ANNALS O F CLINICAL AND LABORATORY SCIEN CE, Vol. 14, No. 4 Copyright © 1984, Institu te for Clinical Science, Inc.

Use of Alkaline Phosphatase Isoenzyme Analysis in the Evaluation of Cholestatic Liver DiseaseEMANUEL EPSTEIN, Ph.D.,* FREDERICK L. KIECHLE, M.D., Ph.D.,* and BENNIE ZAK, PH.D .t

*Department of Clinical Pathology, William Beaumont Hospital,

Royal Oak, MI 48072 and

fDepartment of Pathology, Wayne State University School of Medicine

and Detroit Receiving Hospital/University Health Center,Detroit, MI 48201

ABSTRACTA useful laboratory test for the differentiation of liver, bone, and intes­

tinal alkaline phosphatase (ALP) isoenzymes in serum is presented. Elec­trophoresis in polyacrylamide gel is performed with untreated serum as well as with serum incubated at 56°C for 10 min. The heating step denatures bone isoenzyme which may obscure the liver ALP band when present in large amounts. Visualization of ALP activity is accomplished by the use of buffered p-toluidinium 5-bromo-4-chloro-indolyl phosphate and magne­sium ions. In serum of patients with cholestatic liver disease, the occur­rence of large molecular weight liver cell membrane fragments which contain ALP activity is postulated. These ALP-containing fragments occur at the origin of the electrophoretogram, unable to penetrate the small pore separation gel. Abnormalities involving ALP isoenzymes, such as bone isoenzyme arising from increased osteoblastic activity, may be de­tected. Intestinal isoenzyme, normally present in small amounts in some subjects of blood groups B or O, may be elevated in certain liver diseases, such as cirrhosis. By the use of this m ethod the routine question of whether an ALP found to be increased in a screening procedure is due to liver or bone abnormality may be answered. In addition, the occurrence of abnormal ALP bands arising from cholestatic conditions and the occur­rence of abnormal amounts of intestinal isoenzyme may also be detected.

Introduction laboratory workup of liver disease. Theuse for such a strategy would point to the

It has been suggested that selected site of injury rather than to the categorychemical tests be used for the diagnostic “liver disease.”5,15’23 For example, the

2920091-7370/84/0700-0292 $00.90 © Institute for Clinical Science, Inc.

ALKALINE PHOSPHATASE ISOENZYMES IN CHOLESTASIS 2 9 3

appearance of abnormal amounts of the transam inases, alanine and aspartate aminotransferases (ALT and AST), would indicate cytoplasmic injury and/or mito­chondrial injury to the hepatocyte. Ab­norm al serum concentrations of the “classical” enzymes of obstructive liver disease would suggest cellular m em ­brane injury in the sinusoids and/or the canaliculi. It has been suggested that a panel of liver tests be devised to test all sites of liver injury, the transaminases, p ro te in-syn thetic processes (albumin, serum cholinesterase, prothrombin time) to tes t for hepatocyte injury, and the enzymes alkaline phosphatase (ALP), 7 - glutamyl transferase (7 -GT), leucine ami- nopeptidase (LAP), 5 '-nucleotidase (5'- N), among others, to test for membrane damage of cells lining the canaliculi and sinusoids. No condition, states Burke ,5 is entirely pure as to the serum analytes concentration, but strong inferences can often be drawn concerning the patient’s primary illness. Recent attem pts have been made to study liver function tests by discriminant analysis28 and by com­puter-assisted procedures . 16

The purpose of this paper is to draw attention to the laboratory tests for ALP isoenzymes which may be important in determining the existence of a cholestatic condition. Traditionally, the clinician has ordered laboratory tests to discriminate betw een liver or bone as the organ source of an elevated ALP. While this need is still a viable one in view of the frequent occurrence of an elevated ALP in multi-channel screening procedures, the emphasis often shifts when the cli­nician suspects a liver disorder even in the absence of or with only a slight ele­vation of ALP. The finding of abnormal high molecular weight ALP isoenzymes in serum of patients with liver disease seems to be a test for detection of cho­lestatic liver disease.

Methods21819

The requ ired blood specim en is clotted blood for both total and fraction­ated ALP. Blood collected with calcium and magnesium binding anticoagulants will exhibit falsely low values for total ALP,22 but there will be no interference with electrophoretic separation methods. The electrophoretic method used by us is a disc polyacrylamide gel electropho­retic (PAGE) method which has been ex­tensively pub lished . 10,11,18,19 The sub­strate used for localizing ALP isoenzyme activity is p-toluidinium 5-bromo-4- chloro-indolyl phosphate. Its use as a substrate reagen t for ALP activity in electrophoresis was first reported by us . 11

The procedure used in our laboratory is to devote two gel columns to each pa­tient. One column is sampled with un­treated serum, the second with serum previously heated at 56°C for 10 min to deactivate bone isoenzym e. A 25 jxl serum sample is used per PAGE tube when total ALP is less than 400 IU per L. At higher levels, 10 (xl samples are used. Visual examination is made of each pair of gel tubes, and the existence of bone, liver, abnormal liver, intestinal, or o ther ALP isoenzym e is noted. O ther methods for ALP fractionation using cel­lulose acetate electrophoresis are no­tab le . 13,14,25 A value for total alkaline phosphatase is included in our report as well as the visual interpretation of the electrophoretogram.

Results and DiscussionNormally, a prom inent and compact

liver ALP zone (to be referred to as band) is the predom inant ALP isoenzyme in the electrophoretogram (figure 1-1). Im­m ediately behind the liver band and somewhat overlapping it is the bone iso­

2 9 4 EPSTEIN, KIECHLE, AND ZAK

enzyme. Liver isoenzyme in the PAGE procedure is a little faster than trans­ferrin (in about the same position as ce­ruloplasmin). The intestinal isoezyme, when p resen t, is im m ediately behind transferrin. Transferrin, of course, is not visualized in the ALP electrophoreto- gram. In o rder to do so, it m ust be stained with a total protein stain such as Amido Black B in a separate procedure. For an illustration of the protein bands found in serum protein PAGE, see ref­erence 9.

Bone isoenzyme is elevated in cases of increased osteoblastic activity and is readily seen in children (figure 1-2). Slightly increased bone isoenzyme can also be detected in the serum of people in about the fifth decade of life or greater (figure l-3a). A highly increased adult bone fraction with a normal liver isoen­zyme is shown in figure l-3b.

Intestinal isoenzyme may be seen as a m inor fraction in the serum of people who are of blood types B and O and who are secretors. A somewhat elevated in­testinal band is depicted in figure 1-4. Elevation of the intestinal band may be an indicator of some liver diseases, such as cirrhosis.33

Placental isoenzyme exists in serum as a band, when present, on the slow edge of liver isoenzyme and ordinarily merges with the liver isoenzyme. Placental iso­enzyme as well as some of the variants associated with tumors may be visualized by pre-treating serums by heat treatment at 56°C for two hrs or at 65°C for 15 min which inactivates all other ALP isoen­zymes.

The “cholestatic pattern” is one which indicates a liver disease involving the canaliculi and sinusoids. These patterns show an increased liver isoenzyme in the usual pre-transferrin site as well as ab­normally occurring origin ALP bands in the large pore spacer and sample gels

(see references 2,17,18,19). These ab­normal ALP bands do not readily migrate through the large pore gels and fail to penetrate the small pore separation gel presumably because of their large molec­ular weight. In a non-molecular sieving medium such as cellulose acetate, these abnormal isoenzymes migrate more rap­idly than the norm al liver com ponent and are found in the o^-globulin area.1314,25 The cholestatic-type PAGE electrophoretogram may exist in one of two characteristic patterns: (1) an ab­normal band appears at the top of the spacer gel, and (2) an abnormal band ap­pears at the top of the separation gel as well as on the top of the spacer gel. Griffiths17 refers to these bands as can­alicular and sinusoidal ALP, respec­tively. These electrophoretic types are shown in figures 1-5 and 1-6. Finally, a “m ixed” e lectrophoretic pattern may exist, figure 1-7 w here abnormal liver bands are p resen t along w ith an in­creased bone isoenzyme.

Interest in the high molecular weight ALP was stimulated by Akedo et al in 19671 who stud ied a high molecular w eight m ulti-pro tein lipid-containing complex which was isolated by gel filtra­tion. The high molecular weight material contained ALP as well as leucine ami- nopeptidase activity. These authors spec­ulated that the high molecular weight material, obtainable from serums of sub­jects with hepatic disease and hepatic cancer, represented hepatic membrane fragments because of the occurrence of enzymes which are known to be associ­ated with cellular membranes. In a later paper, Shinkai and Akedo27 reinvesti­gated the subject and found that the high m olecular w eight m aterial and actual liver plasma membranes had similar pro­portions of the enzym es ALP, LAP, Mg++ — ATPase, and 5' —N, as well as antigenic similarities. Fritsche et al in

ALKALINE PHOSPHATASE ISOENZYMES IN CHOLESTASIS 295

1 2 3a 3b

BL It > I» >f It

F i g u r e 1. The eight pairs of polyacrylamide gel electrophoretograms pic­tured have been uniformly produced with untreated serum (left gel) and with serum incubated at 56°Cq for 10 min prior to electro­phoresis (right gel). The origin, denoted by the letter O, is at the interface of large pore gel and the small pore separation gel. Reference to origin bands denotes this interface as well as the zone at the top of the large pore gel. The letters L, B, and I refer to the zones corresponding to liver, bone, and intes­tinal isoenzyme, respec­tively. 1. A normal ALP iso­enzym e pattern. Heat treatment of serum has in­activated the bone isoen­zyme completely, the liver isoenzyme is slightly de­natured. The tubes have been capped with Para- film®. 2. Children who are growing normally yield this pattern. A residue of bone isoenzyme remains in the heated specimen

O ■ i ___ because of the high titer ofbone isoenzym e. Liver isoenzyme either is low or inactivates more readily

I „ than adult liver ALP. 3a. A< * W V I slightly elevated bone iso-enzyme is often seen in

older patients. 3b. An ele­vated hone isoenzyme in an adult subject is d e­picted. 4. In this pair, liver ALP and intestinal ALP occur together with a small amount of origin

ALP. While this intestinal component is more prominent than normal, the intestinal band may occur in subjects who are of blood groups B or O and who are secretors. 5. A small amount of origin ALP is seen in this electrophoretogram which otherwise would present a normal appearance. Even small amounts of the abnormal origin bands are often accompanied by elevated 7 -glutamyltransferase ('y-GT) values. 6. Origin ALP bands are intense in this electrophoretogram pair. An ALP band is seen at the top of the spacer gel; another band is seen at the interface of the large pore gel and the small pore separation gel. Bone isoenzyme is not apparent. A highly elevated 7-GT is found in serum of patients exhibiting this pattern. 7. A mixed pattern of an abnormal liver pattern with origin bands and elevated bone isoenzyme is shown. Bone isoenzyme merges with the main liver band indicating the necessity of differentiation by heat denaturation of the bone fraction.

E 7

197213 and 1974,14 using cellulose acetate e lec tro p h o resis recogn ized th a t a fast (pre-liver) ALP could be dem onstrated in serums from subjects with hepatic cancer and hepatobiliary disease. Rhone e t al in

197325 also called attention to the fast (o^) liver ALP isoenzyme occurring in hepa­tobiliary disease by the use of cellulose acetate electrophoresis.

D eB roe e t al in 19756 and Bröckle-

2 9 6 EPSTEIN, KIECHLE, AND ZAK

hurst in 19763 isolated the high molec­ular w eight m aterial from serum by ultracentrifugation and studied the par­ticulate matter by various enzyme tests and by electron microscopy. Their ob­servations led to the conclusion that the particulate matter was composed of cell membrane fragments which contained the enzymes commonly associated with the liver cell membrane, that is, ALP, LAP, 5' - N, and 7 -GT. A publication in 1978 by us10 involving alteration of the particulate (or slow moving) ALP by de­tergents of the Triton-X group* suggests that the bond between protein and lipid is easily disrupted and may be altered by detergent action. The effect of detergent on the particulate ALP was to increase ALP mobility in the PAGE procedure, indicating that the molecular weight was diminished.

It is thought that shedding of mem­brane fragm ents is a norm al occur­ren ce . 4,7,12 This takes place slowly in health since the abnormal high molecular w eight m aterial is not revealed in a normal serum. It becomes accelerated in hepatobiliary disease, hepatic cancer, ob­structive diseases, etc ., and may be seen in serum in those cases. It has been pro­posed that bile salts accelerate liver cell membrane loss in cholestatic disorders. 26 While the mechanism for the occurrence of high molecular weight ALP activity is still to be elucidated, the value of the laboratory demonstration of the phenom­enon to the clinician seems to be as­sured.Resume of Clinical Interpretations

The greatest utility of ALP PAGE with the finding of abnormal ALP isoenzymes is in obstructive diseases of the liver, pri­mary biliary cirrhosis, and secondary liver damage especially when combined with lipoprotein-X (LP-X)4 assay and an independent liver enzyme test such as

* Rohm and Haas

5'-N or 7 -GT.23,24 Kaplan et al20,21 have shown that in experimental rats ligation of the bile ducts caused induction of ALP, 7 -GT, and 5' — N. The authors con­cluded from their experiments that the increase of these enzymes in cholestasis was due to increased production and not from regurgitation of bile fluid, nor from bile duct proliferation.

Recent interest in the fast ALP isoen­zyme (cellulose acetate electrophoresis) has been shown in the field of oncology in that the abnormal ALP seems to be a sensitive indicator of cancer and of liver metastases. 8.29.30.31

Bone isoenzyme, an indicator of in­creased osteoblastic activity, may be ele­vated in pre-maturity, in bone diseases such as Paget’s disease, in diseases in­volving the vasular system, in vitamin D deficiency, and in liver diseases where metabolism of vitamin D is impaired.

Intestinal isoenzyme, when increased, may be a valuable indicator of liver dis­ease, such as cirrhosis.32,33 Intestinal iso­enzyme normally may appear as a minor fraction in the serum of people with blood types 0 and B and who are secre- tors, but may be induced in all people following a fatty meal.

References1. A k e d o , H., Shi nk ai, K., H o r i u c h i, N ., and

O m o r i, K.: Alkaline phosphatase activity in high molecular compound in serum of hepatobiliary diseases. Ann. Rept. Center Adult Diseases 7:1-7, 1967.

2. A me s C o m p a n y . Quick-Disc QDA Reagent Kit insert directions. Publ. #3510, Elkhart, IN 46514. Information as to commercial kits are available from the current manufacturer of the disc-gel apparatus and alkaline phosphatase PAGE, California Immuno Diagnostics, San Marcos, CA.

3. B r o c k l e h u r s t, D ., L a t h e , G. H., and A paricio,S. R .: Serum alkaline phosphatase, nucleotide pyrophosphatase, 5'-nucleotidase and lipopro- tein-X in cholestasis. Clin. Chim. Acta 67:269— 279, 1976.

4. B r o c k l e h u r s t , D., W i l d e, C. E ., and D o a r , J. W. H.: The incidence and likely origins of serum particulate alkaline phosphatase and li- poprotein-X in liver disease. Clin. Chim. Acta 88:509-515, 1978.

ALKALINE PHOSPHATASE ISOENZYMES IN CHOLESTASIS 2 9 7

5. B u r k e , M. D.: Liver function. Human Pathol. 6:273-286, 1975.

6. D eB ro e , M. E., B o rgers, M., andWiEME, R. J.: T he sep a ra tio n an d ch a ra c te riz a tio n of liv e r plasm a m em brane fragments- circulating in the blood of patients w ith cholestasis. Clin. Chim. Acta 59:369-372, 1975.

7. D e B r o e , M. E., W i e m e, R. J., L o g g h e , G. N., and R o e l s, F.: Spontaneous shedding of plasma membrane fragments by human cells in vivo and in vitro. Clin. Chim. Acta 82:237-245, 1977.

8. E h r m e y e r , S., Jo iner, B. L., K a h a n , L., L a r s o n,F. C., and M e t z e n b e r c, R. L.: A cancer-associ- ated, fast, homoarginine-sensitive electropho­retic form of serum alkaline phosphatase. Cancer Res. 38:599-601, 1978.

9. E pstein, E.: Electrophoresis, Fundamentals of Clinical Chemistry. Tietz, N., ed- Philadelphia, W. B. Saunders, 1976, p. 134.

10. E pstein, E., B aginski, E. S., a n d Z A K , B.: Deter­gent altered alkaline phosphatase patterns of liver disease. Ann. Clin. Lab. Sci. 8:34—41, 1978.

11. E pstein, E., W o l f , P. L., H o r w i t z, J. P., and Z a k , B.: An indigogenic reaction for alkaline phosphatase in disk electrophoresis. Am. J. Clin. Path. 48:530-534, 1967.

12. E v a ns, W. H .: Hepaticplasma-membrane modifi­cations in disease. Clin. Sci. 58:439—444, 1980.

13. F ritsche, H. A., Jr. andADAMS-PARK, H. R.: C el­lulose acetate electrophoresis of alkaline phos­phatase isoenzym es in hum an serum and tissue. Clin. Chem . 28:417-421, 1972.

14. F ri t s c h e, H. A., Jr . and A d a m s -Pa r k , H. R.: High molecular weight isoenzymes of alkaline phosphatase in human serum. Demonstration by cellulose acetate electrophoresis and physico­chemical characterization. Clin. Chim. Acta 52:81-89, 1974.

15. G o t z, W.: Diagnosis of hepatic diseases. Darm­stadt, G-I-T, Verlag Ernst Giebeler, 1980.

16. G o l d b e r g , D. M. and E llis, G.: Mathematical and computer-assisted procedures in the diag­nosis of liver and biliary tract disorders. Adv. Clin. Chem. 20:49-128, 1978.

17. G riffiths, J.: C linical Enzym ology for the Eighties. American Society of Clin. Pathologists, Commission on Continuing Education, Chicago,1982, pp. 82-96.

18. Jo h n s o n , R. B., Jr.: Preferred methods for the isoenzymes of serum alkaline phosphatase. Clin­ical Enzymology. Griffiths, J. C., ed. New York, Masson, 1979, pp. 93-109.

19. Jo h n s o n , R. B., E ll i n g b o e, K. , and G ibbs, P.: A study of various electrophoretic and inhibition techniques for separating serum alkaline phos­phatase isoenzymes. Clin. Chem. 28:110-115, 1972.

20. K a p l a n, M. M., K a n e l , G. C., and Singer, J. A.: Enzyme changes and morphometric analysis of

bile ducts in experimental bile duct obstruction. Clin. Chim. Acta 99:113-119, 1979.

21. K a p l a n , M. M. and R ighetti, A.: Induction of liver alkaline phosphatase by bile duct ligation. Biochim. Biophys. Acta 184:667-669, 1969.

22. L a d e n s o n , J. H . : Nonanalytical sources of vari­ation in clinical chemistry results. Gradwohl’s Clinical Laboratory Methods and Diagnosis, vol.1. Sonnenwirth, A. C., and Jarett, L., eds. St. Louis, Mosby, 1980, p. 154.

23. L e h m a n n , F.-G.: Enzyme and isoenzyme diag­nosis of canceT. Advances in Clinical Enzy­mology. Schmidt, E., Schmidt, F. W . , Traut- schold,-I., and Friedel, R., eds. New York, S. Karger, 1979, pp. 171-195.

24. P rice, C. P. andALBERTi, K. G. M. M. . Biochem­ical assessment of liver function. Liver and Bil­iary Disease. Wright, R., Alberti, K G. M. M., Karran, S., and Millward-Sadler, G. H., eds. Philadelphia, W. B. Saunders, 1979.

25. R h o n e , D. P., W hite, F. M., and G i d a s p o w, H.: Isoenzymes of liver alkaline phosphatase in serum of patients with hepatobiliary disorders. Clin. Chem. 29:1142-1147, 1973.

26. Sc h l a e g e r , R., H a u x , D ., and K a t t e r m a n n , R.: Studies on the mechanism of the increase in serum alkaline phosphatase activity in choles­tasis: Significance of the hepatic bile acid con­centration for the leakage of alkaline phosphatase from rat liver. Enzyme 28:3-13, 1982.

27. Shinkai, K. and A k e d o , H. : A multienzyme com­plex in serum of hepatic cancer. Cancer Res. 32:2307-2313, 1972.

28. So l b e r g , H. E., Sk r e d e , S., andBLOMHOFF, J. P.: Diagnosis of liver diseases by laboratory results and discrim inant analysis. Scand. J. Clin. Lab. Invest. 35:713-721, 1975.

29. V iot, M., Jo u l i n, C., C a m b o n , P., K rebs, B. P., Sc h n e i d e r , M., and L e l a n n e , C. M.: The value of serum alkaline phosphatase a l isoenzyme in the diagnosis of liver metastases. Preliminary re­sults. Biomedicine 32:74- 77, 1979.

30. V iot, M ., T h y s s, A., V iot, G., R a m a i o l i, A., C a m b o n , P., Sc h n e i d e r , M., and L a L a n n e ,C. M.: Comparative study of gamma glutamyl transferase, alkaline phosphatase and its a l iso­enzyme as biological indicators of liver metas­tases. Clin. Chim. Acta 225:349-358, 1981.

31. V iot, M., T hyss, A., Sc h n e i d e r, M., V iot, G., T a- maioli, A., C a m b o n , P., and L a L a n n e , C. M.: a l Isoenzyme of alkaline phosphatases. Clinical importance and value for the detection of liver metastases. Cancer 52:140-145, 1983.

32. W a r n e s, T. W., H ine, P., and K ay, G .: Polyacryl­amide gel disc electrophoresis of alkaline phos­phatase isoenzymes in bone and liver disease. J. Clin. Path. 29:782-787, 1976.

33. W a r n e s, T. W., H ine, P., and K a y, G .: Intestinal alkaline phosphatase in the diagnosis of liver dis­ease. Gut 28:274-278, 1977.