of nestin in the
Downloaded by: University of Arizona Library. Copyrighted material.
assessment of adrenal tumor dignity. This study also contributes to the debate as to whether or not nestin is a true stem cell marker, or whether nestin can reappear in tumors originating from completely differentiated mature cells.
Materials and Methods
&
Materials
The material was obtained from an earlier study [24]. This pro- cedure has been approved by the Ethical Committee of the Med- ical School of Poznan, Poland in the course of another study [25]. Formalin-fixed tissues from human fetal adrenal glands between 13 and 27 weeks of gestation (n=6), normal human adrenal glands (n=9), adrenocortical adenomas (n = 20), and adreno- cortical carcinomas (n=16) were embedded in paraffin. Normal adrenal glands were harvested from patients undergoing surgi- cal kidney removal secondary to renal carcinoma; the adreno- cortical incidentalomas used in this study were surgically removed since they exceeded 4cm in diameter on computed tomography scans. Frozen tissue samples of normal human adrenal glands (NAGs) (n=8), non-hormone-producing adreno- cortical adenomas (NPAs) (n=5), and adrenocortical carcinomas (n=10) were used for mRNA studies, and cultured normal human adrenal cells served as a reference source and for control experiments.
The basis for the classification of the adrenocortical tumors as benign constituted the Weiss classification. Lesions with a score less than two, which showed no recurrence so far and are non- metastatic have been classified as benign. All ACC patients died from their disease and had metastasis. In addition, they had Weiss scores greater than two (Table 1).
The patients’ ages did not differ significantly. However, regarding tumor size and Weiss score, patients with ACCs had larger tumors and higher scores that reached significance (p<0.001 and p<0.001, respectively). For this statistical analysis we employed the GraphPad software Prism4 and for calculations we used ANOVA one-way analysis of variance and Bartlett’s test for equal variances and the unpaired Student’s t-test.
Immunohistology
A rabbit polyclonal antibody directed to the human corticotro- pin (ACTH) receptor (MC2R) (abcam, Cambride, UK) served to characterize adrenocortical cells by immunostaining in paraffin- embedded tissue sections and was used in dilutions of 1:50. We used a polyclonal goat antibody raised against human nestin (Chemicon, Cambridge, UK) for detection of nestin. This anti- body was used in a 1:200 dilution. Specificity of this antibody in
detecting the nestin protein was previously determined by Western blotting [26]. Detection of bound immunoglobulin was achieved by applying a biotinylated secondary swine anti-rabbit antibody (LSAB, DakoCytomation, Hamburg, Germany) for 30 min at room temperature, followed by incubation with per- oxidase-linked streptavidin-biotin complexes (DakoCytomation) for 30 min at room temperature. As a positive control, bound pri- mary antibodies were traced using a goat anti-rabbit (or alterna- tively anti-mouse) antibody- and peroxidase-linked polymer (EnVision, DakoCytomation) with an incubation time of 60 min at room temperature in a second alternative technique. 3-Amino- 9-ethylcarbazole (0.01 % in 0.0003 % H 2 O 2 ) and 0.5 M Tris-HCl buffer) were used for visualization. All sections were counter- stained with hematoxylin (Merck, Darmstadt, Germany). Expres- sion of immunoreactive protein was defined on a qualitative basis. Nestin staining was defined on a qualitative basis and judged to be positive when in minimum five adjacent cells of interest showed a clear staining. Staining of single scattered cells was not defined as being positive.
Laser microdissection, RNA isolation, and reverse- transcription polymerase chain-reaction (RT-PCR) Sections (8um) of all frozen adrenal tissues were fixed in 70% (v/v) ethanol, hematoxylin-stained and dehydrated in rising ethanol concentrations and, eventually, xylene. Adrenocortical cells were separated from chromaffin cells using an AS LMD microscope (Leica Microsystems, Wetzlar, Germany), as described previously [27]. Microdissected material was sub- jected to isolation of RNA using the RNeasy Micro Kit (Qiagen, Hilden, Germany), including a DNase I digestion step. The RNA obtained (1µg) was reversely transcribed to cDNA using the Ready-To-Go T-primed first strand kit (Amersham Biosciences AB, Uppsala, Sweden), skipping reverse transcriptase for nega- tive control reactions. We applied primers and probes for nestin for semiquantitative TaqMan-PCR amplification (forward primer: AGCCCTGACCACTCCAGTTTAG, reverse primer: CCCTCTATGGCT- GTTTCTTTCTCT, probe: TGCTGAACACTCTAGACCCACCGGATTCT), GAP-DH (forward primer: GAAGGTGAAGGTCGGAGTC, reverse primer: GAAGATGGTGATGGGATTTC, probe: CAAGCTCCCGTTCT- CAGCC), and 18S (forward primer: CGGCTACAACATCCAAGGAA, reverse primer: GCTGGAATTACCGCGGCT, probe: TGCTGGCAC- GAGACTTGCCCTC), the latter was used for normalization. PCR conditions were as follows: 40 thermal step cycles of denatura- tion at 95℃ for 15 seconds, and annealing/elongation at 60℃ for 45 seconds. Cultured normal human adrenal cells served as control and results are given in · Fig. 2 (vide infra) as times the basal expression in cultured human adrenal cells. All experi- ments were carried out in triplicates three times on two
| Carcinoma (n=16) | Adenoma (n=21) | Normal (n=9) | ||
|---|---|---|---|---|
| Age | years | 57.5 (+18.5/-25.5) | 52.3 (+20.7/-21.3) | 60.2 (+26.8/-24.2) |
| Sex | % female | ~60% | ~60% | ~80% |
| Size | cm | 9.4 (+9.6/-4.4) | 4.8 (+5.2/-2.8) | not recorded |
| Weiss-Score | Points | 4.8±1.0 | 0.8±1.0 | not recorded |
| Chromogranin A | Staining | ≥15% | 0% | medulla only |
| Synaptophysin | Staining | ~60% | 0% | medulla only |
| Vimentin | Staining | =66% | 0% | 0% |
aThe table summarizes the findings of the immunohistological studies on expression of chromogranin A, synaptophysin, and vimentin. The data are presented as mean (+maximum/-minimum) or as mean ± standard deviation
different ABI-Prism 7700 sequence detector machines. Semi- quantitative PCR results were analyzed by applying the compar- ative CT method to separate tubes as described earlier [25].
Results
&
Immunohistochemistry
Human adrenocortical cells could be identified using the anti- body directed against the human MC2R. All zones of the normal adrenal cortex could be labeled with this antibody with a pre- dominant staining in the zona reticularis and the zona glomeru- losa ( Fig. 1). In fetal adrenal glands, cells of both the fetal zone and the definitive zone were labeled with the anti-MC2R anti- body, whereby staining in the fetal zone was much more intense (· Fig. 1). Also, adrenocortical tumor cells were immunoreac- tive to the MC2R ( · Fig. 1). However, staining of different sec- tions of ACC specimen often revealed a very inhomogeneous staining pattern to the anti-MC2R antibody. Staining using LSAB or EnVision methods led to the same results. The staining of Ley- dig cells did not result in a detectable immunoreactive signal (· Fig. 1).
When assaying for immunoreactive nestin expression, we found positive immunostaining in sections of 1 out of 9 normal adrenal glands (11%), 2 out of 20 adrenocortical adenomas (9%), and 13 out of 16 ACCs (81%) ( Fig. 2). None of the 6 fetal adrenal glands (at 13 weeks of gestation) stained with the anti-nestin antibody ( Fig. 1, panel I), unlike fetal myocytes that also served as a positive control ( · Fig. 1, panel I).
For the separation of ACCs from adenomas by means of nestin staining, we found a sensitivity of 81.2% and a specificity 90% by this method with a positive predicitve value for ACC of 86.7% and a negative predicitive value of 85.7%. The rate of “false posi- tive adenomas” was 10%, the rate of “false negative ACCs” 19%.
mRNA expression studies
Expression of nestin mRNA could be detected in all tissues including in microdissected preparations of normal adrenal cor- tex, adrenocortical adenomas, adrenocortical carcinomas, and cultured normal adrenal cells. Expression of nestin mRNA was highest in preparations of normal adrenal cortices (419× basal expression in cultured human adrenal cells), but higher in ACCs than in NPAs (202 vs. 124x basal expression in cultured human adrenal cells, respectively). Though reproducible within the sample by independent investigators, levels of nestin mRNA expression varied widely between the tissues and between sam- ples of the same tissue and were therefore not correlated to the grade of adrenal gland (de-) differentiation ( · Fig. 2).
Discussion
&
So far, there is no reliable marker to distinguish between benign and malignant adrenocortical lesions to date [8]. This seems to be even more important for a pediatric patient population [28]. There may be also differences in adrenocortical tumorous tis- sues with respect to a sporadic or syndromatic appearance of the lesions [28,29]. Therefore, the issue of more research for new and more specific markers has been raised in a consensus con- ference at the National Institutes of Health, USA [9]. Since nestin is shown to be expressed in several tumors, and adrenocortical carcinomas may gain the ability to express neuroendocrine
A
2
ZG
B
ZF
>
2
100um
ZR
250m
C
D
>
>
zFet
zFet
ZD
>
ŻD
100um
8
200 pin
E
F
>
2
1000m
100km
G
H
2
2
5
50pm
50pm
I
>
K
LCH
500um
Downloaded by: University of Arizona Library. Copyrighted material.
A
B
n
relative expression of nestin mRNA in [x basal]
20
600
500
15
400
10
300
200
5
100
0
ACC
ACA
FAG
NAG
NAC
NPA
ACC
NACC in culture
markers, we studied the expression of nestin in human adrenal tissues. We found expression of nestin protein in 11 % of normal adrenal glands, 9% of adrenocortical adenomas, and in 81% of adrenocortical carcinomas. Of note, expression of nestin mRNA was highest in tissues as compared to cultured human adrenal cells, and expression of nestin mRNA in ACCs was higher than in hormonally silent adenomas. However, in contrast to the immu- nohistological data, expression of nestin mRNA was higher in tissue preparations of normal adrenal cortices than in ACCs. Whether this is due to posttranslational mechanisms, is not answered by our data or by others [26], and therefore, deserves further attention.
Although the adrenal cortex is not part of the diffuse neuro- endocrine system, normal and tumorous adrenocortical cells were shown to exhibit neuroendocrine properties [1-5,30]. According to one hypothesis, remnant cells from the fetal adre- nal cortex may be responsible for this phenomenon [3]. This could also explain why some adrenocortical tumors show expression of neuroendocrine markers while others do not. Thus, neuroendocrine properties were characteristic of adreno- cortical tumors that derive from fetal remnant cells through clonal expansion, but would not be detectable in adrenocortical tumors derived from adult cortical cells. Though we have no ultimate proof, our data do not support this assumption since nestin staining could not be observed in our series of human fetal adrenal glands covering the period between 7 and 27 weeks of gestation. This also means that nestin is not helpful in identi- fying remnant embryonal adrenal progenitor cells or adrenal stem cells.
However, our data suggest that expression of neuroendocrine markers by tumorous adrenocortical cells may be associated with the grade of dedifferentiation in the course of adrenal tumorigenesis. This would explain our observation that all but three carcinomas in our series but only one out of ten adreno- cortical adenomas expressed nestin. However, it does not explain why one normal human adrenal gland also expressed immuno- reactive nestin - a phenomenon reflected by identification of nestin mRNA in all adrenal tissues examined independent of their grade of dedifferentiation.
Inconsistent expression of nestin in rat adrenal glands has led to the hypothesis that the detection of nestin in adrenocortical and adrenomedullary cells depends on their different functional moments [26]. In our series of 9 normal human adrenal glands, nestin expression at protein level could not be observed in the adrenal medulla. In addition, normal human adrenocortical cells may also have neuroendocrine characteristics, namely the capa-
bility of expressing the neural cell adhesion molecule, NCAM [30].
In an earlier small study, expression of nestin was predominantly found in cells of the zona reticularis. In adrenocortical adenomas expression of nestin could be detected in only a few cells, and they were scattered around all areas of the tumors [31]. In contrast to our findings, adrenocortical carcinomas did not show constant expression of nestin. However, the expression of nestin protein in adrenocortical tumors was speculated to be of value in the dif- ferentiation of adrenocortical tumors from renal cell carcinomas that did not exhibit nestin-immunoreactivity [26].
Nestin may be a valuable marker in the characterization of adrenocortical tumors. Other markers have also been described, including the loss of expression of class II molecules of the major histocompatibility complex [32]. Some of them were later found to be of limited benefit. Therefore, evaluation of this antigen reactivity in larger samples and biopsy specimen will be required to further define sensitivity and specificity of nestin as an indi- cator for the dedifferentiation of adrenal cortical tumors. Also, our data support the notion of others that nestin is not specific for stem cells.
Acknowledgements
&
We would like to acknowledge the contribution by W. Wozniak and J. Bocian-Sobkowska from the Department of Anatomy, School of Medicine, Poznan, Poland. This work was supported by a Grant from the University of Düsseldorf to HSW.
Affiliations
1 Department of General, Visceral and Pediatric Surgery, University Hospital Düsseldorf, Düsseldorf, Germany
2 Division of Endocrinology, Medical Department II, University of Freiburg, Freiburg, Germany
3 Department of Endocrinology, Diabetes and Rheumatology, University Hospital Düsseldorf, Düsseldorf, Germany
4 Department of Histology and Embryology, School of Medicine Poznan, Poznan, Poland
5 Department of Surgery I, Lukas Hospital Neuss, Neuss, Germany
6 Department of Medicine III, Medical Faculty of the Technical University Dresden, Germany
References
1 Miettinen M. Neuroendocrine differentiation in adrenocortical carci- noma. New immunohistochemical findings supported by electron microscopy. Lab Invest 1992; 66: 169-174
2 Tartour E, Caillou B, Tenenbaum F, Schroder S, Luciani S, Talbot M, Schlumberger M. Immunohistochemical study of adrenocortical carci-
noma. Predictive value of the D11 monoclonal antibody. Cancer 1993; 72: 3296-3303
3 Haak HR, Fleuren GJ. Neuroendocrine differentiation of adrenocortical tumors. Cancer 1995; 75: 860-864
4 Komminoth P, Roth J, Schroder S, Saremaslani P, Heitz PU. Overlapping expression of immunohistochemical markers and synaptophysin mRNA in pheochromocytomas and adrenocortical carcinomas. Impli- cations for the differential diagnosis of adrenal gland tumors. Lab Invest 1995; 72: 424-431
5 Liu AM, Maeda S, Hosone M, Azuma K, Katayama H, Yokoyama M. Use of electron microscopic evaluation for the diagnosis of adrenal corti- cal carcinoma in fine needle aspiration cytology: a case report and review of the literature. Med Electron Microsc 2001; 34: 190-197
6 Stratakis CA, Carney JA, Kirschner LS, Willenberg HS, Brauer S, Ehrhart- Bornstein M, Bornstein SR. Synaptophysin immunoreactivity in pri- mary pigmented nodular adrenocortical disease: neuroendocrine properties of tumors associated with Carney complex. J Clin Endocri- nol Metab 1999; 84: 1122-1128
7 Bourdeau I, Lacroix A, Schurch W, Caron P, Antakly T, Stratakis CA. Primary pigmented nodular adrenocortical disease: paradoxical responses of cortisol secretion to dexamethasone occur in vitro and are associated with increased expression of the glucocorticoid recep- tor. J Clin Endocrinol Metab 2003; 88: 3931-3937
8 Weiss LM, Medeiros LJ, Vickery AL. Pathologic features of prognostic significance in adrenocortical carcinoma. Am J Surg Pathol 1989; 13: 202-206
9 Grumbach MM, Biller BM, Braunstein GD, Campbell KK, Carney JA, Godley PA, Harris EL, Lee JK, Oertel YC, Posner MC, Schlechte JA, Wieand HS. Management of the Clinically Inapparent Adrenal Mass (“Inciden- taloma”). Ann Intern Med 2003; 138: 424-430
10 Ho CL, Liem RK. Intermediate filaments in the nervous system: impli- cations in cancer. Cancer Metastasis Rev 1996; 15: 483-497
11 Prasad S, Soldatenkov VA, Srinivasarao G, Dritschilo A. Intermediate filament proteins during carcinogenesis and apoptosis. Int J Oncol 1999; 14: 563-570
12 Lendahl U, Zimmerman LB, Mackay RD. CNS stem cells express a new class of intermediate filament protein. Cell 1990; 60: 585-595
13 Zimmerman L, Parr B, Lendahl U, Cunningham M, Mackay R, Gavin B, Mann J, Vassileva G, MacMahon A. Independent regulatory elements in the nestin gene direct transgene expression to neural stem cells or muscle precursors. Neuron 1994; 12: 11-24
14 Dahlstrand J, Lardelli M, Lendahl U. Nestin mRNA expression correlates with the central nervous system progenitor cell state in many, but not all, regions of developing central nervous system. Brain Res Dev Brain Res 1995; 84: 109-129
15 Sejersen T, Lendahl U. Transient expression of the intermediate fila- ment nestin during skeletal muscle development. J Cell Sci 1993; 106: 1291-1300
16 Kachinsky AM, Dominov JA, Miller JB. Myogenesis and the intermediate filament protein, nestin. Dev Biol 1994; 165: 216-228
17 Dahlstrand J, Collins VP, Lendahl U. Expression of the class VI interme- diate filament nestin in human central nervous system tumors. Can- cer Res 1992; 52: 5334-5341
18 Tohyama T, Lee VM, Rorke LB, Marvin M, Mackay RD, Trojanowski JQ. Nestin expression in embryonic human neuroepithelium and in human neuroepithelial tumor cells. Lab Invest 1992; 66: 303-313
19 Tohyama T, Lee VM, Rorke LB, Marvin M, Mackay RD, Trojanowski JQ. Monoclonal antibodies to a rat nestin fusion protein recognize a 220- kDa polypeptide in subsets of fetal and adult human central nervous system neurons and in primitive neuroectodermal tumor cells. Am J Pathol 1993; 143: 258-268
20 Kobayashi M, Sjoberg G, Soderhall S, Lendahl U, Sandstedt B, Sejersen T. Pediatric rhabdomyosarcomas express the intermediate filament nestin. Pediatr Res 1998; 43: 386-392
21 Smits A, van Grieken D, Hartman M, Lendahl U, Funa K, Nister M. Coex- pression of platelet-derived growth factor alpha and beta receptors on medulloblastomas and other primitive neuroectodermal tumors is consistent with an immature stem cell and neuronal derivation. Lab Invest 1996; 74: 188-198
22 Tsujimura T, Makiishi-Shimobayashi C, Lundkvist J, Lendahl U, Nakasho K, Sugihara A, Iwasaki T, Mano M, Yamada N, Yamashita K, Toyosaka A, Terada N. Expression of the intermediate filament nestin in gas- trointestinal stromal tumors and interstitial cells of Cajal. Am J Pathol 2001; 158: 817-823
23 Valtz NL, Hayes TE, Norregaard T, Liu SM, Mackay RD. An embryonic origin for medulloblastoma. New Biol 1991; 3: 364-371
24 Ehrhart-Bornstein M, Breidert M, Guadanucci P, Wozniak W, Bocian- Sobkowska J, Malendowicz LK, Bornstein SR. 17 alpha-Hydroxylase and chromogranin A in 6th week human fetal adrenals. Horm Metab Res 1997; 29: 30-32
25 Haase M, Schott M, Bornstein SR, Malendowicz LK, Scherbaum WA, Willenberg HS. CITED2 is expressed in human adrenocortical cells and regulated by basic fibroblast growth factor. J Endocrinol 2007; 192: 459-465
26 Bertelli E, Regoli M, Lucattelli M, Bastianini A, Fonzi L. Nestin expression in rat adrenal gland. Histochem Cell Biol 2002; 117: 371-377
27 Willenberg HS, Walters R, Bornstein SR. Use of laser microdissection in complex tissue. Methods Enzymol 2002; 356: 216-223
28 Latronico AC, Almeida MQ. The molecular pathogenesis of childhood adrenocortical tumors. Horm Metab Res 2007; 39: 461-466
29 Stratakis CA. Adrenocortical tumors, primary pigmented adrenocorti- cal disease(PPNAD)/Carney complex, and other bilateral hyperplasias: the NIH studies. Horm Metab Res 2007; 39: 467-473
30 Ehrhart-Bornstein M, Hilbers U. Neuroendocrine properties of adreno- cortical cells. Horm Metab Res 1998; 30: 436-439
31 Toti P, Regoli M, Nesi G, Occhini R, Bartolommei S, Fonzi L, Bertelli E. Nestin expression in normal adrenal gland and adrenocortical tumors. Histol Histopathol 2005; 20: 1115-1120
32 Marx C, Wolkersdörfer GW, Brown JW, Scherbaum WA, Bornstein SR. MHC class II expression - a new tool to asses dignity in adrenocorti- cal tumors. J Clin Endocrinol Metab 1996; 81: 4488-4491