Isoenzymes of Liver Alkaline Phosphatase in Serum of Patients with Hepatobiliary Disorders

Transcription

1 clin. CHEM. 9/, 4-47 (973) Isoenzymes of Liver Alkaline Phosphatase in Serum of Patients with Hepatobiliary Disorders Douglas P. Rhone, Florence M. White, and Helene Gidaspow The isoenzymesof alkalinephosphatase(alp; EC 3..3.) present in the.sera of 6 patients with hepatobiliary disease were studied by electrophoresis on cellulose acetate. Individual liver isoenzyme bandswere identified and quantitated by densitometnc scanning. The occurrence and magnitude of increase of the isoenzyme bands of liver ALP (a alone or a and cx) are related to diagnosis. Of the cases studied, 3% showed only the a band of liver ALP. A couplet of liver isoenzyme bands (ai and a) was seen in 7% of the cases studied. The isolated a band occurred in 6 of 36 patients with viral hepatitis and in two patients with hepatic infarction. Most patients with other liver diseases showed the ai and a couplet. The serum pattern of liver ALP is relatedto totalalp, age, ABO blood group, and serum bilirubin concentration. The ai band is markedly increased in patients with granulomatous hepatitis and carcinoma metastatic to the liver. The findings are related to previous work, and their implications are discussed. Additional Keyphrases: diagnostic aid #{49}electrophoresis on cellulose acetate #{49}isoenzymes as indicator of hepatobiliary disease #{49}cervical carcinoma #{49}relation to ABO blood group The alkaline phosphatases (orthophosphoric monoester phosphohydrolase, EC 3..3.) are a group of enzymes that catalyze hydrolysis of phosphate esters at an alkaline ph. The individual tissue and serum isoenzymes of alkaline phosphatase (ALP) differ in physicochemical (, ) and electrophoretic properties (, 3). With a cellulose acetate system in which a substrate-gel imprint technique is used (4), the organ sources of the serum isoenzymes have been verified by comparing mobilities with extracts of tissue ALP and inhibition with heat, urea, and amino acids. From the Section of Clinical Chemistry, Department of Pathology, Illinois Masonic Medical Center, 836 W. Wellington Ave., Chicago, Ill Received April 6, 973; accepted July 3, 973. More recently, we observed two distinct types of patterns of liver ALP in sera of patients with hepatobiliary disorders. In some, a single isoenzyme band of liver ALP appeared, in the a area; others showed a couplet of isoenzyme bands, in the al and a areas (4, 5). Similar isoepzyme bands have been observed by others (6-8), but their incidence and significance have never been explored. In our previous studies, the al band of liver ALP was not demonstrated in the sera of healthy subjects, newborn infants, pregnant females, and patients with primary bone disease (4, 5). The presence of the al liver band is an abnormal finding and appears to indicate disease of the hepatobiliary system. Here, we report the frequency with which the two isoenzyme bands of liver ALP occur in the serum of patients with hepatobiliary disorders. We attempted to make the laboratory study of ALP isoenzymes more useful by relating the presence of the two bands of liver ALP and the magnitude of their elevation to specific disease processes. Materials and Methods Sera. Blood samples were obtained by venipuncture from fasting, healthy laboratory workers and from patients hospitalized with hepatobiliary diseases and who had abnormally high ALP activity in their serum. Diagnoses were verified by the usual clinical and laboratory criteria and in many cases by liver biopsy or autopsy. The serum was separated by centrifugation and its activity measured within 4 h. Alkaline phosphatase. Total ALP activity was determined by the Hansen modification of the Kind- King method (9, ). Isoenzyme electrophoresis. Electrophoresis on cellulose acetate and development of colored isoenzyme bands were done as previously described (4, 5), except for two changes in the original protocol: Concentrations of both fl-naphthyl phosphate and Fast Violet B salt were increased (from 6 mg) to mg per 3 ml of buffer. Also, preparation of the sub- 4 CLINICAL CHEMISTRY, Vol. 9, No., 973

2 strate-gel plates was timed to coincide with the completion of electrophoresis, so that the strips could be imprinted immediately on the gel after electrophoresis. These changes resulted in lower ALP activity being detectable. Isoenzyme bands were classified according to their location when compared with the major serum fractions: albumin, al, a, fi, and -y-globulins. Densitometric scanning. This was done when enough serum was available for us to measure total ALP. The cellulose acetate strips were removed from the third change of distilled water, blotted with a paper towel, and scanned immediately in a densitometer (Helena Laboratories, Beaumont, Tex. 7774) with a green filter. The percentage of total enzyme activity in the sample represented by each isoenzyme band was calculated on a digital computer (Helena Laboratories). These percentages provided a means of calculating the ALP units/ml present in each isoenzyme band. For improved scanning, we attempted to clear the electrophoretic strips by using both various concentrations of aqueous methanol or dimethylformamide ( Sepraclear ; Gelman Instrument Co., Ann Arbor, Mich. 486), but were unsuccessful. For scanning purposes the sera of some controls were concentrated threefold by dialysis against polyvinylpyrrolidone K-3 (Matheson, Coleman, and Bell, Norwood, Ohio 45). Biliru bin determinations. When adequate sample was available, biirubin was measured (). Evaluation and results. Mean values and their standard deviations were calculated for total ALP, percent of total ALP in each isoenzyme band, and units of ALP activity in each isoenzyme band. When applicable, statistical significance of differences between groups was ascertained by means of the Student s t test. Results Healthy subjects. This group consisted of 3 men and women laboratory workers with a mean age of 6.9 ± 4.9 (SD) years. Only one serum showed activity attributable to the liver (a) isoenzyme alone. Table shows the distribution of ALP originating from liver, bone, and intestine in sera from healthy subjects whose serum contained mixtures of isoenzymes. In both mixtures shown, the bone isoenzyme accounted for about 3% of total ALP activity. In those sera containing three isoenzymes, liver ALP represented 57% of the total enzyme activity-significantly less than the 7% seen in those with only two isoenzymes. Intestinal ALP represented 3% of total ALP activity in those sera that contained it. When the ALP units represented by each isoenzyme band were calculated, no statistically significant differences were seen in the absolute activities (units/ml) of bone or liver ALP present in the two mixtures. Patients with hepatobiliary disease. Sera from 6 patients with hepatobiliary diseases were studied. The disorders included are shown in Table, where Table. lsoenzymes of ALP in Sera of Healthy Subjects Enzyme activity in each isoenzyme band: ALP units/isoenzyme bandb,c Eiectrophoretlc pattern n (liver/bone) 7 a/pre-fl/ y (liver/bone/intestine) a pre-( y 3.3±..4±.9 ( ) ( ) 4. ±.. ±.3 (.8-5.7) (.7-4.7)....9 ±.5 (. -.) #{76} Mean SD and (actual range). KK units/mi [a Kind-King unit X 7.9 = U (37#{76}C)]. C calculated Irom densitometric scanning and total ALP. Table. Electrophoretic Patterns of ALP lsoenzymes in Sera of Patients with Hepatobiliary Disorders No. patients showing each electrophoretic pattern#{76} Total Disorder Nutritional cirrhosis Infectious hepatitis Serum hepatitis Metastatic tumors Obstructive jaundice Hepatic infarction Cholecystitis,cholelithiasis Cholangitis Chronic hepatic congestion Hepatocellular carcinoma Granulomatous hepatitis n a/pre-p a/y a/pre-fl/y a, a a, a/pre-9 a a/pre-fl/y % of total a a, and a = liver ALP; pre- = bone ALP; = intestinal ALP. 3 7 CLINICAL CHEMISTRY, Vol.9,No

3 liver Itst,n. /verbo.intest Fig. 3. Densitometric scans of cellulose acetate electrophoretograms of isoenzymes of ALP in hepatobiliary disease Fig.. Electrophoretic patterns of serum isoenzymes of ALP in hepatobiliary disorders and their tissue derivation An arrow marks the point of application. A, serum protein pattern; B, liver (a); c. liver (a,a); D, liver/bone (a/pre-fl); E, pattern in after heat inactivation; F, liver/bone/intestine (a/pre-/ y); G. pattern in F after heat inactivation Fig.. Electrophoretic patterns of serum isoenzymes of ALP in hepatobiliary disorders and their tissue derivation An arrow marks the point of application. A, serum protein pattern; B, liver/bone (a, a/pre-); c, pattern in B after heat inactivation;, liver/ intestine (a/y); E, liver/bone (a,a/pre-$) the electrophoretic patterns are related to diagnosis. Examples of electrophoretic patterns are shown in Figures and, densitometric scans in Figure 3. Some of the electrophoretic patterns illustrated show ALP staining at the origin, which is not true for the fresh specimens upon which the data in this report are based. Activity at the origin appeared after the specimen was frozen and thawed-known to alter total ALP activity in serum. Preliminary further results indicate that all isoenzyme bands are affected similarly by repeated freezing and thawing. Intramolecular rearrangement leading to reversal of partial denaturation of the enzyme, a mechanism proposed for this effect (), may cause immobility of ALP activity under the electrophoretic conditions we used. Thirty-five sera showed the isolated a liver band in combination with the bone and (or) intestinal bands. Twenty-six of the 35 sera with this pattern of liver ALP were from patients with infectious or serum hepatitis, two from patients with hepatic infarction. The remaining seven patients, whose sera showed only the a band of liver ALP, were diagnosed as having nutritional cirrhosis, metastatic carcinoma of the uterine cervix, or obstructive jaundice. Eighty-one sera showed a couplet of liver isoenzymes in the al and a areas. This pattern of liver ALP was seen in 6 of patients with nutritional cirrhosis, in 5 of 7 patients with obstructive jaundice, in all but one case of malignant tumor metastatic to liver, and in all of those with cholecystitis, cholelithiasis, cholangitis, chronic hepatic congestion, and granulomatous hepatitis. One patient with 44 CLINICALCHEMISTRY, Vol.9,No.,973

4 Table 3. Characteristics of Sera Containing ALP lsoenzymes of Liver Origin Electro- ALP, units/lsoenzyme band#{76}.5. Bilirubin, mg/di phoretic pattern n subjects age, yr#{76} Total liver ALPS a, a Direct Indirect a 3.8 ± ± ± ±.4.5 ±. ( ) ( ) ( ) ( - 4.) ( ) a a ± ± 56d 6.6 ± 4.7. ±.7. ±..3 ±.7 (. - 8.) ( ) (.4-5.6) ( ) ( -.) ( - 8.7) a Mean ± SD and (range). KK units/mi. C Calculated from densitometric scanning and total ALP. d P o.ooi. hepatocellular carcinoma also showed the al and a ALP activity in those sera with the isoenzyme couliver couplet. plet is mainly the result of a-band activity. Sera The electrophoretic patterns obtained were stud- with the ai ALP band had lower concentrations of ied in relation to sex, and ABO blood groups. Except all fractions of bilirubin, but in this population of that all of those showing the presence of intestinal patients these differences were not statistically sig- ALP in their serum were of blood groups or B, no nificant. other relation to ABO blood group was seen. Of the The ALP activity represented by each band of the 6 patients studied with hepatobiliary disorders, 53 isoenzyme couplet was calculated and related to were women and 63 were men. Of the 35 patients hepatobiliary disorder (Table 4). The highest activishowing the isolated a liver band, 74% were men ties of total liver ALP and greatest activity in both and 6% were women. Of those showing the couplet isoenzyme bands were seen in those sera from paof liver isoenzymes (al and a), 6% were women tients with granulomatous hepatitis. Sera from paand 38% were men. tients with malignant tumors metastatic to the liver Table 3 shows other factors that we studied in had the second highest total liver AP activities, and relation to the two patterns of liver isoenzymes. The activity in both isoenzyme bands; these activities couplet of liver isoenzymes was encountered in the were significantly higher (P.) than those in sera of older patients. Compared to those sera with sera from patients with cholecystitis, cholelithiasis, only the a liver band, those with the isoenzyme or serum hepatitis. The amount of ALP activity in couplet (al, a) had significantly higher activities of the al band was also significantly higher (P.) total liver ALP. A statistically similar amount of in those with metastatic tumors as compared to ALP activity was seen in the a band of both those with nutritional cirrhosis. Total liver ALP acgroups, which suggests that the higher total liver tivities in sera from patients with obstructive jaun- Table 4. ALP Activity of Liver lsoenzyme Couplet in Hepatobiliary Disordersa ALP activity per Isoenzyme band Disorder No. Total liver ALP a, Band a Band Obstructive jaundice 5 5. ± ±.4.5 ± 5. ( ) (.6-9.) ( ) Cholecystitis,cholelithiasis ±.. ±..8 ±.3 (.3-7.) (.8-4.) ( ) Metastatic tumors ± 6.. ± ± 5.9 (7.6-7.) (.3-47.) ( ) Nutritionalcirrhosis 6.7 ±.6 4. ± ±.3 ( ) (.3 -.6) ( ) Infectioushepatitis ± ± ±. ( ) (. - 5.) ( ) Serum hepatitis ±.7.5 ±.. ±.4 ( ) (.5-5.3) (9. -.8) Granulomatous hepatitisc ( ) (.8-3.) ( ) ChoIangitis ( ) ( ) ( ) a Mean ± SD, and (range). KK units/mi; calculated from total AP and densitometric scanning. C Student s t test not applied. d Only mean values shown. CLINICAL CHEMISTRY, Vol. 9, No.,973 45

5 dice were significantly higher (P.) than those measured in patients with cholecystitis, cholelithiasis, or serum hepatitis Discussion Two distinct electrophoretic patterns of liver ALP were encountered in sera of patients with hepatobiliary disease. In 3% of those studied only an cr band appeared; the remaining 7% showed a couplet of liver isoenzymes (cr and cr). The incidence of these two patterns of liver ALP and their study in relation to age, sex, blood groups, and underlying disease processes in man have not been previously reported. When compared to the occurrence of the liver isoenzyme couplet, the isolated a band was most commonly seen in the sera of younger males and in those sera from patients with serum and infectious hepatitis. This observation probably reflects the higher incidence of hepatitis in younger male patients. In 959 Keiding first described a couplet of liver isoenzymes on starch-gel electrophoretograms of sera from a few patients with liver disease (6). In subsequent studies by use of electrophoretic separations in starch gel (8) and on cellulose acetate (3-5, 7), this pair of fast and slower moving liver isoenzymes was also encountered. The isoenzyme couplet has been described in the sera of a few patients with nutritional cirrhosis (7, 8), chronic passive hepatic congestion (7), fatty metamorphosis of the liver (8), obstructive jaundice (6, 7), malignant tumors metastatic to the liver (3, 6, 7), infectious hepatitis (7), or biliary cirrhosis (7). We encountered the al-a couplet of liver isoenzymes most frequently in the sera of older women with malignant tumors metastatic to the liver, nutritional cirrhosis, obstructive jaundice, or cholangitis. This observation is in harmony with the higher incidence of these disorders in older patients. The cr area has been reported to show marked ALP activity in sera from cases of carcinoma metastatic to the liver and of obstructive jaundice (6, 7). We saw mean cr-alp activities in the sera of patients with hepatobiliary diseases in the following order of decreasing magnitude: granulomatous hepatitis > metastatic tumor > infectious hepatitis > cholangitis > obstructive jaundice > nutritional cirrhosis. The highest single cr-alp activity was seen in a case of tumor metastatic to the liver. High ALP activity in the slow-moving liver isoenzyme band (cr) has been reported in the sera of a few patients with hepatitis (8), nutritional cirrhosis, obstructive jaundice (6), and carcinomas metastatic to the liver (6). In our study, greatest ALP activity in the cr area was seen in those sera containing the crl-a couplet from patients with granulomatous hepatitis. Those patients with malignant tumors metastatic to the liver also showed high ALP activity in the a area. Mean values for the magnitude of a activity in all other hepatobiliary diseases were statistically similar. Kaplan and Righetti (3) induced increased serum ALP activity in rats with acute biliary obstruction, and assayed the serum isoenzymes by polyacrylamide gel electrophoresis. Within 4 h after ligation of the common bile duct a fast-moving liver ALP band appeared on the electrophoretogram, immediately in front of the slower moving band that was normally seen in control animals. This faster moving isoenzyme persisted until the obstruction was relieved (3). Using agar gel electrophoresis, Kang investigated the isoenzymes of ALP appearing in the serum of rats after ligation of the common bile duct, injections of tumors into the liver, or toxic insult to the liver. Two adjacent bands of ALP activity appeared in most of the sera, the faster moving being the more active in those animals with obstructive jaundice or injected tumor. Rats with toxic hepatitis showed an inconstant fast-moving band, which was very light in intensity. The slower-moving band, which was present in normal sera as well, became more intense in all of those animals with hepatobiliary disease (4). The two isoenzyme bands of liver ALP (fast- and slow-moving) described in rats with hepatobiliary disease by both Kaplan and Kang appear to correspond to the isoenzyme couplet studied here in man. The relationship of the isoenzyme pattern of ALP to underlying disease process in man is similar to that seen in the experimental animal. The presence or absence of the a-alp isoenzyme band may have some predictive value in patients with hepatobiliary diseases: because the cr band is present in the serum of only 8% of those patients with serum and infectious hepatitis, absence of the crj band with exaggerated cr-alp activity suggests hepatitis as a diagnostic possibility. Presence of the cr-alp band and its activity can suggest other disease processes: granulomatous hepatitis, metastatic tumor, or obstructive jaundice. Righetti and Kaplan (5) recently demonstrated in rats that impairment of protein synthesis by pretreatment with cycloheximide prevented both the rise in total ALP and the appearance of a fast-moving liver isoenzyme band usually seen in the serum after ligation of the common bile duct. They interpreted these observations to indicate that serum elevations of ALP in biliary obstruction reflect the de novo synthesis of ALP by the liver. The cr-alp band (fast-moving) studied here in man may represent the same liver isoenzyme which appears in rats. Because it is never seen in healthy subjects, the cr- ALP band may be present in serum as a result of de novo synthesis by the liver during hepatobiliary disease. The cr band may appear in serum as a constant or transient expression of significant extrahepatic (obstructive jaundice) or intrahepatic (metastatic tumor, nutritional cirrhosis) biliary obstruction, in contrast to the isolated cr band, which is most commonly seen in those conditions characterized by hepatocellular destruction and enzyme leakage: 46 CLINICAL CHEMI5TRY, Vol. 9. No., 973

6 serum hepatitis, infectious hepatitis, or hepatic infarction. To pursue this question, we are currently studying, sequentially, ALP isoenzyme patterns in patients with serum- or infectious hepatitis. Because sera from these patients occasionally show the couplet of liver isoenzymes, we may be able to relate the individual pattern to elevations of a specific fraction of bilirubin or to a stage in the disease process. We gratefully acknowledge the technical assistance of Marie Ho, Bing Carpizo, and Mary Jo Sanchez-Serrano. The encouragement and guidance of Drs. R. L. Kelsey and G. Gyori are also appreciated. References. Green, S., Cantor, F., Inglis, N. R., and Fishman, W. H., Normal serum alkaline phosphatase isoenzymes examined by acrylamide and starch gel electrophoresis and by isoenzyme analysis using organ-specific inhibitors. Amer. J. Clin. Pathol. 57, 5 (97).. Home, M., Cornish, C. J., and Posen, S., Use of urea denaturation in the identification of human alkaline phosphatase. J. Lab. Clin. Med. 7, 95 (968). 3. Fritsche, H. A., and Adams-Park, H. R., Cellulose acetate electrophoresis of alkaline phosphatase isoenzymes in human serum and tissue. Clin. Chem. 8, 47 (97). 4. Rhone, D. P., and Mizuno, F. M, Separation of isoenzymes of alkaline phosphatase by substrate-gel imprint after electrophoresis on cellulose acetate. Clin. Chem. 8, 66 (97). 5. Rhone, D. P., and Mizuno, F. M., Profiles of alkaline phosphatase isoenzymes in serum using cellulose acetate electrophoresis and organ-specific inhibitors. Amer.,J. Clin. Pathol. 59, 53 (973). 6. Keiding, R. N., Differentiation into three fractions of the serum alkaline phosphatase and the behavior of the fractions in diseases of bone and liver. Scand. J. Clin. Lab. Invest., 6 (959). 7. Korner, N. H., Distribution of alkaline phosphatase in serum protein fractions. J. Clin. Pathot. 5, 95 (96). 8. Taswell, H. F., and Jeffers, D. M., Isoenzymes of serum alkaline phosphatase in hepatobiliary and skeletal disease. Amer. J. Clin. Pat hot. 4, 349 (963). 9. Kind, P. R. N., and King, E. J., Estimation of plasma phosphatase by determination of hydrolysed phenol with aminoantipyrine. J. Clin. Pat hol. 7, 3(954).. Hansen, P. W., A simplification of Kind and King s method for determination of serum phosphatase. Scand. J. Clin. Lab. Invest. 8,353(966).. Jendrassik, L., and Grof, P., Vereinfachte photometrische Methoden zur Bestimmung des Blutbilirubins. Biochem. Z. 97, 8(938).. Brojer, B., and Moss, D. W., Changes in the alkaline phosphatase activity of serum samples after thawing and after reconstitution from the lyophilized state. Clin. Chim. Acta 35, 5 (97). 3. Kaplan, M. M., and Righetti, A., Induction of rat liver alkaline phosphatase: The mechanism of the serum elevation in bile duct obstruction. J. Clin. Invest. 49,58(97). 4. Kang, K. Y., Alkaline phosphatase isoenzyme in rats with damage in hepatobiliary tract. Tohoku J. Exp. Med. 5, 53 (97). 5. Righetti, A. B. B., and Kaplan, M. M., Disparate responses of serum and hepatic alkaline phosphatase and 5-nucleotidase to bile duct obstruction in the rat. Gastroenterology 6, 34(97). CLINICAL CHEMISTRY, Vol.9,No.,973 47