253
Clinica Chimica Actu, 98 (1979) 253-261 @ Elsevier/North-Holland Biomedical Press
CCA 1154
AN ION-EXCHANGE ASSAY FOR HIGH MOLECULAR ALKALINE PHOSPHATASE
PATRICIA
M. CROFTON
Deportment
(Received
of Clinical
June 28th,
WEIGHT
* and A.F. SMITH
Chemistry,
Royal
Infirmory,
Edinburgh
EH3
9YW
(U.K.)
1979)
Summary An ion-exchange method for the measurement of a high molecular weight form of alkaline phosphatase in serum is described. The method is simple, rapid, precise and suitable for processing small batches of samples. Other forms of alkaline phosphatase commonly encountered in serum do not interfere. The correlation between results obtained by this method and those obtained by Sephadex 6B chromatography is discussed. Electrophoretic methods of measurement were also investigated but were found to be both imprecise and inaccurate.
Introduction The alkaline phosphatase (ALP) isoenzymes which appear most commonly in serum are those of liver, bone and intestinal origin. In this context, the term “liver isoenzyme” is used to refer to the low molecular weight (approximately 240 000) ALP from liver. Analysis of these isoenzymes has proved to be a useful diagnostic tool [ 11. In addition to these common isoenzymes which are present in varying proportions in the serum of healthy individuals, there have been reports in recent years of a high molecular weight (mol. wt.) ALP which is not present in health but which appears in the serum of patients with a variety of diseases affecting the liver [ 2,3]. This “abnormal” ALP is excluded from the gel matrix on Sephadex G200 chromatography [ 41, remains at or near the origin on starch gel [ 51 and polyacrylamide gel [ 61 electrophoresis and migrates ahead of the normal liver isoenzyme on cellulose acetate electrophoresis [ 71. A fairly simple method of quantitating the high mol. wt. ALP, based on densitometry of stained isoenzyme bands after electrophoresis, has been described
* To
whom
comespondence
should
be addressed.
254
[8]. However, measurement of ALP activity in the fractions eluted from a Sephadex G200 column must be regarded as the most reliable and accurate method of measurement. The disadvantage of this method is that it is tedious and unsuitable for use with large numbers of samples. This paper presents an alternative method for the measurement of high mol. wt. ALP based on ionexchange chromatography and assesses the performance of both the new method and of a number of methods based on electrophoretic separation of the isoenzymes. Materials and methods Measurement of ALP activity ALP activity in whole serum was measured on the Technicon Sequential Multiple Analysis plus Computer (SMAC) system using as substrate 10 mmol/l p-nitrophenyl phosphate in 0.6 mol/l 2-amino-2-methyl-l-propanol buffer pH 10.2 containing 0.5 mmol/l magnesium chloride. The between batch coefficient of variation was 2.5%. A continuous monitoring (kinetic) method was used to measure ALP activity in column effluents, to obtain greater sensitivity. After mixing the column effluent with an equal volume of 1.8 mol/l diethanolamine buffer pH 10.2 containing 1 mmol/l magnesium chloride, p-nitrophenyl phosphate was added to give a final concentration of 14 mmol/l. The reaction was monitored at 410 nm at 37°C in an LKB 8600 reaction rate analyser. Electrophoretic methods for quantitation of high mol. wt. ALP Electrophoresis of serum ALP was carried out on 7% and 2.5% polyacrylamide gel slabs using a modification [9] of the method of Kaplan and Rogers [lo]. Other electrophoretic media investigated were agarose gel [ 111 and cellulose acetate [ 71. Cellulose acetate electrophoresis was followed by two kinds of staining procedure: (a) using a thin film of solution trapped between the membrane and a glass slide, and (b) using a technique whereby the membrane was imprinted face down on an agar gel containing substrate and dye [7]. In every case, the isoenzyme bands were visualized using 0.75 mmol/l cw-naphthyl phosphate as substrate and 0.5 g/l 4-aminodiphenylamine diazonium sulphate as dye. The isoenzyme bands were quantitated by scanning on a Vitatron densitometer. Ionexchange method for quantitation of high mol. wt. ALP Apparatus. Small columns were employed using 20 ml disposable plastic syringe barrels (internal diameter 2 cm). The gel matrix was supported by small circles of gauze (Into tissues, Robinsons, Chesterfield, U.K.). The upper seal of each column was formed by the original rubber syringe plunger end, through which was passed a small-diameter Teflon tube connected to a peristaltic pump. Reagents 1. DEAE-cellulose pre-cycled and equilibrated with buffer A; 2. Buffer A: 0.1 mol/l NaCl in 0.01 mol/l Tris-HCl buffer pH 7.5; 3. Buffer B: 0.3 mol/l NaCl in 0.01 mol/l Tris-HCl buffer pH 7.5. Assay protocol. 20 ml DEAE-cellulose was poured into each column and
255
washed through with Buffer A for two hours at a flow rate of ‘78 ml/h in order to achieve complete equilibration. The final bed volume was approximately 10 ml. After equilibration the supernatant was removed and replaced with 1 ml serum which had been dialysed overnight against Buffer A. The serum was allowed to soak into the resin and was replaced with approximately 4 ml Buffer A. Buffer A (30 ml) was then pumped through the column, and a single 30 ml fraction collected, designated Fraction I. Pumping was stopped, the supematant removed and replaced with approximately 4 ml Buffer B. A second elution using Buffer B was carried out, a single 30 ml fraction again being collected, designated Fraction II (this fraction contained the high mol. wt. ALP). The ALP activities of the undialysed serum, Fraction I and Fraction II were measured. The activity of high mol. wt. ALP was calculated from the formula: (activity Fraction II) X (activity of serum) U/l_’ = -..-. (activity Fraction I + activity Fraction II) It was possible to analyse up to 8 sera in each batch. The whole assay, including me~urement of activities and c~culation of results, takes about 2 h to perform for each batch. High mol. wt. ALP activity
Validation of conditions Effect of ionic strength.
5 ml serum containing liver was applied to a 16 X 2.5 cm DEAE-cellulose column dient of increasing ionic strength, using NaCl in 0.01 The elution profile showed two main peaks of activity
U
Alkaline
phosphatase
activity liver
___
Serum
and high mol. wt. ALP and eluted with a gramol/l Tris-HCl pH 7.5. (Fig. 1). The fractions
alk.
phos.
protein
0 0 20 I.
i i I f \ f t ,
8
.
0 f
3 40% 5 a Y
,
60 9
\ /
t
f
9
I 80
t
P e
\ :
\ y+-+%100 Elution
Fig. 1. Elution chromatography
200 volume
100
300 (ml)
of a serum containing liver and high mol. wt. ALP obtained by DEAE-cetlulose on a 16 X 2.5 cm column.
pattern
256
300 f BUFFER
A
BUFFER
B
J liver alk. phos.
high M.W.
30
40
50
Elution volume Fig. 2. Elution pattern of a serum containing exchange method. See text for details.
SO
10
80
alk. phos.
so
(ml)
liver and high mol. wt. ALP obtained
by the two-step
ion-
corresponding to each of these peaks were pooled, concentrated and applied to a Sepharose 6B column. They were also subjected to electrophoresis on 7% polyacrylamide gel together with a serum marker containing liver and high mol. wt. alkaline phosphatase. By these means, it was shown that the first ionexchange peak correspond to the lower mol. wt. liver isoenzyme while the second peak corresponded to the high mol. wt. component. The ionic strength co~esponding to the trough between the two peaks was 0.1 mol/l (the ionic strength of Buffer A). Vohne of fractions. The two-step ion-exchange method was carried out on a serum containing liver and high mol. wt. ALP except that, instead of collecting a single 30 ml fraction at each ionic strength, 20 X 2.5 ml fractions were collected. ALP activity was measured in each of these fractions and the elution profile shown in Fig. 2 was obtained. It can be seen that elution of the activity corresponding to each ionic strength was complete when 30 ml had been collected. This volume was therefore chosen for Fractions I and II. Identification of Fractions I and IK; assessment of cross-contamination. A serum containing liver and high mol. wt. ALP only was analysed by the twostep ion exchange method. Fractions I and II were each concentrated and subjected to electrophoresis on 2.5% polyac~lamide gel; this allows penetration of high mol. wt. ALP into the gel matrix. The undialysed serum was also subjected
257
b Fig. 3. ALP electrophoresis patterns on 2.5% polyacrylamide gel. From left to right: (a) serum containing liver and high mol. wt. ALP; (b) Fraction I (concentrated) from this serum: Cc) Fraction II (concentrated) from this senzm. The origin is at the top. The band with highest mobility corresponds to the liver isoenzyme; the band near the origin corresponds to the high mol. wt. component. See text for details.
to electrophoresis as a marker. Fig. 3 shows that Fraction I corresponds to the liver isoenzyme and Fraction II to the high mol. wt. component. There is no detectable cross-contamination or carry-over. Investigation of interference by other ALP isoenzymes. When the two-step ion-exchange method was carried out on 51 sera shown by eiectrophoresis to contain only liver and high mol. wt. ALP in varying proportions, the percentage activity found in Fraction II ranged from 4 to 35% (6 to 318 U/i). In 24 sera, shown by eiectrophoresis to contain undet~tabie levels of high mol. wt. ALP but varying proportions of liver, bone, intestinal and placental isoenzymes, the percentage activity found in Fraction II ranged from 1 to 4% (1 to 9 U/i). One serum contained the rare IgG-ALP complex [12] and gave 4% activity (‘7 U/i) in Fraction II. Another serum contained the Regan variant (Kasahara isoenzyme f 133, a rare cancer-associated isoenzyme, in addition to the liver isoenzyme; this serum gave 9% (16 U/l) activity in Fraction II. ~iectrophoresis of Fraction II demonstrated that it contained only the Regan variant and no high mol. wt. component was present. Precision and yield. Yields from the ion-exchange columns ranged from 85105% with a mean of 95% (N = 11). The between-batch coefficient of variation (including the dialysis step) for high mol. wt. ALP measured by the two-step ion-exchange method was 3.4%
;0 High M.W.
Alk.
Phos.
;0
(SC) by Sepharose
;0 6B chromatoRraphy
between high mol. wt. ALP activities measured by Sepharose 6B chromatography (X-axis) and ion-exchange chromatography (Y-axis) in 6 subjects. - - - 45* line.
Fig. 4. Comparison
(N = 30, mean activity comparison
102 U/l, range 6 to 350 U/l),
of con-exchange
with Sepharose 6B chromatography. Six sera with varying proportions of high mol. wt. component were analysed by the two-step ion-exchange method and by Sepharose 6B chromatography. ALP activities were measured in the fractions corresponding to the liver and high mol. wt. component peaks respectively. Correspondence between the two methods is good (Fig. 4). Although the ion-exchange method appears to give a slightly higher estimate of the high mol. wt. component this is probably within the limits of analytical error for the two methods. Value of electrophoretic methods for quan titation. Preliminary experiments using sera containing various amounts of high mol. wt. ALP, and using partially purified preparations of liver and high mol. wt. ALP respectively, revealed that the high mol. wt. component which eluted in the void volume during Sepharose 63 chromatography was identical to: (1) the ALP peak which eluted at an ionic strength of 135 mmol/l during salt gradient DEAE-cellulose chromatography; (2) the enzyme activity present in Fraction II in the two-step ion-exchange assay described;
259
(3) the activity remaining at the origin after electrophoresis in 7% polyacrylamide gel; (4) the enzyme band which migrated a little way into the gel during electrophoresis in 2.5% polyacrylamide gel; (5) the activity band which migrated immediately behind the liver isoenzyme during electrophoresis in 1% agarose gel, and between the intestinal isoenzyme and the origin in 2.2% agarose gel; and (6) the enzyme band which migrated ahead of the liver isoenzyme in the LY ,-globulin position during electrophoresis in cellulose acetate. The precision of the electrophoretic methods was assessed initially by repeated within-run analysis of single specimens. Coefficients of variation (CV) of about. 4% were obtained for polyacrylamide and agarose gels, whereas cellulose acetate gave much poorer precision and was therefore not studied further. Between batch precision was assessed by repeat analysis of sera from patients; for 2.2% agarose the CV was 27.5% (N = 14, mean: 16.4%, range: 2.7-40.5%) and for 2.5% polyacrylamide gel the CV was 14.6% (N = 9, mean: 9.7%, range:’ 2.4-17.7%). Electrophoresis of six sera was carried out in various media and the high mol. wt. ALP estimated by scanning densitomet~ (Table I). With the exception of the agar imprinting technique used with cellulose acetate strips, roughly comparable results for the different methods were obtained. The overlay technique gave results which were about twice as high as those with other methods. This was attributed to the slower rate of diffusion of the high mol. wt. ALP into the agarose, leaving more of this enzyme on the strip which was subsequently scanned for quantitation of the isoenzymes: the overlay, on the other hand, showed denser staining in the region of the lower mol. wt. ALP. Because of these problems of variable diffusion rates and poor reproducibility neither agarose gel nor cellulose acetate were considered suitable media for quantitation. Polyacrylamide gel, with its property of reducing diffusion to a minimum and its better precision, appeared to be more suitable for routine measurements. However, comparison with the reference method of Sepharose 6B chromatography showed that electrophoretid estimates of the high mol. wt.
TABLE I PERCENTAGE OF HIGH MOL. WT. ALP IN 6 SERA, %STIMATED BY SCANNING FOLLOWING ELECTROPHORESIS AND STAINING IN THREE MEDIA SerLlm
Total ALP activity (U/l)
DENSITOMETRY,
High mol. wt. ALP (W total) 2.2% Agarose gel
2.5% Polyacrylamide gel
Cellulose acetate Thin-film staining
Agar template staining
1 2 3
144 304 544
26 32 17
20 30 21
22 32 33
48 56 49
4 5 6
216 568 520
24 3 8
17 4 5
15 14 14
50 15 24
__
260
ALP gave values which, especially for low activity samples, were often less than half those obtained by Sepharose 6B chromatography. Further investigation revealed that this difference was caused by the different substrate specificities of liver and high mol. wt. ALP. cu-Naphthyl phosphate (used in the electrophoretic methads) gave lower activities of the high mol. wt. ALP, relative to the liver isoenzyme, than did p-nitrophenyl phosphate (used in the kinetic methods). Discussion There are three principal methods at present available for measuring high mol. wt. ALP. The first is gel chromato~aphy of a single sample [4], followed by measurement of activities in a large number of fractions, an accurate reference method but slow and tedious. The second is elution from sequential strips of electrophoresis media, usually polyacrylamide [6] or starch gel, followed by measurement of activities in each strip, again a time-consuming procedure. Furthermore, recovery of enzyme activity from macerated gel may vary from isoenzyme to isoenzyme, from serum to serum and from gel to gel. In particular, it seems likely that large molecules will be more difficult to recover from the gel than small molecules and that the proportion may be critically dependent on the characteristics of the gel, such as its strength and degree of crosspolymerisation, as well as the conditions of elution. A third method of quantitation has used electrophoresis on cellulose acetate strips followed by densitometry [ 81. We found this technique to be very imprecise, presumably owing to diffusion and variable background staining. In addition, the agar overlay technique was inaccurate, since it overestimated the amount of high mol. wt. ALP in the serum. Electrophoresis on agarose gel was also rather imprecise whereas the precision with polyacrylamide gel was adequate. However, the results obtained with the electrophoretic methods were much lower than those obtained by Sepharose 6B chromatography. This could be partly attributed to the different substrate specificities of the high and low mol. wt. isoenzymes. Electrophoretic methods have the additional disadvantage that hydrolysis of a-naphthyl phosphate by ALP is inhibited by most of the coupling dyes used to visualize the isoenzymes; this may cause non-line~ity of colour development. We therefore considered the simple electrophoretic techniques to be unsuitable for accurate and precise measurement of high mol. wt. ALP. The ion-exchange method presented in this paper separates and quantitates high mol. wt. ALP, not on the basis of its high mol. wt., but on the basis of its difference in charge from most other ALP isoenzymes. The common isoenzymes of ALP, namely liver, bone, intestinal and placental, and the IgG-ALP complex [12] which is occasionally encountered, do not appear to interfere with quantitation of the high mol. wt. component by this method. Only the Regan variant [13], as might be expected from its high electrophoretic mobility and presumably greater charge, elutes in Fraction II. This is a rare cancer-associated isoenzyme, much less commonly encountered than the better known Regan isoenzyme, and the proportion of cases in which its presence would interfere with the method described here is likely to be very
261
small. To detect such cases, 7% polyacrylamide gel electrophoresis could be performed on all sera to be analysed. If the Regan variant were to be found, the ion-exchange method could not be used to quantitate the high mol. wt. component, and Sephadex G200 (or Sepharose 6B) chromatography should be used instead. There is fairly good correlation between estimates of high mol. wt. ALP obtained by the ion-exchange method and by Sepharose 6B chromatography, considering that the two methods are based on different properties of the high mol. wt. component. Although the slightly higher results obtained by the ionexchange method could be due to a small amount of carry-over, the evidence presented in this paper fails to support this. Another possible explanation is that some degree of denaturation of high mol. wt. ALP may occur on the Sepharose 6B column, since the enzyme is in contact with the column matrix for a longer period of time and more dilution takes place. However, it is also possible that the slight discrepancy between the two methods is within the limits of their analytical error. In conclusion, the method presented here provides a simple, relatively rapid, reliable and precise measurement of the high mol. wt. component of ALP in small batches of sera. References 1
Warnes,
2
Fennelly,
T.W.
3
Price.
(1972)
J.J.,
C.P.
4
Estborn.
5
DuBuisson,
6
Price,
M.X.
and
H.G.
(1974)
B. (1964)
Z. Klin.
J. and and
7
Rhone,
D.P.
8
Rhone.
D.P.,
9
Warwick,
13.926-937
Sammons.
and
C.P.
Gut
Fitzgerald.
Pepler,
Sammons.
and
Mizuno,
White,
R.R.G..
F.M.
Chem.
McGeeney,
K.
J. Clin.
(1966)
S. Afr.
H.G.
(1976)
J. Clin.
(1973)
Am.
F.M.
Shearman.
Gut
27,
10.45-51
392-398
2, 53-54
W.J.
and
(1969)
Pathol.
Gidaspow, D.J.C.,
Med.
J. Suppl.,
Pathol.
J. Clin.
H. (1973)
Percy-Robb.
29.
Pathol. Clin. I.W.
696498
976-980 59,
531-541
Chem.
and
19.
Smith,
175 10
Kaplan,
11
Ewen,
M.M.
12
Crofton,
P.M.
13
Crofton.
P.M.
L.M.
and (1974)
Rogers.
L. (1969)
Lancet
Pathol.
ii, 1029-1031
Am.
J. Clin.
61,142-154
and
Smith,
A.F.
(1978)
Clin.
Chim.
Acta
83,
235-247
and
Smith,
A.F.
(1978)
Clin.
Chim.
Acta
86,
81-88
1142-1147 A.F.
(1972)
Scot.
Med.
J. 17.
172-
Clinica Chimica Actu, 98 (1979) 253-261 @ Elsevier/North-Holland Biomedical Press
CCA 1154
AN ION-EXCHANGE ASSAY FOR HIGH MOLECULAR ALKALINE PHOSPHATASE
PATRICIA
M. CROFTON
Deportment
(Received
of Clinical
June 28th,
WEIGHT
* and A.F. SMITH
Chemistry,
Royal
Infirmory,
Edinburgh
EH3
9YW
(U.K.)
1979)
Summary An ion-exchange method for the measurement of a high molecular weight form of alkaline phosphatase in serum is described. The method is simple, rapid, precise and suitable for processing small batches of samples. Other forms of alkaline phosphatase commonly encountered in serum do not interfere. The correlation between results obtained by this method and those obtained by Sephadex 6B chromatography is discussed. Electrophoretic methods of measurement were also investigated but were found to be both imprecise and inaccurate.
Introduction The alkaline phosphatase (ALP) isoenzymes which appear most commonly in serum are those of liver, bone and intestinal origin. In this context, the term “liver isoenzyme” is used to refer to the low molecular weight (approximately 240 000) ALP from liver. Analysis of these isoenzymes has proved to be a useful diagnostic tool [ 11. In addition to these common isoenzymes which are present in varying proportions in the serum of healthy individuals, there have been reports in recent years of a high molecular weight (mol. wt.) ALP which is not present in health but which appears in the serum of patients with a variety of diseases affecting the liver [ 2,3]. This “abnormal” ALP is excluded from the gel matrix on Sephadex G200 chromatography [ 41, remains at or near the origin on starch gel [ 51 and polyacrylamide gel [ 61 electrophoresis and migrates ahead of the normal liver isoenzyme on cellulose acetate electrophoresis [ 71. A fairly simple method of quantitating the high mol. wt. ALP, based on densitometry of stained isoenzyme bands after electrophoresis, has been described
* To
whom
comespondence
should
be addressed.
254
[8]. However, measurement of ALP activity in the fractions eluted from a Sephadex G200 column must be regarded as the most reliable and accurate method of measurement. The disadvantage of this method is that it is tedious and unsuitable for use with large numbers of samples. This paper presents an alternative method for the measurement of high mol. wt. ALP based on ionexchange chromatography and assesses the performance of both the new method and of a number of methods based on electrophoretic separation of the isoenzymes. Materials and methods Measurement of ALP activity ALP activity in whole serum was measured on the Technicon Sequential Multiple Analysis plus Computer (SMAC) system using as substrate 10 mmol/l p-nitrophenyl phosphate in 0.6 mol/l 2-amino-2-methyl-l-propanol buffer pH 10.2 containing 0.5 mmol/l magnesium chloride. The between batch coefficient of variation was 2.5%. A continuous monitoring (kinetic) method was used to measure ALP activity in column effluents, to obtain greater sensitivity. After mixing the column effluent with an equal volume of 1.8 mol/l diethanolamine buffer pH 10.2 containing 1 mmol/l magnesium chloride, p-nitrophenyl phosphate was added to give a final concentration of 14 mmol/l. The reaction was monitored at 410 nm at 37°C in an LKB 8600 reaction rate analyser. Electrophoretic methods for quantitation of high mol. wt. ALP Electrophoresis of serum ALP was carried out on 7% and 2.5% polyacrylamide gel slabs using a modification [9] of the method of Kaplan and Rogers [lo]. Other electrophoretic media investigated were agarose gel [ 111 and cellulose acetate [ 71. Cellulose acetate electrophoresis was followed by two kinds of staining procedure: (a) using a thin film of solution trapped between the membrane and a glass slide, and (b) using a technique whereby the membrane was imprinted face down on an agar gel containing substrate and dye [7]. In every case, the isoenzyme bands were visualized using 0.75 mmol/l cw-naphthyl phosphate as substrate and 0.5 g/l 4-aminodiphenylamine diazonium sulphate as dye. The isoenzyme bands were quantitated by scanning on a Vitatron densitometer. Ionexchange method for quantitation of high mol. wt. ALP Apparatus. Small columns were employed using 20 ml disposable plastic syringe barrels (internal diameter 2 cm). The gel matrix was supported by small circles of gauze (Into tissues, Robinsons, Chesterfield, U.K.). The upper seal of each column was formed by the original rubber syringe plunger end, through which was passed a small-diameter Teflon tube connected to a peristaltic pump. Reagents 1. DEAE-cellulose pre-cycled and equilibrated with buffer A; 2. Buffer A: 0.1 mol/l NaCl in 0.01 mol/l Tris-HCl buffer pH 7.5; 3. Buffer B: 0.3 mol/l NaCl in 0.01 mol/l Tris-HCl buffer pH 7.5. Assay protocol. 20 ml DEAE-cellulose was poured into each column and
255
washed through with Buffer A for two hours at a flow rate of ‘78 ml/h in order to achieve complete equilibration. The final bed volume was approximately 10 ml. After equilibration the supernatant was removed and replaced with 1 ml serum which had been dialysed overnight against Buffer A. The serum was allowed to soak into the resin and was replaced with approximately 4 ml Buffer A. Buffer A (30 ml) was then pumped through the column, and a single 30 ml fraction collected, designated Fraction I. Pumping was stopped, the supematant removed and replaced with approximately 4 ml Buffer B. A second elution using Buffer B was carried out, a single 30 ml fraction again being collected, designated Fraction II (this fraction contained the high mol. wt. ALP). The ALP activities of the undialysed serum, Fraction I and Fraction II were measured. The activity of high mol. wt. ALP was calculated from the formula: (activity Fraction II) X (activity of serum) U/l_’ = -..-. (activity Fraction I + activity Fraction II) It was possible to analyse up to 8 sera in each batch. The whole assay, including me~urement of activities and c~culation of results, takes about 2 h to perform for each batch. High mol. wt. ALP activity
Validation of conditions Effect of ionic strength.
5 ml serum containing liver was applied to a 16 X 2.5 cm DEAE-cellulose column dient of increasing ionic strength, using NaCl in 0.01 The elution profile showed two main peaks of activity
U
Alkaline
phosphatase
activity liver
___
Serum
and high mol. wt. ALP and eluted with a gramol/l Tris-HCl pH 7.5. (Fig. 1). The fractions
alk.
phos.
protein
0 0 20 I.
i i I f \ f t ,
8
.
0 f
3 40% 5 a Y
,
60 9
\ /
t
f
9
I 80
t
P e
\ :
\ y+-+%100 Elution
Fig. 1. Elution chromatography
200 volume
100
300 (ml)
of a serum containing liver and high mol. wt. ALP obtained by DEAE-cetlulose on a 16 X 2.5 cm column.
pattern
256
300 f BUFFER
A
BUFFER
B
J liver alk. phos.
high M.W.
30
40
50
Elution volume Fig. 2. Elution pattern of a serum containing exchange method. See text for details.
SO
10
80
alk. phos.
so
(ml)
liver and high mol. wt. ALP obtained
by the two-step
ion-
corresponding to each of these peaks were pooled, concentrated and applied to a Sepharose 6B column. They were also subjected to electrophoresis on 7% polyacrylamide gel together with a serum marker containing liver and high mol. wt. alkaline phosphatase. By these means, it was shown that the first ionexchange peak correspond to the lower mol. wt. liver isoenzyme while the second peak corresponded to the high mol. wt. component. The ionic strength co~esponding to the trough between the two peaks was 0.1 mol/l (the ionic strength of Buffer A). Vohne of fractions. The two-step ion-exchange method was carried out on a serum containing liver and high mol. wt. ALP except that, instead of collecting a single 30 ml fraction at each ionic strength, 20 X 2.5 ml fractions were collected. ALP activity was measured in each of these fractions and the elution profile shown in Fig. 2 was obtained. It can be seen that elution of the activity corresponding to each ionic strength was complete when 30 ml had been collected. This volume was therefore chosen for Fractions I and II. Identification of Fractions I and IK; assessment of cross-contamination. A serum containing liver and high mol. wt. ALP only was analysed by the twostep ion exchange method. Fractions I and II were each concentrated and subjected to electrophoresis on 2.5% polyac~lamide gel; this allows penetration of high mol. wt. ALP into the gel matrix. The undialysed serum was also subjected
257
b Fig. 3. ALP electrophoresis patterns on 2.5% polyacrylamide gel. From left to right: (a) serum containing liver and high mol. wt. ALP; (b) Fraction I (concentrated) from this serum: Cc) Fraction II (concentrated) from this senzm. The origin is at the top. The band with highest mobility corresponds to the liver isoenzyme; the band near the origin corresponds to the high mol. wt. component. See text for details.
to electrophoresis as a marker. Fig. 3 shows that Fraction I corresponds to the liver isoenzyme and Fraction II to the high mol. wt. component. There is no detectable cross-contamination or carry-over. Investigation of interference by other ALP isoenzymes. When the two-step ion-exchange method was carried out on 51 sera shown by eiectrophoresis to contain only liver and high mol. wt. ALP in varying proportions, the percentage activity found in Fraction II ranged from 4 to 35% (6 to 318 U/i). In 24 sera, shown by eiectrophoresis to contain undet~tabie levels of high mol. wt. ALP but varying proportions of liver, bone, intestinal and placental isoenzymes, the percentage activity found in Fraction II ranged from 1 to 4% (1 to 9 U/i). One serum contained the rare IgG-ALP complex [12] and gave 4% activity (‘7 U/i) in Fraction II. Another serum contained the Regan variant (Kasahara isoenzyme f 133, a rare cancer-associated isoenzyme, in addition to the liver isoenzyme; this serum gave 9% (16 U/l) activity in Fraction II. ~iectrophoresis of Fraction II demonstrated that it contained only the Regan variant and no high mol. wt. component was present. Precision and yield. Yields from the ion-exchange columns ranged from 85105% with a mean of 95% (N = 11). The between-batch coefficient of variation (including the dialysis step) for high mol. wt. ALP measured by the two-step ion-exchange method was 3.4%
;0 High M.W.
Alk.
Phos.
;0
(SC) by Sepharose
;0 6B chromatoRraphy
between high mol. wt. ALP activities measured by Sepharose 6B chromatography (X-axis) and ion-exchange chromatography (Y-axis) in 6 subjects. - - - 45* line.
Fig. 4. Comparison
(N = 30, mean activity comparison
102 U/l, range 6 to 350 U/l),
of con-exchange
with Sepharose 6B chromatography. Six sera with varying proportions of high mol. wt. component were analysed by the two-step ion-exchange method and by Sepharose 6B chromatography. ALP activities were measured in the fractions corresponding to the liver and high mol. wt. component peaks respectively. Correspondence between the two methods is good (Fig. 4). Although the ion-exchange method appears to give a slightly higher estimate of the high mol. wt. component this is probably within the limits of analytical error for the two methods. Value of electrophoretic methods for quan titation. Preliminary experiments using sera containing various amounts of high mol. wt. ALP, and using partially purified preparations of liver and high mol. wt. ALP respectively, revealed that the high mol. wt. component which eluted in the void volume during Sepharose 63 chromatography was identical to: (1) the ALP peak which eluted at an ionic strength of 135 mmol/l during salt gradient DEAE-cellulose chromatography; (2) the enzyme activity present in Fraction II in the two-step ion-exchange assay described;
259
(3) the activity remaining at the origin after electrophoresis in 7% polyacrylamide gel; (4) the enzyme band which migrated a little way into the gel during electrophoresis in 2.5% polyacrylamide gel; (5) the activity band which migrated immediately behind the liver isoenzyme during electrophoresis in 1% agarose gel, and between the intestinal isoenzyme and the origin in 2.2% agarose gel; and (6) the enzyme band which migrated ahead of the liver isoenzyme in the LY ,-globulin position during electrophoresis in cellulose acetate. The precision of the electrophoretic methods was assessed initially by repeated within-run analysis of single specimens. Coefficients of variation (CV) of about. 4% were obtained for polyacrylamide and agarose gels, whereas cellulose acetate gave much poorer precision and was therefore not studied further. Between batch precision was assessed by repeat analysis of sera from patients; for 2.2% agarose the CV was 27.5% (N = 14, mean: 16.4%, range: 2.7-40.5%) and for 2.5% polyacrylamide gel the CV was 14.6% (N = 9, mean: 9.7%, range:’ 2.4-17.7%). Electrophoresis of six sera was carried out in various media and the high mol. wt. ALP estimated by scanning densitomet~ (Table I). With the exception of the agar imprinting technique used with cellulose acetate strips, roughly comparable results for the different methods were obtained. The overlay technique gave results which were about twice as high as those with other methods. This was attributed to the slower rate of diffusion of the high mol. wt. ALP into the agarose, leaving more of this enzyme on the strip which was subsequently scanned for quantitation of the isoenzymes: the overlay, on the other hand, showed denser staining in the region of the lower mol. wt. ALP. Because of these problems of variable diffusion rates and poor reproducibility neither agarose gel nor cellulose acetate were considered suitable media for quantitation. Polyacrylamide gel, with its property of reducing diffusion to a minimum and its better precision, appeared to be more suitable for routine measurements. However, comparison with the reference method of Sepharose 6B chromatography showed that electrophoretid estimates of the high mol. wt.
TABLE I PERCENTAGE OF HIGH MOL. WT. ALP IN 6 SERA, %STIMATED BY SCANNING FOLLOWING ELECTROPHORESIS AND STAINING IN THREE MEDIA SerLlm
Total ALP activity (U/l)
DENSITOMETRY,
High mol. wt. ALP (W total) 2.2% Agarose gel
2.5% Polyacrylamide gel
Cellulose acetate Thin-film staining
Agar template staining
1 2 3
144 304 544
26 32 17
20 30 21
22 32 33
48 56 49
4 5 6
216 568 520
24 3 8
17 4 5
15 14 14
50 15 24
__
260
ALP gave values which, especially for low activity samples, were often less than half those obtained by Sepharose 6B chromatography. Further investigation revealed that this difference was caused by the different substrate specificities of liver and high mol. wt. ALP. cu-Naphthyl phosphate (used in the electrophoretic methads) gave lower activities of the high mol. wt. ALP, relative to the liver isoenzyme, than did p-nitrophenyl phosphate (used in the kinetic methods). Discussion There are three principal methods at present available for measuring high mol. wt. ALP. The first is gel chromato~aphy of a single sample [4], followed by measurement of activities in a large number of fractions, an accurate reference method but slow and tedious. The second is elution from sequential strips of electrophoresis media, usually polyacrylamide [6] or starch gel, followed by measurement of activities in each strip, again a time-consuming procedure. Furthermore, recovery of enzyme activity from macerated gel may vary from isoenzyme to isoenzyme, from serum to serum and from gel to gel. In particular, it seems likely that large molecules will be more difficult to recover from the gel than small molecules and that the proportion may be critically dependent on the characteristics of the gel, such as its strength and degree of crosspolymerisation, as well as the conditions of elution. A third method of quantitation has used electrophoresis on cellulose acetate strips followed by densitometry [ 81. We found this technique to be very imprecise, presumably owing to diffusion and variable background staining. In addition, the agar overlay technique was inaccurate, since it overestimated the amount of high mol. wt. ALP in the serum. Electrophoresis on agarose gel was also rather imprecise whereas the precision with polyacrylamide gel was adequate. However, the results obtained with the electrophoretic methods were much lower than those obtained by Sepharose 6B chromatography. This could be partly attributed to the different substrate specificities of the high and low mol. wt. isoenzymes. Electrophoretic methods have the additional disadvantage that hydrolysis of a-naphthyl phosphate by ALP is inhibited by most of the coupling dyes used to visualize the isoenzymes; this may cause non-line~ity of colour development. We therefore considered the simple electrophoretic techniques to be unsuitable for accurate and precise measurement of high mol. wt. ALP. The ion-exchange method presented in this paper separates and quantitates high mol. wt. ALP, not on the basis of its high mol. wt., but on the basis of its difference in charge from most other ALP isoenzymes. The common isoenzymes of ALP, namely liver, bone, intestinal and placental, and the IgG-ALP complex [12] which is occasionally encountered, do not appear to interfere with quantitation of the high mol. wt. component by this method. Only the Regan variant [13], as might be expected from its high electrophoretic mobility and presumably greater charge, elutes in Fraction II. This is a rare cancer-associated isoenzyme, much less commonly encountered than the better known Regan isoenzyme, and the proportion of cases in which its presence would interfere with the method described here is likely to be very
261
small. To detect such cases, 7% polyacrylamide gel electrophoresis could be performed on all sera to be analysed. If the Regan variant were to be found, the ion-exchange method could not be used to quantitate the high mol. wt. component, and Sephadex G200 (or Sepharose 6B) chromatography should be used instead. There is fairly good correlation between estimates of high mol. wt. ALP obtained by the ion-exchange method and by Sepharose 6B chromatography, considering that the two methods are based on different properties of the high mol. wt. component. Although the slightly higher results obtained by the ionexchange method could be due to a small amount of carry-over, the evidence presented in this paper fails to support this. Another possible explanation is that some degree of denaturation of high mol. wt. ALP may occur on the Sepharose 6B column, since the enzyme is in contact with the column matrix for a longer period of time and more dilution takes place. However, it is also possible that the slight discrepancy between the two methods is within the limits of their analytical error. In conclusion, the method presented here provides a simple, relatively rapid, reliable and precise measurement of the high mol. wt. component of ALP in small batches of sera. References 1
Warnes,
2
Fennelly,
T.W.
3
Price.
(1972)
J.J.,
C.P.
4
Estborn.
5
DuBuisson,
6
Price,
M.X.
and
H.G.
(1974)
B. (1964)
Z. Klin.
J. and and
7
Rhone,
D.P.
8
Rhone.
D.P.,
9
Warwick,
13.926-937
Sammons.
and
C.P.
Gut
Fitzgerald.
Pepler,
Sammons.
and
Mizuno,
White,
R.R.G..
F.M.
Chem.
McGeeney,
K.
J. Clin.
(1966)
S. Afr.
H.G.
(1976)
J. Clin.
(1973)
Am.
F.M.
Shearman.
Gut
27,
10.45-51
392-398
2, 53-54
W.J.
and
(1969)
Pathol.
Gidaspow, D.J.C.,
Med.
J. Suppl.,
Pathol.
J. Clin.
H. (1973)
Percy-Robb.
29.
Pathol. Clin. I.W.
696498
976-980 59,
531-541
Chem.
and
19.
Smith,
175 10
Kaplan,
11
Ewen,
M.M.
12
Crofton,
P.M.
13
Crofton.
P.M.
L.M.
and (1974)
Rogers.
L. (1969)
Lancet
Pathol.
ii, 1029-1031
Am.
J. Clin.
61,142-154
and
Smith,
A.F.
(1978)
Clin.
Chim.
Acta
83,
235-247
and
Smith,
A.F.
(1978)
Clin.
Chim.
Acta
86,
81-88
1142-1147 A.F.
(1972)
Scot.
Med.
J. 17.
172-
