- Email: [email protected]
Isoelectric focusing of serum alkaline phosphatase isoenzymes
157
Clinica Chimica Acta, 129 (1983) 157-164 Elsevier Biomedical Press
CCA 2473
Isoelectric Josefa
focusing of serum alkaline phosphatase isoenzymes
Blum-Skolnik
a,*, Fabio
Pace b, Gerold
Mi.inst
b and Willi
Minder
a
0 Central Laboratory and b Department of Medicine, Stadtspital Triemli. Zurich (Switzerland) (Received
June 30th; revision December
16th, 1982)
Summary We present a new method for the separation of alkaline phosphatase isoenzymes based on isoelectric focusing of serum proteins. Sera of healthy subjects studied by this method present seven groups of bands within a pH range of 3-9. To date, the organ source of six out of seven groups has been identified directly or indirectly. Serum of patients with histologically proven malignant neoplasms showed an additional group in the pH range 3.6-4.2. This fraction seems to be heterogeneous with respect to the pattern and stability to heat, and to r_-phenylalanine and L-homoarginine inhibition. It was present in all but two of the proven malignant tumour patients but also in 26 out of 70 with benign disorders.
Introduction The clinical importance of the determination of alkaline phosphatase isoenzymes (API) has been well recognized for more than two decades. Methods such as physical and chemical inhibition, starch electrophoresis as well as disc isoelectric focusing have been described and reviewed by Fishman [l]. He discussed the attempts to characterise the physical and biochemical properties of these isoenzymes. None of them are quite satisfactory from the routine clinical point of view [2]. The most suitable method for clinical laboratory use was shown to be cellulose acetate membrane electrophoresis [3]. But even this method shows (a) poor separation of bands, (b) no visible bands at normal AP level, and (c) no ‘cancer band’ in particular. A method based on isoelectric focusing [4] was therefore developed.
* To whom correspondence
OOOS-8981/83/OOWJ-0000/$3.00
should be addressed.
0 1983 Elsevier Science Publishers
158
Materials Organs For identification mortem.
of API,
specimens
of human
organs
were
obtained
post
Subjects Three groups were studied: (1) 50 healthy volunteers; (2) 70 patients from medical wards with final diagnoses of benign diseases; (3) 70 patients with histologically proven malignant neoplasms. Chemicals and reagents Desaga Polyfix 1000 was used to silylanise the glass plates; for isoelectric focusing, the following Bio-Rad polyacrylamide gel-forming reagents were used: acrylamide ultra pure, BIS, riboflavin-5-phosphate, ammonium persulphate, TEMED (Bio-Rad Laboratories AG, Glattbru~-Zurich, Switzerland) with l-naphthyl-phosphate monosodium salt (Fluka AG, Buchs, Switzerland) as substrate, variamin blue B salt (Fluka AG) for visualisation of the isoenzymes. Absolute ethanol, acetone and boric acid used were analytical grade. Isoelectric focusing was performed using a Bio-Rad Model 1415 A and pH gradient was measured with a surface electrode (Dr. W. Ingold AG, 8902 Urdorf, Switzerland). Isoenzyme patterns were evaluated with the Integraph-CH densitometer (Bender Hobein, Zurich, Switzerland) and a Hi-PAD Digitizer of Houston Instruments, USA using an integration programme on a microcomputer (CBM-3000, Commodore, USA). Methods Total AP was determined by Bessey’s method [5] adapted to the GSA II selective analyser (Greiner Electronics AG, Langenthal, Switzerland). In order to identify the source of the API bands, human organ homogenates were prepared as described by Morton, and as modified by Smith [6], and added to normal serum in equal parts. For further identification, heat inactivation at 56°C for 10 min, ~-phenyIalanine (5 mmol/l) and L-homoarginine (8 mmol/l) inhibition were applied. (A) Preparation of polyacrylamide gel Glass plates are degreased with acetone and silylanised [7] in a 0.2% solution of Polyfix-1000 in distilled water: ethanol (1 : 1) for 5-10 min and dried at room temperature. 1.5 ml carrier ampholyte, pH range 3-9 (composed of 0.85 ml 3/5 with 20% dry weight and 0.65 ml 4/9 with 40% dry weight resp. 0.75 ml 3/5 and 0.75 ml 3/10), 6 ml 25% w/v glycerol, 6 ml monomer cont., and 16.5 ml distilled water are mixed and degassed for 15 min. Then 150 ~1 riboflavin-5-phosphate sodium salt dihydrate * (1 g/l water), 75 ~1 TEMED (100 pl/lO ml water) and 30 ~1 ammonium persulphate * (100 mg/S ml water) are added, and the glass plates immediately covered with this mixture. The gel is then allowed to set.
159
(B) Gel jiocusing The gel is focussed for 60 min with 1 mol/l H,PO, at the anode and with 1 moi/l NaOH at the cathode, 3 W per 45 mm X 120 mm plate. The pH range generated is measured and the ‘running curve’ is determined. The samples are applied (15 ~1) at for 60 min. the pH range of - 8.0 and focusing continued (C) incubation and elation The plates are immersed in a 1 g/l solution of I-naphthylphosphate salt *, in 0.06 mol borate buffer/acetone (1 : 1, v/v) and incubated room temperature.
monosodium for 120 min at
(D) Staining and destaining Staining is performed in a solution * of 1 mg variamin blue + 5 mg CuSO, - 5H,O per ml 0.06 mol borate buffer, pH 9.7, for 30-60 min. For destaining, the plates are washed twice in a solution of 5 g/l CuSO, - 5H,O in distilled water/ethanol (1 : 1, v/v) for about 30 min each time, and then in water/ethanol (1 : 1, v/v) overnight. The next day the plates are immersed in a solution of glycerol/distilled water/ethanol (10 : 45 : 45, v/v/v) and covered with cellophane previously wetted with the above mixture. The plates are dried at room temperature. (E) Evaluation Evaluation is performed
using a densitometer.
Results
Seven groups of bands, each consisting of 1-3 lines, were identified in sera of healthy individuals. Comparison of extracts of organ homogenates enabled the identification of the renal, hepatic- 1, intestinal 1, -2, -3, and hepatic-2 groups (Fig. 1). Neither pulmonary nor bone marrow extracts yielded any visible fraction after isoefectric focussing although their AP levels were 200 U/l and 230 U/l respectively. The bone group No. 3 was indirectly identifiable by L-phenylalanine inhibition and heat inactivation (PI: 4.5). Another group No. 4 (PI: 4.6-4.8) behaving identically to the bone group with respect to heat and L-phenylalanine inhibition was established on clinical grounds to be of pulmonary origin (Table I, Fig. 2). The 140 patients showed 69 correct negative, two false negative and 14 false positive results. The intestinal group (5a, 5b, 5c) identified by means of jejunum extract shows three well separated bands in the pH range 5.0-5.9. We found the 6th group to be a second hepatic one, identified by means of a liver extract. In serum of healthy persons it seems to be responsible for about half of the total activity of AP. We also observed a 7th group, p1 7.0-7.5 whose organ source could not be identified. Angellis et al [ 171 have reported the appearance of a B form of purified placental type at this pH range. Since this group was present in almost all * These reagents
must be prepared
just before use.
160 ‘NO of the fraction
1 2 3 4
renal liver bone’ pulmonary’”
1 4.2-4.5 x
5 a, b. c intestinal
5b
5c
r
4.6-4.8
_
6
6 liver
7
7.0
7,
Fig. 1.pH distribution of the bands in the serum of healthy subjects. Total AP. 36 U/If 6. * Identified by means of heat inactivation only. ** Found to be increased in patients with lung diseases.
of our serum, it cannot be only of placental origin. Table II shows the percentage distribution of the total AP activity in the serum of healthy controls. Placental homogenates yielded a strong band in the pH range 3.6-4.2 designated by us ‘pre-1’. In all but two of the 70 patients with proven malignant neoplasms this additional band was found. It was never observed in healthy subjects but in serum of pregnant women in the last trimester. Its presence was independent of both the nature of the neoplasms and the level of AP. The pronounced heterogeneity observed in the pre-1 group is therefore not surprising. The appearance of this TABLE
I
PATIENTS
WITH LUNG
DISEASES
SHOWING
AN INCREASED
Carcinoma of the lung Pneumonia Emphysema Pulmonary embolism Chronic bronchitis Sarcoidosis
14
Total
33
1
1 4 6 1
‘PULMONARY’
FRACTION
161
PH
range
F1-l
r-
P-l
I_
;zI 14.2 -4.5
24.2-4.5 34.6
14.6-4.8
4.2-4.5
41
-4.8 5a 14.9 -5.4
5a
4.9-5.4
5
1
5b/
35.5-5.7
5cL-
7
5
/m-m
5c
5.8-6.0
6
6.0-6.8
6.0 -6.8
6.0-6.8
6
5b
i5.8 -6.0
5
i
5.5 -5.7
~
_I
7 i
7.0
7.0 -
-l
7
C
A Fig. 2. Bands and their pH range from: (A) a patient with carcinoma (total AP, 107 U/1); (B) a patient with carcinoma of the pancreas U/l); (C) a patient with bronchogenic cancer (total AP 37 U/l).
of the prostate and bone metastases and liver metastases (total AP 750
fraction varies from two or more fine bands to a smudged area (Fig. 2). It exhibits a different sensitivity to heat and to L-phenylalanine and L-homoarginine inhibition (probably related to different organ sources). We observed this group in sera of 26 of 70 patients with benign diseases (Table III). The reproducibility of the fractions in the serum of a patient with normal AP level and another one with an elevated level, is shown in Table IV. TABLE
II
THE PERCENTAGE ISOENZYME DISTRIBUTION SERUM OF 50 HEALTHY SUBJECTS Total AP x+
I
2.5 1.3
TOTAL
AP ACTIVITY
IN THE
SD: 36 f 6 U/l.
Renal
x SD*
OF THE
2 Liver
4.3 1.6
3 Bone
5.8 2.2
4 Pulmona~
5.4 0.65
5 Intestinal a
b
c
6.33 2.7
10.17 5.3
13.6 2.5
6 Liver-2
7
50.9 5.3
3.9 3
162 TABLE
III
TOTAL
SERUM
Note presence
AP IN SERUM
of pre-I
OF PATIENTS
WITH
BENIGN
group. pre-I present
pre- 1 absent
Total serum AP
Malignant proven Benign proven
TABLE
DISEASES
Total serum AP
elevated
normal
elevated
normal
43 18
25 8
0 22
2 22
IV
REPRODUCIBILITY LEVEL AND FROM
I. AP: 31 U/l n=8 x(g) SD+ II. AP: 121 U/l n = 14 X(S) SD-I:
OF THE FRACTION IN SERA FROM A PATIENT A PATIENT WITH AN ELEVATED LEVEL
p-l
1
2
0
1.9 0.9
2.1 0.1
7.9 1.1
3.9 0.7
6.0 1.1
3
WITH
A NORMAL
4
S(a+b+c)
6
4.0 0.7
4.2 0.9
30.2 4.3
48.4 3.1
8.8 3.1
15.6 2.6
7.9 1.2
17.8 2.5
37.9 2.8
15.3 28.2
AP
7
Discussion Separation of organ-specific API fractions has been attempted since 1959 by the application of different electrophoretic techniques. In Fishman’s [l] review article these experiments and results are discussed in detail. None of these methods are, however, really suitable for clinical use. Their common problem is that they may show similar characteristics for isoenzymes produced by different organs, and even genuine differences did not show satisfactory separation. For instance the method of Siede et al [3] can produce three, or under certain circumstances four different fractions, (liver-l, bone, liver-2 and possibly a pathological bile fraction). All the API are distributed in these fractions. Results of different methods are thus often contradictory. Working with serum there is no real possibility of comparison with the other methods because of the much higher solution-capacity of the IEF. This, on the other hand, leads to additional problems of correct source allocation. As described above this method produces 10-15 bands within seven groups in normal serum and an additional ‘pre-I’ in pathological serum. Six of the seven groups have been traced to organ sources either directly or indirectly as described. The 7th normal group has not been shown to be due to any organ source tested so far. We could identify the bone fraction only by means of heat inactivation of serum of healthy subjects and of patients with bone diseases or
163
bone metastases. The bone marrow itself did not yield any group although its total activity was 200 U/l (reference range: 10-50). The amount of this isoenzyme is much lower as usually reported because the usual electrophoretic methods with serum are unable to separate other isoenzymes (in part liver, placental and probably this one which we call pulmonary) from that of the bone. Similarly the heat inactivation method cannot differentiate between the loss of bone and liver fractions, in which the loss of the latter is estimated to be 60%. We had only 5 patients with Paget’s disease. They were not severe cases (AP levels of 36-130 U/l) but the average value of their bone fraction was 11.75% + 3.1; healthy controls have 5% f 2. Serum from patients with pulmonary diseases, benign or malignant (Fig. 2C) show 2 or 3 bands in the pH range 4.8 + 5.0 which are barely visible in the serum of healthy subjects (see Fig. 1). We presume therefore that these bands originate mainly in the lungs. Both of these groups will be further investigated, by combining them with immunoprecipitation methods. A particular problem is encountered with ‘tumour-specific’ API groups. Such an API called ‘Regan isoenzyme’ was described by Fishman et al [ 181 and another one called Nagao described by Nakayama et al [ 191. Physical and chemical properties showed the Regan isoenzyme to be identical with the isoenzymes prepared from placenta [9]. It is heat stable and L-phenylalanine sensitive as distinct from our ‘pre-1’ fraction. Our ‘pre-1’ fraction seems to be similar but not identical to placental and Regan isoenzymes because ‘pre-1’ was never found in healthy subjects. Usategui-Gomez et al demonstrated Regan isoenzymes in 50 healthy subjects [20]. ‘Pre-1’ is present as Regan isoenzyme in placental homogenates and in the serum of pregnant women in the last trimester, but its heat stability and L-phenylalanine sensitivity varies. Iglis et al stated that Regan isoenzyme is not exclusively a property of cancer patients, but has been identified at low levels in populations with other diseases [8,9]. We have also found the ‘pre-1’ group as described above, in sera of patients with benign diseases (Table III). All these findings are probably due to the different types of isoenzymes in this particular pH range [lo- 131. For clarification, further investigations are necessary. In accord with earlier publications [ 1,14-161 the hepatic-2 group is responsible for approximately half of the total AP activity. In serum from patients with liver metastases this band always becomes very faint (Fig. 2B). The reproducibility of the results critically depends on the carrier ampholyte. It is very important in new batches to check its behaviour with known serum. It is possible to adjust it by changes in its composition. Our method can be recommended for clinical laboratory use because it is simple to perform. The time and money involved are reasonable, and the sample size is low. Acknowledgements The authors thank Prof. Dr. R. Siebenmann, Chief of Department of Pathology, Stadtspital Triemli, Zurich, for supplying materials for this work. This work was supported in part by the Swiss National Foundation Grant, No. 3.940.0.80.
164
References 1 Fishman WH. Perspectives on alkaline phosphatase isoenzymes. Am J Med 1974; 56: 617-650. 2 Bauer John D. Clinical laboratory methods, 9th Edition. St. Louis, Toronto, London: C.V. Mosby Company, 1982: 582. 3 Siede WH, Seiffert UB. Quantitative alkaline phosphatase isoenzyme determination by electrophoresis on cellulose acetate membranes. Clin Chem 1977; 23/l: 28-34. 4 Righetti P, Drysdale JW. Isoelectric focusing in polyacrylamide gels Biochem Biophys Acta 197 I ; 236: 17-28. 5 Bessey OA, Lowry OH, Brock ME. Determination of alkaline phosphatase in human serum. J Biol Chem 1964; 164: 321. 6 Smith I, Lightstone PJ, Perry JO. Purification of alkaline phosphatase. Clin Chim Acta 1968; 19: 499-505. 7 Radola BJ. Electrophoresis 1980; 1: 43-56. 8 Anstiss CL, Green S, Fishman WH. An automated technique for segregation populations with a high incidence of Regan isoenzyme in serum. Clin Chim Acta 1974; 33: 279. 9 Inglis NR, Kirley S, Stolbach LL, Fishman WH. Phenotypes of the Regan isoenzyme and identity between the placental D variant and the Nagao isoenzyme. Cancer Res 1973; 33: 1657. 10 Harkness DR. Studies on human placental alkaline phosphatase. Arch Biochem Biophys 1968; 126: 513. 11 Fishman WH, Sie HG. Organ specific inhibition of human alkaline phosphatase isoenzymes of liver, bone, intestine and placenta, r_-phenylalanine, L-tryptophan and L-homoarginine. Enzymologia 1971; 41: 141. 12 Damle SR, Shelly PA, Jussavalla D, Bhide SV, Baxi AJ. Occurrence of heat stabile Regan type of alkaline phosphatase in hematopoietic tumours. Int J Cancer 1979; 24/4: 398-401. 13 Fishman WH, Driscoll S, Miyayama H, Fishman L. Human tumour non-Regan alkaline phosphatase isoenzyme. Proc Am Assoc Cant Res 1975; 16: 195. 14 Gutman AB. Serum alkaline phosphatase activity in diseases of the skeletal and hepatobihary systems. Am J Med 1959; 27: 875. 15 Posen S, Neele FC, Clubb JS. Heat inactivation in the study of human alkaline phosphatases. Am Int Med 1965; 62: 1234. 16 Posen S. Alkaline phosphatase. Am Int Med 1967; 67: 189. 17 Angellis D, Inglis NR, Fishman WH. Isoelectric focusing of alkaline phosphatase isoenzymes in polyacrylamide gels. Am J Clin Path01 1976; 66: 929-934. 18 Fishman WH, Inglis NR, Stolbach LL, Krant MJ. Cancer Res 1968; 28: 150. 19 Nahayama T, Yoshida M, Kitamura M. Clin Chim Acta 1970; 30: 546. 20 Usategui-Gomez M, Yeager FM, Fernandez de Costa A. Cancer Res 1974; 34: 2544.
Clinica Chimica Acta, 129 (1983) 157-164 Elsevier Biomedical Press
CCA 2473
Isoelectric Josefa
focusing of serum alkaline phosphatase isoenzymes
Blum-Skolnik
a,*, Fabio
Pace b, Gerold
Mi.inst
b and Willi
Minder
a
0 Central Laboratory and b Department of Medicine, Stadtspital Triemli. Zurich (Switzerland) (Received
June 30th; revision December
16th, 1982)
Summary We present a new method for the separation of alkaline phosphatase isoenzymes based on isoelectric focusing of serum proteins. Sera of healthy subjects studied by this method present seven groups of bands within a pH range of 3-9. To date, the organ source of six out of seven groups has been identified directly or indirectly. Serum of patients with histologically proven malignant neoplasms showed an additional group in the pH range 3.6-4.2. This fraction seems to be heterogeneous with respect to the pattern and stability to heat, and to r_-phenylalanine and L-homoarginine inhibition. It was present in all but two of the proven malignant tumour patients but also in 26 out of 70 with benign disorders.
Introduction The clinical importance of the determination of alkaline phosphatase isoenzymes (API) has been well recognized for more than two decades. Methods such as physical and chemical inhibition, starch electrophoresis as well as disc isoelectric focusing have been described and reviewed by Fishman [l]. He discussed the attempts to characterise the physical and biochemical properties of these isoenzymes. None of them are quite satisfactory from the routine clinical point of view [2]. The most suitable method for clinical laboratory use was shown to be cellulose acetate membrane electrophoresis [3]. But even this method shows (a) poor separation of bands, (b) no visible bands at normal AP level, and (c) no ‘cancer band’ in particular. A method based on isoelectric focusing [4] was therefore developed.
* To whom correspondence
OOOS-8981/83/OOWJ-0000/$3.00
should be addressed.
0 1983 Elsevier Science Publishers
158
Materials Organs For identification mortem.
of API,
specimens
of human
organs
were
obtained
post
Subjects Three groups were studied: (1) 50 healthy volunteers; (2) 70 patients from medical wards with final diagnoses of benign diseases; (3) 70 patients with histologically proven malignant neoplasms. Chemicals and reagents Desaga Polyfix 1000 was used to silylanise the glass plates; for isoelectric focusing, the following Bio-Rad polyacrylamide gel-forming reagents were used: acrylamide ultra pure, BIS, riboflavin-5-phosphate, ammonium persulphate, TEMED (Bio-Rad Laboratories AG, Glattbru~-Zurich, Switzerland) with l-naphthyl-phosphate monosodium salt (Fluka AG, Buchs, Switzerland) as substrate, variamin blue B salt (Fluka AG) for visualisation of the isoenzymes. Absolute ethanol, acetone and boric acid used were analytical grade. Isoelectric focusing was performed using a Bio-Rad Model 1415 A and pH gradient was measured with a surface electrode (Dr. W. Ingold AG, 8902 Urdorf, Switzerland). Isoenzyme patterns were evaluated with the Integraph-CH densitometer (Bender Hobein, Zurich, Switzerland) and a Hi-PAD Digitizer of Houston Instruments, USA using an integration programme on a microcomputer (CBM-3000, Commodore, USA). Methods Total AP was determined by Bessey’s method [5] adapted to the GSA II selective analyser (Greiner Electronics AG, Langenthal, Switzerland). In order to identify the source of the API bands, human organ homogenates were prepared as described by Morton, and as modified by Smith [6], and added to normal serum in equal parts. For further identification, heat inactivation at 56°C for 10 min, ~-phenyIalanine (5 mmol/l) and L-homoarginine (8 mmol/l) inhibition were applied. (A) Preparation of polyacrylamide gel Glass plates are degreased with acetone and silylanised [7] in a 0.2% solution of Polyfix-1000 in distilled water: ethanol (1 : 1) for 5-10 min and dried at room temperature. 1.5 ml carrier ampholyte, pH range 3-9 (composed of 0.85 ml 3/5 with 20% dry weight and 0.65 ml 4/9 with 40% dry weight resp. 0.75 ml 3/5 and 0.75 ml 3/10), 6 ml 25% w/v glycerol, 6 ml monomer cont., and 16.5 ml distilled water are mixed and degassed for 15 min. Then 150 ~1 riboflavin-5-phosphate sodium salt dihydrate * (1 g/l water), 75 ~1 TEMED (100 pl/lO ml water) and 30 ~1 ammonium persulphate * (100 mg/S ml water) are added, and the glass plates immediately covered with this mixture. The gel is then allowed to set.
159
(B) Gel jiocusing The gel is focussed for 60 min with 1 mol/l H,PO, at the anode and with 1 moi/l NaOH at the cathode, 3 W per 45 mm X 120 mm plate. The pH range generated is measured and the ‘running curve’ is determined. The samples are applied (15 ~1) at for 60 min. the pH range of - 8.0 and focusing continued (C) incubation and elation The plates are immersed in a 1 g/l solution of I-naphthylphosphate salt *, in 0.06 mol borate buffer/acetone (1 : 1, v/v) and incubated room temperature.
monosodium for 120 min at
(D) Staining and destaining Staining is performed in a solution * of 1 mg variamin blue + 5 mg CuSO, - 5H,O per ml 0.06 mol borate buffer, pH 9.7, for 30-60 min. For destaining, the plates are washed twice in a solution of 5 g/l CuSO, - 5H,O in distilled water/ethanol (1 : 1, v/v) for about 30 min each time, and then in water/ethanol (1 : 1, v/v) overnight. The next day the plates are immersed in a solution of glycerol/distilled water/ethanol (10 : 45 : 45, v/v/v) and covered with cellophane previously wetted with the above mixture. The plates are dried at room temperature. (E) Evaluation Evaluation is performed
using a densitometer.
Results
Seven groups of bands, each consisting of 1-3 lines, were identified in sera of healthy individuals. Comparison of extracts of organ homogenates enabled the identification of the renal, hepatic- 1, intestinal 1, -2, -3, and hepatic-2 groups (Fig. 1). Neither pulmonary nor bone marrow extracts yielded any visible fraction after isoefectric focussing although their AP levels were 200 U/l and 230 U/l respectively. The bone group No. 3 was indirectly identifiable by L-phenylalanine inhibition and heat inactivation (PI: 4.5). Another group No. 4 (PI: 4.6-4.8) behaving identically to the bone group with respect to heat and L-phenylalanine inhibition was established on clinical grounds to be of pulmonary origin (Table I, Fig. 2). The 140 patients showed 69 correct negative, two false negative and 14 false positive results. The intestinal group (5a, 5b, 5c) identified by means of jejunum extract shows three well separated bands in the pH range 5.0-5.9. We found the 6th group to be a second hepatic one, identified by means of a liver extract. In serum of healthy persons it seems to be responsible for about half of the total activity of AP. We also observed a 7th group, p1 7.0-7.5 whose organ source could not be identified. Angellis et al [ 171 have reported the appearance of a B form of purified placental type at this pH range. Since this group was present in almost all * These reagents
must be prepared
just before use.
160 ‘NO of the fraction
1 2 3 4
renal liver bone’ pulmonary’”
1 4.2-4.5 x
5 a, b. c intestinal
5b
5c
r
4.6-4.8
_
6
6 liver
7
7.0
7,
Fig. 1.pH distribution of the bands in the serum of healthy subjects. Total AP. 36 U/If 6. * Identified by means of heat inactivation only. ** Found to be increased in patients with lung diseases.
of our serum, it cannot be only of placental origin. Table II shows the percentage distribution of the total AP activity in the serum of healthy controls. Placental homogenates yielded a strong band in the pH range 3.6-4.2 designated by us ‘pre-1’. In all but two of the 70 patients with proven malignant neoplasms this additional band was found. It was never observed in healthy subjects but in serum of pregnant women in the last trimester. Its presence was independent of both the nature of the neoplasms and the level of AP. The pronounced heterogeneity observed in the pre-1 group is therefore not surprising. The appearance of this TABLE
I
PATIENTS
WITH LUNG
DISEASES
SHOWING
AN INCREASED
Carcinoma of the lung Pneumonia Emphysema Pulmonary embolism Chronic bronchitis Sarcoidosis
14
Total
33
1
1 4 6 1
‘PULMONARY’
FRACTION
161
PH
range
F1-l
r-
P-l
I_
;zI 14.2 -4.5
24.2-4.5 34.6
14.6-4.8
4.2-4.5
41
-4.8 5a 14.9 -5.4
5a
4.9-5.4
5
1
5b/
35.5-5.7
5cL-
7
5
/m-m
5c
5.8-6.0
6
6.0-6.8
6.0 -6.8
6.0-6.8
6
5b
i5.8 -6.0
5
i
5.5 -5.7
~
_I
7 i
7.0
7.0 -
-l
7
C
A Fig. 2. Bands and their pH range from: (A) a patient with carcinoma (total AP, 107 U/1); (B) a patient with carcinoma of the pancreas U/l); (C) a patient with bronchogenic cancer (total AP 37 U/l).
of the prostate and bone metastases and liver metastases (total AP 750
fraction varies from two or more fine bands to a smudged area (Fig. 2). It exhibits a different sensitivity to heat and to L-phenylalanine and L-homoarginine inhibition (probably related to different organ sources). We observed this group in sera of 26 of 70 patients with benign diseases (Table III). The reproducibility of the fractions in the serum of a patient with normal AP level and another one with an elevated level, is shown in Table IV. TABLE
II
THE PERCENTAGE ISOENZYME DISTRIBUTION SERUM OF 50 HEALTHY SUBJECTS Total AP x+
I
2.5 1.3
TOTAL
AP ACTIVITY
IN THE
SD: 36 f 6 U/l.
Renal
x SD*
OF THE
2 Liver
4.3 1.6
3 Bone
5.8 2.2
4 Pulmona~
5.4 0.65
5 Intestinal a
b
c
6.33 2.7
10.17 5.3
13.6 2.5
6 Liver-2
7
50.9 5.3
3.9 3
162 TABLE
III
TOTAL
SERUM
Note presence
AP IN SERUM
of pre-I
OF PATIENTS
WITH
BENIGN
group. pre-I present
pre- 1 absent
Total serum AP
Malignant proven Benign proven
TABLE
DISEASES
Total serum AP
elevated
normal
elevated
normal
43 18
25 8
0 22
2 22
IV
REPRODUCIBILITY LEVEL AND FROM
I. AP: 31 U/l n=8 x(g) SD+ II. AP: 121 U/l n = 14 X(S) SD-I:
OF THE FRACTION IN SERA FROM A PATIENT A PATIENT WITH AN ELEVATED LEVEL
p-l
1
2
0
1.9 0.9
2.1 0.1
7.9 1.1
3.9 0.7
6.0 1.1
3
WITH
A NORMAL
4
S(a+b+c)
6
4.0 0.7
4.2 0.9
30.2 4.3
48.4 3.1
8.8 3.1
15.6 2.6
7.9 1.2
17.8 2.5
37.9 2.8
15.3 28.2
AP
7
Discussion Separation of organ-specific API fractions has been attempted since 1959 by the application of different electrophoretic techniques. In Fishman’s [l] review article these experiments and results are discussed in detail. None of these methods are, however, really suitable for clinical use. Their common problem is that they may show similar characteristics for isoenzymes produced by different organs, and even genuine differences did not show satisfactory separation. For instance the method of Siede et al [3] can produce three, or under certain circumstances four different fractions, (liver-l, bone, liver-2 and possibly a pathological bile fraction). All the API are distributed in these fractions. Results of different methods are thus often contradictory. Working with serum there is no real possibility of comparison with the other methods because of the much higher solution-capacity of the IEF. This, on the other hand, leads to additional problems of correct source allocation. As described above this method produces 10-15 bands within seven groups in normal serum and an additional ‘pre-I’ in pathological serum. Six of the seven groups have been traced to organ sources either directly or indirectly as described. The 7th normal group has not been shown to be due to any organ source tested so far. We could identify the bone fraction only by means of heat inactivation of serum of healthy subjects and of patients with bone diseases or
163
bone metastases. The bone marrow itself did not yield any group although its total activity was 200 U/l (reference range: 10-50). The amount of this isoenzyme is much lower as usually reported because the usual electrophoretic methods with serum are unable to separate other isoenzymes (in part liver, placental and probably this one which we call pulmonary) from that of the bone. Similarly the heat inactivation method cannot differentiate between the loss of bone and liver fractions, in which the loss of the latter is estimated to be 60%. We had only 5 patients with Paget’s disease. They were not severe cases (AP levels of 36-130 U/l) but the average value of their bone fraction was 11.75% + 3.1; healthy controls have 5% f 2. Serum from patients with pulmonary diseases, benign or malignant (Fig. 2C) show 2 or 3 bands in the pH range 4.8 + 5.0 which are barely visible in the serum of healthy subjects (see Fig. 1). We presume therefore that these bands originate mainly in the lungs. Both of these groups will be further investigated, by combining them with immunoprecipitation methods. A particular problem is encountered with ‘tumour-specific’ API groups. Such an API called ‘Regan isoenzyme’ was described by Fishman et al [ 181 and another one called Nagao described by Nakayama et al [ 191. Physical and chemical properties showed the Regan isoenzyme to be identical with the isoenzymes prepared from placenta [9]. It is heat stable and L-phenylalanine sensitive as distinct from our ‘pre-1’ fraction. Our ‘pre-1’ fraction seems to be similar but not identical to placental and Regan isoenzymes because ‘pre-1’ was never found in healthy subjects. Usategui-Gomez et al demonstrated Regan isoenzymes in 50 healthy subjects [20]. ‘Pre-1’ is present as Regan isoenzyme in placental homogenates and in the serum of pregnant women in the last trimester, but its heat stability and L-phenylalanine sensitivity varies. Iglis et al stated that Regan isoenzyme is not exclusively a property of cancer patients, but has been identified at low levels in populations with other diseases [8,9]. We have also found the ‘pre-1’ group as described above, in sera of patients with benign diseases (Table III). All these findings are probably due to the different types of isoenzymes in this particular pH range [lo- 131. For clarification, further investigations are necessary. In accord with earlier publications [ 1,14-161 the hepatic-2 group is responsible for approximately half of the total AP activity. In serum from patients with liver metastases this band always becomes very faint (Fig. 2B). The reproducibility of the results critically depends on the carrier ampholyte. It is very important in new batches to check its behaviour with known serum. It is possible to adjust it by changes in its composition. Our method can be recommended for clinical laboratory use because it is simple to perform. The time and money involved are reasonable, and the sample size is low. Acknowledgements The authors thank Prof. Dr. R. Siebenmann, Chief of Department of Pathology, Stadtspital Triemli, Zurich, for supplying materials for this work. This work was supported in part by the Swiss National Foundation Grant, No. 3.940.0.80.
164
References 1 Fishman WH. Perspectives on alkaline phosphatase isoenzymes. Am J Med 1974; 56: 617-650. 2 Bauer John D. Clinical laboratory methods, 9th Edition. St. Louis, Toronto, London: C.V. Mosby Company, 1982: 582. 3 Siede WH, Seiffert UB. Quantitative alkaline phosphatase isoenzyme determination by electrophoresis on cellulose acetate membranes. Clin Chem 1977; 23/l: 28-34. 4 Righetti P, Drysdale JW. Isoelectric focusing in polyacrylamide gels Biochem Biophys Acta 197 I ; 236: 17-28. 5 Bessey OA, Lowry OH, Brock ME. Determination of alkaline phosphatase in human serum. J Biol Chem 1964; 164: 321. 6 Smith I, Lightstone PJ, Perry JO. Purification of alkaline phosphatase. Clin Chim Acta 1968; 19: 499-505. 7 Radola BJ. Electrophoresis 1980; 1: 43-56. 8 Anstiss CL, Green S, Fishman WH. An automated technique for segregation populations with a high incidence of Regan isoenzyme in serum. Clin Chim Acta 1974; 33: 279. 9 Inglis NR, Kirley S, Stolbach LL, Fishman WH. Phenotypes of the Regan isoenzyme and identity between the placental D variant and the Nagao isoenzyme. Cancer Res 1973; 33: 1657. 10 Harkness DR. Studies on human placental alkaline phosphatase. Arch Biochem Biophys 1968; 126: 513. 11 Fishman WH, Sie HG. Organ specific inhibition of human alkaline phosphatase isoenzymes of liver, bone, intestine and placenta, r_-phenylalanine, L-tryptophan and L-homoarginine. Enzymologia 1971; 41: 141. 12 Damle SR, Shelly PA, Jussavalla D, Bhide SV, Baxi AJ. Occurrence of heat stabile Regan type of alkaline phosphatase in hematopoietic tumours. Int J Cancer 1979; 24/4: 398-401. 13 Fishman WH, Driscoll S, Miyayama H, Fishman L. Human tumour non-Regan alkaline phosphatase isoenzyme. Proc Am Assoc Cant Res 1975; 16: 195. 14 Gutman AB. Serum alkaline phosphatase activity in diseases of the skeletal and hepatobihary systems. Am J Med 1959; 27: 875. 15 Posen S, Neele FC, Clubb JS. Heat inactivation in the study of human alkaline phosphatases. Am Int Med 1965; 62: 1234. 16 Posen S. Alkaline phosphatase. Am Int Med 1967; 67: 189. 17 Angellis D, Inglis NR, Fishman WH. Isoelectric focusing of alkaline phosphatase isoenzymes in polyacrylamide gels. Am J Clin Path01 1976; 66: 929-934. 18 Fishman WH, Inglis NR, Stolbach LL, Krant MJ. Cancer Res 1968; 28: 150. 19 Nahayama T, Yoshida M, Kitamura M. Clin Chim Acta 1970; 30: 546. 20 Usategui-Gomez M, Yeager FM, Fernandez de Costa A. Cancer Res 1974; 34: 2544.