- Email: [email protected]
Evaluation of serum bone alkaline phosphatase activity in patients with liver disease: Comparison between electrophoresis and chemiluminescent enzyme immunoassay
Evaluation of serum bone alkaline phosphatase activity in patients with liver disease: Comparison between electrophoresis and chemiluminescent enzyme immunoassay Fangjie Zhan, Yoshihisa Watanabe, Aya Shimoda, Etsuko Hamada, Yoshimasa Kobayashi, Masato Maekawa PII: DOI: Reference:
S0009-8981(16)30253-4 doi: 10.1016/j.cca.2016.06.008 CCA 14398
To appear in:
Clinica Chimica Acta
Received date: Revised date: Accepted date:
18 March 2016 8 June 2016 8 June 2016
Please cite this article as: Zhan Fangjie, Watanabe Yoshihisa, Shimoda Aya, Hamada Etsuko, Kobayashi Yoshimasa, Maekawa Masato, Evaluation of serum bone alkaline phosphatase activity in patients with liver disease: Comparison between electrophoresis and chemiluminescent enzyme immunoassay, Clinica Chimica Acta (2016), doi: 10.1016/j.cca.2016.06.008
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Evaluation of serum bone alkaline phosphatase activity in patients with liver disease:
IP
T
comparison between electrophoresis and chemiluminescent enzyme immunoassay
SC R
Fangjie Zhana, Yoshihisa Watanabea, Aya Shimodab, Etsuko Hamadab, Yoshimasa Kobayashic, Masato Maekawaa, b,*
Department of Laboratory Medicine, Hamamatsu University School of Medicine.
b
Department of Clinical Laboratory, Hamamatsu University Hospital.
MA
c
NU
a
Second Division, Department of Internal Medicine, Hamamatsu University School of
D
Medicine.
CE P
TE
1-20-1, Handayama, Higashi-ku, Hamamatsu, 431-3192, Japan.
*Corresponding author at: Department of Laboratory Medicine, Hamamatsu University
AC
School of Medicine, Hamamatsu 431-3192, Japan. Tel: +81-53-435-2721; Fax: +81-53-435-2096.
E-mail address: [email protected] (M. Maekawa).
Abbreviations: ALP, alkaline phosphatase; BAP, bone-type alkaline phosphatase; CLEIA, chemiluminescent enzyme immunoassay; TC, total cholesterol; CH, cholesterol; TG, triglyceride; γ-GTP, γ-glutamyltransferase; LAP, leucine aminopeptidas
1
ACCEPTED MANUSCRIPT Abstract Background: Serum bone alkaline phosphatase (ALP) is a marker of bone formation and
IP
T
metabolism. However, existing methods for measuring it have their limitations and their
SC R
accuracy has not been determined.
Methods: We measured serum bone ALP activity in 127 patients with liver disease using 2 methods: electrophoresis and chemiluminescent enzyme immunoassay (CLEIA). The results
NU
of these 2 methods were compared and analyzed according to gender, age and several serum
MA
biochemical markers.
Results: When ALP3 (%; bone-type isozyme activity as a percentage of total ALP activity)
D
values were high, the 2 methods showed good correlation. However, with a decrease in ALP3
TE
(%) levels, the correlation coefficient (R) also decreased. Starting with ALP3 (%) < 23, R
CE P
values markedly decreased to <0.50 (p > 0.05). Five outliers displayed low ALP3 (%) activity levels. Furthermore, in regard to genders, there were significant differences in total
AC
cholesterol (TC), γ-glutamyltransferase (γ-GTP), ALP and ALP3 (%) levels (p < 0.05). Conclusions: When serum ALP3 (%) levels were high in patients with liver disease, the accuracy of electrophoresis was comparable to that of CLEIA. However, the accuracy of electrophoresis needs to be evaluated with further when patient samples under certain conditions. Keywords: Alkaline phosphatase isozyme; Bone-type alkaline phosphatase; Liver disease; Electrophoresis; Chemiluminescent enzyme immunoassay
2
ACCEPTED MANUSCRIPT 1. Introduction Alkaline phosphatase (ALP) is a membrane-bound metalloenzyme that is found in nearly
IP
T
all species [1,2]. Human ALP consists of a group of isozymes, placental, intestinal and three
SC R
major tissue-nonspecific isoforms [3], which are encoded by 4 main gene loci: tissue-nonspecific, intestinal, placental, and germ-cell ALP [4-6]. Tissue-nonspecific bone-
NU
and liver-type ALP are formed through posttranslational modifications and differences in carbohydrate composition, leading to different clinical phenotypes [4]. ALP isozymes have
MA
been used as biomarkers for liver and bone disease in the clinic [7-10]. Hence, the measurement of ALP isozymes activity in serum has a high clinical value in the differential
D
diagnosis and analysis of some diseases.
TE
Bone ALP is an extracellular enzyme that is found anchored to the osteoblast membrane
CE P
and which reflects whole skeletal remodeling [11]; it is one of the most commonly used biochemical markers for osteoblastic bone formation analyzed in the routine clinical
AC
chemistry laboratory [12,13]. Metastatic bone tumors, diabetes, hyperthyroidism, fracture recovery period and other conditions lead to the hyperfunction of bone metabolism or formation, which, in turn, increases serum bone ALP levels. In addition, serum bone ALP varies with age and gender. Humans in infancy and at puberty show 2 physiological blood bone ALP concentration peaks, with troughs in mid-childhood and at the end of adolescence [14,15]. Furthermore, for postmenopausal women with high bone turnover, bone ALP concentrations were found to be significantly higher than for those in premenopausal [16-18]. Presently, the predominant methods for separating and measuring bone ALP are
3
ACCEPTED MANUSCRIPT electrophoresis and chemiluminescent enzyme immunoassay (CLEIA). The separation of proteins by gel electrophoresis occurs according to their differences in net charge, isoelectric
IP
T
point and molecular weights, resulting in different migration patterns. Migrating in order
SC R
from the anode to cathode, ALPs were named ALP1 (a high molecular mass, liver-type ALP), ALP2 (liver-type ALP), ALP3 (bone-type ALP), ALP4 (placental-type ALP), ALP5 (intestinal-type ALP) and ALP6 (immunoglobulin binding-type ALP). However, the
NU
migration of bone ALP by electrophoresis is also influenced by the presence of other types of
MA
ALP isozymes.
The CLEIA technique involves samples binding to mouse monoclonal anti-bone ALP
D
antibodies that, in turn, bind magnetic particles coated with goat anti-mouse polyclonal
TE
antibodies. The chemiluminescent substrate (lumigen PPD) emits light according to the
CE P
activity of bone ALP, with the amount of light emitted reflecting the bone ALP concentration of the sample [19]. CLEIA is widely used in the clinic due to its high sensitivity and
AC
specificity, and is simple and rapid [20]. However, the CLEIA method also has disadvantages such as cross-immunological reactivity between bone- and liver-type ALPs, as well as interference by liver-type ALP. CLEIA also does not yield information on other isozymes and isoforms of ALP present within a sample. In fact, the 2 methods have their own distinct limitations when detecting bone ALP.
Materials and methods 2.1. Subjects
4
ACCEPTED MANUSCRIPT The subjects of this study were 127 patients (25 – 83 y; males 54, females 73) with liver disease who were treated at Hamamatsu University Hospital.
Of these, 52 patients had
IP
T
primary biliary cirrhosis, 8 were affected by cholestatic cirrhosis, 13 had cholestatic hepatic
SC R
disorder, and the rest were diagnosed with other liver diseases. Both agarose gel electrophoresis and CLEIA were used to detect bone ALP in serum samples. Other patient biochemical analyses were also undertaken, such as tests for total cholesterol (TC),
NU
triglyceride (TG), γ-glutamyltransferase (γ-GTP) and leucine aminopeptidase (LAP), as well
MA
as others. The ethics committee of Hamamatsu University School of Medicine approved this
TE
D
study.
CE P
1.2.Electrophoretic separation of ALP isozymes For the analysis of ALP isozymes, electrophoresis using agarose gels was undertaken according to the manufacturer’s operating instructions. Quick Gel ALP and Quick ALP
AC
reagent (Quick ALP; Helena Labs) have been commercially developed for the electrophoretic analysis of ALP [21]. In order to separate liver- and bone-type ALP completely, samples were pretreated with neuraminidase (ALP separator; Helena) [6]. In addition, samples were pretreated with protease (Quick ALP; Helena) to separate bone- and intestinal-type ALP. Following electrophoresis, a 3-indoxyl phosphate disodium salt was used as a substrate and nitrotetrazolium blue was used as a chromogenic agent. Bands were scanned by densitometry at a wavelength of 570 nm [21]. ALP isozyme activity was expressed in U/l and as a percentage of total ALP activity. Samples from 3 healthy individuals were measured using 5
ACCEPTED MANUSCRIPT three lots of reagents, indicating the number of assays for reproducibility analysis totaled nine times. The coefficients of variation (CVs) for ALP3 activities were 1.22–3.00%. However,
IP
T
these were not the CVs for ALP2 and ALP3 percentages, which are quite different samples.
SC R
Serum samples and protease were mixed using a ratio of 5:1 at room temperature for 30 min; both serum and protease-treated samples, respectively, were mixed with neuraminidase at a ratio of 7:1, and left at room temperature for 12 min. Control, serum, separator-treated,
NU
protease-treated and separator + protease-treated samples were loaded in order on the gel
MA
plate. Electrophoresis was performed at 250 V for 20 min. The gel was coated in sample reagent and stained for one min before incubation at 37ºC for 35 min. The gel was soaked in
D
70% methanol for 10 min and washed with water twice for 10 min each wash. Bands were
1.3. CLEIA
CE P
TE
then scanned by densitometry at a wavelength of 570 nm.
AC
A 1-step immunoenzymatic assay was used in this study. An automated chemiluminescence enzyme immunoassay device, known as an Access Immunoanalyzer (Beckman Coulter), and a special bone-type ALP (BAP) kit (Access Ostase; Beckman Coulter) were used [19,22]. Beckman BAP Access Ostase operating instructions were followed. Samples and reagents were mixed, and all operations were automatically performed as long as the appropriate operating items were chosen. Each 25 μl sample was mixed with 50 μl of a solid phase reagent (magnetic particles coated with goat anti-mouse polyclonal antibodies) and 15 μl of an antibody reagent (mouse monoclonal anti-BAP antibodies), and 6
ACCEPTED MANUSCRIPT then washed after reacting at 37℃ for 14.4 min. Next, a 200 μl substrate solution (lumigen PPD) was added and samples incubated at 37℃ for 5 min. After each reaction was complete,
IP
T
its luminous intensity (λ max = 540 nm) was assayed by immunoanalyzer and the
SC R
concentration of BAP in the sample was automatically calculated using a standard curve. Assays were repeated 3 times in reproducibility analysis, for which intraday and interday CVs for BAP were 3.82–5.79% and 2.60–5.25% for a 10.39 μg/l concentration, 3.77–5.66%
D
2.4. Serum biochemical examinations
MA
NU
and 3.33–5.80% for 28.07 μg/l.
TE
Basic biochemical examinations of serum samples, using a chemistry analyzer (Labospect 008; Hitachi Ltd), were undertaken. Cholesterol (CH), TG and γ-GTP assays were performed
CE P
according to the manufacturer’s operating instructions. To determine CH and TG fractions in serum, serum lipoproteins were separated by electrophoresis using agarose gels. After
AC
electrophoresis, reagents (Choletricombo CH/TG; Helena) within the agarose reacted with CH and TG in lipoprotein fractions, and the formazan consequently produced was scanned by densitometry at a wavelength of 570 nm. For the measurement of serum γ-GTP isozymes, a cellulose acetate membrane was used as a support and soaked in Tris-barbital buffer (pH 8.8) before serum γ-GTP isozymes were separated by electrophoresis. After electrophoresis, reagents (Titan γ-GTP; Helena) within the acetate membrane generated azodyes that were scanned by densitometry at a wavelength of 570 nm.
7
ACCEPTED MANUSCRIPT 2.5. Statistical analysis All statistical analyses were performed using SPSS software (v18.0, SPSS Inc.,). Data are
IP
T
expressed as mean ± SD. Correlations between ALP3 (bone ALP determined by
SC R
electrophoresis) and BAP (bone ALP measured by CLEIA) were determined using Pearson’s correlation coefficient. Differences in means between methods were evaluated by Student’s
NU
t-test. A p <0.05 was considered to indicate statistical significance.
MA
3. Results
D
3.1. Determination of serum bone-type ALP by 2 methods
TE
We used 2 methods, electrophoresis and CLEIA (Suppl. Fig. 1), to detect bone ALP in
CE P
the serum samples of 127 patients with liver disease (Suppl. Table 1). A comparison of the 2 methods was made, with CLEIA used as a control method and electrophoresis evaluated in order to assess its accuracy in the determination of bone ALP. Fig. 1 shows how bone ALP
AC
samples analyzed by electrophoresis (ALP3) compared with those analyzed by CLEIA (BAP). The linear equation was calculated as y ＝ 8.63 x - 44.9, with a correlation coefficient (R) of 0.886; The 2 methods showed a statistically significant correlation (p < 0.001). Furthermore, results for some samples deviated from those of most samples (black triangles), with all outliers below the line.
3.2. Further stratified analysis of all samples We grouped all samples according to their percentage of ALP3 in 20% increments: > 80% 8
ACCEPTED MANUSCRIPT (n = 8), 60–80% (n = 18), 40–60% (n = 42), 20–40% (n = 41) and < 20% (n = 18), in a total of five groups as shown in Fig. 2. The corresponding R values were 0.988, 0.969, 0.947,
IP
T
0.892 and 0.350. Although 4 groups (Figs. 2A–D) showed a good correlation between the 2
SC R
detection methods, concomitant with a decrease in ALP3 (%) the R values for the 2 methods progressively decreased. For the < 20% group, R = 0.350 was significantly lower than the average of the other 4 groups (p < 0.001; Fig. 2E). The slopes of linear equations for all 5
NU
groups also progressively became reduced (Figs. 2A to E).
MA
Furthermore, we performed a more refined stratification between the 2 groups, 20 – 40% and < 20% (Fig. 3A), which were subsequently divided into 8 groups: ALP3 (%) < 27, < 26,
D
< 25, < 24, < 23, < 22, < 21 and < 20. We found that R values markedly dropped to less than
TE
0.50 from ALP3 (%) < 23 onwards, corresponding to a bad correlation between the 2
CE P
methods (p > 0.05). We compared the “all samples” group, ALP3 (%) > 23, and < 23 groups (Fig. 3B). For the “all samples” group, R = 0.886 (p < 0.001). For ALP3 (%) > 23, R = 0.951
AC
(p < 0.001). However, the ALP3 (%) < 23 group showed a poor correlation, with R = 0.405 and p > 0.05. Thus the ALP3（%）< 23 group exhibited the worse correlation between the 2 kinds of detection methods compared with the other 2 groups. When we removed these 5 deviating samples from the rest of the samples, we found that the corresponding R values of the 2 groups, 20 < ALP3 (%) < 40 and ALP3 (%) < 20, were 0.920 and 0.772, respectively, which were significantly higher than those of samples containing outliers (Figs. 2D–E and 4). However, the R value of the < 20% group, after the removal of outliers, was also significantly lower than the average of the other 4 groups (p <
9
ACCEPTED MANUSCRIPT 0.001).
IP
T
3.3. Clinical profiles of patients with liver disease
SC R
We assayed other patient biochemical data using detection methods according to Japan Society for Clinical Chemistry (JSCC) standards. The baseline characteristics of the 127 liver disease patients are shown in Table 1. Mean levels of serum BAP, ALP, γ-GTP and LAP for
NU
all samples were all above the respective normal upper limit. Due to gender differences, TG,
MA
LAP, γ-GTP, ALP and BAP levels for males were higher than for females, although only γ-GTP (p < 0.001) and ALP (p < 0.010) were significant. However, for TC, HDL-C, LDL-C
D
and ALP3 (%), levels for males were lower than those for females, although only TC (p <
CE P
TE
0.002) and ALP3 (%；p < 0.022) were significant.
3.4. Sample deviations
AC
Five outliers (Fig. 1; black triangles) were noted with common characteristics. Table 2 shows that total serum ALP activity was clearly raised at > 800 U/l for each; ALP2 was the predominant ALP isozyme, and all outliers exhibited high levels of ALP1 (> 175 U/l). LAP and γ-GTP were significantly increased but HDL was reduced. Since all outliers were below the line, bone ALP values determined by electrophoresis were lower than those determined by CLEIA. The outliers showed low ALP3 (%) levels.
10
ACCEPTED MANUSCRIPT 4. Discussion The main challenge of bone ALP quantification is to differentiate bone- and liver-type
IP
T
ALPs since these are encoded by the same gene and have an identical amino acid sequence
SC R
[23]. Electrophoresis is, at present, considered unable to completely distinguish between bone- and liver-type ALP in some liver diseases. Therefore, the accuracy of detection of
NU
serum bone ALP by electrophoresis in patients with liver disease is worthy of further, more
MA
detailed study.
4.1. Correlation of the 2 methods
D
In this study, we used CLEIA as a control method with which to compare the
TE
determination of bone ALP by electrophoresis, and to find the limitations of electrophoresis
CE P
when measuring bone ALP. The overall aim is to avoid such limitations in future clinical examinations, and to select the best detection methods under different situations, in order to
AC
provide accurate and reliable data for clinical medicine. Using all samples, the 2 methods showed a good correlation with R = 0.886 (p < 0.001), indicating the reliability of the determination of bone ALP by electrophoresis in the majority of cases. However, the regression of all samples was a curve, the slope of the curve line increased with the increase of bone ALP activity. The reason for this may be related to the CLEIA method has cross-immunological reactivity between bone- and liver-type ALPs. When liver-type ALP predominated, the BAP interference by liver-type ALP was serious, which led to fact that the results measured by CLEIA was higher than those of electrophoresis. And when ALP3
11
ACCEPTED MANUSCRIPT extremely high and low, it is difficult to quantify by visually determining the inflection point between fractions on densitometry when ALP isozymes are analyzed by electrophoresis, the
4.2. Exploration of the limitations of electrophoresis
SC R
IP
T
measurement errors exist.
All serum samples were then divided into 5 groups according to the percentage of ALP3
NU
activity. Although the 2 methods correlated well for the first 4 groups showing the highest
MA
percentage of ALP3 activity (p < 0.001), R values became progressively smaller from the first to the fourth group in descending order of ALP3（%）. The last group, with ALP3 (%) <
D
20, showed R = 0.350 (p > 0.05), meaning that correlation between the 2 methods was poor.
TE
In other words, the lower the ALP3 (%), the worse the correlation was. Because the R value
CE P
of the ALP3 (%) < 20 group was significantly lower than those of the other 4 groups, we performed a more refined stratification between the 20 – 40% and < 20% groups in order to
AC
identify the limit for correlation between the 2 methods. Starting with an ALP3 (%) < 23 downwards, correlation between the 2 methods worsened, with all displaying R < 0.50 and p > 0.05. The reason for this may be related to the proportion of samples containing deviation samples becoming higher. The accuracy of the determination of bone ALP by electrophoresis in some special cases requires further, more detailed analysis. After removing the 5 outliers from the groups, 20 < ALP3 (%) < 40 and ALP3 (%) < 20, we found that the 2 groups’ corresponding R values were 0.920 and 0.772, respectively, which were significantly higher than that of samples containing outliers. However, the R value of the ALP3 (%) < 20 group
12
ACCEPTED MANUSCRIPT after removal of the outliers was also significantly lower than the average of the other 4 groups, and with a decrease in ALP3 (%) levels, the correlation coefficient also decreased. It
IP
T
can be concluded that from the first to the fifth group, the correlation worsened and the trend
SC R
did not change after the 5 outliers were removed.
Fig. 2 shows that the slope of the linear equation decreased with each decrease in ALP3 (%). A shallow slope signified that results measured by electrophoresis were lower than those
NU
of CLEIA. In this regard, the existence of 5 outliers is particularly concerning. It may be that
MA
the lower the ALP3 (%) level, the more difficult it is to quantify by visually determining the inflection point between fractions on densitometry when ALP isozymes are analyzed by
D
electrophoresis. Therefore, when liver-type ALP predominated (ALP2 % > ALP3 %), we
TE
considered the error in the measurement of bone-type ALP activity by electrophoresis to be
CE P
large. In contrast, when bone-type ALP predominated (ALP2 % ≤ ALP3 %), the correlation between these 2 methods was good, and so data from electrophoresis were considered to be
AC
relatively reliable (see Suppl. Fig. 2).
4.3. Biochemical data showing significant gender differences As mentioned above, the measurement of bone ALP was interfered with by liver ALP. The various liver diseases, which can lead to different levels of total ALP (tALP), and changes in bone ALP levels also differed. So we combined this with liver disease-related biochemical data to analyze the ratio of liver- and bone-type ALPs and the degree of interference by liver ALP in the serum of patients with different diseases in order to
13
ACCEPTED MANUSCRIPT determine the cause of abnormal increases or decreases of bone ALP. These biochemical data play an important role in determining the reliability of the results of electrophoresis.
IP
T
In our study, tALP activity in males was significantly higher than that for females. For
SC R
males, tALP was higher than that of women of the corresponding age, up until 60 years of age, by which time levels became almost equal [24]. In addition, since all samples used in this study were from patients with liver disease, the observed gender difference may be
NU
related to differences in liver diseases and the severity of disease. From all samples, the
MA
ALP3 (%) level for women were significantly higher than that for men, which may be related to a postmenopausal increase in ALP levels because of high bone turnover, with elevated
D
serum ALP derived from bone tissue [25,26]. TC in women was significantly higher than in
TE
men in our study. Previous research highlighted a difference between males and females in
CE P
the causes of liver disease, with TC synthesis representing a key point of gender difference [27]. In addition, γ-GTP values of all samples were increased, with those in men significantly
AC
higher than those in women. In fact, γ-GTP levels showed large individual differences, which significantly differed according to gender and drinking history [28], while it has been suggested that lower γ-GTP levels in females is probably physiological [29]. Although serum bone ALP changes with age and sex, in our study these had no serious impact on the overall correlation between the 2 methods (Suppl. Figs. 3, 4).
4.4. Common characteristics of deviating samples We found 5 outliers with common characteristics. As a result, we inferred that outlier 14
ACCEPTED MANUSCRIPT samples came from patients who may have been affected by an intrahepatic space-occupying lesion or cholangitis. The bone ALP values of such samples as determined by electrophoresis
IP
T
were lower than those determined by CLEIA, and therefore deviated from the overall
SC R
linearity displayed by most samples. When we removed these 5 deviating samples from the rest of the samples, we found that for the 2 groups, 20 < ALP3 (%) < 40 and ALP3 (%) < 20, the correlation between the 2 methods improved. This is indicates that the main reason for the
NU
deterioration in the correlation for the 20 < ALP3 (%) < 40 and < 20% groups were these 5
MA
outlier samples. From < 23%, the 2 method correlation deteriorated and the reason why may be related to the increasing proportion of deviating samples. Therefore, the 5 outliers showed
D
a certain representation that was not accidental; with the determination of such samples by
TE
electrophoresis, the accuracy was questionable.
CE P
The outliers were very low in ALP3 (%), thus indicating the predominance of liver-type ALP. The questionable accuracy of densitometry after electrophoresis may be a major factor
AC
for such variation. Bone and liver ALP were quantified by visually determining the inflection point between the 2 ALP fractions on densitometry, so when ALP3 (%) is low, quantification is more prone to errors in judgement. From this, we have doubts concerning the accuracy of the determination of serum bone ALP by electrophoresis in patients with an intrahepatic space-occupying lesion or cholangitis, when the ALP3 (%) level is very low.
4.5. Limitations There are some limitations in this study. First, the samples were obtained from a general
15
ACCEPTED MANUSCRIPT population of different liver disease patients. It could lead to different levels of the activity of ALP and bone-type ALP, as well as the larger variation of the results of other biochemical
IP
T
tests; Although we grouped all samples according to their percentage of ALP3 in each 20%
SC R
increments, the number of samples of each group was not the same: very high and low ALP3 (%) groups were less than the middle level groups. Therefore, the R values statistical reliability was worth thinking about, especially the group of too small sample size. Second,
NU
when electrophoresis was used to analyze the ALP isozymes required by visually determining
MA
the inflection point between fractions on densitometry, this may cause the measurement
D
errors.
TE
5. Conclusion
CE P
Our study suggests that for whole and bone-type ALP predominating (ALP2 % ≤ oALP3 %) samples, electrophoresis and CLEIA methods show good correlation. However, with a decrease in ALP3 (%) levels, the correlation coefficient also decreased. The 5 outlier
AC
samples lead to start with ALP3 (%) < 23, the R value decreased to less than 0.50 and no statistical correlation between the 2 methods was observed (p > 0.05). Therefore, the accuracy of electrophoresis needs to be evaluated further when the patient is affected by an intrahepatic space-occupying lesion or cholangitis, and shows a low ALP3 (%) level. As aforementioned, although electrophoresis exhibits measurement limitations, it provides useful information on several different isozymes. Validation requirements have increased with further developments and progress in clinical laboratory technology. A major necessity is to uncover the limitations of currently used detection methods, to avoid such 16
ACCEPTED MANUSCRIPT errors in clinical applications and to maximally exploit the advantages of the methods used. Finally, in a similar manner, we can uncover limitations in electrophoresis measurements with
IP
T
regard to other proteases, and can review and evaluate further detection methods based on the
SC R
same principles (Suppl. Fig. 5).
Disclosure
MA
NU
The authors wish to declare that they have no conflicts of interest.
Acknowledgment
D
The authors wish to thank Mr Shinichi Miyashita of the Department of Helena
References
CE P
TE
Laboratories for technical guidance in Choletricombo CH/TG and γ-GTP experiments.
AC
[1] Sharma U, Pal D, Prasad R. Alkaline phosphatase: an overview. Indian J Clin Biochem 2014; 29: 269-278.
[2] Khan KN, Tsutsumi T, Nakata K, Nakao K, Kato Y, Nagataki S. Regulation of alkaline phosphatase gene expression in human hepatoma cells by bile acids. J Gastroenterol Hepatol 1998; 13:643-650. [3] Khan KN, Tsutsumi T, Nakata K, Kato Y. Sodium butyrate induces alkaline phosphatase gene expression in human hepatoma cells. J Gastroenterol Hepatol 1999; 14:156-162. [4] Van Hoof VO, De Broe ME. Interpretation and clinical significance of alkaline
17
ACCEPTED MANUSCRIPT phosphatase isoenzyme patterns. Crit Rev Clin Lab Sci 1994; 31:197-293. [5] Kiffer-Moreira T, Sheen CR, Gasque KC, et al. Catalytic signature of a heat-stable,
IP
T
chimeric human alkaline phosphatase with therapeutic potential. PloS one 2014; 9:e89374.
SC R
[6] Nollet E, Van Craenenbroeck EM, Martinet W, et al. Bone matrix vesicle-bound alkaline phosphatase for the assessment of peripheral blood admixture to human bone marrow aspirates. Clin Chim Acta 2015; 446:253-260.
NU
[7] Tsumura M, Ueno Y, Kinouchi T, Koyama I, Komoda T. Atypical alkaline phosphatase
MA
isozymes in serum and urine of patients with renal failure. Clin Chim Acta 2001; 312:169-178.
D
[8] Kang KY, Hong YS, Park SH, Ju JH. Increased serum alkaline phosphatase levels
TE
correlate with high disease activity and low bone mineral density in patients with axial
CE P
spondyloarthritis. Semin Arthritis Rheum 2015: 202-207. [9] Woitge HW, Seibel MJ, Ziegler R. Comparison of total and bone-specific alkaline
AC
phosphatase in patients with nonskeletal disorder or metabolic bone diseases. Clin Chem 1996; 42:1796-1804.
[10] Nayeem F, Anderson KE, Nagamani M, Grady JJ, Lu LJ. Alkaline phosphatase and percentage body fat predict circulating C-reactive protein in premenopausal women. Biomarkers 2010; 15:663-670. [11] Lau WL, Kalantar-Zadeh K, Kovesdy CP, Mehrotra R. Alkaline phosphatase: Better than PTH as a marker of cardiovascular and bone disease? Hemodial Int 2014; 18:720-724. [12] Dziedziejko V, Safranow K, Slowik-Zylka D, et al. Characterisation of rat and human
18
ACCEPTED MANUSCRIPT tissue alkaline phosphatase isoforms by high-performance liquid chromatography and agarose gel electrophoresis. Biochimie 2009; 91:445-452.
IP
T
[13] Pathomwichaiwat T, Ochareon P, Soonthornchareonnon N, Ali Z, Khan IA,
SC R
Prathanturarug S. Alkaline phosphatase activity-guided isolation of active compounds and new dammarane-type triterpenes from Cissus quadrangularis hexane extract. J Ethnopharmacol 2015; 160:52-60.
NU
[14] Shimose S, Kubo T, Fujimori J, Furuta T, Ochi M. A novel assessment method of serum
MA
alkaline phosphatase for the diagnosis of osteosarcoma in children and adolescents. J Orthop Sci 2014; 19:997-1003.
D
[15] Rauchenzauner M, Schmid A, Heinz-Erian P, et al. Sex- and age-specific reference
TE
curves for serum markers of bone turnover in healthy children from 2 months to 18 years. J
CE P
Clin Endocrinol Metab 2007; 92:443-449. [16] Atalay S, Elci A, Kayadibi H, Onder CB, Aka N. Diagnostic utility of osteocalcin,
AC
undercarboxylated osteocalcin, and alkaline phosphatase for osteoporosis in premenopausal and postmenopausal women. Ann Lab Med 2012; 32:23-30. [17] Shan PF, Wu XP, Zhang H, et al. Age-related changes of serum bone alkaline phosphatase and cross-linked C-telopeptides of type I collagen and the relationship with bone mineral density in Chinese women. Clin Chim Acta 2006; 366:233-238. [18] Takahashi M, Kushida K, Hoshino H, Miura M, Ohishi T, Inoue T. Comparison of bone and total alkaline phosphatase activity on bone turnover during menopause and in patients with established osteoporosis. Clin Endocrinol (Oxf) 1997; 47:177-183.
19
ACCEPTED MANUSCRIPT [19] Cavalier E, Rozet E, Carlisi A, et al. Analytical validation of serum bone alkaline phosphatase (BAP OSTASE) on Liaison. Clin Chem Lab Med 2010; 48:67-72.
IP
T
[20] Sekiguchi S, Kohno H, Yasukawa K, Inouye K. Chemiluminescent enzyme immunoassay
SC R
for measuring leptin. Biosci Biotechnol Biochem 2011; 75:752-756.
[21] Ooi K, Shiraki K, Morishita Y, Nobori T. High-molecular intestinal alkaline phosphatase in chronic liver diseases. J Clin Lab Anal 2007; 21:133-139.
NU
[22] Schafer AL, Vittinghoff E, Ramachandran R, Mahmoudi N, Bauer DC. Laboratory
MA
reproducibility of biochemical markers of bone turnover in clinical practice. Osteoporos Int 2010; 21:439-445.
D
[23] Broyles DL, Nielsen RG, Bussett EM, et al. Analytical and clinical performance
TE
characteristics of Tandem-MP Ostase, a new immunoassay for serum bone alkaline
CE P
phosphatase. Clin Chem 1998; 44:2139-2147. [24] Shimizu Y, Ichihara K, Asia-Pacific Federation of Clinical B. Sources of variation
AC
analysis and derivation of reference intervals for ALP, LDH, and amylase isozymes using sera from the Asian multicenter study on reference values. Clin Chim Acta 2015; 446:64-72. [25] Mukaiyama K, Kamimura M, Uchiyama S, Ikegami S, Nakamura Y, Kato H. Elevation of serum alkaline phosphatase (ALP) level in postmenopausal women is caused by high bone turnover. Aging Clin Exp Res 2015; 27:413-418. [26] Bhattarai T, Bhattacharya K, Chaudhuri P, Sengupta P. Correlation of common biochemical markers for bone turnover, serum calcium, and alkaline phosphatase in post-menopausal women. Malays J Med Sci 2014; 21:58-61.
20
ACCEPTED MANUSCRIPT [27] Lewinska M, Juvan P, Perse M, et al. Hidden disease susceptibility and sexual dimorphism in the heterozygous knockout of Cyp51 from cholesterol synthesis. PloS one
IP
T
2014; 9:e112787.
SC R
[28] Pintus F, Mascia P. Distribution and population determinants of
gamma-glutamyltransferase in a random sample of Sardinian inhabitants. Eur J Epidemiol 1996; 12:71-76.
NU
[29] Skurtveit S, Tverdal A. Sex differences in gamma-glutamyltransferase in people aged
AC
CE P
TE
D
MA
40-42 years in two Norwegian counties. Drug Alcohol Depend 2002; 67:95-98.
21
ACCEPTED MANUSCRIPT Figure legends Fig. 1. Correlation between electrophoresis and CLEIA. Bone ALP in serum samples from
IP
T
127 patients was subjected to electrophoresis (ALP3) and CLEIA (BAP). The correlation
SC R
coefficient (R) between ALP3 and BAP is 0.886, and the linear equation is y ＝ 8.63 x － 44.9. Black triangles (▲)are outliers.
NU
Fig. 2. The correlation between the 2 methods after stratification. According to ALP3 (%)
MA
values, samples were divided into 5 groups: (A) ALP3 (%) > 80; (B) 60 < ALP3 (%) < 80; (C) 40 < ALP3 (%) < 60; (D) 20 < ALP3 (%) < 40; and (E) ALP3 (%) < 20. ALP3 was then
TE
D
compared with BAP for each group.
CE P
Fig. 3. The critical point for good correlation between the 2 methods. (A) A detailed analysis was made on a combination of the 20 < ALP3 (%) < 40 and ALP3 (%) < 20 groups: ALP3 (%)
AC
< 27, < 26, < 25, < 24, < 23, < 22, < 21 and < 20. Points and lines on the chart represent R and p values comparing each group. (B) Data points and lines represent R and p values comparing “all samples”, ALP3 (%) > 23, and < 23 groups.
Fig. 4. Correlation between electrophoresis and CLEIA after the removal of 5 outliers. (A) Correlation between the 2 methods after the removal of one outlier sample from the 20 < ALP3 (%) < 40 group. (B) Correlation between the 2 methods after 4 outliers were removed from the ALP3 (%) < 20 group.
22
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
IP
T
Table 1. Patient characteristics and serum biochemistries. Subjects All samples: Mean ± SD Male: Mean ± SD Female: Mean ± SD p value Reference ranges Age (years) 61.8 ± 12.2 60.3 ± 14.2 62.8 ± 10.5 0.266 TG (mg/dL) 126.8 ± 76.8 141.2 ± 74.8 116.0 ± 77.0 0.069 0－150 TC (mg/dL) 189.2 ± 42.1 175.7 ± 37.0 199.3 ± 43.1 0.002 150－219 HDL-C (mg/dL) 59.7 ± 31.7 54.7 ± 41.7 63.5 ± 20.6 0.126 M: 40－80, F: 40－100 LDL-C (mg/dL) 106.0 ± 33.5 99.3 ± 30.0 111.0 ± 35.1 0.056 65－139 LAP (U/L) 86.0 ± 51.5 89.8 ± 49.2 83.3 ± 53.3 0.484 30－70 γ-GTP (U/L) 111.4 ± 129.4 161.3 ± 161.1 74.6 ± 83.5 0.001 M: 0－50, F: 0－30 ALP (U/L) 467.9 ± 347.1 572.8 ± 489.2 390.4 ± 142.4 0.01 115－359 # BAP (μg/L) 26.3 ± 16.3 28.8 ± 20.7 24.5 ± 11.8 0.169 10.4－24.6 ALP3 (%) 42.8 ± 20.9 37.8 ± 21.2 46.4 ± 20.1 0.022 25.1－59.9 # Male: 3.7－20.9 μg/L；Premenopausal women: 2.9－14.5 μg/L；Postmenopausal women: 3.8－22.6 μg/L. The results are expressed as mean ± standard deviation (SD). p < 0.05 indicates the statistical significance of differences between male and female. TG, triglyceride; TC, total cholesterol; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; γGTP, γ-glutamyltransferase; LAP, leucine aminopeptidase; ALP, alkaline phosphatase; BAP, bone-type alkaline phosphatase (by CLEIA); ALP3 (%), percentage of bone-type alkaline phosphatase isozyme activity as a proportion of total ALP activity (by electrophoresis).
23
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
IP
T
Table 2. Serum biochemical characteristics of five patient outliers. Sex Age HDL-C LAP γ-GTP ALP BAP ALP1 ALP3 ALP3 Sample ALP iso (%) ALP1，ALP5 (M/F) (years) (mg/dL) (U/L) (U/L) (U/L) (μg/L) (U/L) (U/L) (%) D1 M 66 ALP2>>>ALP3 ALP1 33 241 768 2931 54.9 454 70 2.4 D2 M 25 ALP2>>>ALP3 ALP1 24 170 356 1146 25.2 358 36 3.1 D3 M 55 ALP2>>>ALP3 ALP1, 5 38 246 548 2301 59.7 366 97 4.1 D4 M 79 ALP2>>>ALP3 ALP1 19 74 119 833 26.5 293 40 4.8 D5 F 73 ALP2>>ALP3 ALP1 26 133 121 811 48.7 176 191 23.6 Sex/M, male; Sex/F, female; ALP1, high molecular mass liver-type ALP isoform (by electrophoresis); ALP2, liver-type alkaline phosphatase (by electrophoresis); ALP3, bone-type alkaline phosphatase (by electrophoresis); ALPiso (%), comparison of the percentage of bone- and livertype alkaline phosphatase isozymes activity in total alkaline phosphatase; [ ALP1, ALP5], whether these two types of isozymes were present; HDL-C, high-density lipoprotein cholesterol; LAP, leucine aminopeptidase; γ-GTP, γ-glutamyltransferase; ALP, alkaline phosphatase; BAP, bone-type alkaline phosphatase (by CLEIA); ALP3 (%), bone-type alkaline phosphatase activity as a percentage of total alkaline phosphatase activity (by electrophoresis).
24
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
Fig. 1
25
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
IP
Fig. 2
26
SC R
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
Fig. 3
27
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
Fig. 4
28
ACCEPTED MANUSCRIPT Highlights
IP
T
1. Bone alkaline phosphatase (ALP) is a marker of bone formation and metabolism;
SC R
2. Bone ALP was measured by electrophoresis & chemiluminescent enzyme immunoassay; 3. When ALP3 isozyme predominated, the two methods showed good correlation; 4. From ALP3 (%) < 23, correlation between the two methods deteriorated;
AC
CE P
TE
D
MA
NU
5. In some special cases, the accuracy of the electrophoresis is questionable.
29
S0009-8981(16)30253-4 doi: 10.1016/j.cca.2016.06.008 CCA 14398
To appear in:
Clinica Chimica Acta
Received date: Revised date: Accepted date:
18 March 2016 8 June 2016 8 June 2016
Please cite this article as: Zhan Fangjie, Watanabe Yoshihisa, Shimoda Aya, Hamada Etsuko, Kobayashi Yoshimasa, Maekawa Masato, Evaluation of serum bone alkaline phosphatase activity in patients with liver disease: Comparison between electrophoresis and chemiluminescent enzyme immunoassay, Clinica Chimica Acta (2016), doi: 10.1016/j.cca.2016.06.008
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Evaluation of serum bone alkaline phosphatase activity in patients with liver disease:
IP
T
comparison between electrophoresis and chemiluminescent enzyme immunoassay
SC R
Fangjie Zhana, Yoshihisa Watanabea, Aya Shimodab, Etsuko Hamadab, Yoshimasa Kobayashic, Masato Maekawaa, b,*
Department of Laboratory Medicine, Hamamatsu University School of Medicine.
b
Department of Clinical Laboratory, Hamamatsu University Hospital.
MA
c
NU
a
Second Division, Department of Internal Medicine, Hamamatsu University School of
D
Medicine.
CE P
TE
1-20-1, Handayama, Higashi-ku, Hamamatsu, 431-3192, Japan.
*Corresponding author at: Department of Laboratory Medicine, Hamamatsu University
AC
School of Medicine, Hamamatsu 431-3192, Japan. Tel: +81-53-435-2721; Fax: +81-53-435-2096.
E-mail address: [email protected] (M. Maekawa).
Abbreviations: ALP, alkaline phosphatase; BAP, bone-type alkaline phosphatase; CLEIA, chemiluminescent enzyme immunoassay; TC, total cholesterol; CH, cholesterol; TG, triglyceride; γ-GTP, γ-glutamyltransferase; LAP, leucine aminopeptidas
1
ACCEPTED MANUSCRIPT Abstract Background: Serum bone alkaline phosphatase (ALP) is a marker of bone formation and
IP
T
metabolism. However, existing methods for measuring it have their limitations and their
SC R
accuracy has not been determined.
Methods: We measured serum bone ALP activity in 127 patients with liver disease using 2 methods: electrophoresis and chemiluminescent enzyme immunoassay (CLEIA). The results
NU
of these 2 methods were compared and analyzed according to gender, age and several serum
MA
biochemical markers.
Results: When ALP3 (%; bone-type isozyme activity as a percentage of total ALP activity)
D
values were high, the 2 methods showed good correlation. However, with a decrease in ALP3
TE
(%) levels, the correlation coefficient (R) also decreased. Starting with ALP3 (%) < 23, R
CE P
values markedly decreased to <0.50 (p > 0.05). Five outliers displayed low ALP3 (%) activity levels. Furthermore, in regard to genders, there were significant differences in total
AC
cholesterol (TC), γ-glutamyltransferase (γ-GTP), ALP and ALP3 (%) levels (p < 0.05). Conclusions: When serum ALP3 (%) levels were high in patients with liver disease, the accuracy of electrophoresis was comparable to that of CLEIA. However, the accuracy of electrophoresis needs to be evaluated with further when patient samples under certain conditions. Keywords: Alkaline phosphatase isozyme; Bone-type alkaline phosphatase; Liver disease; Electrophoresis; Chemiluminescent enzyme immunoassay
2
ACCEPTED MANUSCRIPT 1. Introduction Alkaline phosphatase (ALP) is a membrane-bound metalloenzyme that is found in nearly
IP
T
all species [1,2]. Human ALP consists of a group of isozymes, placental, intestinal and three
SC R
major tissue-nonspecific isoforms [3], which are encoded by 4 main gene loci: tissue-nonspecific, intestinal, placental, and germ-cell ALP [4-6]. Tissue-nonspecific bone-
NU
and liver-type ALP are formed through posttranslational modifications and differences in carbohydrate composition, leading to different clinical phenotypes [4]. ALP isozymes have
MA
been used as biomarkers for liver and bone disease in the clinic [7-10]. Hence, the measurement of ALP isozymes activity in serum has a high clinical value in the differential
D
diagnosis and analysis of some diseases.
TE
Bone ALP is an extracellular enzyme that is found anchored to the osteoblast membrane
CE P
and which reflects whole skeletal remodeling [11]; it is one of the most commonly used biochemical markers for osteoblastic bone formation analyzed in the routine clinical
AC
chemistry laboratory [12,13]. Metastatic bone tumors, diabetes, hyperthyroidism, fracture recovery period and other conditions lead to the hyperfunction of bone metabolism or formation, which, in turn, increases serum bone ALP levels. In addition, serum bone ALP varies with age and gender. Humans in infancy and at puberty show 2 physiological blood bone ALP concentration peaks, with troughs in mid-childhood and at the end of adolescence [14,15]. Furthermore, for postmenopausal women with high bone turnover, bone ALP concentrations were found to be significantly higher than for those in premenopausal [16-18]. Presently, the predominant methods for separating and measuring bone ALP are
3
ACCEPTED MANUSCRIPT electrophoresis and chemiluminescent enzyme immunoassay (CLEIA). The separation of proteins by gel electrophoresis occurs according to their differences in net charge, isoelectric
IP
T
point and molecular weights, resulting in different migration patterns. Migrating in order
SC R
from the anode to cathode, ALPs were named ALP1 (a high molecular mass, liver-type ALP), ALP2 (liver-type ALP), ALP3 (bone-type ALP), ALP4 (placental-type ALP), ALP5 (intestinal-type ALP) and ALP6 (immunoglobulin binding-type ALP). However, the
NU
migration of bone ALP by electrophoresis is also influenced by the presence of other types of
MA
ALP isozymes.
The CLEIA technique involves samples binding to mouse monoclonal anti-bone ALP
D
antibodies that, in turn, bind magnetic particles coated with goat anti-mouse polyclonal
TE
antibodies. The chemiluminescent substrate (lumigen PPD) emits light according to the
CE P
activity of bone ALP, with the amount of light emitted reflecting the bone ALP concentration of the sample [19]. CLEIA is widely used in the clinic due to its high sensitivity and
AC
specificity, and is simple and rapid [20]. However, the CLEIA method also has disadvantages such as cross-immunological reactivity between bone- and liver-type ALPs, as well as interference by liver-type ALP. CLEIA also does not yield information on other isozymes and isoforms of ALP present within a sample. In fact, the 2 methods have their own distinct limitations when detecting bone ALP.
Materials and methods 2.1. Subjects
4
ACCEPTED MANUSCRIPT The subjects of this study were 127 patients (25 – 83 y; males 54, females 73) with liver disease who were treated at Hamamatsu University Hospital.
Of these, 52 patients had
IP
T
primary biliary cirrhosis, 8 were affected by cholestatic cirrhosis, 13 had cholestatic hepatic
SC R
disorder, and the rest were diagnosed with other liver diseases. Both agarose gel electrophoresis and CLEIA were used to detect bone ALP in serum samples. Other patient biochemical analyses were also undertaken, such as tests for total cholesterol (TC),
NU
triglyceride (TG), γ-glutamyltransferase (γ-GTP) and leucine aminopeptidase (LAP), as well
MA
as others. The ethics committee of Hamamatsu University School of Medicine approved this
TE
D
study.
CE P
1.2.Electrophoretic separation of ALP isozymes For the analysis of ALP isozymes, electrophoresis using agarose gels was undertaken according to the manufacturer’s operating instructions. Quick Gel ALP and Quick ALP
AC
reagent (Quick ALP; Helena Labs) have been commercially developed for the electrophoretic analysis of ALP [21]. In order to separate liver- and bone-type ALP completely, samples were pretreated with neuraminidase (ALP separator; Helena) [6]. In addition, samples were pretreated with protease (Quick ALP; Helena) to separate bone- and intestinal-type ALP. Following electrophoresis, a 3-indoxyl phosphate disodium salt was used as a substrate and nitrotetrazolium blue was used as a chromogenic agent. Bands were scanned by densitometry at a wavelength of 570 nm [21]. ALP isozyme activity was expressed in U/l and as a percentage of total ALP activity. Samples from 3 healthy individuals were measured using 5
ACCEPTED MANUSCRIPT three lots of reagents, indicating the number of assays for reproducibility analysis totaled nine times. The coefficients of variation (CVs) for ALP3 activities were 1.22–3.00%. However,
IP
T
these were not the CVs for ALP2 and ALP3 percentages, which are quite different samples.
SC R
Serum samples and protease were mixed using a ratio of 5:1 at room temperature for 30 min; both serum and protease-treated samples, respectively, were mixed with neuraminidase at a ratio of 7:1, and left at room temperature for 12 min. Control, serum, separator-treated,
NU
protease-treated and separator + protease-treated samples were loaded in order on the gel
MA
plate. Electrophoresis was performed at 250 V for 20 min. The gel was coated in sample reagent and stained for one min before incubation at 37ºC for 35 min. The gel was soaked in
D
70% methanol for 10 min and washed with water twice for 10 min each wash. Bands were
1.3. CLEIA
CE P
TE
then scanned by densitometry at a wavelength of 570 nm.
AC
A 1-step immunoenzymatic assay was used in this study. An automated chemiluminescence enzyme immunoassay device, known as an Access Immunoanalyzer (Beckman Coulter), and a special bone-type ALP (BAP) kit (Access Ostase; Beckman Coulter) were used [19,22]. Beckman BAP Access Ostase operating instructions were followed. Samples and reagents were mixed, and all operations were automatically performed as long as the appropriate operating items were chosen. Each 25 μl sample was mixed with 50 μl of a solid phase reagent (magnetic particles coated with goat anti-mouse polyclonal antibodies) and 15 μl of an antibody reagent (mouse monoclonal anti-BAP antibodies), and 6
ACCEPTED MANUSCRIPT then washed after reacting at 37℃ for 14.4 min. Next, a 200 μl substrate solution (lumigen PPD) was added and samples incubated at 37℃ for 5 min. After each reaction was complete,
IP
T
its luminous intensity (λ max = 540 nm) was assayed by immunoanalyzer and the
SC R
concentration of BAP in the sample was automatically calculated using a standard curve. Assays were repeated 3 times in reproducibility analysis, for which intraday and interday CVs for BAP were 3.82–5.79% and 2.60–5.25% for a 10.39 μg/l concentration, 3.77–5.66%
D
2.4. Serum biochemical examinations
MA
NU
and 3.33–5.80% for 28.07 μg/l.
TE
Basic biochemical examinations of serum samples, using a chemistry analyzer (Labospect 008; Hitachi Ltd), were undertaken. Cholesterol (CH), TG and γ-GTP assays were performed
CE P
according to the manufacturer’s operating instructions. To determine CH and TG fractions in serum, serum lipoproteins were separated by electrophoresis using agarose gels. After
AC
electrophoresis, reagents (Choletricombo CH/TG; Helena) within the agarose reacted with CH and TG in lipoprotein fractions, and the formazan consequently produced was scanned by densitometry at a wavelength of 570 nm. For the measurement of serum γ-GTP isozymes, a cellulose acetate membrane was used as a support and soaked in Tris-barbital buffer (pH 8.8) before serum γ-GTP isozymes were separated by electrophoresis. After electrophoresis, reagents (Titan γ-GTP; Helena) within the acetate membrane generated azodyes that were scanned by densitometry at a wavelength of 570 nm.
7
ACCEPTED MANUSCRIPT 2.5. Statistical analysis All statistical analyses were performed using SPSS software (v18.0, SPSS Inc.,). Data are
IP
T
expressed as mean ± SD. Correlations between ALP3 (bone ALP determined by
SC R
electrophoresis) and BAP (bone ALP measured by CLEIA) were determined using Pearson’s correlation coefficient. Differences in means between methods were evaluated by Student’s
NU
t-test. A p <0.05 was considered to indicate statistical significance.
MA
3. Results
D
3.1. Determination of serum bone-type ALP by 2 methods
TE
We used 2 methods, electrophoresis and CLEIA (Suppl. Fig. 1), to detect bone ALP in
CE P
the serum samples of 127 patients with liver disease (Suppl. Table 1). A comparison of the 2 methods was made, with CLEIA used as a control method and electrophoresis evaluated in order to assess its accuracy in the determination of bone ALP. Fig. 1 shows how bone ALP
AC
samples analyzed by electrophoresis (ALP3) compared with those analyzed by CLEIA (BAP). The linear equation was calculated as y ＝ 8.63 x - 44.9, with a correlation coefficient (R) of 0.886; The 2 methods showed a statistically significant correlation (p < 0.001). Furthermore, results for some samples deviated from those of most samples (black triangles), with all outliers below the line.
3.2. Further stratified analysis of all samples We grouped all samples according to their percentage of ALP3 in 20% increments: > 80% 8
ACCEPTED MANUSCRIPT (n = 8), 60–80% (n = 18), 40–60% (n = 42), 20–40% (n = 41) and < 20% (n = 18), in a total of five groups as shown in Fig. 2. The corresponding R values were 0.988, 0.969, 0.947,
IP
T
0.892 and 0.350. Although 4 groups (Figs. 2A–D) showed a good correlation between the 2
SC R
detection methods, concomitant with a decrease in ALP3 (%) the R values for the 2 methods progressively decreased. For the < 20% group, R = 0.350 was significantly lower than the average of the other 4 groups (p < 0.001; Fig. 2E). The slopes of linear equations for all 5
NU
groups also progressively became reduced (Figs. 2A to E).
MA
Furthermore, we performed a more refined stratification between the 2 groups, 20 – 40% and < 20% (Fig. 3A), which were subsequently divided into 8 groups: ALP3 (%) < 27, < 26,
D
< 25, < 24, < 23, < 22, < 21 and < 20. We found that R values markedly dropped to less than
TE
0.50 from ALP3 (%) < 23 onwards, corresponding to a bad correlation between the 2
CE P
methods (p > 0.05). We compared the “all samples” group, ALP3 (%) > 23, and < 23 groups (Fig. 3B). For the “all samples” group, R = 0.886 (p < 0.001). For ALP3 (%) > 23, R = 0.951
AC
(p < 0.001). However, the ALP3 (%) < 23 group showed a poor correlation, with R = 0.405 and p > 0.05. Thus the ALP3（%）< 23 group exhibited the worse correlation between the 2 kinds of detection methods compared with the other 2 groups. When we removed these 5 deviating samples from the rest of the samples, we found that the corresponding R values of the 2 groups, 20 < ALP3 (%) < 40 and ALP3 (%) < 20, were 0.920 and 0.772, respectively, which were significantly higher than those of samples containing outliers (Figs. 2D–E and 4). However, the R value of the < 20% group, after the removal of outliers, was also significantly lower than the average of the other 4 groups (p <
9
ACCEPTED MANUSCRIPT 0.001).
IP
T
3.3. Clinical profiles of patients with liver disease
SC R
We assayed other patient biochemical data using detection methods according to Japan Society for Clinical Chemistry (JSCC) standards. The baseline characteristics of the 127 liver disease patients are shown in Table 1. Mean levels of serum BAP, ALP, γ-GTP and LAP for
NU
all samples were all above the respective normal upper limit. Due to gender differences, TG,
MA
LAP, γ-GTP, ALP and BAP levels for males were higher than for females, although only γ-GTP (p < 0.001) and ALP (p < 0.010) were significant. However, for TC, HDL-C, LDL-C
D
and ALP3 (%), levels for males were lower than those for females, although only TC (p <
CE P
TE
0.002) and ALP3 (%；p < 0.022) were significant.
3.4. Sample deviations
AC
Five outliers (Fig. 1; black triangles) were noted with common characteristics. Table 2 shows that total serum ALP activity was clearly raised at > 800 U/l for each; ALP2 was the predominant ALP isozyme, and all outliers exhibited high levels of ALP1 (> 175 U/l). LAP and γ-GTP were significantly increased but HDL was reduced. Since all outliers were below the line, bone ALP values determined by electrophoresis were lower than those determined by CLEIA. The outliers showed low ALP3 (%) levels.
10
ACCEPTED MANUSCRIPT 4. Discussion The main challenge of bone ALP quantification is to differentiate bone- and liver-type
IP
T
ALPs since these are encoded by the same gene and have an identical amino acid sequence
SC R
[23]. Electrophoresis is, at present, considered unable to completely distinguish between bone- and liver-type ALP in some liver diseases. Therefore, the accuracy of detection of
NU
serum bone ALP by electrophoresis in patients with liver disease is worthy of further, more
MA
detailed study.
4.1. Correlation of the 2 methods
D
In this study, we used CLEIA as a control method with which to compare the
TE
determination of bone ALP by electrophoresis, and to find the limitations of electrophoresis
CE P
when measuring bone ALP. The overall aim is to avoid such limitations in future clinical examinations, and to select the best detection methods under different situations, in order to
AC
provide accurate and reliable data for clinical medicine. Using all samples, the 2 methods showed a good correlation with R = 0.886 (p < 0.001), indicating the reliability of the determination of bone ALP by electrophoresis in the majority of cases. However, the regression of all samples was a curve, the slope of the curve line increased with the increase of bone ALP activity. The reason for this may be related to the CLEIA method has cross-immunological reactivity between bone- and liver-type ALPs. When liver-type ALP predominated, the BAP interference by liver-type ALP was serious, which led to fact that the results measured by CLEIA was higher than those of electrophoresis. And when ALP3
11
ACCEPTED MANUSCRIPT extremely high and low, it is difficult to quantify by visually determining the inflection point between fractions on densitometry when ALP isozymes are analyzed by electrophoresis, the
4.2. Exploration of the limitations of electrophoresis
SC R
IP
T
measurement errors exist.
All serum samples were then divided into 5 groups according to the percentage of ALP3
NU
activity. Although the 2 methods correlated well for the first 4 groups showing the highest
MA
percentage of ALP3 activity (p < 0.001), R values became progressively smaller from the first to the fourth group in descending order of ALP3（%）. The last group, with ALP3 (%) <
D
20, showed R = 0.350 (p > 0.05), meaning that correlation between the 2 methods was poor.
TE
In other words, the lower the ALP3 (%), the worse the correlation was. Because the R value
CE P
of the ALP3 (%) < 20 group was significantly lower than those of the other 4 groups, we performed a more refined stratification between the 20 – 40% and < 20% groups in order to
AC
identify the limit for correlation between the 2 methods. Starting with an ALP3 (%) < 23 downwards, correlation between the 2 methods worsened, with all displaying R < 0.50 and p > 0.05. The reason for this may be related to the proportion of samples containing deviation samples becoming higher. The accuracy of the determination of bone ALP by electrophoresis in some special cases requires further, more detailed analysis. After removing the 5 outliers from the groups, 20 < ALP3 (%) < 40 and ALP3 (%) < 20, we found that the 2 groups’ corresponding R values were 0.920 and 0.772, respectively, which were significantly higher than that of samples containing outliers. However, the R value of the ALP3 (%) < 20 group
12
ACCEPTED MANUSCRIPT after removal of the outliers was also significantly lower than the average of the other 4 groups, and with a decrease in ALP3 (%) levels, the correlation coefficient also decreased. It
IP
T
can be concluded that from the first to the fifth group, the correlation worsened and the trend
SC R
did not change after the 5 outliers were removed.
Fig. 2 shows that the slope of the linear equation decreased with each decrease in ALP3 (%). A shallow slope signified that results measured by electrophoresis were lower than those
NU
of CLEIA. In this regard, the existence of 5 outliers is particularly concerning. It may be that
MA
the lower the ALP3 (%) level, the more difficult it is to quantify by visually determining the inflection point between fractions on densitometry when ALP isozymes are analyzed by
D
electrophoresis. Therefore, when liver-type ALP predominated (ALP2 % > ALP3 %), we
TE
considered the error in the measurement of bone-type ALP activity by electrophoresis to be
CE P
large. In contrast, when bone-type ALP predominated (ALP2 % ≤ ALP3 %), the correlation between these 2 methods was good, and so data from electrophoresis were considered to be
AC
relatively reliable (see Suppl. Fig. 2).
4.3. Biochemical data showing significant gender differences As mentioned above, the measurement of bone ALP was interfered with by liver ALP. The various liver diseases, which can lead to different levels of total ALP (tALP), and changes in bone ALP levels also differed. So we combined this with liver disease-related biochemical data to analyze the ratio of liver- and bone-type ALPs and the degree of interference by liver ALP in the serum of patients with different diseases in order to
13
ACCEPTED MANUSCRIPT determine the cause of abnormal increases or decreases of bone ALP. These biochemical data play an important role in determining the reliability of the results of electrophoresis.
IP
T
In our study, tALP activity in males was significantly higher than that for females. For
SC R
males, tALP was higher than that of women of the corresponding age, up until 60 years of age, by which time levels became almost equal [24]. In addition, since all samples used in this study were from patients with liver disease, the observed gender difference may be
NU
related to differences in liver diseases and the severity of disease. From all samples, the
MA
ALP3 (%) level for women were significantly higher than that for men, which may be related to a postmenopausal increase in ALP levels because of high bone turnover, with elevated
D
serum ALP derived from bone tissue [25,26]. TC in women was significantly higher than in
TE
men in our study. Previous research highlighted a difference between males and females in
CE P
the causes of liver disease, with TC synthesis representing a key point of gender difference [27]. In addition, γ-GTP values of all samples were increased, with those in men significantly
AC
higher than those in women. In fact, γ-GTP levels showed large individual differences, which significantly differed according to gender and drinking history [28], while it has been suggested that lower γ-GTP levels in females is probably physiological [29]. Although serum bone ALP changes with age and sex, in our study these had no serious impact on the overall correlation between the 2 methods (Suppl. Figs. 3, 4).
4.4. Common characteristics of deviating samples We found 5 outliers with common characteristics. As a result, we inferred that outlier 14
ACCEPTED MANUSCRIPT samples came from patients who may have been affected by an intrahepatic space-occupying lesion or cholangitis. The bone ALP values of such samples as determined by electrophoresis
IP
T
were lower than those determined by CLEIA, and therefore deviated from the overall
SC R
linearity displayed by most samples. When we removed these 5 deviating samples from the rest of the samples, we found that for the 2 groups, 20 < ALP3 (%) < 40 and ALP3 (%) < 20, the correlation between the 2 methods improved. This is indicates that the main reason for the
NU
deterioration in the correlation for the 20 < ALP3 (%) < 40 and < 20% groups were these 5
MA
outlier samples. From < 23%, the 2 method correlation deteriorated and the reason why may be related to the increasing proportion of deviating samples. Therefore, the 5 outliers showed
D
a certain representation that was not accidental; with the determination of such samples by
TE
electrophoresis, the accuracy was questionable.
CE P
The outliers were very low in ALP3 (%), thus indicating the predominance of liver-type ALP. The questionable accuracy of densitometry after electrophoresis may be a major factor
AC
for such variation. Bone and liver ALP were quantified by visually determining the inflection point between the 2 ALP fractions on densitometry, so when ALP3 (%) is low, quantification is more prone to errors in judgement. From this, we have doubts concerning the accuracy of the determination of serum bone ALP by electrophoresis in patients with an intrahepatic space-occupying lesion or cholangitis, when the ALP3 (%) level is very low.
4.5. Limitations There are some limitations in this study. First, the samples were obtained from a general
15
ACCEPTED MANUSCRIPT population of different liver disease patients. It could lead to different levels of the activity of ALP and bone-type ALP, as well as the larger variation of the results of other biochemical
IP
T
tests; Although we grouped all samples according to their percentage of ALP3 in each 20%
SC R
increments, the number of samples of each group was not the same: very high and low ALP3 (%) groups were less than the middle level groups. Therefore, the R values statistical reliability was worth thinking about, especially the group of too small sample size. Second,
NU
when electrophoresis was used to analyze the ALP isozymes required by visually determining
MA
the inflection point between fractions on densitometry, this may cause the measurement
D
errors.
TE
5. Conclusion
CE P
Our study suggests that for whole and bone-type ALP predominating (ALP2 % ≤ oALP3 %) samples, electrophoresis and CLEIA methods show good correlation. However, with a decrease in ALP3 (%) levels, the correlation coefficient also decreased. The 5 outlier
AC
samples lead to start with ALP3 (%) < 23, the R value decreased to less than 0.50 and no statistical correlation between the 2 methods was observed (p > 0.05). Therefore, the accuracy of electrophoresis needs to be evaluated further when the patient is affected by an intrahepatic space-occupying lesion or cholangitis, and shows a low ALP3 (%) level. As aforementioned, although electrophoresis exhibits measurement limitations, it provides useful information on several different isozymes. Validation requirements have increased with further developments and progress in clinical laboratory technology. A major necessity is to uncover the limitations of currently used detection methods, to avoid such 16
ACCEPTED MANUSCRIPT errors in clinical applications and to maximally exploit the advantages of the methods used. Finally, in a similar manner, we can uncover limitations in electrophoresis measurements with
IP
T
regard to other proteases, and can review and evaluate further detection methods based on the
SC R
same principles (Suppl. Fig. 5).
Disclosure
MA
NU
The authors wish to declare that they have no conflicts of interest.
Acknowledgment
D
The authors wish to thank Mr Shinichi Miyashita of the Department of Helena
References
CE P
TE
Laboratories for technical guidance in Choletricombo CH/TG and γ-GTP experiments.
AC
[1] Sharma U, Pal D, Prasad R. Alkaline phosphatase: an overview. Indian J Clin Biochem 2014; 29: 269-278.
[2] Khan KN, Tsutsumi T, Nakata K, Nakao K, Kato Y, Nagataki S. Regulation of alkaline phosphatase gene expression in human hepatoma cells by bile acids. J Gastroenterol Hepatol 1998; 13:643-650. [3] Khan KN, Tsutsumi T, Nakata K, Kato Y. Sodium butyrate induces alkaline phosphatase gene expression in human hepatoma cells. J Gastroenterol Hepatol 1999; 14:156-162. [4] Van Hoof VO, De Broe ME. Interpretation and clinical significance of alkaline
17
ACCEPTED MANUSCRIPT phosphatase isoenzyme patterns. Crit Rev Clin Lab Sci 1994; 31:197-293. [5] Kiffer-Moreira T, Sheen CR, Gasque KC, et al. Catalytic signature of a heat-stable,
IP
T
chimeric human alkaline phosphatase with therapeutic potential. PloS one 2014; 9:e89374.
SC R
[6] Nollet E, Van Craenenbroeck EM, Martinet W, et al. Bone matrix vesicle-bound alkaline phosphatase for the assessment of peripheral blood admixture to human bone marrow aspirates. Clin Chim Acta 2015; 446:253-260.
NU
[7] Tsumura M, Ueno Y, Kinouchi T, Koyama I, Komoda T. Atypical alkaline phosphatase
MA
isozymes in serum and urine of patients with renal failure. Clin Chim Acta 2001; 312:169-178.
D
[8] Kang KY, Hong YS, Park SH, Ju JH. Increased serum alkaline phosphatase levels
TE
correlate with high disease activity and low bone mineral density in patients with axial
CE P
spondyloarthritis. Semin Arthritis Rheum 2015: 202-207. [9] Woitge HW, Seibel MJ, Ziegler R. Comparison of total and bone-specific alkaline
AC
phosphatase in patients with nonskeletal disorder or metabolic bone diseases. Clin Chem 1996; 42:1796-1804.
[10] Nayeem F, Anderson KE, Nagamani M, Grady JJ, Lu LJ. Alkaline phosphatase and percentage body fat predict circulating C-reactive protein in premenopausal women. Biomarkers 2010; 15:663-670. [11] Lau WL, Kalantar-Zadeh K, Kovesdy CP, Mehrotra R. Alkaline phosphatase: Better than PTH as a marker of cardiovascular and bone disease? Hemodial Int 2014; 18:720-724. [12] Dziedziejko V, Safranow K, Slowik-Zylka D, et al. Characterisation of rat and human
18
ACCEPTED MANUSCRIPT tissue alkaline phosphatase isoforms by high-performance liquid chromatography and agarose gel electrophoresis. Biochimie 2009; 91:445-452.
IP
T
[13] Pathomwichaiwat T, Ochareon P, Soonthornchareonnon N, Ali Z, Khan IA,
SC R
Prathanturarug S. Alkaline phosphatase activity-guided isolation of active compounds and new dammarane-type triterpenes from Cissus quadrangularis hexane extract. J Ethnopharmacol 2015; 160:52-60.
NU
[14] Shimose S, Kubo T, Fujimori J, Furuta T, Ochi M. A novel assessment method of serum
MA
alkaline phosphatase for the diagnosis of osteosarcoma in children and adolescents. J Orthop Sci 2014; 19:997-1003.
D
[15] Rauchenzauner M, Schmid A, Heinz-Erian P, et al. Sex- and age-specific reference
TE
curves for serum markers of bone turnover in healthy children from 2 months to 18 years. J
CE P
Clin Endocrinol Metab 2007; 92:443-449. [16] Atalay S, Elci A, Kayadibi H, Onder CB, Aka N. Diagnostic utility of osteocalcin,
AC
undercarboxylated osteocalcin, and alkaline phosphatase for osteoporosis in premenopausal and postmenopausal women. Ann Lab Med 2012; 32:23-30. [17] Shan PF, Wu XP, Zhang H, et al. Age-related changes of serum bone alkaline phosphatase and cross-linked C-telopeptides of type I collagen and the relationship with bone mineral density in Chinese women. Clin Chim Acta 2006; 366:233-238. [18] Takahashi M, Kushida K, Hoshino H, Miura M, Ohishi T, Inoue T. Comparison of bone and total alkaline phosphatase activity on bone turnover during menopause and in patients with established osteoporosis. Clin Endocrinol (Oxf) 1997; 47:177-183.
19
ACCEPTED MANUSCRIPT [19] Cavalier E, Rozet E, Carlisi A, et al. Analytical validation of serum bone alkaline phosphatase (BAP OSTASE) on Liaison. Clin Chem Lab Med 2010; 48:67-72.
IP
T
[20] Sekiguchi S, Kohno H, Yasukawa K, Inouye K. Chemiluminescent enzyme immunoassay
SC R
for measuring leptin. Biosci Biotechnol Biochem 2011; 75:752-756.
[21] Ooi K, Shiraki K, Morishita Y, Nobori T. High-molecular intestinal alkaline phosphatase in chronic liver diseases. J Clin Lab Anal 2007; 21:133-139.
NU
[22] Schafer AL, Vittinghoff E, Ramachandran R, Mahmoudi N, Bauer DC. Laboratory
MA
reproducibility of biochemical markers of bone turnover in clinical practice. Osteoporos Int 2010; 21:439-445.
D
[23] Broyles DL, Nielsen RG, Bussett EM, et al. Analytical and clinical performance
TE
characteristics of Tandem-MP Ostase, a new immunoassay for serum bone alkaline
CE P
phosphatase. Clin Chem 1998; 44:2139-2147. [24] Shimizu Y, Ichihara K, Asia-Pacific Federation of Clinical B. Sources of variation
AC
analysis and derivation of reference intervals for ALP, LDH, and amylase isozymes using sera from the Asian multicenter study on reference values. Clin Chim Acta 2015; 446:64-72. [25] Mukaiyama K, Kamimura M, Uchiyama S, Ikegami S, Nakamura Y, Kato H. Elevation of serum alkaline phosphatase (ALP) level in postmenopausal women is caused by high bone turnover. Aging Clin Exp Res 2015; 27:413-418. [26] Bhattarai T, Bhattacharya K, Chaudhuri P, Sengupta P. Correlation of common biochemical markers for bone turnover, serum calcium, and alkaline phosphatase in post-menopausal women. Malays J Med Sci 2014; 21:58-61.
20
ACCEPTED MANUSCRIPT [27] Lewinska M, Juvan P, Perse M, et al. Hidden disease susceptibility and sexual dimorphism in the heterozygous knockout of Cyp51 from cholesterol synthesis. PloS one
IP
T
2014; 9:e112787.
SC R
[28] Pintus F, Mascia P. Distribution and population determinants of
gamma-glutamyltransferase in a random sample of Sardinian inhabitants. Eur J Epidemiol 1996; 12:71-76.
NU
[29] Skurtveit S, Tverdal A. Sex differences in gamma-glutamyltransferase in people aged
AC
CE P
TE
D
MA
40-42 years in two Norwegian counties. Drug Alcohol Depend 2002; 67:95-98.
21
ACCEPTED MANUSCRIPT Figure legends Fig. 1. Correlation between electrophoresis and CLEIA. Bone ALP in serum samples from
IP
T
127 patients was subjected to electrophoresis (ALP3) and CLEIA (BAP). The correlation
SC R
coefficient (R) between ALP3 and BAP is 0.886, and the linear equation is y ＝ 8.63 x － 44.9. Black triangles (▲)are outliers.
NU
Fig. 2. The correlation between the 2 methods after stratification. According to ALP3 (%)
MA
values, samples were divided into 5 groups: (A) ALP3 (%) > 80; (B) 60 < ALP3 (%) < 80; (C) 40 < ALP3 (%) < 60; (D) 20 < ALP3 (%) < 40; and (E) ALP3 (%) < 20. ALP3 was then
TE
D
compared with BAP for each group.
CE P
Fig. 3. The critical point for good correlation between the 2 methods. (A) A detailed analysis was made on a combination of the 20 < ALP3 (%) < 40 and ALP3 (%) < 20 groups: ALP3 (%)
AC
< 27, < 26, < 25, < 24, < 23, < 22, < 21 and < 20. Points and lines on the chart represent R and p values comparing each group. (B) Data points and lines represent R and p values comparing “all samples”, ALP3 (%) > 23, and < 23 groups.
Fig. 4. Correlation between electrophoresis and CLEIA after the removal of 5 outliers. (A) Correlation between the 2 methods after the removal of one outlier sample from the 20 < ALP3 (%) < 40 group. (B) Correlation between the 2 methods after 4 outliers were removed from the ALP3 (%) < 20 group.
22
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
IP
T
Table 1. Patient characteristics and serum biochemistries. Subjects All samples: Mean ± SD Male: Mean ± SD Female: Mean ± SD p value Reference ranges Age (years) 61.8 ± 12.2 60.3 ± 14.2 62.8 ± 10.5 0.266 TG (mg/dL) 126.8 ± 76.8 141.2 ± 74.8 116.0 ± 77.0 0.069 0－150 TC (mg/dL) 189.2 ± 42.1 175.7 ± 37.0 199.3 ± 43.1 0.002 150－219 HDL-C (mg/dL) 59.7 ± 31.7 54.7 ± 41.7 63.5 ± 20.6 0.126 M: 40－80, F: 40－100 LDL-C (mg/dL) 106.0 ± 33.5 99.3 ± 30.0 111.0 ± 35.1 0.056 65－139 LAP (U/L) 86.0 ± 51.5 89.8 ± 49.2 83.3 ± 53.3 0.484 30－70 γ-GTP (U/L) 111.4 ± 129.4 161.3 ± 161.1 74.6 ± 83.5 0.001 M: 0－50, F: 0－30 ALP (U/L) 467.9 ± 347.1 572.8 ± 489.2 390.4 ± 142.4 0.01 115－359 # BAP (μg/L) 26.3 ± 16.3 28.8 ± 20.7 24.5 ± 11.8 0.169 10.4－24.6 ALP3 (%) 42.8 ± 20.9 37.8 ± 21.2 46.4 ± 20.1 0.022 25.1－59.9 # Male: 3.7－20.9 μg/L；Premenopausal women: 2.9－14.5 μg/L；Postmenopausal women: 3.8－22.6 μg/L. The results are expressed as mean ± standard deviation (SD). p < 0.05 indicates the statistical significance of differences between male and female. TG, triglyceride; TC, total cholesterol; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; γGTP, γ-glutamyltransferase; LAP, leucine aminopeptidase; ALP, alkaline phosphatase; BAP, bone-type alkaline phosphatase (by CLEIA); ALP3 (%), percentage of bone-type alkaline phosphatase isozyme activity as a proportion of total ALP activity (by electrophoresis).
23
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
IP
T
Table 2. Serum biochemical characteristics of five patient outliers. Sex Age HDL-C LAP γ-GTP ALP BAP ALP1 ALP3 ALP3 Sample ALP iso (%) ALP1，ALP5 (M/F) (years) (mg/dL) (U/L) (U/L) (U/L) (μg/L) (U/L) (U/L) (%) D1 M 66 ALP2>>>ALP3 ALP1 33 241 768 2931 54.9 454 70 2.4 D2 M 25 ALP2>>>ALP3 ALP1 24 170 356 1146 25.2 358 36 3.1 D3 M 55 ALP2>>>ALP3 ALP1, 5 38 246 548 2301 59.7 366 97 4.1 D4 M 79 ALP2>>>ALP3 ALP1 19 74 119 833 26.5 293 40 4.8 D5 F 73 ALP2>>ALP3 ALP1 26 133 121 811 48.7 176 191 23.6 Sex/M, male; Sex/F, female; ALP1, high molecular mass liver-type ALP isoform (by electrophoresis); ALP2, liver-type alkaline phosphatase (by electrophoresis); ALP3, bone-type alkaline phosphatase (by electrophoresis); ALPiso (%), comparison of the percentage of bone- and livertype alkaline phosphatase isozymes activity in total alkaline phosphatase; [ ALP1, ALP5], whether these two types of isozymes were present; HDL-C, high-density lipoprotein cholesterol; LAP, leucine aminopeptidase; γ-GTP, γ-glutamyltransferase; ALP, alkaline phosphatase; BAP, bone-type alkaline phosphatase (by CLEIA); ALP3 (%), bone-type alkaline phosphatase activity as a percentage of total alkaline phosphatase activity (by electrophoresis).
24
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
Fig. 1
25
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
IP
Fig. 2
26
SC R
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
Fig. 3
27
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
Fig. 4
28
ACCEPTED MANUSCRIPT Highlights
IP
T
1. Bone alkaline phosphatase (ALP) is a marker of bone formation and metabolism;
SC R
2. Bone ALP was measured by electrophoresis & chemiluminescent enzyme immunoassay; 3. When ALP3 isozyme predominated, the two methods showed good correlation; 4. From ALP3 (%) < 23, correlation between the two methods deteriorated;
AC
CE P
TE
D
MA
NU
5. In some special cases, the accuracy of the electrophoresis is questionable.
29