Clinical Science (1992) 83, 683-687 (Printed in Great Britain)
683
Glucosedphosphatase in normal adult human intestinal mucosa John PEARS', Roland T. JUNG', John JANKOWSKI', Ian
D. WADDELLZ and Ann BURCHELLl
'Department of Medicine, and 2Centre for Research into Human Development, Department of Obstetrics and Gynaecology and Department of Child Health, Ninewells Hospital and Medical School, Dundee. U.K. (Received 6 February/E July 1992; accepted 29 July 1992)
1. The existence of specific glucose-6-phosphatase activity in human intestinal mucosa has been somewhat controversial. 2. We have demonstrated the presence of low levels of specific glucose-6-phosphatase activity in normal human adult intestinal mucosa. Activity was found in oesophagus, stomach, duodenum and colon. 3. Immunoblot analysis using antibodies monospecific for the 36.5 kDa liver glucose-6-phosphatase catalytic subunit demonstrated that intestinal mucosa contains low levels of the glucose-6-phosphatase enzyme protein. 4. The low levels of activity together with problems of proteolysis make human intestinal biopsies unsuitable for use in the diagnosis of type 1 glycogen-storage disease.
which allow the substrates and products of the enzyme (glucose 6-phosphate, phosphate and glucose) to cross the endoplasmic reticulum membrane [3-51 (Fig. 1). A deficiency of any of the protein components of hepatic glucose-6-phosphatase will impair both hepatic glucose output and glucose-6phosphatase activity and result in type la, laSP, lb, lc or Id glycogen-storage disease [3-51. Recently, it has been realized that there is an association between type l b glycogen-storage disease and Crohn's-like colitis [9, 101, which suggests the possibility that there might also be a specific glucose-6phosphatase in intestinal mucosa. This possibility is further supported by a number of reports of histochemical detection of glucose-6phosphatase activity in the intestine of mouse and rat [l l-141 and reports of biochemical detection of low levels of glucose-6-phosphatase activity in the
INTRODUCTION
The liver plays an important role in the regulation of blood glucose homoeostasis [l, 21. Whenever blood glucose levels fall, the liver releases glucose into the bloodstream for use by other tissues which cannot make significant levels of glucose (e.g. brain) (Fig. 1). The two pathways by which glucose is made in the liver are glycogenolysis and gluconeogenesis (Fig. 1). The enzyme glucose-6phosphatase (EC 3.1.3.9) catalyses the terminal step of both pathways (Fig. 1) [l, 21. The importance of hepatic glucose-6-phosphatase in the maintenance of blood glucose levels first became obvious from the studies of patients with type 1 glycogen-storage disease (glucose-6-phosphatase deficiency or Von Gierke's disease) [3-51. The microsomal glucose-6phosphatase system has been unequivocally demonstrated to be present also in kidney [ 6 ] , pancreatic islets [7] and gall bladder [S], although its role in these tissues is not so clear. Glucose-6-phosphatase is multicomponent and consists of a catalytic subunit with its active site inside the lumen of the endoplasmic reticulum, a regulatory protein and three transport systems
I A
Blood
Fig. I. Schematic representation of the pathways of liver glucose production. The bold arrows indicate the pathways by which glucose is produced in the liver. Abbreviations: ER, endoplasmic reticulum; G-6-P, glucose &phosphate; G4-Pase. glucose-6-phosphatase; TI, microsomal glucose-6-phosphatetransportsystem; T2, microsomal phosphatetransport system; T3 (or G L U V ) , microsomal glucose-transport protein; GLUTI-5 plasma membrane facilitative glucose-transport proteins.
Key words: gIucose&phosphatase, intestinal mucosa. Abbreviations: Hepes, 4-(2-hydroxyethyl)-l-pipernin~thanesulphonic acid; PMSF, phenylmethanerulphonyl fluoride. Correspondence: Dr A. Burchell, Department of Obstetrics and Gynaecology, Ninewells Hospital and Medical School, Dundee D D I 9SY, U.K.
684
J. Pears et al.
intestine of rat, mouse, rabbit, guinea pig, chicken and man [15-201. Unfortunately, however, these reports all failed to unequivocally establish whether the activity measured was due to the presence of the microsomal glucose-6-phosphatase system or merely due to the non-specific hydrolysis of glucose 6phosphate by other non-specific phosphatases. In addition, there have also been contrasting reports that glucose-6-phosphatase is not present in rat intestine [19, 21, 221 or human intestine [22]. We have therefore used microassay techniques, together with monospecific antibodies to the catalytic subunit of liver microsomal glucose-6phosphatase [6], to determine unequivocally whether specific microsomal glucose-6-phosphatase is present in normal human intestinal mucosa. MATERIALS A N D METHODS Materials
The monosodium salt of glucose 6-phosphate was obtained from BDH Chemicals, Poole, Dorset, U.K. Phenylmethanesulphonyl fluoride (PMSF), trypsinchymotrypsin inhibitor, 1, 10-o-phenanthroline, pepstatin A, streptavidin/horseradish peroxidase complex, histone 2A and 4-chloro-1-naphthol were purchased from Sigma Chemicals, Poole, Dorset, U.K. The cocktail of protease inhibitors [containing 50 p g of antipain dihydrochloride/ml, 40 pg of PMSF/ml, 1Opg of aprotinin/ml, 40pg of bestatin/ml, 1OOpg of chymostatin/ml, 0.5 mg of E-64/ml, 0.5 mg of EDTA (di-sodium salt)/ml, 0.5 pg of leupeptin/ml, 0.7 pg of pepstatin/ml and 330 pg of phosphoramidon/ml] used to thaw the fast-frozen mucosal samples was obtained from Boehringer Mannheim, London, U.K. Nitrocellulose was purchased from Schleicher and Schuell, Dassel, Germany, and biotin-labelled secondary antibody was obtained from Amersham International, Amersham, Bucks, U.K. All other reagents were of Analar grade or better. Patients
Patients attending the Department of Clinical Measurement, Ninewells Hospital and Medical School, Dundee, for routine endoscopy were asked to give informed consent for four pinch biopsies of endoscopically normal mucosa to be taken from a single area of gut. No individual was sampled from more than one site or on more than one occasion. No patient with malignant disease of contraindication to biopsy was approached. No complications other than mild surface bleeding were observed. The study was approved by Tayside Health Board Committee on Medical Ethics. Samples
Pinch mucosal biopsies (wet weight 3 M O mg) were treated in one of two ways. Biopsies used for
measuring specific glucose-6-phosphatase activity were placed immediately into ice-cold 0.25 mol/l sucrose/5 mmol/l 4-(2-hydroxyethyl)- l-piperazineethanesulphonic acid (Hepes), pH 7.4 (SH buffer) containing 10mmol/l 1, 10-o-phenanthroline, 0.02% PMSF, 0.01 mg of trypsin/chymotrypsin inhibitor/ml and 0.1 mmol/l pepstatin A. The samples were quickly homogenized by hand in a Jencons glass/ teflon homogenizer in SH buffer containing protease inhibitors to a final volume of 10 times the original wet weight of tissue before being centrifuged for 1 min at 10 OOOg. The supernatant was separated immediately for rapid assay (see below). Biopsies for Western immunoblot analysis were wrapped in aluminium foil and frozen in liquid nitrogen immediately on being taken and were stored at -70°C. The samples were thawed in SH buffer containing the commercially available ‘cocktail’ of protease inhibitors described above and were homogenized rapidly by hand in a Jencons glass/teflon homogenizer to a final volume of 10 times the original frozen weight. Aliquots of homogenate were immediately boiled for 5min in sample buffer containing 2% (w/v) SDS and 1 mmol/l mercaptoethanol. Assays
Glucose-6-phosphatase activity was measured as previously described [23] using 30 mmol/l glucose 6phosphate as substrate. Specific glucose-6phosphatase activity was calculated as nmol of Pi produced min- mg - of protein. Non-specific phosphatase activity was measured by pretreatment of the sample of supernatant at pH 5 and 37°C for 15 min to inactivate the glucose6-phosphatase. The pH was returned to 6.5 with 0.4 mol/l Hepes before assay with 30 mmol/l glucose 6-phosphate as substrate as described above. The non-specific phosphatase assay is based on the observation that incubation under the conditions used completely destroys the ability of the specific glucose-6-phosphatase enzyme to hydrolyse glucose 6-phosphate without any loss of non-specific phosphatase activities (for a recent review, see [4]). Protein concentrations were measured by the method of Lowry et al. as modified by Peterson [24] using BSA as standard. SDS/PAGE and immunoblot analysis
Mucosal homogenates were subjected to electrophoresis on 7-16% SDS/polyacrylamide gels as described by Laemmli [25] to separate proteins. The separated proteins were then transferred on to nitrocellulose by the method of Towbin et al. [26] using an LKB semi-dry blot apparatus and were detected with sheep antibodies previously shown to be monospecific for the catalytic subunit of glucose6-phosphatase [6, 71. A biotin-streptavidin/ horseradish peroxidase detection system was used with 4-chloro-1-naphthol/H2O,as substrate.
685
Glucose6-phosphatase in normal intestine
t
1
2
3
4
5
6
7
8
9 Da
I80 000
Il6WO
84 000 58000 48 000
t I
36 500
' 26 600
0
0 Oesophagus Stomach Duodenum Ascending Descending colon colon
Fig. 2. Specific glucosdphosphatase activity measured in samples from differing parts of the human gastrointestinal tract
RESULTS
The specific glucose-6-phosphatase activities measured in mucosal samples from 20 individuals are shown in Fig. 2. The non-specific phosphatase activity measurements which were subtracted from the total phosphatase activities (see the Methods section) for each site were in the ranges: of protein; oesophagus, 0-30 nmol min- '-Yg-' stomach 0-130nmol min-' of protein; duodenum, 5-10 nmol min-' 'm ; of protein; ascending colon, (r6.2 nmol min- mg- of protein; descending colon, 0-100 nmol min- mgof protein. Median values were 0, 2.5, 6.9, 2.8 and 2.0 nmol min-l mg- of protein, respectively. The level of specific glucose-6-phosphatase activity found in the mucosal samples was variable. In all cases the level of activity in the gut mucosa was much lower than the average specific activity of 490 nmolmin-' mg- of microsomal protein previously reported for the glucose-6-phosphatase enzyme in control adult human liver [27], but it was in the same range as that previously found in human gall bladder mucosa [S]. Microsomes isolated from rat liver and human gut mucosa samples were subjected to immunoblot analysis (see Figs. 3 and 4) using a sheep antihepatic microsomal glucose-6-phosphatase antiserum previously shown to be monospecific for the 36 500 Da polypeptide which contains the active site of glucose-6-phosphatase [6, 71. Fig. 3 (lanes 1-8) and Fig. 4 (lane 4) show that oesophagus, stomach, duodenum and colon mucosa all contain the liverspecific glucose-6-phosphatase protein. However, the most common result with gut mucosa was the
'
'
Fig. 3. lmmunoblot analysis of adult human gastrointestinal tract. Lanes: I, oesophagus (5Opg); 2, stomach (50pg); 3, stomach (different subject) (50pg); 4, duodenum (5Opg); 5, ascending colon (5Opg); 6, transverse colon (50pg); 7, descending colon (5Opg); 8, sigmoid colon (50pg); 9, rat hepatic microsomes as control (9pg of protein).
I I
2
3
4
I
I
I
5 I
6
1
8
I
I
I Da I80000
'
I l6OOO 84000 58 OOO '
48000
*
36500
m26600
Fig. 4. lmmunoblot analysis of adult human gastrointestinal tract. Lanes: I, stomach (56pg); 2, stomach (64pg); 3, stomach (20pg); 4, duodenum (34pg); 5, ascending colon (54pg): 6, transverse colon (6pg); 7, descending colon (68pg); 8, sigmoid colon (46pg).
686
1. Pean et al.
appearance of smaller-molecular-mass immunoreactive proteolytic breakdown products of the glucose-6-phosphatase protein (Fig. 4, lanes 2, 3, 6, 7 and 8) or occasionally no immunoreactive peptides in badly degraded samples (e.g. Fig. 4, lanes 1 and 5) even in the presence of ‘cocktails’ of protease inhibitors. DISCUSSION
The role of the liver glucose-6-phosphatase enzyme in hepatic glucose production has been understood since the 1950s (for an early review, see [2]). In contrast, the role of glucose-6-phosphatase in other tissues (e.g. pancreatic islet /3-cells and intestinal mucosa) is much less clear. The importance of glucose production by the intestinal mucosa has not been well defined nor have differences in glucose metabolism in the various regions of the gut been clearly established. In intestinal mucosa, glutamine has been shown to be a quantitatively more important fuel than glucose [28]. Intestinal mucosa has been shown to contain several of the enzymes of gluconeogenesis [29, 301, and both gluconeogenesis and intestinal glucose production increase during starvation [17, 311. Glycogen levels in intestinal mucosa have also been shown to vary depending on the metabolic state of the animal studied [32], but the glycogenolytic enzymes of intestinal mucosa have not been extensively investigated. These studies suggested that glucose-6-phosphatase must be present in intestinal mucosa as glucose-6-phosphatase is the terminal step of gluconeogenesis (Fig. 1). Enzymic analysis of glucose-6-phosphatase activity in adult human intestinal biopsy samples showed that there are low levels of specific glucose6-phosphatase activity in all the areas of intestine that were tested (Fig. 2). However, the activities measured were very variable and unstable (in a time-dependent manner), as indeed were the nonspecific phosphatase levels that we measured. Immunoblot analysis (Fig. 3) confirmed that the 36.5 kDa specific glucose-6-phosphatase catalytic subunit is present in intestinal mucosa. It also revealed evidence of proteolysis in the majority of biopsies (Fig. 4) even when a variety of ‘cocktails’ of protease inhibitors were used to inhibit proteolysis. The proteolysis that occurs in human intestinal biopsy samples is the most likely explanation for the variable measurements that we obtained and for the controversy in the literature concerning whether glucose-6-phosphatase is present or absent in human intestinal mucosa, e.g. [17, 201. The low activity and proteolysis also make it difficult to carry out metabolic studies of glucose-6phosphatase in intestinal mucosa. The current methods of unequivocally diagnosing glucose-6-phosphatase deficiencies (glycogen-storage disease types la, laSP, lb, lc and Id) all rely on enzymic analysis of glucose-6-phosphatase activity
in liver biopsy samples (for recent reviews see [3-51). The presence of the specific glucose-6phosphatase system in normal human intestinal mucosa would potentially allow diagnosis of the type 1 glycogen-storage diseases using intestinal biopsies rather than the more invasive liver wedge or needle biopsies, and it has been suggested that it is possible to diagnose type 1 glycogen-storage disease using intestinal biopsies [19, 203. However, unfortunately, the high levels of glucose-6-phosphatase proteolysis and low levels of activity in human intestinal mucosa mean that intestinal biopsies cannot be used to unequivocally diagnose genetic deficiencies of the glucose-6phosphatase system because the most common cause of low or absent activity is proteolysis not glycogen-storage disease. ACKNOWLEDGMENTS
This work was funded by grants from the Scottish Home and Health Department and the Scottish Hospital Endowments Research Trust to A.B., and from Novo Nordisk to A.B. and R.T.J. and by equipment grants from the Research Trust for Metabolic Diseases in Children and the Anonymous Trust to A.B. and from Tenovus (Scotland) to A.B. and I.D.W. J.P. is a Novo Research Fellow. A.B. is a Lister Institute Research Fellow. We thank Dr G. Barclay for providing some of the biopsy material, and the staff of the Department of Clinical Measurement, Ninewells Hospital and Medical School, Dundee, for their co-operation.
REFERENCES I. Nordlie, R.C. Fine tuning of blood glucose concentrations. Trends Biochem. Sci. 1985; 10, 70-5. 2. Ashmore, J.& Weber, G. The role of hepatic glucos&phosphatase in the regulation of carbohydrate metabolism. In: Harris, R.S. & Marrion, G.F., eds. Vitamins and hormones. New York: Academic Press, 1959 91-132. 3. Burchell, A. The molecular pathology of glucore-bphosphatase. FASEB J. 19w; 4, 2 9 1 M . 4. Burchell, A. & Waddell, I.D. The molecular basis of the hepatic microromal glucose-bphosphatase system. Biochim. Biophys. Acta 1991; 1092, 129-37. 5. Burchell, A. The molecular basis of the type I glycogen storage diseases. Bioassays 1992; 14, 395-400. 6. Burchell, A., Waddell. I.D., Countaway, J.L., Arion, W.J. & Hume. R. Identification of the human hepatic microromal glucor&phosphatase enzyme. FEBS Lett. 1988; 242, 153-6. 7. Waddell. I.D. & Burchell. A. The microsomal glucor+horphatase enzyme of pancreatic islets. Biochem. J. 1988; 255, 471-6. 8. Hill, A., Waddell, I.D.. Hopwood, D. & Burchell, A. The microromal glucose-bphosphatase enzyme of human gall bladder. J. Pathol. 1989; 158, 53-6. 9. Roe, T.F., Thomas, D.W., Gilsanz. V., Isaacs, H. & Atkinson, 1.6. Inflammatory bowel disease in glycogen storage disease type I.J. Pediatr. 1986; 109, 55-9. 10. Sanderon, I.R., Bisset, W.M., Millar, P.J. & Leonard, J.V. Chronic inflammatory bowel disease in glycogen storage disease, Type Ib. J. Inher. Metab. Dis. 1991; 14, 771-6. I I. Hugon, J.S., Maestracci, D. & Menard, D. Glucore-bphosphatase activity in intestinal epithelium of the mouse. J. Histochem. Cytochem. 1971; 19, 515-25. 12. Hugon, IS.,Borgiers, M. & Maestracci, D. Glucose-bphosphatase and thiamine pyrophosphatase activities in the jeiunal epithelium of the mouse. J. Histochem. Cytochem. 1970; IS, 3614.
687
Glucore4phosphatase in normal intestine 13. Chabot, J.G., Menard, D. & Hugon, J.S. Organ culture of adult mouse intestine: IV. Stimulation of gIucose4phosphatase in vitro. Histochemistry 1978; 57,3345. 14. Calvert, D., Malkon, D. & Menard, D. Developmental pattern of glucosebphosphatse activity in the small intestine of the mouse fetus. Histochemistry 197% 63, 209-20. 15. Colilla, W., Jwgenson, R.A. & Nordlie, R.C. Mammalian carbamyl phosphate: glucose phosphotransferase and glucose-bphosphatase phosphohydrolase: extended tissue distribution. Biochim. Biophp. Acta 1971; 377, 117-25. 16. Strimntter, C.F. Phosphatate activities of chicken liver and duodenum: characteristics, level during development, and hydrocortisone induced changes. Biochim. Biophp. Aaa 1972.284, 183-95. 17. Anderson, J.W. Glucose metabolism in jejunal mucosa of fed, fasted and streptozotocindiabetic rats. Am. J. Physiol. 1974; 226, 226-9. 18. Lygre, D.G. & Nordlie, R.C. Phosphohydrolase and phosphotransferase activities of intestinal glucosebphosphatase. Biochemistry 1969; 7, 3219-26. 19. Ockerman, P.A. Glucosebphosphatasein human jejunal mucosa. Lack of activity in glycogenolysis of Cori's type 1. Clin. Chim. Acta 1964; 9, 151-6. 20. Williams, H.E., Johnson, P.L., Fenshaw, L.F., Lester, L. & Field, R.A. Intestinal glucose+hosphame in control subjects and relatives of a patient with glycogen storage disease. Metab. Clin. Exp. 12, 2354. 21. Hers, H.G. & DeDuve, C. Le systeme hexosephorphitasique. II. Repartion de I'activite glucosebphosphasiquedanr les tissues. Bull. Soc. Chim. Biol. 1950; 32, 20-29. 22. Ginsberg V. & Hers, H.C. On the conversion of fructose to glucose by guinea pig intestine. Biochim. Biophp. Acta 1960; 38, 427-31.
23. Burchell, A., Hume, R. & Burchell, B. A new microtechnique for the analysis of the human hepatic microsomal glucosebphosphacase system. Clin. Chim. Acta 1988; 173, 183-92. 24. Peterson, G.L. A simplification of the protein assay method of Lowry et 01. which is generally more applicable?Anal. Biochem. 63, W 5 6 . 25. Laemmli, U.K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 1970; 227, 680-5. 26. Towbin, H., Staehelin, T. & Gordon, J. Electrophoretic transfer of proteins from polyxrylamide gels to nitro-cellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. U.S.A. 197%76, 43504. 27. Burchell, A., lung, R.T., Lang, C.C., Bennet, W. & Shepherd, A.N. Diagnosis of type l a and type I c glycogen storage disease in adults. Lancet 1987; i, 105%2. 28. Windmueller, H.G. Metabolism of vascular and luminal glutamine by intestinal mucosa in viva. In: Hanssinger, E. & Sies, H., edr. Glutamine metabolism in mammalian tissues. Berlin: Springer-Verslag, 1981: 61-77. 29. Anderson, J.W. Pyruvate carboxylase and phosphoenolpyruvate carboxykinase in rat intestinal muma. Biochim. Biophys. Acta 1970 165-7. 30. Anderson, J.W. & Raaendall, A.F. Gluconeogenesis in jejunal mucosa of guinea pig. Biochim. Biophys. Acta 1973; 304,W-8. 31. Anderson, J.W. & Zakim, D. The influence of alloxandiabetes and fasting on glycolytic and gluconeogenic enzyme activities of rat intestinal mucosa and liver. Biochim. Biophys. Acta 1970; 201, 236-41. 32. Anderson, J.W. & Jones, A.L. Biochemical and ultrastructural study of glycogen in jejunal mucora of diabetic rat. Proc. Soc. Exp. Bml. Med. 1974; 145,260-72.
19n
a,
683
Glucosedphosphatase in normal adult human intestinal mucosa John PEARS', Roland T. JUNG', John JANKOWSKI', Ian
D. WADDELLZ and Ann BURCHELLl
'Department of Medicine, and 2Centre for Research into Human Development, Department of Obstetrics and Gynaecology and Department of Child Health, Ninewells Hospital and Medical School, Dundee. U.K. (Received 6 February/E July 1992; accepted 29 July 1992)
1. The existence of specific glucose-6-phosphatase activity in human intestinal mucosa has been somewhat controversial. 2. We have demonstrated the presence of low levels of specific glucose-6-phosphatase activity in normal human adult intestinal mucosa. Activity was found in oesophagus, stomach, duodenum and colon. 3. Immunoblot analysis using antibodies monospecific for the 36.5 kDa liver glucose-6-phosphatase catalytic subunit demonstrated that intestinal mucosa contains low levels of the glucose-6-phosphatase enzyme protein. 4. The low levels of activity together with problems of proteolysis make human intestinal biopsies unsuitable for use in the diagnosis of type 1 glycogen-storage disease.
which allow the substrates and products of the enzyme (glucose 6-phosphate, phosphate and glucose) to cross the endoplasmic reticulum membrane [3-51 (Fig. 1). A deficiency of any of the protein components of hepatic glucose-6-phosphatase will impair both hepatic glucose output and glucose-6phosphatase activity and result in type la, laSP, lb, lc or Id glycogen-storage disease [3-51. Recently, it has been realized that there is an association between type l b glycogen-storage disease and Crohn's-like colitis [9, 101, which suggests the possibility that there might also be a specific glucose-6phosphatase in intestinal mucosa. This possibility is further supported by a number of reports of histochemical detection of glucose-6phosphatase activity in the intestine of mouse and rat [l l-141 and reports of biochemical detection of low levels of glucose-6-phosphatase activity in the
INTRODUCTION
The liver plays an important role in the regulation of blood glucose homoeostasis [l, 21. Whenever blood glucose levels fall, the liver releases glucose into the bloodstream for use by other tissues which cannot make significant levels of glucose (e.g. brain) (Fig. 1). The two pathways by which glucose is made in the liver are glycogenolysis and gluconeogenesis (Fig. 1). The enzyme glucose-6phosphatase (EC 3.1.3.9) catalyses the terminal step of both pathways (Fig. 1) [l, 21. The importance of hepatic glucose-6-phosphatase in the maintenance of blood glucose levels first became obvious from the studies of patients with type 1 glycogen-storage disease (glucose-6-phosphatase deficiency or Von Gierke's disease) [3-51. The microsomal glucose-6phosphatase system has been unequivocally demonstrated to be present also in kidney [ 6 ] , pancreatic islets [7] and gall bladder [S], although its role in these tissues is not so clear. Glucose-6-phosphatase is multicomponent and consists of a catalytic subunit with its active site inside the lumen of the endoplasmic reticulum, a regulatory protein and three transport systems
I A
Blood
Fig. I. Schematic representation of the pathways of liver glucose production. The bold arrows indicate the pathways by which glucose is produced in the liver. Abbreviations: ER, endoplasmic reticulum; G-6-P, glucose &phosphate; G4-Pase. glucose-6-phosphatase; TI, microsomal glucose-6-phosphatetransportsystem; T2, microsomal phosphatetransport system; T3 (or G L U V ) , microsomal glucose-transport protein; GLUTI-5 plasma membrane facilitative glucose-transport proteins.
Key words: gIucose&phosphatase, intestinal mucosa. Abbreviations: Hepes, 4-(2-hydroxyethyl)-l-pipernin~thanesulphonic acid; PMSF, phenylmethanerulphonyl fluoride. Correspondence: Dr A. Burchell, Department of Obstetrics and Gynaecology, Ninewells Hospital and Medical School, Dundee D D I 9SY, U.K.
684
J. Pears et al.
intestine of rat, mouse, rabbit, guinea pig, chicken and man [15-201. Unfortunately, however, these reports all failed to unequivocally establish whether the activity measured was due to the presence of the microsomal glucose-6-phosphatase system or merely due to the non-specific hydrolysis of glucose 6phosphate by other non-specific phosphatases. In addition, there have also been contrasting reports that glucose-6-phosphatase is not present in rat intestine [19, 21, 221 or human intestine [22]. We have therefore used microassay techniques, together with monospecific antibodies to the catalytic subunit of liver microsomal glucose-6phosphatase [6], to determine unequivocally whether specific microsomal glucose-6-phosphatase is present in normal human intestinal mucosa. MATERIALS A N D METHODS Materials
The monosodium salt of glucose 6-phosphate was obtained from BDH Chemicals, Poole, Dorset, U.K. Phenylmethanesulphonyl fluoride (PMSF), trypsinchymotrypsin inhibitor, 1, 10-o-phenanthroline, pepstatin A, streptavidin/horseradish peroxidase complex, histone 2A and 4-chloro-1-naphthol were purchased from Sigma Chemicals, Poole, Dorset, U.K. The cocktail of protease inhibitors [containing 50 p g of antipain dihydrochloride/ml, 40 pg of PMSF/ml, 1Opg of aprotinin/ml, 40pg of bestatin/ml, 1OOpg of chymostatin/ml, 0.5 mg of E-64/ml, 0.5 mg of EDTA (di-sodium salt)/ml, 0.5 pg of leupeptin/ml, 0.7 pg of pepstatin/ml and 330 pg of phosphoramidon/ml] used to thaw the fast-frozen mucosal samples was obtained from Boehringer Mannheim, London, U.K. Nitrocellulose was purchased from Schleicher and Schuell, Dassel, Germany, and biotin-labelled secondary antibody was obtained from Amersham International, Amersham, Bucks, U.K. All other reagents were of Analar grade or better. Patients
Patients attending the Department of Clinical Measurement, Ninewells Hospital and Medical School, Dundee, for routine endoscopy were asked to give informed consent for four pinch biopsies of endoscopically normal mucosa to be taken from a single area of gut. No individual was sampled from more than one site or on more than one occasion. No patient with malignant disease of contraindication to biopsy was approached. No complications other than mild surface bleeding were observed. The study was approved by Tayside Health Board Committee on Medical Ethics. Samples
Pinch mucosal biopsies (wet weight 3 M O mg) were treated in one of two ways. Biopsies used for
measuring specific glucose-6-phosphatase activity were placed immediately into ice-cold 0.25 mol/l sucrose/5 mmol/l 4-(2-hydroxyethyl)- l-piperazineethanesulphonic acid (Hepes), pH 7.4 (SH buffer) containing 10mmol/l 1, 10-o-phenanthroline, 0.02% PMSF, 0.01 mg of trypsin/chymotrypsin inhibitor/ml and 0.1 mmol/l pepstatin A. The samples were quickly homogenized by hand in a Jencons glass/ teflon homogenizer in SH buffer containing protease inhibitors to a final volume of 10 times the original wet weight of tissue before being centrifuged for 1 min at 10 OOOg. The supernatant was separated immediately for rapid assay (see below). Biopsies for Western immunoblot analysis were wrapped in aluminium foil and frozen in liquid nitrogen immediately on being taken and were stored at -70°C. The samples were thawed in SH buffer containing the commercially available ‘cocktail’ of protease inhibitors described above and were homogenized rapidly by hand in a Jencons glass/teflon homogenizer to a final volume of 10 times the original frozen weight. Aliquots of homogenate were immediately boiled for 5min in sample buffer containing 2% (w/v) SDS and 1 mmol/l mercaptoethanol. Assays
Glucose-6-phosphatase activity was measured as previously described [23] using 30 mmol/l glucose 6phosphate as substrate. Specific glucose-6phosphatase activity was calculated as nmol of Pi produced min- mg - of protein. Non-specific phosphatase activity was measured by pretreatment of the sample of supernatant at pH 5 and 37°C for 15 min to inactivate the glucose6-phosphatase. The pH was returned to 6.5 with 0.4 mol/l Hepes before assay with 30 mmol/l glucose 6-phosphate as substrate as described above. The non-specific phosphatase assay is based on the observation that incubation under the conditions used completely destroys the ability of the specific glucose-6-phosphatase enzyme to hydrolyse glucose 6-phosphate without any loss of non-specific phosphatase activities (for a recent review, see [4]). Protein concentrations were measured by the method of Lowry et al. as modified by Peterson [24] using BSA as standard. SDS/PAGE and immunoblot analysis
Mucosal homogenates were subjected to electrophoresis on 7-16% SDS/polyacrylamide gels as described by Laemmli [25] to separate proteins. The separated proteins were then transferred on to nitrocellulose by the method of Towbin et al. [26] using an LKB semi-dry blot apparatus and were detected with sheep antibodies previously shown to be monospecific for the catalytic subunit of glucose6-phosphatase [6, 71. A biotin-streptavidin/ horseradish peroxidase detection system was used with 4-chloro-1-naphthol/H2O,as substrate.
685
Glucose6-phosphatase in normal intestine
t
1
2
3
4
5
6
7
8
9 Da
I80 000
Il6WO
84 000 58000 48 000
t I
36 500
' 26 600
0
0 Oesophagus Stomach Duodenum Ascending Descending colon colon
Fig. 2. Specific glucosdphosphatase activity measured in samples from differing parts of the human gastrointestinal tract
RESULTS
The specific glucose-6-phosphatase activities measured in mucosal samples from 20 individuals are shown in Fig. 2. The non-specific phosphatase activity measurements which were subtracted from the total phosphatase activities (see the Methods section) for each site were in the ranges: of protein; oesophagus, 0-30 nmol min- '-Yg-' stomach 0-130nmol min-' of protein; duodenum, 5-10 nmol min-' 'm ; of protein; ascending colon, (r6.2 nmol min- mg- of protein; descending colon, 0-100 nmol min- mgof protein. Median values were 0, 2.5, 6.9, 2.8 and 2.0 nmol min-l mg- of protein, respectively. The level of specific glucose-6-phosphatase activity found in the mucosal samples was variable. In all cases the level of activity in the gut mucosa was much lower than the average specific activity of 490 nmolmin-' mg- of microsomal protein previously reported for the glucose-6-phosphatase enzyme in control adult human liver [27], but it was in the same range as that previously found in human gall bladder mucosa [S]. Microsomes isolated from rat liver and human gut mucosa samples were subjected to immunoblot analysis (see Figs. 3 and 4) using a sheep antihepatic microsomal glucose-6-phosphatase antiserum previously shown to be monospecific for the 36 500 Da polypeptide which contains the active site of glucose-6-phosphatase [6, 71. Fig. 3 (lanes 1-8) and Fig. 4 (lane 4) show that oesophagus, stomach, duodenum and colon mucosa all contain the liverspecific glucose-6-phosphatase protein. However, the most common result with gut mucosa was the
'
'
Fig. 3. lmmunoblot analysis of adult human gastrointestinal tract. Lanes: I, oesophagus (5Opg); 2, stomach (50pg); 3, stomach (different subject) (50pg); 4, duodenum (5Opg); 5, ascending colon (5Opg); 6, transverse colon (50pg); 7, descending colon (5Opg); 8, sigmoid colon (50pg); 9, rat hepatic microsomes as control (9pg of protein).
I I
2
3
4
I
I
I
5 I
6
1
8
I
I
I Da I80000
'
I l6OOO 84000 58 OOO '
48000
*
36500
m26600
Fig. 4. lmmunoblot analysis of adult human gastrointestinal tract. Lanes: I, stomach (56pg); 2, stomach (64pg); 3, stomach (20pg); 4, duodenum (34pg); 5, ascending colon (54pg): 6, transverse colon (6pg); 7, descending colon (68pg); 8, sigmoid colon (46pg).
686
1. Pean et al.
appearance of smaller-molecular-mass immunoreactive proteolytic breakdown products of the glucose-6-phosphatase protein (Fig. 4, lanes 2, 3, 6, 7 and 8) or occasionally no immunoreactive peptides in badly degraded samples (e.g. Fig. 4, lanes 1 and 5) even in the presence of ‘cocktails’ of protease inhibitors. DISCUSSION
The role of the liver glucose-6-phosphatase enzyme in hepatic glucose production has been understood since the 1950s (for an early review, see [2]). In contrast, the role of glucose-6-phosphatase in other tissues (e.g. pancreatic islet /3-cells and intestinal mucosa) is much less clear. The importance of glucose production by the intestinal mucosa has not been well defined nor have differences in glucose metabolism in the various regions of the gut been clearly established. In intestinal mucosa, glutamine has been shown to be a quantitatively more important fuel than glucose [28]. Intestinal mucosa has been shown to contain several of the enzymes of gluconeogenesis [29, 301, and both gluconeogenesis and intestinal glucose production increase during starvation [17, 311. Glycogen levels in intestinal mucosa have also been shown to vary depending on the metabolic state of the animal studied [32], but the glycogenolytic enzymes of intestinal mucosa have not been extensively investigated. These studies suggested that glucose-6-phosphatase must be present in intestinal mucosa as glucose-6-phosphatase is the terminal step of gluconeogenesis (Fig. 1). Enzymic analysis of glucose-6-phosphatase activity in adult human intestinal biopsy samples showed that there are low levels of specific glucose6-phosphatase activity in all the areas of intestine that were tested (Fig. 2). However, the activities measured were very variable and unstable (in a time-dependent manner), as indeed were the nonspecific phosphatase levels that we measured. Immunoblot analysis (Fig. 3) confirmed that the 36.5 kDa specific glucose-6-phosphatase catalytic subunit is present in intestinal mucosa. It also revealed evidence of proteolysis in the majority of biopsies (Fig. 4) even when a variety of ‘cocktails’ of protease inhibitors were used to inhibit proteolysis. The proteolysis that occurs in human intestinal biopsy samples is the most likely explanation for the variable measurements that we obtained and for the controversy in the literature concerning whether glucose-6-phosphatase is present or absent in human intestinal mucosa, e.g. [17, 201. The low activity and proteolysis also make it difficult to carry out metabolic studies of glucose-6phosphatase in intestinal mucosa. The current methods of unequivocally diagnosing glucose-6-phosphatase deficiencies (glycogen-storage disease types la, laSP, lb, lc and Id) all rely on enzymic analysis of glucose-6-phosphatase activity
in liver biopsy samples (for recent reviews see [3-51). The presence of the specific glucose-6phosphatase system in normal human intestinal mucosa would potentially allow diagnosis of the type 1 glycogen-storage diseases using intestinal biopsies rather than the more invasive liver wedge or needle biopsies, and it has been suggested that it is possible to diagnose type 1 glycogen-storage disease using intestinal biopsies [19, 203. However, unfortunately, the high levels of glucose-6-phosphatase proteolysis and low levels of activity in human intestinal mucosa mean that intestinal biopsies cannot be used to unequivocally diagnose genetic deficiencies of the glucose-6phosphatase system because the most common cause of low or absent activity is proteolysis not glycogen-storage disease. ACKNOWLEDGMENTS
This work was funded by grants from the Scottish Home and Health Department and the Scottish Hospital Endowments Research Trust to A.B., and from Novo Nordisk to A.B. and R.T.J. and by equipment grants from the Research Trust for Metabolic Diseases in Children and the Anonymous Trust to A.B. and from Tenovus (Scotland) to A.B. and I.D.W. J.P. is a Novo Research Fellow. A.B. is a Lister Institute Research Fellow. We thank Dr G. Barclay for providing some of the biopsy material, and the staff of the Department of Clinical Measurement, Ninewells Hospital and Medical School, Dundee, for their co-operation.
REFERENCES I. Nordlie, R.C. Fine tuning of blood glucose concentrations. Trends Biochem. Sci. 1985; 10, 70-5. 2. Ashmore, J.& Weber, G. The role of hepatic glucos&phosphatase in the regulation of carbohydrate metabolism. In: Harris, R.S. & Marrion, G.F., eds. Vitamins and hormones. New York: Academic Press, 1959 91-132. 3. Burchell, A. The molecular pathology of glucore-bphosphatase. FASEB J. 19w; 4, 2 9 1 M . 4. Burchell, A. & Waddell, I.D. The molecular basis of the hepatic microromal glucose-bphosphatase system. Biochim. Biophys. Acta 1991; 1092, 129-37. 5. Burchell, A. The molecular basis of the type I glycogen storage diseases. Bioassays 1992; 14, 395-400. 6. Burchell, A., Waddell. I.D., Countaway, J.L., Arion, W.J. & Hume. R. Identification of the human hepatic microromal glucor&phosphatase enzyme. FEBS Lett. 1988; 242, 153-6. 7. Waddell. I.D. & Burchell. A. The microsomal glucor+horphatase enzyme of pancreatic islets. Biochem. J. 1988; 255, 471-6. 8. Hill, A., Waddell, I.D.. Hopwood, D. & Burchell, A. The microromal glucose-bphosphatase enzyme of human gall bladder. J. Pathol. 1989; 158, 53-6. 9. Roe, T.F., Thomas, D.W., Gilsanz. V., Isaacs, H. & Atkinson, 1.6. Inflammatory bowel disease in glycogen storage disease type I.J. Pediatr. 1986; 109, 55-9. 10. Sanderon, I.R., Bisset, W.M., Millar, P.J. & Leonard, J.V. Chronic inflammatory bowel disease in glycogen storage disease, Type Ib. J. Inher. Metab. Dis. 1991; 14, 771-6. I I. Hugon, J.S., Maestracci, D. & Menard, D. Glucore-bphosphatase activity in intestinal epithelium of the mouse. J. Histochem. Cytochem. 1971; 19, 515-25. 12. Hugon, IS.,Borgiers, M. & Maestracci, D. Glucose-bphosphatase and thiamine pyrophosphatase activities in the jeiunal epithelium of the mouse. J. Histochem. Cytochem. 1970; IS, 3614.
687
Glucore4phosphatase in normal intestine 13. Chabot, J.G., Menard, D. & Hugon, J.S. Organ culture of adult mouse intestine: IV. Stimulation of gIucose4phosphatase in vitro. Histochemistry 1978; 57,3345. 14. Calvert, D., Malkon, D. & Menard, D. Developmental pattern of glucosebphosphatse activity in the small intestine of the mouse fetus. Histochemistry 197% 63, 209-20. 15. Colilla, W., Jwgenson, R.A. & Nordlie, R.C. Mammalian carbamyl phosphate: glucose phosphotransferase and glucose-bphosphatase phosphohydrolase: extended tissue distribution. Biochim. Biophp. Acta 1971; 377, 117-25. 16. Strimntter, C.F. Phosphatate activities of chicken liver and duodenum: characteristics, level during development, and hydrocortisone induced changes. Biochim. Biophp. Aaa 1972.284, 183-95. 17. Anderson, J.W. Glucose metabolism in jejunal mucosa of fed, fasted and streptozotocindiabetic rats. Am. J. Physiol. 1974; 226, 226-9. 18. Lygre, D.G. & Nordlie, R.C. Phosphohydrolase and phosphotransferase activities of intestinal glucosebphosphatase. Biochemistry 1969; 7, 3219-26. 19. Ockerman, P.A. Glucosebphosphatasein human jejunal mucosa. Lack of activity in glycogenolysis of Cori's type 1. Clin. Chim. Acta 1964; 9, 151-6. 20. Williams, H.E., Johnson, P.L., Fenshaw, L.F., Lester, L. & Field, R.A. Intestinal glucose+hosphame in control subjects and relatives of a patient with glycogen storage disease. Metab. Clin. Exp. 12, 2354. 21. Hers, H.G. & DeDuve, C. Le systeme hexosephorphitasique. II. Repartion de I'activite glucosebphosphasiquedanr les tissues. Bull. Soc. Chim. Biol. 1950; 32, 20-29. 22. Ginsberg V. & Hers, H.C. On the conversion of fructose to glucose by guinea pig intestine. Biochim. Biophp. Acta 1960; 38, 427-31.
23. Burchell, A., Hume, R. & Burchell, B. A new microtechnique for the analysis of the human hepatic microsomal glucosebphosphacase system. Clin. Chim. Acta 1988; 173, 183-92. 24. Peterson, G.L. A simplification of the protein assay method of Lowry et 01. which is generally more applicable?Anal. Biochem. 63, W 5 6 . 25. Laemmli, U.K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 1970; 227, 680-5. 26. Towbin, H., Staehelin, T. & Gordon, J. Electrophoretic transfer of proteins from polyxrylamide gels to nitro-cellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. U.S.A. 197%76, 43504. 27. Burchell, A., lung, R.T., Lang, C.C., Bennet, W. & Shepherd, A.N. Diagnosis of type l a and type I c glycogen storage disease in adults. Lancet 1987; i, 105%2. 28. Windmueller, H.G. Metabolism of vascular and luminal glutamine by intestinal mucosa in viva. In: Hanssinger, E. & Sies, H., edr. Glutamine metabolism in mammalian tissues. Berlin: Springer-Verslag, 1981: 61-77. 29. Anderson, J.W. Pyruvate carboxylase and phosphoenolpyruvate carboxykinase in rat intestinal muma. Biochim. Biophys. Acta 1970 165-7. 30. Anderson, J.W. & Raaendall, A.F. Gluconeogenesis in jejunal mucosa of guinea pig. Biochim. Biophys. Acta 1973; 304,W-8. 31. Anderson, J.W. & Zakim, D. The influence of alloxandiabetes and fasting on glycolytic and gluconeogenic enzyme activities of rat intestinal mucosa and liver. Biochim. Biophys. Acta 1970; 201, 236-41. 32. Anderson, J.W. & Jones, A.L. Biochemical and ultrastructural study of glycogen in jejunal mucora of diabetic rat. Proc. Soc. Exp. Bml. Med. 1974; 145,260-72.
19n
a,
