JOURNAL OF BONE AND MINERAL RESEARCH Volume 7, Number 7, 1992 Mary A n n Liebert. Inc., Publishers

Quantitation of Skeletal Alkaline Phosphatase Isoenzyme Activity in Canine Serum JOHN R. FARLEY,'-' SUSAN L. HALL,' CANDACE RITCHIE,'~*SANDRA HERRING,' CHRISTOPHER ORCUTT,' and BARBARA E. MILLER4

ABSTRACT Pursuing the hypothesis that quantitation of skeletal alkaline phosphatase (ALP) activity in canine serum would provide an index of the rate of bone formation, we compared three methods for isoenzyme-specific identification of skeletal ALP activity in canine serum: heat inactivation, wheat germ agglutinin (WGA) precipitation, and concanavalin A (ConA) precipitation. ALP isoenzyme activities were extracted from canine bone, intestine, and liver, diluted into heat-inactivated canine serum (i.e., serum without ALP activity), and used as calibrators of ALP isoenzyme activities. Differential sensitivity to inhibition by 10 mM L-homoarginine was used to distinguish intestinal ALP activity from hepatic and skeletal ALP activities (i.e., 9, 80, and 72% inhibition, respectively). To allow resolution of skeletal ALP activity from hepatic ALP activity, we tested two established methods (heat inactivation and WGA precipitation) and a novel method, ConA precipitation. The organ-derived skeletal and hepatic ALP isoenzyme activities were used to compare these three methods with respect to linearity, isoenzyme separation, and precision. All three methods were linear, but the WGA and ConA methods afforded greater isoenzyme separation and precision. The relative extent of isoenzyme separation (i.e., the difference in percentage remaining skeletal and hepatic ALP isoenzyme activities) averaged 23, 40, and 47% remaining ALP activity for the heat, WGA, and ConA methods, respectively. However, when these methods were applied to the quantitation of skeletal ALP activity in sera from 10 young and 10 adult beagles, the WGA method was found to be unacceptable because most of the results fell outside the range of the WGA assay calibrators (i.e., >100% skeletal ALP activity). The heat and ConA methods showed that the amount of skeletal ALP activity in the beagle sera decreased with age, both as ALP activity per liter and as percentage of total serum ALP activity 0, c 0.001 for each). Skeletal ALP activity levels determined by ConA were correlated with values determined by heat inactivation ( r = 0.87, p < 0.001) but not with WGA-determined levels ( r = 0.26). Intestinal ALP activity was detected in only 1 of these 20 sera. We conclude that ConA precipitation can be used for quantitation of skeletal ALP activity in beagle serum.

INTRODUCTION LTHOUGH THE PRECISE BIOCHEMICAL FUNCTION of alkaline phosphatase (ALP) in bone cells is unknown, the enzyme is present on osteoblasts (i.e., as an ectoenzyme) and has been shown to play a role in bone Previous studies indicate that quantitation of skeletal ALP activity in serum can provide an index of the rate of bone formation. ( 7 - 1 3 )

A

The practical problem has been to design selective assay methods that can distinguish and quantitate skeletal A L P isoenzyme activity in the serum isoenzyme mixture of ALP activities derived from bone, liver, and intestine (and, during pregnancy, placenta). In previous studies, skeletal ALP activity has been distinguished from other ALP isoenzyme activities in human serum by means of characteristic differences in electrophoretic mobility,'"' sensitivity to chemical inhibitors and inactivation by heat,"5-201 precipitation

'Department of Medicine, Loma Linda University, Loma Linda, California. 'Department of Biochemistry, Loma Linda University, Loma Linda, California. 'Jerry L. Pettis Memorial Veterans' Hospital, Loma Linda, California. 'Norwich Eaton Pharmaceuticals, Inc., a Proctor and Gamble Company, Norwich, New York. 'Current address: Endocrine Research Unit, Mayo Clinic, Rochester, Minnesota.

779

780

FARLEY ET AL.

with wheat germ agglutinin (WGA),(13,21-23) and immunoaffinity. ( 2 4 - 2 6 ) A number of assays have been developed that exploit these differences for selective quantitation of skeletal ALP isoenzyme activity (or skeletal ALP immunoreactive protein) in human serum.(12.13,*2-”’ Under normal circumstances, canine serum also contains ALP isoenzyme activities derived from liver, intestine, and bone. (30-41) [Under conditions of steroid excess, canine serum can also contain a second hepatic ALP isoenzyme activity, which is specifically induced by exposure to corticosteroids. The current studies were intended to adapt existing human serum ALP isoenzyme assay methods to allow quantitation of skeletal ALP activity in canine serum. Specifically, we wanted to define an assay protocol(s) that combined selective chemical inhibition (with L-phenylalanine, L-homoarginine, or levamisole) with either differential heat inactivation or differential precipitation with lectin [i.e., WGA from Triticum vulguris or concanavalin A ( C o d ) from Cunuvaliu ensiformis] and then apply that protocol to quantitate skeletal ALP activity in young and adult beagle serum. Our presumption in these studies was that such a serum assay would provide a convenient and accurate method for serial estimations of the rate of bone formation in a large animal model that has proved to be useful for mechanistic studies of skeletal metabolism and remodeling boine physiology.

Measurement of ALP activity As in previous s t u d i e ~ , ( ~ ’ -ALP ’ ~ ~ activities were determined colorometrically in 96-place microtiter wells. The standard reaction mixture contained 10 mM PNPP, 1 mM MgCI,, and a source of ALP activity in a total volume of 0.3 ml of 0.15 M NatC03 buffer (pH 10.3). Assays were initiated by addition of the substrate (PNPP). ALP activities were measured as the time-dependent increases in absorbance at 410 nm (i.e., p-nitrophenolate production) using an automatic recording microtiter plate spectrophotometer (SLT Lablnstruments model EAR 400 AT, Hillsborough, NC). All assays were performed in duplicate. Serum A L P activities are reported as units/liter of serum, where 1 unit isoenzyme activity represents 1 pm of p-nitrophenolate produced per minute at room temperature (22”C), using a calculated value of 6.0 for the millimolar extinction coefficient of p-nitrophenolate. This value was determined by experiment and applies specifically to our reaction protocol, with a total volume of 0.3 ml in a 96-well microtiter plate, corresponding to a light path (i.e., liquid cylinder height in the well) of approximately 0.3 cm.

Preparation of heat-inactivated canine serum (HIS) A commercial preparation of pooled, nonsterile canine serum was incubated overnight at 52°C in a heated water bath. During the incubation, the total amount of ALP activity decreased from 39.6 units/liter to less than 0.3 unitslliter (i.e., less than 1% remaining ALP activity).

MATERIALS AND METHODS Chemicals and supplies

Preparation of ALP isoenzyme assay calibrators

p-Nitrophenylphosphate (PNPP), levamisole, L-phenylalanine, L-homoarginine, Triton X-100, wheat germ agglutinin, and concanavalin A (Sigma Type IV) were obtained from Sigma (St. Louis, MO). Type I11 Aquacide was obtained from CalBiochem (San Diego, CA). Dog serum (500 ml, pooled, nonsterile, nonhemolyzed) was obtained from Pel-Freeze Biologicals (Rogers, AR). Multiwell (96-place) dishes (Corning microtiter plates for enzyme-linked immunosorbent assay, ELISA). Nalgene 0.22 pm sterile filters (100 ml capacity), and 10 x 75 mm polypropylene test tubes were obtained from Fisher Scientific (Tustin, CA). Bio-Rad protein assay reagent was obtained from BioRad Laboratories (Richmond, CA).

Intestinal, hepatic, and skeletal ALP isoenzyme activities were prepared from tissue extracts. Adult beagle tissues (bone, liver, and jejunum) were cut into sections and rinsed in phosphate-buffered saline (pH 7.2, containing 0.01 070 sodium azide) to remove residual serum. The rinsed tissues were extracted in 25 mM NaHCO, buffer (pH 8.0, containing 20% (vol/vol) n-butanol) with stirring for 72 h at 4°C. The extracts were centrifuged to separate the butano1 (upper) and aqueous phases. The ALP-containing aqueous phases were transferred to dialysis tubing, concentrated (type 111 Aquacide), and dialyzed extensively against 25 mM NaHCO, buffer (pH 8.0, containing 0.01 ‘3’0 sodium azide). The dialyzed solutions were centrifuged to remove insoluble material, and the concentrations of ALP isoenzyme activity and were determined. ALP isoenzyme activity levels were measured with 10 mM P N P P as the substrate in 0.15 M NaHCO, buffer (pH 10.3, containing 1 mM MgCI,). The solutions were reconcentrated and dialyzed, as necessary, to obtain ALP isoenzyme activities in excess of 230 units/liter, then filter sterilized and stored as 5-7 ml aliquots at -20°C. Typical yields of ALP isoenzyme activity (pooled extracts of tissues from three dogs) were 110 units from 430 g (wet weight) liver, 456 units from 110 g (wet weight) jejunum, and 5.2 units from six tibiae and six femora, with specific isoenzyme activities of 0.018, 0.586, and 0.014 units/mg protein, respectively. Isoenzymes to be used for assay Cali-

Approved use of animal subjects and derived materials All animals were maintained in an accredited facility, and all procedures (including collection of serum and euthanasia) were consistent with guidelines established by the American Veterinary Medical Association. Adult beagle tissues (i.e., bone, liver, and jejunum) were extracted to prepare isoenzyme standards of ALP activity, and serum samples were obtained from fed and (overnight) fasted young ( 2 years) beagles and used to evaluate the isoenzyme assays. The samples were stored at -20°C.

78 1

CANINE SERUM SKELETAL ALKALINE PHOSPHATASE bration were prepared by diluting aliquots of each tissuederived ALP isoenzyme activity into heat-inactivated canine serum, to yield final activities of 20-40 units/liter. As with human tissue-derived ALP isoenzymes,(291the maximum activities of the canine organ-derived ALP isoenzymes were increased (1.4 to 2-fold) by dilution into canine HIS.

ALP isoenzyme assay protocols Heat Inactivation Assay: The standard assay protocol was adapted from previous studies. ( 1 8 , ’ 9 ”-”) In polypropylene test tubes, 0.15 ml serum was combined with 0.15 ml of 25 mM NaHCO, buffer (pH 8.0, with 0.01% sodium azide) and 0.15 ml deionized water. Similar solutions of organ-derived ALP isoenzyme activities (to be used as assay calibrators) were prepared in the same NaHCO, buffer by combining 0.15 ml HIS with 0.15 ml deionized water and 0.15 ml of an appropriate dilution of each isoenzyme activity. Two sets of duplicate 0.05 ml aliquots of each solution (i.e., both samples and calibrators) were transferred to microtiter wells to determine the amounts of total ALP activity and L-homoarginine-insensitive ALP activity. The remaining volumes of diluted sera (i.e., 0.25 ml) were transferred to a heated water bath for a 50 minute incubation at 48-49°C and then cooled with ice to 22°C or less (i.e., ambient room temperature or colder). Duplicate 0.05 ml aliquots were transferred to a microtiter plate to determine the amount of heat-insensitive ALP activity. (ALP activity levels were determined in Na,CO, buffer, pH 10.3, containing 10 mM P N P P and 1 mM MgCI,.) The amount of L-homoarginine-insensitive ALP activity was determined in the presence of 10 mM L-homoarginine. Negative controls, containing HIS but without added ALP, were included in each assay to correct for nonenzymatic hydrolysis of PNPP. Organ-derived ALP isoenzyme activities were also included in each assay for internal calibration of L-homoarginine sensitivity and heat inactivation. The results were screened for evidence of intestinal ALP isoenzyme activity. The amount of skeletal ALP isoenzyme activity in each sample was then calculated, with two or three unknowns as appropriate (i.e., either skeletal and hepatic ALP or skeletal, hepatic, and intestinal ALP isoenzyme activities). These calculation method^(^'-'^) are described in Results. Preliminary studies assessed the significance of the following four variables as determinants of assay sensitivity and error: (1) the concentration of serum, (2) the concentration of NaHCO, buffer (i.e., in the initial serum mixtures), (3) the incubation temperature, and (4) the length of incubation. These studies are described in Results. WGA Precipitation Assay: The standard assay protocol was adapted from previous studies.(13,21-*3) In polypropylene test tubes, 0.2 ml serum was mixed with 0.25 ml deionized water and 0.05 ml Triton X-100 (20 glliter, in deionized water) and incubated for 30 minutes at 37°C. Organ-derived ALP isoenzyme activities (in HIS) to be used as assay calibrators were diluted in water and Triton X-100to yield identical mixtures and similarly incubated.

A 0.2 ml aliquot of each solution was then combined with 0.2 ml deionized water, and a second 0.2 ml aliquot was combined with 0.2 ml WGA ( 5 mg/ml in deionized water containing 0.01% sodium azide). The mixtures were agitated on a vortex mixer, incubated 30 minutes at 37”C, and centrifuged for 10 minutes at 2500 rpm in a Beckman TJ-6, nonrefrigerated, tabletop centrifuge. Two sets of duplicate 0.05 ml aliquots of the mixtures without WGA were transferred to microtiter wells to determine the amounts of total ALP activity and L-homoarginine-insensitive ALP activity. Duplicate 0.05 ml aliquots of the mixtures with WGA were transferred to microtiter wells to determine the amount of WGA-soluble ALP activity. Negative controls, containing HIS but without added ALP (fWGA), were included in each assay to correct for nonenzymatic hydrolysis of PNPP. Organ-derived ALP isoenzymes were included in each assay for internal calibration of L-homoarginine sensitivity and WGA precipitation. The results were screened for evidence of intestinal ALP isoenzyme activity. The amount of skeletal ALP isoenzyme activity in each serum sample was calculated with two or three unknowns, as described in Results. Preliminary studies assessed the significance of the following six variables as determinants of assay sensitivity and error: (1) the amount of serum in the initial incubation, (2) the amount of ALP isoenzyme activity in the initial incubation, (3) the glucose concentration in the initial incubation, (4) the length of the initial incubation, ( 5 ) the amount of WGA in the second incubation, and (6) the length of the second incubation. These studies are described in Results. ConA Precipitation Assay: The standard assay protocol for ConA precipitation was identical to the WGA precipitation assay protocol, except that 0.2 ml ConA (6 mg/ml in deionized water, with 0.01% sodium azide) was substituted for WGA in the second incubation. Preliminary studies assessed the significance of the following six variables as determinants of assay sensitivity and error: (1) the amount of serum in the initial incubation, (2) the amount of ALP isoenzyme activity in the initial incubation, (3) the glucose concentration in the initial incubation, (4) the length of the initial incubation, (5) the amount of ConA in the second incubation, and (6) the length of the second incubation. These studies are described in the Results.

Statistical analyses All data are shown as the average of replicates (mean f standard deviation, SD, or mean f standard error of the mean, SEM). Correlation coefficients were determined by linear regression. Comparisons were made by analysis of variance (ANOVA) and Student’s t-test.

RESULTS Initial characterizations of organ-derived A L P isoenzymes Organ-derived standards of canine ALP isoenzyme activities showed responses to

FARLEY ET AL.

782

L-phenylalanine, L-homoarginine, levamisole, heat, and variation in pH (Table 1). On the basis of these data, differential sensitivity to inhibition by 10 mM L-homoarginine was selected as a means to distinguish intestinal ALP isoenzyme activity from skeletal ALP isoenzyme activity and hepatic ALP isoenzyme activity. Assay development Heat Inactivation Studies: Canine skeletal ALP isoenzyme activity was more sensitive to heat inactivation than canine hepatic A L P isoenzyme activity, which was in turn more sensitive than canine intestinal ALP isoenzyme activity (Fig. 1). To determine the heating conditions that would yield the maximum difference between skeletal and hepatic ALP isoenzyme activities, half-lives were determined (from heat inactivation rates) for 33% serum solutions of each tissue-derived standard of ALP isoenzyme activity at temperatures ranging from 45 to 53°C and NaHCO, buffer concentrations of 8.3 and 16.7 mM. The results of those studies (Table 2) showed that skeletal ALP isoenzyme activity was more sensitive to heat inactivation than hepatic ALP isoenzyme activity under all tested conditions, with a maximum 34% difference in remaining ALP isoenzyme activity after 50 minutes in 8.3 mM N a H C 0 3 buffer at 47-49°C. Under these conditions, the organ-derived standards of skeletal, hepatic, and intestinal ALP isoenzyme activities were irreversibly inactivated by 61, 27, and 14%, respectively (i.e., 39, 73, and 86% remaining ALP isoenzyme activity). WGA Precipitation Studies: Canine skeletal ALP isoenzyme activity was more sensitive to WGA precipitation than canine hepatic ALP isoenzyme activity (Fig. 2), with a maximum 37% difference in the amount of WGA-soluble A L P isoenzyme activity at a WGA concentration of 2.5 mg/ml. Canine intestinal ALP isoenzyme activity showed intermediate sensitivity to precipitation with WGA (i.e., 70, 58, and 33% remaining intestinal ALP isoenzyme ac-

tivity at WGA concentrations of 2.0, 2.5, and 3.0 mg/ml, respectively; data not shown). The sensitivity of canine skeletal ALP isoenzyme activity to WGA precipitation at 2.5 mg/ml was not affected by changes in serum (HIS) volume or the length of incubation with WGA (Table 3), but the assay was sensitive to changes in the length of the initial (Triton X-100) incubation. Sensitivity of skeletal ALP isoenzyme activity to WGA precipitation was decreased by longer Triton X-100 incubations (i.e., between 0 and 90 minutes), but this phenomenon was not observed with hepatic ALP isoenzyme activity (Fig. 3). This suggested that assay sensitivity (which is primarily determined by the absolute difference between percentage remaining skeletal and hepatic A L P isoenzyme activities) would be decreased with longer Triton X-100 incubations. The assay was also dependent on the total amount of skeletal ALP isoenzyme activity ( n = 7, r = -0.98, and p < 0.001), but the slope of that inverse relationship was small (i.e., the amount of WGA-sensitive skeletal A L P isoenzyme activity changed by only 2.5% across a 100-fold range of skeletal ALP isoenzyme activity). The slope of the inverse relationship with glucose concentration was also small (i.e., the amount of WGA-insensitive skeletal ALP isoenzyme activity changed by only 0.09% for every 1 mg/ dl of glucose added to the assay mixture). Thus, we would predict that a change in serm glucose of 25 mg/dl (which would translate to an increase of 10 mg/dl in the assay mixture) would cause a 0.9% change in WGA-insensitive skeletal ALP isoenzyme activity, or 2.8% of the observed difference between the skeletal and hepatic standards. The sensitivity of the hepatic ALP isoenzyme activity standard to WGA precipitation was unaffected by glucose. ConA Precipitation Studies: In contrast to WGA, ConA was more effective at precipitating canine hepatic ALP isoenzyme activity than canine skeletal A L P isoenzyme activity. ConA titration (Fig. 4) indicated a maximum 45% separation of hepatic and skeletal ALP isoen-

TABLE1 . INITIALCHARACTERIZATION OF ORGAN-DERIVED ALP

~ O E N Z Y M EACTIVITIES

% Remaining ALP isoenzyme activitya

Effectorb L-Phenylalanine (10 mM) L-Homoarginine (10 mM) Levamisole ( 1 mM) Heat (10 minutes at 54°C) Assay pH

Assay pH

Skeletal

Hepatic

Intestinal

10.3 10.3 10.3 10.3 9.3

70 28 7 36 65

71 20 5 56 57

31 91 82 76 90

Control (pretreatment) levels of ALP isoenzyme activity (at pH 10.3, with 10 mM PNPP and 1 mM MgCI2)were 19, 17, and 26 units/liter for the organ-derived preparations of skeletal, hepatic, and intestinal ALP isoenzymes, respectively. The isoenzyme activities were prepared in solutions of 33% (vol/vol) heat-inactivated (canine) serum in 16.7 mM carbonate buffer, pH 8.0. Data shown are means of triplicate determinations. bShown in parentheses are final concentration of effectors in the assay mixture or conditions of preassay heat inactivation.

783

CANINE SERUM SKELETAL ALKALINE PHOSPHATASE

a

m

zyme activities at 3 mg/ml. Canine intestinal ALP isoenzyme activity was less sensitive to precipitation with ConA than canine skeletal ALP isoenzyme activity (i.e., 87, 83, and 67% remaining intestinal ALP isoenzyme activity at ConA concentrations of 2, 3, and 4 mg/ml, respectively; data not shown). The sensitivity of canine skeletal ALP isoenzyme activity to precipitation with ConA at 3 mg/ml was not affected by changes in incubation times (i.e., incubations with Triton X-100 and ConA), the amount of ALP activity, or the amount of added glucose (Table 3). Although the amount of skeletal ALP isoenzyme activity precipitated by ConA was not correlated with the length of incubation with ConA, a zero incubation time decreased the amount of hepatic ALP isoenzyme activity that was precipitated by ConA from 84 to 53%, thus decreasing assay sensitivity because it decreased separation between isoenzyme activities. This effect was not observed with ConA incubation times of 15-90 minutes. Although the efficiency of ConA precipitation was dependent on the serum (HIS) volume for both skeletal ALP isoenzyme activity and hepatic A L P isoenzyme activity, the difference between percentage remaining skeletal and hepatic ALP isoenzyme activities was not affected (Fig. 5 ) .

75--

.-

S

.-Q5

E 50-.

B c

$ 2

25-

a"

I

I

r

0

L

I

1

i

15

20

25

30

1

5

10

Time

(min)

FIG. 1. Heat inactivation of organ-derived ALP isoenzyme activities. Percentage of remaining ALP isoenzyme activity is shown as a function of time for organ-derived skeletal (circles), hepatic (black-and-white squares), and intestinal (solid squares) ALP isoenzyme activities. All samples were incubated in NaHC03 buffer (16.7 mM, pH 8.0) containing 33% (vol/vol) HIS in a heated water bath at 52°C. Basal skeletal, hepatic, and intestinal ALP isoenzyme activity levels were 19, 17, and 26 units/liter, respectively. Data shown are means of triplicate determinations. Heat inactivation rate constants were calculated from linear replots of logarithmic percentage remaining activity versus time (slope = - k / 2 . 3 ) ; half-lives were calculated from the relationship t,,2 = 0.693/k.

Comparisons of assay methods The results of these studies suggested that any one of the three methods (heat inactivation, 50 minutes at 47-49°C; WGA precipitation, 2.5 mg/ml of WGA; or ConA precipitation, 3 mg/ml of ConA) could be used to distinguish ca-

TABLE2. SENSITWITY OF SKELETAL AND HEPATIC ALP ISOENZYME ACTIVITIES TO INACTIVATION BY HEAT Haw-lif e (minutes) of A L P isoenzyme activity a Temperature 45-47 47-49 49-5 1 51-53 45-47 47-49 49-5 1 51-53

("C)b

NaHCO, buffer (mM) 8.3 8.3 8.3 8.3 16.7 16.7 16.7 16.7

Maximum difference in Skeletal

Hepatic

100 31 15 8 50 24

> 120

10

5

74 33 14 70 41 22 8

% remaining A L P activityc

7, 34, 28, 24, 14, 22, 26, 20,

30 minutes 50 minutes 50 minutes 15 minutes 25 minutes 30 minutes 25 minutes 10 minutes

=Half-lives of ALP isoenzyme activity were determined from the slopes of semilogarithmic plots of logarithmic percentage remaining ALP activity versus time. bOrgan-derived skeletal and hepatic ALP isoenzyme activities in 8.3 or 16.7 mM NaHCO, buffer containing 33% (vol/vol) heat-inactivated canine serum were incubated at the indicated temperatures. Percentage remaining ALP isoenzyme activities were determined at 5 minute intervals for a total of 30 minutes at temperatures below 49°C and at 10 minute intervals for a total of 60 minutes at temperatures above 49°C. Before the incubations, the skeletal and hepatic ALP isoenzyme solutions contained 17 and 16 units/liter of ALP activity, respectively. CData are maximum percentage differences between percentage of remaining skeletal ALP isoenzyme activity and percentage of remaining hepatic ALP isoenzyme activity observed at the indicated time. Data shown as means of duplicate determinations.

FARLEY ET AL.

784

nine skeletal ALP isoenzyme activity from hepatic ALP isoenzyme activity. To compare these methods, artificial mixtures (i.e., calibration solutions) of organ-derived skeletal and hepatic ALP isoenzyme activities were prepared in HIS. These mixtures approximated ALP isoenzyme compositions of 100, 80, 60,40, 20, and 0% skeletal ALP isoenzyme activity, with corresponding levels of hepatic ALP isoenzyme activity of 0, 20, 40,60, 80, and 100V0,respectively. Each of the three assay methods was tested four times on each artificial mixture, and each test was assessed in two ways. First, the data were tested for dose-dependent linearity by correlating the observed percentage remaining ALP activity (after heating or precipitation with lectin) with the known percentages of skeletal ALP isoenzyme activity in the artificial mixtures. Second, the two extreme mixtures (i.e., 100% skeletal ALP isoenzyme activity and 100% hepatic ALP isoenzyme activity) were used to calibrate the assays (i.e., by construction of a standard curve that was applied to calculate the compositions of the other mixtures). Assay precision was calculated by comparing these values with the known compositions of the mixtures. The results of these comparisons are shown in Table 4. All three assay methods were linear; the correlation coefficients determined for correlations of percentage skeletal ALP isoenzyme activity (which was predetermined in the artificial isoenzyme mixtures) versus percentage remaining total ALP activity (the observed, dependent variable) ranged from -0.943 to -0.987, -0.986 to -0.993, and 0.987 to 0.999 for the heat inactivation, WGA precipitation, and ConA precipitation assays, respectively. However, the maximum separation between the calibration

1OO-l-k

c

I?

251 0 1 0

0.5

1

I

I

I

I

I

I

1

1.5

2

2.5

I

I 3

WGA (mghl) FIG. 2. Characterization of WGA assay. Dose-dependent precipitation of skeletal ALP isoenzyme activity and hepatic ALP isoenzyme activity by WGA. Percentage remaining ALP isoenzyme activity is shown as a function of WGA concentration. Organ-derived skeletal (circles) and hepatic (squares) ALP isoenzyme activities in HIS were incubated with Triton X-100 and then mixed with WGA to produce the indicated final concentrations. Basal skeletal and hepatic ALP isoenzyme activity levels were 38 and 30 units/liter, respectively. Data shown are means of two separate assays, with duplicate determinations in each. Note: An intestinal ALP isoenzyme activity calibrator preparation showed 70, 58, and 33% remaining ALP activity at WGA concentrations of 2.0, 2.5, and 3.0 mg/ml (data not shown).

TABLE3. LECTINPRECIPITATION ASSAYS FOR SKELETAL ALP ISOENZYME ACTIVITY: TO ASSAY VARIABLES SENSITIVITY Correlation roefficientsa (range Yo remaining A L P activity) WGA assay

Variable Initial (Triton X-100) incubation Serum (HIS) volumeb Total skeletal ALP activityd Glucose concentratione Length of incubationf Second (lectin) incubation Lectin concentrationh Length of incubation'

ConA assay

-0.46 -0.98C -0.96C 0.928

(42-56) (34-31) (41-35) (28-47)

0.97C 0.07 0.60 -0.18

0.95C -0.51

( 10- 100)

-

-

(28-37)

0.99C 0.22

(16-77) (52-60)

(58-63) (56-61) (24- 100) (63-66)

aData shown are correlation coefficients for correlations of variable versus percentage remaining skeletal ALP isoenzyme activity and the range of remaining skeletal ALP isoenzyme activity, with the latter values in parentheses. &rum volume is 20-80% of the initial incubation volume, n = 8. CIndicates a significant correlation, p < 0.001. dActivity is 0.1-10 times the basal level of 26 units/liter, n = 7. eGlucose at 0-60 mg/dl added to the initial incubation volume, n = 8. fLength of 0-90 minutes, n = 6 . Elndicates a significant correlation, p < 0.01. hLectin at 0-3 mg/ml of WGA, n = 9, and 0-5 mg/ml of ConA, n = 9. ilncubation for 0-90 minutes. n = 6 .

785

CANINE SERUM SKELETAL ALKALINE PHOSPHATASE standards of 100% skeletal ALP isoenzyme activity and 100% hepatic ALP isoenzyme activity was greater with the WGA precipitation and ConA precipitation methods than with the heat inactivation method.

Because each serum sample was considered a three-component mixture of skeletal, hepatic, and intestinal ALP isoenzyme activities, the amount of skeletal ALP isoenzyme activity could be calculated by simultaneous solution of three equations with three unknowns:

Application of the assays to determination of skeletal ALP activity in young and adult beagle sera

Total serum ALP activity = skeletal ALP isoenzyme activity + hepatic ALP isoenzyme activity + intestinal ALP isoenzyme activity

Table 5 shows the results of two comparison assays in which the heat inactivation, WGA precipitation, and ConA precipitation methods were applied (along with L-homoarginine sensitivity) to determine the amount of skeletal ALP isoenzyme activity in the serum of 10 young ( < 7 months) and 10 adult ( > years) beagles. Half the dogs in each group had been fasted overnight. Organ-derived preparations of intestinal ALP isoenzyme activity, hepatic ALP isoenzyme activity, and skeletal ALP isoenzyme activity in HIS and a 50:50 mixture of skeletal ALP and hepatic ALP isoenzyme activities in HIS were used to calibrate each assay. An aliquot of control canine serum (i.e., the unheated source of HIS) was also included as an additional control.

L-Homoarginine-insensitive ALP activity = ( a x skeletal ALP isoenzyme activity) + ( b x hepatic ALP isoenzyme activity) + c x intestinal ALP isoenzyme activity) Heat-insensitive ALP activity/WGA-insensitive ALP activity/ConA-insensitive ALP activity = ( d x skeletal ALP isoenzyme activity) + ( e x hepatic ALP isoenzyme activity) + cf x intestinal ALP isoenzyme activity)

100 h

.-> 4-

'G 80

u

a L

a

< 60

rn

0,

I

60

P I L

':

Lz: c 8

40

.-8 .-8

t

z

40

Q,

Lz: 4-

8

Q,

20

2

2 I

0 I

i

60

90

1

I

0

30 Time

(min)

FIG. 3. Characterization of WGA assay. Dependence on Triton X-100 incubation time. Percentage remaining ALP isoenzyme activity is shown as a function of initial (Triton X-100) incubation time. Organ-derived skeletal (circles) and hepatic (squares) ALP isoenzyme activities were incubated with Triton X-100 (0.2 mg/ml), HIS, and deionized water for 0-90 minutes at 37°C; aliquots of those solutions were used to measure the amounts of WGA-insensitive (i.e., soluble) ALP isoenzyme activity. WGA was used at a final concentration of 2.5 mg/ml. Data shown are means of duplicate determinations; r = 0.92 (p < 0.01) and r = 0.04 for skeletal and hepatic ALP activities, respectively.

1

2

1

I

i

3

4

5

coNcANAV A L I N - A ( mg/m I) FIG. 4. Characterization of ConA assay: Dose-dependent precipitation of skeletal ALP isoenzyme activity and hepatic ALP isoenzyme activity by ConA. Percentage remaining ALP isoenzyme activity is shown as a function of ConA concentration. Organ-derived skeletal (circles) and hepatic (squares) ALP isoenzyme activities, in HIS, incubated with Triton X-100 and then mixed with ConA to produce the indicated final concentrations. Basal skeletal and hepatic ALP isoenzyme activity levels were 33 and 31 units/liter, respectively. Data shown are means of two separate assays, with duplicate determinations in each. Note: An intestinal ALP isoenzyme activity calibrator preparation showed 86, 83, 67, and 45% remaining ALP activity at ConA concentrations of 2, 3, 4,and 5 mg/ml (data not shown).

FARLEY ET AL.

786

where the coefficients of isoenzyme sensitivity (i.e., constants a-f, were defined by the isoenzyme activity calibrators in each assay and the amount of skeletal ALP isoenzyme activity was c a l ~ u l a t e d ~ ~as ' - ~a ~percentage ' of total serum ALP activity and as an absolute value (i.e., units/ liter serum). Before applying this method, however, we screened the data for evidence of intestinal ALP isoenzyme activity. Because 10 mM L-homoarginine inhibited 70-80% of the ALP isoenzyme activities in the skeletal and hepatic standards but only 5-10% of the ALP isoenzyme activity in the intestinal standard, we assumed that sera that resembled the skeletal ALP and hepatic ALP isoenzyme activity calibrators with respect to L-homoarginine sensitivity (i.e., within 15%) did not contain significant amounts of intestinal ALP isoenzyme activity. By this criterion of L-homoarginine sensitivity, only 1 of the 20 serum samples that were tested for these comparisons was found to contain in" testinal ALP isoenzyme activity (21% of the total serum t I I I i ALP activity). Overall, during the course of these studies, 0 20 40 60 80 ALP activity was analyzed in serum samples obtained Serum Volume ' p e r c e n t ' from 40 different dogs (i.e., the 20 used for comparison FIG. 5. Characterization of ConA assay. Dependence on studies and 20 additional animals), and only 3 (which were serum volume assayed. Percentage remaining ALP isoen- all from adult, fasted dogs) contained measurable amounts zyme activity is shown as a function of the serum (HIS) of intestinal ALP isoenzyme activity. These findings are volume in the initial (Triton X-100) incubation mixture. consistent with previous reports that canine intestinal ALP Organ-derived skeretal (circles) and hepatic (squares) ALP isoenzyme activity and canine hepatic ALP isoenzyme acisoenzyme activities were incubated with Triton X-100 (0.2 tivity have serum half-lives of less than 6 minutes and 3 mg/ml), varying amounts of HIS, and deionized water; days, the half-life of canine skeletal ALP aliquots of those solutions were used to measure ConA-in- isoenzyme activity is also presumed to be days. In supplesensitive (i.e., soluble) ALP isoenzyme activity levels. mental studies, the 20 serum samples that were used for ConA was used at a final concentration of 3 mg/ml. Data shown are means of duplicate determinations; r = 0.97 comparing the methods (Tables 5 through 7) were reand r = 0.96 for skeletal ALP isoenzyme activity and he- screened for intestinal isoenzyme ALP activity with 1 mM patic ALP isoenzyme activity, respectively, with p < 0.001 levamisole (in place of 10 mM L-homoarginine) as the isoenzyme-specific inhibitor: the results were unchanged. The for each (n = 8 without zero serum data point).

looT

#

TABLE 4. COMPARISON OF ASSAYS: QUANTITATION OF SKELETAL ALP ISOENZYME ACTIVITY IN ARTIFICIAL ISOENZYMEMIXTURES~ Heat inactivation (Yo)

WGA precipitation (Yo)

Test

Separation

Variation

Separation

1

25 15 33 20

28 f 26 19 f 14 18 f 11 41 f 28

49 37 33 39

2 3 4

Variation 15

f

6

13 f 8

20 f 13 24 f 16

ConA precipitation (Yo) Separation

Variation

46 51 49 42

23 + 7 8 f 4 13 f 8 10 f 2

aThe heat inactivation, WGA precipitation, and ConA precipitation assays were each tested four times, and six isoenzyme calibration mixtures were included in each test. These mixtures, in HIS, consisted of 100% skeletal ALP isoenzyme activity, 100% hepatic ALP isoenzyme activity, and four intermediate mixtures containing 80:20,60:40,40:60, and 20:SO distributions of skeletal to hepatic ALP isoenzyme activity. Basal ALP isoenzyme activities were 30 and 32 units/liter in the 100% skeletal and 100% hepatic standards, respectively. Separation indicates the difference between percentage remaining ALP activity in the 100% skeletal ALP and 100% hepatic ALP isoenzyme activity mixtures that were used to calibrate the assays. Coefficients of variation (CVs) were determined as (actual - calculated)/actual values for skeletal ALP isoenzyme activity (as percentage of total ALP activity) in each of the four calibration mixtures of intermediate composition. CVs are reported as averages (mean f SD) for these four artificial standards. CVs were largest for the artificial standard that contained the lowest (percentage) levels of skeletal ALP isoenzyme activity, regardless of the assay used. Correlation coefficients determined for (known) percentage skeletal ALP isoenzyme activity in the six artificial standards versus (observed) percentage remaining ALP activity ranged from -0.94 to -0.99, -0.99 to -0.99, and 0.99 to 1 .OO for the heat inactivation, WGA precipitation, and ConA precipitation assays, respectively.

CANINE SERUM SKELETAL ALKALINE PHOSPHATASE

787

to construct regression lines of percentage remaining ALP activity versus the known percentage composition of skeletal ALP isoenzyme activity in the calibrator, and these were used to estimate the amount of skeletal ALP isoenzyme activity in the serum samples as a percentage of total serum ALP activity. Interassay variations were determined for each method by analysis of the results for individual canine serum samples in the two comparison assays (the same 20 samples in each assay). Interassay variation for measurement of total serum ALP activity was 4 f 3.7% (range 0.3-14.8%) and 2.5 f 2.4% (0.4-8.3%) for the heat inactivation and lectin precipitation methods, respectively. Interassay variation

same single serum sample (adult, fasted dog) was identified as containing intestinal ALP isoenzyme activity (i.e., corresponding to about 30% of the total serum ALP activity and consistent with the previous estimation of 21%). The sample containing intestinal ALP isoenzyme activity was analyzed as a three-component mixture by the preceding calculation, and the other 19 samples were analyzed as two-component mixtures of skeletal ALP isoenzyme activity and hepatic ALP isoenzyme activity. For the latter analyses, the three calibration solutions of skeletal ALP isoenzyme activity and hepatic ALP isoenzyme activity (i.e., 100% skeletal ALP isoenzyme activity, 100% hepatic ALP isoenzyme activity, and the 50:50 mixture) were used

TABLE5. APPLICATION OF ISOENZYMEASSAYS: SKELETAL ALP ISOENZYMEACTIVITY IN YOUNG AND ADULTCANINE SERA

Group (n

= S)

Young fasted Young fed Young fasted Young fed Adult fasted Adult fed Adult fasted Adult fed

Measurement

Heat inactivation

Total ALP Total ALP Skeletal ALP Skeletal ALP Total ALP Total ALP Skeletal ALP Skeletal ALP

226 111 185 98 24 31 0 0.1

f 42 f 8b f f f f

60 22d 8e 10e Oc3e

f

0.2c.e

WGA precipitation 242 f 45 110 f 9b 242 f 45c 110 f gblc 24 f 9e 31 f Ile 20 f 9c.e 30 f 9c.e

ConA precipitation 242 110 213 82 24 31 0.5 1.1

45 9b 58 9b f 9e f f f f

f Ile f 0.5e f 2.2e

Calculations assume that total serum ALP activity = skeletal ALP isoenzyme activity + hepatic ALP isoenzyme activity, except in one sample from an adult, fasted dog, which contained 21% intestinal ALP, for which the calculations assumed that total serum ALP activity = skeletal ALP isoenzyme activity + hepatic ALP isoenzyme activity + intestinal ALP isoenzyme activity. Data shown are group means f SD (average values from two assays, each in duplicate). bDifferent from values in fasted dogs, corresponding age and method, p < 0.001. CData indicating > 100% and 2 years) beagles, with 5 fed and 5 fasted dogs in each group. Consistent with the results of previous studies, intestinal ALP isoenzyme activity was not a major source of serum ALP in young or adult dogs (either fed or fasted). This observation allowed us to calculate the distribution of ALP isoenzyme activities in terms of two components (skeletal ALP isoenzyme activity and hepatic ALP isoenzyme activity), instead of three, and this probably increased precision.(28) All three methods indicated age-dependent decreases in skeletal ALP isoenzyme activity (and total serum ALP activity). All three methods also showed that feeding decreased skeletal ALP isoenzyme activity (and total serum ALP activity) in young but not adult dogs. Both these observations were again consistent with previous findings. Additional comparisons of skeletal ALP isoenzyme activity data between the assay methods revealed significant differences, with consistently higher results for the WGA precipitation method. These differences were most apparent with the adult canine serum, in which the WGA precipitation method gave values for skeletal ALP isoenzyme activities of 20-30 units/liter, compared with 0-0.1 units/liter by the heat inactivation method and 0.5-1.1 units/liter by the ConA precipitation method. When all the data were compared, skeletal ALP isoenzyme activity levels determined by the ConA precipitation method were correlated with those determined by the heat inactivation method ( r = 0.87,p < 0.001) but not with the levels determined by the WGA precipitation method ( r = 0.26). On closer inspection, the data suggested that the WGA precipitation method was systematically overestimating skeletal ALP isoenzyme activity levels because the ALP isoenzyme activities in young and adult beagle serum were more susceptible to WGA precipitation than the organ-derived skeletal ALP isoenzyme activity that was used to calibrate the assay. In other words, these studies showed that bone-derived skeletal ALP isoenzyme activity was not an appropriate standard (i.e., assay calibrator) for the WGA precipitation method. A similar but quantitatively less significant problem was also observed with the heat inactivation method, which may have underestimated skeletal ALP isoenzyme activity levels in adult but not young beagle serum, because serum ALP activity was more heat stable than the assay calibrator prepared from organ-derived hepatic ALP isoenzyme activity. These findings raise questions. What could account for

FARLEY ET AL.

790 these differences? Why should organ-derived skeletal ALP isoenzyme activity serve to calibrate the heat inactivation and ConA precipitation methods but not the WGA precipitation method? T o answer these questions, we need to consider the structural (biochemical) differences between skeletal and hepatic ALP isoenzyme activities that are responsible for differences in sensitivity to inactivation by heat as well as differences in lectin binding affinities. Human skeletal and hepatic ALP isoenzymes are believed to have identical amino acid sequences and differ only with respect to secondary, posttranslation glycosylation. ‘1,5,6,461 Recent indicate that the human hepatic ALP isoenzyme is 33% carbohydrate by weight, with glycosylation at multiple sites. Together, these data indicate that the physicochemical differences between human skeletal and hepatic ALP isoenzyme activities reflect differences in carbohydrate structure, and we can hypothesize analagous sugar-chain differences for canine skeletal and hepatic ALP isoenzyme activities. Although the structures of these carbohydrate moieties are undetermined, the observed affinities of skeletal and hepatic ALP isoenzyme activities for ConA and WGA indicate the presence of a-D-mannose/a-D-glucose and sialic acid/Nacetylneuraminic acid residues, respectively. Thus, we can hypothesize that bone-derived canine skeletal ALP isoenzyme activity (diluted into HIS) can be used to calibrate the heat inactivation and ConA precipitation methods because it is effectively identical to the skeletal ALP isoenzyme activity in canine serum, at least with respect to overall glycosylation (i.e., the determinant of heat inactivation) and accessible a-D-mannose/a-D-glucose residues (i.e., the determinants of C o d affinity). Conversely, we can hypothesize that bone-derived canine skeletal ALP isoenzyme activity (diluted into HIS) cannot be used to calibrate the WGA precipitation method because it differs significantly from the skeletal ALP isoenzyme activity in canine serum with respect to accessible sialic acid/N-acetylneuraminic acid residues (i.e., the determinants of WGA affinity). Consistent with the latter premise, previous studies have reported microheterogeneity of carbohydrate moieties on human skeletal ALP isoenzyme activities prepared from skeletal tissues obtained from children, adults, and patients with Paget’s disease.(484 9 ) Those studies used sequential lectin-affinity chromatography to show that essentially all the bone-derived ALP isoenzyme activity that was extracted from each tissue sample was removed from solution by ConA, but not by WGA. Sequential lectin affinity studies have also provided evidence that circulating skeletal ALP isoenzyme activity can be similarly heterogeneous, in particular with respect to glycosylation and sialic acid content.(501 Additional studies are required to resolve this issue and to address the derivative question of potential age-dependent differences in the sialic acid/N-acetylneuraminic acid content of skeletal ALP isoenzyme activity in canine serum. With respect to our original hypothesis that quantitation of skeletal alkaline phosphatase activity in canine serum would provide a useful index of the (systemic average) rate of bone formation, the current studies indicate that a combination of L-homoarginine inhibition and ConA precipitation will allow selective quantitation of skeletal A L P iso-

enzyme activity in young and adult beagle serum. Additional studies are required to determine whether these assay methods can be applied to other breeds of dogs. Additional studies are also required to define the limits of applicability of this isoenzyme activity assay as a systemic index of the rate of and/or capacity for bone formation. If this assay can provide serial estimates of systemic bone formation during long-term studies, it could be a useful complement to skeletal histomorphometry, which gives local information by methods that can only be applied (to the same animal) infrequently.

ACKNOWLEDGMENTS The authors thank the technical and secretarial staff of the Mineral Metabolism Unit of the Loma Linda University Department of Medicine, which is located in the Jerry L. Pettis Memorial Veterans’ Hospital and directed by David J. Baylink, M.D., for their assistance in these studies. The authors also thank the Medical Media Service of the Jerry L. Pettis Memorial Veterans’ Hospital and P . J . Davis, Scientific Editor, Editorial Projects Group, Norwich Eaton Pharmaceuticals, Inc. for assistance in the preparation of this manuscript. These investigations were supported by Norwich Eaton Pharmaceuticals, Inc., the Veterans’ Administration, NIH Training Grant #AR 0754304, and the Loma Linda University Department of Medicine.

REFERENCES 1 . McComb RB, Bowers GN Jr, Posen S 1979 Alkaline Phosphatase. Plenum Press, New York. 2. Robison R 1923 The possible significance of hexose-phosphoric esters in ossification. Biochem J 17:286-294. 3. Siffert RS 1951 The role of alkaline phosphatase in osteogenesis. J Exp Med 93:415-422. 4. Weiss MJ. Cole DE, Ray K, et al. 1988 A missense mutation in the human liver/bone/kidney alkaline phosphatase gene causing a lethal form of hypophosphatasia. Proc Natl Acad Sci USA 897666-7669. 5. Harris H 1989 The human alkaline phosphatases: What we know and what we don’t know. Clin Chim Acta 186:133150. 6. Fishman WH 1990 Alkaline phosphatase isozymes: Recent progress. Clin Biochem 23:99-104. 7. Clark LC, Beck E 1950 Serum alkaline phosphatase and growth in adolescent children. J Pediatr 36:335-341. 8. Lauffenburger T, Olah A J , Dambacher J , et al. 1977 Bone remodeling and calcium metabolism: A correlated histomorphometric, calcium kinetic, and biochemical study in patients with osteoporosis and Paget’s disease. Metabolism 26:589597. 9 . Marie PJ. Travers R 1983 Continuous infusion of 1,25-dihydroxy-vitamin D, stimulates bone turnover in the normal young mouse. Calcif Tissue In1 36:418-425. 10. Farley J R , Baylink DJ 1986 Skeletal alkaline phosphatase activity as a bone formation index in vitro. Metabolism 35: 563-571. 11. Farley SM, Wergedal JE, Smith LC, Lundy MW, Farley J , Baylink DJ 1987 Fluoride therapy for osteoporosis: Characterization of the skeletal response by serial measurements of

CANINE SERUM SKELETAL ALKALINE PHOSPHATASE serum alkaline phosphatase activity. Metabolism 36:211-218. 12. Duda RJ, O’Brien J F Jr, Katzmann JS, Peterson JM, Mann KG, Riggs BL 1988 Concurrent assays of circulating bone gla-protein and bone alkaline phosphtase: Effects of sex, age, and metabolic bone disease. J Clin Endocrinol Metab 66:95 1-957. 13. Brixen K, Nielsen HK, Eriksen EF, Charles P , Mosekilde L 1989 Efficacy of wheat germ lectin-precipitated alkaline phosphatase in serum as an estimator of bone mineralization rate: Comparison to serum total alkaline phosphatase and serum bone Gla-protein. Calcif Tissue Int 44:93-98. 14. Beyers DA, Fernley HN, Walker PG 1972 Studies on alkaline phosphatase. Eur J Biochem 29: 197-204. IS. VanBelle H 1976 Alkaline phosphatase. 1 . Kinetics and inhibition by levamisole of purified isoenzymes from humans. Clin Chem 22:972-976. 16. Lin CW, Fishman W H 1972 L-Homoarginine: Organ-specific uncompetitive inhibitor of human liver and bone alkaline phosphatases. J Biol Chem 247:3082-3087. 17. Lum G , Catrou P , Liuzza G , et al. 1983 Clinical assessment of the electrophoretic separation of alkaline phosphatase isoenzymes. Am J Clin Pathol 80:682-685. 18. Green S, Antiss CL, Fishman WH 1971 Automated differential isoenzyme analysis 11. Enzymologia 41:9-26. 19. Fitzgerald MNM, Fennelly J J , McGenney K 1969 The value of differential alkaline phosphatase thermostability in clinical diagnosis. Am J Clin Pathol 51:194-201. 20. Posen S, Neale FC, Clubb JS 1965 Heat inactivation in the study of human alkaline phosphatase. Ann Intern Med 62: 1234-1239. 21. Lehmann F 1980 Differentiation of human alkaline phosphatases by lectin affinity binding. Klin Wochenschr 58:947951. 22. Rosalki SB, Foo AY 1984 Two new methods for separating and quantifying bone and liver alkaline phosphatase isoenzymes in plasma. Clin Chem 30:1182-1186. 23. Behr W, Barnert J 1986 Quantification of bone alkaline phosphatase in serum by precipitation with wheat germ lectin: A simplified method and its clinical plausability. Clin Chem 32: 1960-1966. 24. Lawson GM, Katzmann JA, Kimlinger TK, O’Brien J F 1985 Isolation and preliminary characterization of a monoclonal antibody that interacts preferentially with the liver isoenzyme of human alkaline phosphatase isoenzymes. Clin Chem 31: 381-385. 25. Hirano K, Matsumoto H , Tanaka T, et al. 1987 Specific assays for human alkaline phosphatase isoenzymes. Clin Chim Acta 166:265-273. 26. Hill CS, Wolfert RL 1990 The preparation of monoclonal antibodies which react preferentially with human bone alkaline phosphatase and not liver alkaline phosphatase. Clin Chim Acta 186:315-320. 27. Farley JR, Chesnut C H 111, Baylink DJ 1981 Improved method for determination of serum alkaline phosphatase of skeletal origin. Clin Chem 27:2002-2007. 28. Valenzuela GJ, Munson LA, Tarbaux NM, Farley J R 1987 Time-dependent changes in bone, placental, intestinal, and hepatic alkaline phosphatase activities in serum during human pregnancy. Clin Chem 33:1801-1806. 29. Farley J , Kyeyune-Nyombi E, Tarbaux N, Hall S, Strong D 1989 Alkaline phosphatase from human osteosarcoma cell line SaOS-2: An isoenzyme standard for quantifying skeletal alkaline phosphatase activity in serum. Clin Chem 35:223229.

791

30. Bodansky A 1934 Phosphatase studies. VII. Inorganic phosphorous and phosphatase of the serum in newborn puppies. J Biol Chem 104:717-725. 31. Bodansky A 1934 Phosphatase studies. VI. Non-osseous origins of serum alkaline phosphatase by the liver and pancreas of normal dogs. J Biol Chem 104:473-482. 32. Drill VA, Annegers J H , Ivy AC 1944 Effect of cyanide, fluoride and magnesium on serum phosphatase activity during hepatic damage. J Biol Chem 152:339-347. 33. Nagode LA, Koestner A, Steinrneyer C L 1969 The effects of purification and serum proteins on the organ-identifying properties of alkaline phosphatases from canine tissues. Clin Chim Acta 2655-61, 34. Van Belle H 1972 Kinetics and inhibition of alkaline phosphatases from canine tissues. Biochim Biophys Acta 289: 158-168. 35. Marcuson RW, Tomlinson B 1972 L-Phenylalanine inhibition as a method for measuring the intestinal component of alkaline phosphatase in dog serum. Clin Chim Acta 42:245253. 36. Hoffman W, Dorner J L 1975 Separation of isoenzymes of canine alkaline phosphatase by cellulose acetate electrophoresis. J Am Anim Hosp Assoc 11:283-285. 37. Hoffman W, Dorner J L 1977 Disappearance rates of intravenously-injected canine alkaline phosphatase isoenzymes. Am J Vet Res 38:1553-1556. 38. Saini PK, Peavy GM, Hauser DE, et al. 1978 Diagnostic evaluation of canine serum alkaline phosphatase by immunochemical means and interpretation of results. Am J Vet Res 39: 15 14- 15 18. 39. Eckersall PD, Nash AS 1983 Isoenzymes of canine plasma alkaline phosphatase: An investigation using isoelectric focusing and related to diagnosis. Res Vet Sci 34:310-314. 40. Amacher DE, Smith DJ, Martz LK, Hoffmann WE 1987 Characterization of alkaline phosphatase in canine serum. Enzyme 37: 141- 149. 41. Amacher DE, Higgins CV, Schomaker SJ, Clay RJ 1989 Further characterization of serum alkaline phosphatase from male and female beagle dogs. Enzyme 42:l-7. 42. Dorner JL, Hoffmann WE, Long GB 1974 Corticosteroid induction of an isoenzyme of alkaline phosphatase in the dog. Am J Vet Res 35:1457-1458. 43. Wellman ML, Hoffmann WE, Dorner JL, Mock RE 1982 Comparison of the steroid-induced, intestinal, and hepatic isoenzymes of alkaline phosphatase in the dog. Am J Vet Res 43: 1204- 1207. 44. Wellman ML, Hoffmann WE, Dorner JL, Mock RE 1982 Immunoassay for the steroid-induced isoenzyme of alkaline phosphatase in the dog. Am J Vet Res 43:1200-1203. 45. Bradford MM 1976 A rapid and sensitive method for the quantitation of microgram amounts of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248-255. 46. Weiss MJ, Harthorn PS, Lafferty MA, Slaughter C , Raducha M, Harris H 1986 Isolation and characterization of a cDNA encoding a human liver/bone/lidney-type alkaline phosphatase. Proc Natl Acad Sci USA 83:7182-7186. 47. Garattini E, Hua J-C, Pan Y-CE, Udenfriend S 1986 Human liver alkaline phosphatase, purification and partial sequencing: Homology with the placental isoenzyme. Arch Biochem Biophys 245331-337. 48. Koyarna 1, Miura M, Matsuzaki H, Sakagishi Y, Komoda T 1987 Sugar chain heterogeneity of human alkaline phosphatases: Differences between normal and tumour-associated isoenzymes. J Chromatog 413:65-78.

FARLEY ET AL.

792 49. Miura M, Matsuzaki H, Bailyes EM, Koyama I , Sakagishi Y , Sekine T, Komoda T 1989 Differences between human liverand bone-type alkaline phosphatases. Clin Chim Acta 180: 177-1 88. 50. Kuwana T, Sugita 0, Yakata M 1989 Sugar-chain analysis of

alkaline phosphatase by lectin affinity chromatography in serum from a patient with osteosarcoma. Clin Chem 35:179.

Address reprint requests to:

John R. Farley Research Service (151) Jerry L . Pettis Memorial Veterans’ Hospital 11201 Benton Street Loma Linda, CA 92374 Received for publication August 7, 1991; in revised form February 4, 1992; accepted February 5, 1992.

Quantitation of skeletal alkaline phosphatase isoenzyme activity in canine serum.

Pursuing the hypothesis that quantitation of skeletal alkaline phosphatase (ALP) activity in canine serum would provide an index of the rate of bone f...
1MB Sizes 0 Downloads 0 Views