Alkaline phosphatase activity in chronic streptozotocin-induced insulin deficiency in the rat: Effect of insulin replacement

Alkaline Phosphatase Insulin Deficiency Stephen Alterations insulin

diabetic

phosphatase following animals alkaline

l-lough, Louis V. Avioli, Steven L. Teitelbaum,

in circulating alkaline phosphatase

deficiency.

induced

Activity in Chronic Streptozotocin-Induced in the Rat: Effect of Insulin Replacement

We and

activity

insulin

evaluated

diabetic

was

elevated

markedly

administration.

phosphatase

insulin therapy

replacement. phosphatase

rats,

alkaline seven

in the

elevated

plasma

activity

was

significantly

higher

for 36 hr prior to sacrifice resulted to those observed

activity was decreased Neither

after

0.01)

of diabetes.

(p KY0.001) activity

and

diabetic

streptorotocin-

Circulating

completely

observed

of the intestinal

in the

animal with chronic

control,

in the

insulin

isoenzyme.

animals,

but

in an abrupt

in the insulin-deficient

rise in both plasma and intestinal

state. In contrast

nor insulin replacement

resulted

in any significant

comparable

in the

withholding

alkaline phosphatase skeletal

was corrected

alkaline

by insulin

changes in the hepatic alkaline

isoenzyme.

D

IABETES MELLITUS in both man and the experimental animal may be associated with high circulating levels of alkaline phosphatase.‘-8 This is not surprising as diabetic complications may directly or indirectly involve liver, bone and intestine; tissues which contribute to blood levels of this enzyme. However, the pathophysiology, tissue origin and influence of insulin replacement on altered serum and tissue alkaline phosphatase levels in diabetes still remain ill defined. This study was therefore undertaken to examine, in greater detail, the nature of the alterations in circulating alkaline phosphatase which is characteristic of the experimentally induced chronic insulin deficient state. MATERIALS

AND

tory chow (Ca

water.

1.2% P 0.8%, Fat S.O’%) and had free access to tap

All animals were regularly

METHODS

urine volumes were measured in 8 animals from each group. Sixteen diabetic rats received NPH-insulin tion of the study. Sixteen on the forty-eighth

housed in stainless steel hanging metabolic cages, fed rodent labora-

and fourteen

control animals

Insulin was withheld on the day prior to the

rest of the group

received

diabetic rats. while the

insulin approximately

8 hr prior

to

sacrifice. Animals were killed the next morning (0900-l guination

from the abdominal

samples from all animals were analyzed um,”

phosphate”

100 hr) by exsan-

aorta under ether anesthesia.

Blood

for plasma glucose,’ calci-

and magnesium. ” Plasma alkaline

phosphatase

by p-nikro-phenyl

phosphate hydrolysis at pH 10.2.

From each animal, approximately

2 g of liver and the entire small

bowel (approximately

8 g) were rinsed in saline and homogenized in

a Polytron homogenizer

1 cm

(Beckmann)

with 3 ml of cold saline. The

portion of the femurs from each animal were dissected

free, split longitudinally, porcelain mortar

washed with a stream of cold saline to

minced with bone scissors and ground in a

with washed sand and cold saline. Tissue homo-

genates were allowed to stand overnight at 4°C. centrifuged

(3000

in the supernatant

200 ~1 of diluted supernatant mixture

(pH _ 10.3,

propranol mM

(Sigma),

MgCI,). nM,

4.0 mM

0.84

plotted

absorbance was calculated graphic

M

2-amino-2-methyl-l-

para-nitrophenylphosphate,

The reaction medium was maintained and

representation.”

phosphatase

by a kinetic method.”

was added to 5.8 ml buffered reaction

containing

absorbance was measured manually at 404

then shaken and

x g for 45 minutes) at 4OC. Alkaline

activity was determined

From the Division of Bone and Mineral Metabolism, Department of Medicine, The Jewish Hospital of St. Louis, Washington Vniversity School of Medicine. St. Louis. Missouri, and the Department of Pathology and Laboratory Medicine, The Jewish Hospital of St. Louis. Washington University School of Medicine. St. Louis, Missouri. Supported in part by National Institutes of Health Grants AMI 1674 and AM2052I; The Saint Louis Shriners Hospital for Crippled Children and a post-doctoral South African Medical Research Council grant to Stephen Hough. Received for publication December 24. 1980. Address reprint requests to Michael D. Fallon, M.D., Department of Pathology, The Jewish Hospital of St. Louis, 216 South Kings Highway, St. Louis. Missouri 631 IO. 0 1981 by Grune & Stratton, Inc. 0026-0495/81/3012-0008801.00/0

SC.)

dilution fluid only. Before sacrifice

overnight fast in eight of the insulin-treated

remove the marrow,

were individually

units/day.

until the comple-

day, all animals were fasted for 24 hr, but had

free access to water.

normal

rats. Animals

diabetic

received daily injections of NPH

distal

freely-fed

(3-5

from the fifth day after injection of streptozotocin

Wistar-Lewis rats, weighing approximately 300 g, by the intravenous injection, of 65 mg/kg streptozotocin (Upjohn Co., Kalamazoo, Michigan) freshly dissolved in citrate buffer (pH 4.5). A control group consisted of sham-injected,

weighed and, during the three

days prior to sacrifice, total food and water consumption and 24 hr

was determined

Experimental diabetes was induced in male

1190

deficient

Small intestinal

to these observations,

0.01) and this abnormality

alkaline

normalized

activity was found to be strikingly insulin-sensitive;

in the insulin deficient animal (p i

insulin deficiency

animal

typical

in freely-fed

induction

phosphatase

a pattern (p i

levels

the

deficient

alkaline

and phenylalanine-sensitive,

D. Fallon

in both man and the experimental

phosphatase

weeks

insulin

and control rats. The intestinal isoenzyme

values comparable phosphatase

The

have been described

and tissue

insulin-treated

was heat-resistant

insulin-replaced

plasma

and Michael

every 30-60

versus time.

The

and 0.5

at 37OC while

set for 5 minutes. rate

of change

of

for a three minute linear segment of the Assays were always performed

in tripli-

inhibition of plasma and tissue alkaline

phospha-

cate. Phenylalanine

tase activity was performed by incorporating the reaction media. Heat inactivation was carried out at 56’?

5 mM L-phenylaline

in

of plasma and tissue samples

for 10 min. Protein was determined

in tissue

homogenate supernatants by the method of Lowry.14 Statistical

assessment of the data was made using analysis of

variance and Scheffe’s multiple comparison test of the meansI

Merabolism,

Vol. 30, No. 12 (December). 1991

1191

ALKALINE PHDSPHATASE AND DIABETES

Table 2. Plasma Values of Glucose, Calcium Ka), Phosphate (Pi)

RESULTS

Untreated diabetic animals lost weight during the seven week study period, while insulin treated rats gained weight (Table 1). Even though the diabetic animals failed to grow, their food consumption was 1% times that of the control animals; there was a lo-fold increase in urine output and a water intake which approached 6 times that of the control animals. Insulin treatment markedly reduced the water intake and daily urine output, but only moderately altered food consumption (Table 1). As noted in Table 2, untreated diabetic animals were markedly hyperglycemic. Significant hypercalcemia and hyperphosphatemia were also noted in the diabetic animals, although plasma magnesium levels were comparable to those observed in the control animals. Plasma Alkaline Phosphatase Activity

Plasma alkaline phosphatase activity was strikingly elevated in the untreated diabetic rats and completely normalized following insulin administration (Fig. 1A). The elevated plasma alkaline phosphatase activity observed in the insulin deficient animals was heatresistant and phenylalanine-sensitive, a pattern typical of the intestinal isoenzyme (Tables 3 and 4). Enzyme activity in the plasma of both control and insulintreated diabetic rats, however, was heat-sensitive and phenylalanine-resistant; a pattern consistent with the properties of the skeletal isoenzyme. Withholding insulin for approximately 36 hours resulted in plasma values of heat-resistant, phenylalanine-sensitive alkaline phosphatase activity that were comparable to that observed in the insulin deficient state (Table 4). Tissue Alkaline Phosphatase Activity

Small intestinal alkaline phosphatase activity was significantly higher (p < 0.01) in the diabetic animals, but comparable in the insulin-treated and Table 1.

BodyWeights,

Food and Water Consumption and Urine

Output in Control, Diabetic and Insulin Treated Animals 6odv\NT (Q) Initial

Final

Food Int&e gf24 hr

water Intake ml/24 hr

Urine Volume ml/24 hr

3Ok

13 + 1

COlttd (n-8)

312

* 8

438

+ 12

23.8

+0.6

1

Diabetic In - 8)

313 f 6

283 + 7*

37.0

f 1.4*

165 f 7*

135 k 6.

307 * 7

378

30.5

f 1.6t

67 f 8*

47 * 8*

Insulin Treated fn - 8)

f 8*

Data ere presented as mean f SEM. *Significantly different TSignificantly different

( p -c Cr.00 1) from ( p -c 0.005) from

control animals. control animals.

and Magnesium (Mg) in Control, Diabetic and Insulin Treated Animals

me/d1

rnQ/d

Pi mg/dl

147 f 6

9.7 kO.1

5.7 f 0.1

1.5 * 0.05

529 + 26’

10.2 f O.lT

6.4 * 0.2T

1.5 f 0.20

11

9.6 f 0.1

6.5 k 0.2T

1.4 + 0.04

521 +_31.

9.7 f 0.2

6.5 f 0.2T

1.4 f 0.04

Glucose

Ca

MQ mQ/dl

control bl -

12)

Diabetic (n -

16)

Treated (n - 8)

130*

Insulin$ Withheld (n = 81

Data are presented as mean f SEM. *Significantly different 1 p < 0.00 1) from control animals. TSignificantly different ( p < 0.05) from control animals. $lnsulin withheld for 36 hr prior to sacrifice.

control rats (Fig. 1B). However, withdrawing insulin therapy 36 hr prior to sacrifice resulted in an abrupt rise in alkaline phosphatase activity to levels observed in the insulin-deficient state (Fig. 1B). Bone alkaline phosphatase activity was decreased in the diabetic rats and this abnormality was corrected by insulin treatment (Fig. 1C). Withholding insulin 36 hr prior to sacrifice did not alter enzyme activity. As noted in Fig. lD, neither insulin deficiency nor insulin replacement resulted in any significant changes in the hepatic alkaline phosphatase isoenzyme. DISCUSSION

The function(s) of alkaline phosphatase remain poorly understood. Osteoblasts are rich in an alkaline phosphatase isoenzyme which has been implicated in calcification mechanisms.‘“” In the intestine, bile passages and kidney tubules, the distribution of alkaline phosphatase in surface epithelium strongly suggests a functional role in transport mechanisms, and the transcellular movement of calcium,*~** phosphate,’ water’ and lipids23 have been linked to the enzyme. In humans, the majority of the circulating alkaline phosphatase activity is derived from hepatic and skeletal sources with only a minor intestinal component, in contrast to the normal rat where the main contribution to the circulating alkaline phosphatase activity is from the intestine. Elevated levels of serum alkaline phosphatase have been reported in both human and experimental diabetes mellitus.‘” Hyperphosphatasemia occurs in 7% 44% of cases in clinical diabetes mellitus and has generally been assumed to result from associated liver disease.‘* However, in a study of more than 300 diabetics, no relationship between the elevation in alkaline phosphatase and the duration, complications

HOUGH ET AL.

1192

I

Plosmo

1000

--____.

c

900

_--____

ntestine

p<.OOl’ 800 < 700

p<.Ol’

p<.Ol' -------

L

z4

2

k

~

3 r 3

-2

1

B

A

0.7

6C

iver

0.6

50 z

NS

0.5

z E n. e 3

0.4

40

NS

g

30

g

0.3

__-----

____--p<.Ol.

F . 2 20

0.2 10 0.1

C

D

Fig. 1. (A-D) Plasma and tissue alkaline phosphatase activity. A = Plasma alkaline phosphatase activity, B -i Intestinal homogenate alkaline phosphatase activity, C = Bone homogenate alkaline phosphatase activity, D = Liver homogenate alkaline phosphatase activity, For Fig. 1A through 1D: C = control, D = diabetic, I = insulin treated and I- = insulin withheld for 36 hours prior to sacrifice. Bars indicate mean + SEM for each group of 8-16 animals. (‘1 = significantly different from control.

(including hepatic involvement) or treatment of the disease could be documented. It was suggested that elevated serum alkaline phosphatase in diabetes mellitus represented an “intrinsic feature of the diabetic Animal studies have demonstrated a condition.“4 progressive rise, over several weeks, in serum alkaline

phosphatase levels after alloxan injection.’ Furthermore, a 70% increase in both intestinal and hepatic. without any appreciable change in skeletal alkaline phosphatase has been shown to attend alloxan diabetes in the rat.’ The present study revealed a striking elevation in

1193

ALKALINE PHOSPHATASE AND DIABETES

Table 3. Effect of Heat and Phenylalanine on Tissue

Table 4. Effect of Heat and Phenylalanine on Plasma From

Homogenates From Normal Rats*

Control, Diabetic and Insulin Treated Animals*

Heat (56°C

Intestine Bone Liver

92.8

x 10 mid

+ 3.7

4.3 + 0.4 62.2

f 14.8

L-Phenylalanina (5

Heat

mM)

(56°C

x 10 mid

L-Phenylalanine

(5 mMl

34.6

* 4.1

(n = 131

Control

17.7 f 2.3

62.3

+ 5.2

(n = 6)

84.8

+ 8.6

(n = 14)

Diabetic

84.4

+ 3.2

31.6

f 7.1

(n = 6)

+ 10.2

(n = 121

Insulin18.9 + 4.6

55.5

f 2.8

(n = 6)

79.8 f 6.2

29.2

+ 4.4

fn = 6)

67.7

*The data are expressed as mean percent residual activity + 1 SD.

Treated InsulinWithheldt

plasma alkaline phosphatase (hyperphosphatasemia) in longstanding experimental diabetes. Insulin replacement was attended by the complete normalization of hyperphosphatasemia. The results of heatdenaturation and L-phenylalanine inhibitor studies suggested that the hyperphosphatasemia was of intestinal origin; this proved to be the case by direct analysis of tissue alkaline phosphatase content. The intestinal alkaline phosphatase activity was found to be strikingly insulin-sensitive. Withholding insulin for thirty-six hours resulted in plasma and intestinal tissue values of alkaline phosphatase activity comparable to those observed in the insulin-deficient state. This rapid return of alkaline phosphatase activity of pretreatment levels may reflect the short half-life of rat intestinal alkaline phosphatase.24 In contrast to a previous observation’ in alloxan diabetic animals, hepatic alkaline phosphatase was found to be similar in diabetic and control rats and uninfluenced by insulin therapy. Whether this discrepancy relates to a difference in animal strain or to the fact that we used streptozotocin rather than alloxan to induce insulin deficiency is not readily apparent at this time. Skeletal alkaline phosphatase is generally accepted as a parameter of bone formation and purely resorptive processes, such as multiple myeloma, generally do not alter alkaline phosphatase activity.” Our chronically diabetic rats were characterized by a significant reduction in skeletal alkaline phosphatase; this observation may relate to previous reports of decreased bone formation and turnover in both experimental” and human*‘j diabetes. We have previously shown that insulin therapy results in a marked stimulation of bone turnover in experimental diabetes” and this may relate to the observed insulin-mediated increase in skeletal alkaline phosphatase. The present study documents a marked increment in the intestinal alkaline phosphatase isoenzyme activity in chronically diabetic rats, which accounts for the striking hyperphosphatasemia that characterizes this disease. To date, the exact cause(s) of the enhanced intestinal enzyme activity remains to be resolved. The close correlation between intestinal calcium accumulation and enhanced gut alkaline phosphatase activity supports the view that the enzyme is concerned with calcium absorption.2S22 Recently, the temporal rela-

*The data are expressed as mean percent residual activity f 1 SD. tlnsulin withheld for 36 hr prior to sacrifice.

tionship between duodenal calcium accumulation and the increase in alkaline phosphtase activity after vitamin D administration has, however, cast some doubt as to a direct cause-effect correlation between these parameters.** We have previously demonstrated that intestinal calcium absorption is significantly augmented in the rat with chronic insulin deficiency.*’ This observation could relate to the fact that experimental diabetes is characterized by intestinal hypertrophy, increased mucosal cell proliferation and stimulation of several membrane transport systems.27-29 Long-term insulin replacement corrects the hyperabsorption of calcium in chronically diabetic animals” and, in marked contrast to our observation of intestinal alkaline phosphatase activity, this occurred despite withholding insulin for 36 hr. Thus, it is apparent that whereas the enhanced intestinal enzyme activity may represent an epiphenomenon of augmented calcium absorption that occurs in rats with chronic insulin deficiency, a direct cause and effect correlation appears unlikely. It is not inconceivable that the enhanced intestinal alkaline phosphatase activity, observed in the chronic insulin deficient animal, results from the intestinal adaptations that attend this disease. The possibility that intestinal alterations in experimental diabetes could, however, result from a direct streptozotocin effect on the intestine or as a consequence of the altered food consumption of the diabetic animal should be considered. These seem unlikely as they are prevented by insulin therapy*‘-** and do not occur in control animals pair-fed with diabetics.*’ Intestinal alkaline phosphatase in the rat is enhanced by a 25% fat diet,3 but this is unlikely to account for the marked stimulation noted in our animals who are reared on a 5% fat diet. Moreover, quantitative differences in food consumption in these animals were only slightly altered by insulin therapy (e.g., fat intake decreased from 1.85 g/24 hr to 1.53 g/24 hr) whereas alkaline phosphatase levels were normalized. Hormonal alterations may also elevate serum alkaline phosphatase activity. For example, the intestinal isoenzyme is stimulated by 1,25_dihydroxyvitamin

HOUGH ET AL.

1194

D levels have D. 2’.30However, 1,2Sdihydroxyvitamin been shown to be decreased in animals with experimentally induced insulin deficiency.2S.3’ Hypercorticosteronism also occurs in experimental diabetes;25V3’ while glucocorticoid induction of intestinal alkaline

phosphatase has been reported in the weanling mouse, 33corticosteroids do not alter duodenal alkaline phosphatase activity of adult rats14 and are unlikely to account for the marked hyperphosphatasemia noted in our animals.

REFERENCES 1. McComb RB, Bowers GN, Posen S: Alkaline Phosphatase. New York, Plenum Press, 1979, pp 394408, 5388679, 809-814. 865-891 2. Cantor MM, Tuba J, Capsey PA: Serum phosphatases and alloxan diabetes. Science 105:476477, 1947 3. Madsen NB, Tuba J: On the source of the alkaline phosphatase in rat serum. J Biol Chem 195:741-750, 1952 4. Goldberg DM, Martin JV, Knight AH: Elevation of serum alkaline phosphatase activity and related enzymes in diabetes mellitus. Clin Biochem 10:8-I I, 1977 5. Belfiore F, Lo Vecchio L, Napoli E: Serum enzymes in diabetes mellitus. Clin Chem 19:447452. 1973 6. Frankel JJ, Ashbury CE, Baker LA: Hepatic insufficiency and cirrhosis in diabetes mellitus. Arch Intern Med 86:376-390, 1950 7. Bogoch A, Casselman WGB, Kaplan A, et al: Studies of hepatic function in diabetes mellitus, portal cirrhosis and other liver diseases. Am J Med 18:354-384, 1955 8. Camerini-Davalos R, Marble A. Muench H: Liver function in diabetes mellitus. N Eng J Med 266:134991354, 1962 9. Gochman N, Schmitz JM: Application of a new peroxide indicator reaction to the specific automated determination of glucose with glucose oxidase. Clin Chem 18:9433950. 1972 IO. Pybus JF, Feldman J, Bowers GN: Measurement of total calcium in serum by atomic absorption spectrophotometry with use of strontium internal reference. Clin Chem 16:99881007, 1970 I I. Chen PS, Toribara TY, Warner H: Microdetermination of phosphorus. Anal Chem 28:1756-1758, 1956 12. Dawson JB, Heaton FW: The determination of magnesium of biological materials by atomic absorption spectrophotometry. Biochem J 80:99-106, 1961 13. Kachmar JF, Moss DW: Fundamentals of Clinical Chemistry. Philadelphia, WB Sauders Co., 1976, pp 606-609 14. Lowry OH, Rosebrough NJ, Farr AL, et al: Protein measurement with folin phenol reagent. J Biol Chem 193:2655275, 1951 15. Pollard JH: A Handbook of Numerical and Statistical Techniques with Examples Mainly from the Life Sciences, Cambridge University Press, Cambridge, 1977, pp 175-l 77 16. Robison R: The possible significance of hexosephosphoric esters in ossification. Biochem J 17:286-293, 1923 17. Gomori G: Calcification and phosphatase. Am J Pathol 19:197-207, 1943 IS. Fallon MD, Whyte MP, Teitelbaum SL: Stereospecific inhibition of alkaline phosphatase by I-tetramisole prevents in vitro cartilage calcification. Lab Invest 43:489494, I98 I 19. Posen S, Cornish C, Kleerekoper M: Alkaline phosphatase and metabolic bone disease, in Avioli LV and Krane SM (eds.): Metabolic Bone Disease, vol I. New York, Academic Press, 1977, pp 141-174 20. Russell RGG, Monod A, Bonjour JP et al: Relation between

alkaline phosphatase and Ca*+-ATPase in calcium transport. Nature (London) New Biol240:126-127. 1972 21. Haussler M, Nagode L. Rasmussen J: Induction of intestinal brush border alkaline phosphatase by vitamin D and identify with Ca-ATPase. Nature 228:119991201, 1970 22. Morrissey RL, Zolock DT, Bikle DD, et al: Intestinal response to la,25-dihydroxycholecalciferol. RNA polymerase. alkaline phosphatase. calcium and phosphorus uptake in vitro, and in vivo calcium transport and accumulation. Biochim Biophys Acta 538:23-33.1978 23. Linscheer WC, Malagelada JR, Fishman WH: Diminished oleic acid absorption in man by I -phenylalanine inhibition of an intestinal phosphohydrolase. Nature (New Biology) 23 I :I I661 17. 1971 24. Saini PK, Posen S: The origin of serum alkaline phosphatase in the rat. Biochim Biophys Acta 177:42-49. 1969 25. Hough FS. Teitelbaum SL. Russell J, et al: Normalization ol deranged mineral homeostasis and skeletal morphology in the chronically diabetic rat by insulin therapy. Clin Res 28:50A, 1980 26. Lander0 0. Frost HF: Radial rate of osteon closure measured by means of tetracycline labelling. Henry Ford Hosp Med Bull I2:4999505, 1964 27. Olsen WA. Korsmo H: The intestinal brush membrane in diabetes. .I Clin Invest 60:18lll88, 1977

border

28. Caspary WF, Rhein AM, Creutzfeldt W: Increase of intestinal brush-border hydrolases in mucosa of streptozotocin-diabetic rats. Diabetologia 8:4 I224 14, I972 29. Caspary WF: Effect of insulin and experimental diabetes mellitus on the digestive-absorptive function of the small intestime. Digestion 9:248-263, 1973 30. Norman AW, MirchelTAK, Adams TH, et al: Studies on the mechanism of action of calciferol III. Vitamin D mediated increase of intestinal brush border alkaline phosphatase activity. Biochem Biophys Acta 215:348-359, 1970 3 I. Schneider LE. Schedl HP, McCain T, et al: Experimental diabetes reduces circulating 1,25-dihydroxyvitamin D in the rat. Science 196:1452-1453, 1977 32. DeNicola AF, Fridman 0. Del Castillo EJ, et al: kbnormal regulation of adrenal function in rats with streptozotocin diabetes. Horm Metab Res 9:4699473, 1977 33. Moog F: Corticosteroids and enzymic maturation of the intestinal surface: Alkaline phosphatase, leucyl naphthylamidase and sucrase in, Hamburgh H, Barrington EJ (eds): Hormones in Development, New York, Appleton-Century-Crofts. 197 I. pp 1433 160 34. Schraier M, Mazure PA, Lozzio B, et al: Les phosphatases alcalines du tube digestif du rat: Action de la surrenalectomie et de la cortisone. Rev Fr Etud Clin Biol 8: 177-l 84, 1963