Sensitive enzymatic assay for erythrocyte creatine with production of methylene blue

¹ Department of Laboratory Medicine and
² First Department of Internal Medicine, Kochi Medical School, Oko-cho, Nankoku 783-8505, Japan.

³ Diagnostic Research and Production Department, Kyowa Medix Company, Ltd., Shimotogari, Shizuoka 411-0943, Japan.

⁴ Department of Medical Informatics, Faculty of Medicine, Osaka City University, Osaka 545-8586, Japan.
^a Author for correspondence. Fax 81-888-80-2462; e-mail okumiyat/KMS{at}kochi-ms.ac.jp.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
We developed a new, highly sensitive enzymatic method for quantifying creatine in erythrocytes, which comprises creatine amidinohydrolase, sarcosine oxidase, and peroxidase. In the present method, an N-methylcarbamoyl derivative of methylene blue, 10-N-methylcarbamoyl-3,7-bis(dimethylamino)phenothiazine (MCDP), was used as a sensitive chromogenic compound. Potassium ferrocyanide was used to prevent nonspecific oxidation of MCDP. The enzymatic method exhibited good analytical performance: precision, within-run CVs <1.0% and between-day CVs <2.0%; average analytical recovery, 99.3% ± 1.8%; detection limit, 1.0 µmol/L in hemolysate; and linearity, at least up to 500 µmol/L as creatine concentration in hemolysate. Excellent agreement was observed between the present method (y) and HPLC (x), y = 1.029x - 0.002 µmol/g hemoglobin, r = 0.9998, S_yx = 0.053 µmol/g hemoglobin (n = 110). No significant interference was produced by various compounds, including guanidino compounds, amino acids, and reducing materials. The reference intervals (mean ± 2 SD) for erythrocyte creatine obtained from 60 males and 60 females were (in µmol/g hemoglobin) 1.18 ± 0.52 (0.66–1.70) for males and 1.35 ± 0.49 (0.86–1.84) for females. Using this method, we documented changes in erythrocyte creatine in patients with various hemolytic conditions, including hemolytic anemia, liver cirrhosis, renal insufficiency, and chronic renal failure treated with hemodialysis with or without the administration of erythropoietin. We conclude that the use of MCDP allows sensitive measurement of erythrocyte creatine and that MCDP with potassium ferrocyanide can improve the sensitivity of assays that use peroxidase for detection of H₂O₂.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
A decrease in the erythrocyte survival time is regarded as an essential diagnostic feature of hemolytic disorders. For determination of the erythrocyte survival time, in vivo labeling of erythrocytes with Cr has been performed as a standard method (1). However, the Cr-labeling method is not suitable for routine tests in clinical laboratories because it requires exclusive equipment for radioactive materials and a prolonged examination period for a series of blood drawings from patient. The reticulocyte count has also been used as an indirect marker for hemolytic disorders. The number of reticulocytes is known to reflect the rate of erythrocyte production (erythropoiesis) (2). In practice, however, the number of reticulocytes is influenced by other factors, including the length of the reticulocyte growth time in peripheral blood and the severity of anemia (3). Thus, it is now known that the reticulocyte count is not a good indicator for the mean age of the erythrocyte population (3). In addition, the reticulocyte survival time in peripheral blood seems to be too short to detect a slight decrease in the mean age of the erythrocyte population (4)(5). Other laboratory tests, including ones for hemoglobin, indirect bilirubin, and haptoglobin in serum, are widely used for detection of the presence of hemolytic processes, but these tests are not specific to hemolytic episodes.

In 1967, Griffiths and Fitzpatrick (6) suggested that erythrocyte creatine was a sensitive indicator of the mean age of the erythrocyte population, because young erythrocytes contain at least six- to ninefold higher creatine than old erythrocytes. Opalinski and Beutler (7) also showed that erythrocyte creatine was chiefly related to the mean age of erythrocytes rather than to the degree of anemia. Fehr and Knob (4) demonstrated that the erythrocyte creatine content was closely correlated with the erythrocyte survival time, as measured with the Cr-labeling technique, in several hemolytic patients. Thereafter, various applications of erythrocyte creatine were reported: e.g., for efficient detection of erythrocyte enzyme deficiencies in reticulocytosis (8), for estimation of neonatal erythrocyte age using cord blood (9), for monitoring of erythropoiesis in patients after renal transplantation (10), and for detection of slight hemolysis in long-distance runners (11).

A colorimetric method based on the diacetyl--naphthol reaction has been widely used for measurement of erythrocyte creatine (4)(5)(6)(7)(8)(9)(10)(11)(12). The diacetyl--naphthol method, however, exhibits positive interference from various guanidino compounds and amino acids in erythrocytes (e.g., arginine, creatinine, guanidine, and guanidinoacetic acid) (13)(14), and is not appropriate for automated analyzers commonly used in clinical laboratories because of instability of the reagents (12)(15). Recently, some enzymatic methods for erythrocyte creatine involving two major principles have been reported: the creatine kinase (CK; EC 2.7.3.2)¹ /pyruvate kinase (PK; EC 2.7.1.40)/lactate dehydrogenase (LDH; EC 1.1.1.27) system (15)(16), and the creatine amidinohydrolase (CTase; EC 3.5.3.3)/sarcosine oxidase (SOX; EC 1.5.3.1)/peroxidase (POD; EC 1.11.1.7) system (14)(17). These enzymatic methods are specific to creatine in erythrocytes. The former method, however, has low sensitivity because of the modest molar absorptivity of NADH. Because an erythrocyte sample for creatine measurement must be highly diluted in the process of sample pretreatment for both hemolysis and deproteinization, a much more sensitive enzymatic method is desirable.

In this study, we developed a new, highly sensitive enzymatic method for the measurement of erythrocyte creatine comprising the CTase/SOX/POD system coupled with an N-methylcarbamoyl derivative of methylene blue, 10-N-methylcarbamoyl-3,7-bis(dimethylamino)phenothiazine (MCDP) (18), as a chromogenic compound. Using this method, we measured erythrocyte creatine in patients with various hemolytic conditions to evaluate its potential for clinical usefulness as an index of the erythrocyte mean age.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
apparatus
We used a UVIDEC-610C spectrophotometer (JASCO) and a COBAS MIRA® automatic analyzer (Roche) to establish the optimal conditions for the present method. The COBAS MIRA analyzer was also used for the enzymatic measurement of creatine in erythrocyte samples. HPLC (LC4A/CTO-2A/RF-530; Shimadzu) was performed on an ion-exchange column (ISC-05/S0504, Shimadzu). Hematological examinations, including erythrocyte count, hemoglobin concentration, platelet count, and mean corpuscular hemoglobin concentration in blood samples were carried out with an automated analyzer, Sysmex SE-9000; and reticulocyte counts were performed with an automated analyzer, Sysmex R-3000 (both from TOA Medical Electronics).

enzymes and chemicals
CTase (from Actinobacillus sp.), SOX (from Arthrobactor sp.), and POD (from horseradish) were purchased from Toyobo. Catalase (from bovine liver) was purchased from Sigma Chemical Co. Ascorbate oxidase (from Cucurbita sp.) was purchased from Wako Pure Chemical. The methylene blue derivative, MCDP, which was synthesized by Kyowa Medix, was used as a chromogenic compound. The structure of MCDP (C₁₈H₂₂N₄OS; molecular weight, 342.4) is shown in Fig. 1 . Triethylenetetraminehexaacetic acid (TTHA), potassium ferrocyanide, sodium azide, glutathione (reduced and oxidized forms), ascorbate, and various amino acids and guanidino compounds including creatine were obtained from Sigma. Other chemicals used were of reagent grade unless otherwise stated.

View larger version (14K): [in this window] [in a new window]   Figure 1. Structural organization of MCDP, and the reaction between MCDP and H2O2 catalyzed by POD.

reaction principle for the enzymatic method
The principle of the present method can be summarized as follows:

In the first reaction, endogenous sarcosine is eliminated from a sample with SOX and catalase. In the second reaction, catalase is completely inhibited with sodium azide, and then creatine present in the sample is detected with the CTase/SOX/POD system coupled with MCDP oxidation (Fig. 1 ). The MCDP oxidation yields methylene blue, which leads to an increase in the absorbance at 660 nm in proportion to the creatine concentration.

procedure
Pretreatment of specimens.
Blood was collected in EDTA-containing tubes and then subjected to hematological examinations. Each blood sample was centrifuged for 10 min at 1500g to remove the plasma and buffy coat. The erythrocytes were hemolyzed with an approximately sixfold volume of 1.0 g/L saponin and then stored for 10 min at room temperature to accomplish complete hemolysis. Because the amount of creatine in the trapped plasma was negligible because of the low concentration of plasma creatine (approximately one-tenth of that of erythrocyte creatine) (4), the erythrocyte sample was not washed in the above procedure. An aliquot (50 µL) of the hemolysate was then mixed with 100 µL of 0.15 mol/L Ba(OH)₂ and 100 µL of 0.15 mol/L ZnSO₄ for deproteinization. The supernatant, obtained on centrifugation for 5 min at 10 000g and filtration, was used for creatine measurement by both the present method and HPLC. The hemolysate was also subjected to hemoglobin measurement for conversion of the measured creatine value (µmol/L in supernatant) to micromoles of creatine per gram of hemoglobin.

Reagents and analytical conditions for the present method.
A 4.0 mmol/L MCDP solution was first made by dissolution in methanol and then diluted to an appropriate concentration, according to the description below, with 50 mmol/L Tris-HCL buffer (pH 8.0) containing 1.0 g/L Triton^® X-100. The enzymatic creatine measurement was performed with two reagents, as follows: reagent 1, comprising 15.4 kU/L SOX, 86 kU/L catalase, 3.0 kU/L ascorbate oxidase, 0.8 mmol/L TTHA, 1.0 g/L Triton X-100, and 150 µmol/L MCDP in 50 mmol/L Tris-HCL buffer (pH 8.0); and reagent 2, comprising 85 kU/L CTase, 11 kU/L POD, 1.0 g/L Triton X-100, 72 µmol/L potassium ferrocyanide, and 18 mmol/L sodium azide in 50 mmol/L Tris-HCL buffer (pH 8.0). To avoid nonspecific oxidation of MCDP, reagent 1 was put into a light-proof bottle and was then set on the COBAS MIRA analyzer. The two reagents were stored in light-proof bottles at 4 °C and could be used for at least 3 weeks.

The analytical conditions for erythrocyte creatine on the COBAS MIRA analyzer were as follows: wavelength, 660 nm; temperature, 37 °C; sample volume, 30 µL (washing H₂O volume, 20 µL); reagent 1 volume, 130 µL; reagent 2 volume, 70 µL; and calculation mode, endpoint assay with a reagent blank. The timing (25 s per one cycle) for sample and reagent additions was: sample and reagent 1, cycle 1; and reagent 2, cycle 12. The timing for readings was: first, cycle 11; and last, cycle 30. The reaction time after the addition of reagent 2 was 7 min 55 s. As a calibrator for measurement of erythrocyte creatine, we used 100 µmol/L creatine dissolved in water, which can be used for at least 1 month when stored at 4 °C.

Comparison methods.
To compare the sensitivity of the present method with that of other enzymatic methods for creatine measurement, we measured 0, 25, 50, 75, and 100 µmol/L creatine calibrators in triplicate with the present method and two other enzymatic methods (each an endpoint assay system), involving the CK/PK/LDH/NADH system (16) and the CTase/SOX/POD system coupled with 3-hydroxy-2,4,6-triiodobenzoic acid (HTIB) as a chromogenic compound (17), respectively, according to previous reports, except that the sample dilution ratio (sample volume/total reaction mixture volume) was 0.12. For the correlation study, HPLC with fluorescence detection based on the alkaline-ninhydrin reaction was performed as described previously (19) with minor modifications (20).

Optimization of the components for the present method.
The optimal concentration (or activity) for each component was investigated by assaying the 100 µmol/L creatine calibrator with the analytical conditions described above. The observed absorbance was expressed as relative reactivity in comparison with the maximal absorbance. The optimal activity of catalase for elimination of endogenous sarcosine was also investigated by assaying an aqueous 100 µmol/L sarcosine.

Effect of potassium ferrocyanide on the linearity.
The effect of potassium ferrocyanide on the linearity of the present method was assessed with 0, 6.25, 12.5, 25, 50, and 100 µmol/L creatine calibrators in the presence of various concentrations of potassium ferrocyanide (i.e., the reaction mixture did not contain erythrocyte sample).

Interference.
Various substances, including guanidino compounds and amino acids with chemical structures related to that of creatine and reducing agents, were investigated as to their possible interference with creatine measurement by the present method. The creatine concentration was measured in solutions containing 100 µmol/L creatine to which each substance (100 µmol/L) had been added, and the measured value was examined for the percentage of cross-reactivity, defined as [(the measured value/the value for creatine only) - 1] x 100.

specimens
To determine the reference intervals, we measured creatine in 120 blood samples obtained from 60 healthy males (age, 40.5 ± 18.9 years; mean ± SD) and 60 healthy females (age, 40.9 ± 19.2 years). One hundred and ten patient blood samples, which were submitted routinely to our laboratory, were used for comparison with HPLC. For clinical evaluation, erythrocyte creatine was measured in various patient groups summarized in Table 1 . These blood samples were stored at 4 °C and were subjected to treatment for hemolysis and deproteinization within 3 days. The supernatants were stored at -20 °C until creatine measurement.

These samples were prepared and analyzed in accordance with the ethical recommendations of the hospital's responsible committee.

expression units
The measured value of erythrocyte creatine was expressed as both micromoles per gram of hemoglobin (µmol/g Hb) and micromoles per liter of packed erythrocytes [µmol/L red blood cells (RBCs)] unless otherwise indicated: µmol/g Hb, [creatine concentration in the supernatant (µmol/L) x 5]/individual hemoglobin concentration in the hemolysate (g/L); and µmol/L RBCs, erythrocyte creatine (µmol/g Hb) x individual mean corpuscular hemoglobin concentration (g/L).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
optimization of the present method
Effects of various components on the enzymatic reaction.
The optimal conditions for the main reaction are as follows: buffer, 50 mmol/L Tris-HCL (pH 8.0); enzymes (activity in the final reaction mixture): CTase, 25 kU/L; SOX, 8 kU/L; and POD, 3 kU/L; other chemicals (concentration in the final reaction mixture): MCDP, 60 µmol/L; TTHA, 0.4 mmol/L; and Triton X-100, 1.0 g/L. The effects of the buffer (pH), the activities of the key enzymes, and the concentration of MCDP on the enzymatic reaction are shown in Fig. 2 . The presence of 60 kU/L catalase in the first reaction mixture was sufficient to eliminate 100 µmol/L sarcosine in a sample within 4 min 35 s (the time for the first reaction); the catalase was then promptly and completely inhibited by the presence of 5.0 mmol/L sodium azide in the final reaction mixture. However, the creatine assay was not affected by the concentration of sodium azide.

View larger version (28K): [in this window] [in a new window]   Figure 2. Optimization of pH (A), CTase activity (B), SOX activity (C), and MCDP concentration (D) for the present method. The pH, enzyme activities, and MCDP concentration were varied as indicated, the enzymes and MCDP being expressed as the activity and concentration in the final reaction mixture, respectively. The results are expressed as relative reactivity in comparison with the maximal absorbance.

Effect of potassium ferrocyanide on the linearity of the creatine measurement.
The effect of potassium ferrocyanide on the linearity of the present method is shown in Fig. 3 . The presence of 20 µmol/L potassium ferrocyanide in the final reaction mixture most efficiently improved the linearity for quantifying creatine.

View larger version (21K): [in this window] [in a new window]   Figure 3. Effect of potassium ferrocyanide on the linearity of the present method. (A) Measured creatine vs theoretical creatine; (B) measured/theoretical ratio vs theoretical creatine. Various concentrations of potassium ferrocyanide were added to the second reagent; concentrations in the final reaction mixtures were () 0 µmol/L, () 10 µmol/L, () 20 µmol/L, () 40 µmol/L, () 80 µmol/L, and () 160 µmol/L.

analytical evaluation
Precision.
The within-run imprecision of the present method was estimated by repeated analysis of three different hemolysates containing 76.8, 157.6, and 296.5 µmol/L creatine, respectively. The within-run CVs (n = 20 for each hemolysate) were 0.9%, 0.7%, and 0.8%, respectively. To estimate the between-day imprecision, the three hemolysates were stored at -20 °C and then assayed with the present method over 20 days. The between-day CVs (n = 20 for each hemolysate) were 1.5%, 1.3%, and 1.4%, respectively.

Linearity.
To examine the linearity of the calibration curve, nine different creatine solutions, ranging from 0 to 100 µmol/L in 12.5 µmol/L increments, were measured with the present method in triplicate. The mean values were estimated by linear regression analysis of the theoretical (x) and measured values (y). Good linearity was observed at least up to 100 µmol/L in a solution (i.e., equivalent to 500 µmol/L in a hemolysate): correlation coefficient, 0.9998; slope, 1.0066 ± 0.0081 (mean ± SD); intercept, 0.514 ± 0.480 (mean ± SD) µmol/L; and S_yx, 0.661 µmol/L. Because the erythrocyte samples, after the removal of the plasma and buffy coat, were diluted sevenfold or more with 1.0 g/L saponin in the sample pretreatment, the measurable upper limit was estimated to be ~11 µmol/g Hb, or 3500 µmol/L RBCs.

Detection limit.
To examine the detection limit of the present method, a creatine free-hemolysate, which was prepared from hemolysate (from a healthy subject) by incubation with 25 kU/L CTase at 37 °C for 1 h, was measured 10 times after deproteinization. The detection limit, defined as mean 3 SD of the measured values, was 1.0 µmol/L in hemolysate (~0.02 µmol/g Hb, or 7 µmol/L RBCs).

Analytical recovery rate.
In this experiment, three different hemolysates, to which various concentrations of creatine had been added, were assayed five times, respectively. As shown in Table 2 , the analytical recovery rates varied from 95.5% to 101.4% (mean ± SD, 99.3% ± 1.8%).

Interference.
No cross-reactivity (within ± 1.0%) was observed with various substances, including sarcosine, creatinine, creatine phosphate, guanidinoacetic acid, guanidinosuccinic acid, -guanidinobutyric acid, ß-guanidin-opropionic acid, methylguanidine, arginine, arginosuccinic acid, alanine, aspartic acid, ornithine, citrulline, proline, cysteine, tryptophan, methionine, glutathione (reduced and oxidized forms), and ascorbic acid, respectively, each at a concentration of 100 µmol/L.

Comparison of the sensitivity of the three enzymatic methods for creatine measurement.
As shown Fig. 4 , the present method exhibited higher sensitivity than the other two enzymatic methods. The relative sensitivities (present method = 1.00) were 0.079 for the CK/PK/LDH/NADH method and 0.36 for the CTase/SOX/POD/HTIB method, respectively, as estimated with the slope of the curve.

View larger version (21K): [in this window] [in a new window]   Figure 4. Comparison of the sensitivity of the three enzymatic methods for creatine measurement. The values are COBAS MIRA-generated absorbances at 660 nm for the present method (), at 500 nm for the CTase/SOX/POD/HTIB method (), and at 340 nm for the CK/PK/LDH/NADH method (), respectively.

Correlation between the present method and HPLC.
Erythrocyte creatine was measured in 110 different blood samples with the present method (y) and HPLC (x), and the results were estimated by linear regression analysis. Excellent agreement was observed between the two methods (Fig. 5 ).

View larger version (19K): [in this window] [in a new window]   Figure 5. Correlation between the present method and HPLC for measurement of erythrocyte creatine. The measured erythrocyte creatine is expressed as µmol/g Hb (A) and µmol/L RBCs (B), respectively.

reference intervals for erythrocyte creatine
The results obtained on measuring erythrocyte creatine in both 60 healthy males and 60 healthy females fitted a gaussian distribution, as judged with the coefficients of skewness and kurtosis, respectively. The reference intervals (mean ± 2 SD) were: 1.18 ± 0.52 (0.66–1.70) µmol/g Hb and 384 ± 168 (216–552) µmol/L RBCs for males; and 1.35 ± 0.49 (0.86–1.84) µmol/g Hb and 431 ± 150 (281–581) µmol/L RBCs for females. The mean value of erythrocyte creatine was compared between males and females, using the two-sample independent-groups t-test. Erythrocyte creatine in the females was slightly but significantly (P <0.001) higher than that in the males with both expression units.

clinical evaluation
The mean value of erythrocyte creatine was compared between healthy control and patient groups in each gender group, using the two-sample independent-groups t-test (Table 3 ). Significantly higher erythrocyte creatine concentrations were observed in patients with hemolytic anemia, liver cirrhosis, and renal insufficiency, and hemodialysis patients, respectively. Hemodialysis patients undergoing erythropoietin therapy showed significantly higher erythrocyte creatine than those not undergoing the therapy. No significant difference in erythrocyte creatine was observed between the renal insufficiency patients and the hemodialysis patients without erythropoietin therapy.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
In human erythrocytes, there are various cell age-related markers other than creatine, including 2,3-diphosphoglycerate (7)(11), potassium (21), hemoglobin (21), ATP (21)(22), and a number of enzymes, e.g., hexokinase (7)(22), aspartate aminotransferase (7)(23), glucose-6-phosphate dehydrogenase (21)(22)(23), cholinesterase (21)(24), and pyruvate kinase (25), which change in content with advancing age of the erythrocytes. These markers, however, have various disadvantages for estimating the erythrocyte age: 2,3-Diphosphoglycerate is affected by the severity of anemia (7); potassium, hemoglobin, and ATP are not sensitive enough for quantitative estimation of erythrocyte aging because of minor differences in their contents between young and old cells (21); and the erythrocyte enzymes are unstable on sample storage, and the activities are likely to be interfered by other factors, such as possible contamination by the leukocyte and platelet enzymes, and erroneous results because of possible interference by non-cell age-dependent disease processes. On the other hand, erythrocyte creatine is thought to be the most promising marker for the mean age of the erythrocyte population for the reasons we described earlier. However, only a few methods for erythrocyte creatine have been reported. Recently, we established an enzymatic method for erythrocyte creatine, using a commercially available kit that had been developed for serum and urine creatine, involving the CTase/SOX/POD/HTIB system (17). This method was postulated to have higher sensitivity for erythrocyte creatine as compared with the CK/PK/LDH/NADH method, because the molar absorptivity of NADH at 340 nm is fivefold higher than that of a HTIB-derived quinone-imine dye at 515 nm.

In this study, we further developed a new highly sensitive enzymatic method for erythrocyte creatine comprising the CTase/SOX/POD system coupled with MCDP as a chromogenic compound. MCDP is quite suitable for detection of a very small amount of analytes, because the methylene blue derived from MCDP exhibits higher molar absorptivity (96 x 10 L · mol^-1 · cm^-1 at 666 nm per mole of H₂O₂) than that of other chromogenic compounds used in assays based on enzymatic reaction coupled with POD, e.g., p-chlorophenol (15.0 x 10 at 500 nm), N-ethyl-N-(2-hydroxy-3-sulfopropyl)-3,5-di-methoxyaniline (17.5 x 10 at 593 nm), N-ethyl-N-(2-hydroxy-3-sulfopropyl)-m-toluidine (33.0 x 10 at 555 nm), and HTIB (31.2 x 10 at 515 nm). In the course of establishing the assay system, however, there was a serious problem: the MCDP oxidation derived from the enzymatic reaction was not proportional to creatine concentration (Fig. 3 ; the curve for 0 µmol/L potassium ferrocyanide). We sought a substance to eliminate the nonspecific oxidation of MCDP, and found that potassium ferrocyanide inhibited nonspecific oxidation in the enzymatic reaction. Satisfactory linearity was observed in the presence of 20 µmol/L potassium ferrocyanide (Fig. 3 ). The sensitivity of the present method was 12.7-fold higher than that of the CK/PK/LDH/NADH method and 2.8-fold higher than that of the CTase/SOX/POD/HTIB method, respectively, suggesting that the present method makes it possible to quantify erythrocyte creatine using blood samples taken in a microcapillary for neonatal patients and that MCDP with potassium ferrocyanide can improve the sensitivity of POD-coupled assays. The present method showed good precision, within-run CVs <1.0%, and between-day CVs <2.0%, an acceptable analytical recovery rate that averaged 99.3% ± 1.8%, and high specificity. As compared with other methods, this method is not likely to be affected by the sample matrix, because the volume of an erythrocyte sample is minimal in the final reaction mixture. Excellent correlation was also observed between the present method and HPLC. These data indicate that the present method exhibits favorable analytical performance in sensitivity, precision, and accuracy.

Using the present method, we measured erythrocyte creatine in healthy subjects and patients with various hemolytic conditions. In healthy subjects, the mean of the erythrocyte creatine in females was slightly but significantly (P <0.001) higher than that in males, in agreement with other reports (6)(9)(12)(14)(16)(17). In patients with hemolytic anemia, extremely increased erythrocyte creatine concentrations were observed, in accordance with previous reports (4)(5)(6)(7)(16). A significant increase in erythrocyte creatine was also observed in patients with liver cirrhosis, suggesting accelerated hemolysis because of hypersplenism and a possible abnormality of the erythrocyte membrane in cirrhotic patients (26). For renal anemia, in the patients with renal insufficiency or chronic renal failure undergoing hemodialysis, a significant increase in erythrocyte creatine was observed in accordance with other reports (4)(14)(16). In addition, hemodialysis patients undergoing erythropoietin therapy showed significantly higher erythrocyte creatine concentrations than those not undergoing the therapy. These findings suggest that renal anemia is caused in part by a shortened erythrocyte survival time and by insufficient erythropoietin production as a result of destruction of the renal mass, and that the administration of erythropoietin compensates for the reduced erythropoiesis and leads to an increase in young erythrocytes (i.e., a decrease in the mean age of erythrocytes). Thus, erythrocyte creatine measurement provides us with useful information for the evaluation of hemolytic processes not only in patients with hemolytic anemia but also in patients with liver and renal disorders.

In conclusion, the present method is highly sensitive and specific to creatine in erythrocytes and has favorable characteristics for routine work in clinical laboratories because of its applicability to commonly used automated analyzers.

	Acknowledgments

 
We thank Kohichi Saika, Shigeo Yamanaka, and Tomoyo Fujita of Kochi Medical School Hospital (Oko-cho, Nankoku, Japan) for excellent technical support, and Eiji Tsubosaki and Hiroshi Matsumura of Kochi Kenshin Clinic (Kochi, Japan) for supplying the blood samples from healthy volunteers. This work was supported in part by a grant from the Kurozumi Medical Foundation.

	Footnotes

 
¹ Nonstandard abbreviations: CK, creatine kinase; PK, pyruvate kinase; LDH, lactate dehydrogenase; CTase, creatine amidinohydrolase; SOX, sarcosine oxidase; POD, peroxidase; MCDP, 10-N-methylcarbamoyl-3,7-bis (dimethylamino) phenothiazine; HTIB, 3-hydroxy-2,4,6-triiodobenzoic acid; TTHA, triethylenetetraminehexaacetic acid; and RBC, red blood cell.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Lee GR. The hemolytic disorders: general considerations. Lee GR Bithell TC Foerster J Athens JW Lukens JN eds. 9th ed. Wintrobe's clinical hematology 1993;Vol. 1:944-964 Lea & Febiger Philadelphia, PA. .
2. Cline MJ, Berlin NI. The reticulocyte count as an indicator of the rate of erythropoiesis. Am J Clin Pathol 1963;39:121-128.
3. Hillman RS, Finch CA. The misused reticulocyte. Br J Haematol 1969;17:313-315. [ISI][Medline] [Order article via Infotrieve]
4. Fehr J, Knob M. Comparison of red cell creatine and reticulocyte count in appraising the severity of hemolytic processes. Blood 1979;53:966-976. [Abstract/Free Full Text]
5. Smith BJ, Mohler DN, Wills MR, Savory J. Erythrocyte creatine levels in anemia. Ann Clin Lab Sci 1982;12:439-446. [Abstract]
6. Griffiths WJ, Fitzpatrick M. The effect of age on the creatine in red cells. Br J Haemat 1967;13:175-180. [ISI][Medline] [Order article via Infotrieve]
7. Opalinski A, Beutler E. Creatine, 2,3-diphosphoglycerate, and anemia. New Eng J Med 1971;285:483-486.
8. Brewster MA, Berry DH. Detection of G6PD and pyruvate kinase deficiencies in reticulocytosis by reference to erythrocyte creatine. Clin Biochem 1981;14:132-134. [ISI][Medline] [Order article via Infotrieve]
9. Smith BJ, Butzner JD, Wills MR, Saibo H, Savory J. Erythrocyte creatine in cord blood. Ann Clin Lab Sci 1993;13:439-443.
10. Rejiman ASM, Grimes AJ, Cotes PM, Mansell MA, Joekes AM. Correction of anaemia following renal transplantation: serial changes in serum immunoreactive erythropoietin, absolute reticulocyte count and red-cell creatine levels. Br J Haemat 1985;61:421-431. [ISI][Medline] [Order article via Infotrieve]
11. Ricci G, Masotti M, De Paoli Vitali E, Vedovato M, Zanotti G. Effects of exercise on haematologic parameters, serum iron, serum ferritin, red cell 2,3-diphosphoglycerate and creatine contents, and serum erythropoietin in long-distance runners during basal training. Acta Haemat 1988;80:95-98. [ISI][Medline] [Order article via Infotrieve]
12. Li PK, Lee JT, Li CS, Deshpande G. Improved method for determining erythrocyte creatine by the diacetyl--naphthol reaction: elimination of endogenous glutathione interference. Clin Chem 1982;28:92-96. [Abstract/Free Full Text]
13. Wong T. Studies on creatine determination by -naphthol-diacetyl reaction. Anal Biochem 1971;40:18-28. [ISI][Medline] [Order article via Infotrieve]
14. Buysse AM, Delanghe JR, De Buyzere ML, De Scheerder IK, De Mol AM, Noens L. Enzymatic erythrocyte creatine determinations as an index for cell age. Clin Chim Acta 1990;187:155-162. [ISI][Medline] [Order article via Infotrieve]
15. Cramer H, Dauwalder H, Meier H, Colombo JP. Enzymatic determination of red cell creatine as an index of hemolysis. Clin Biochem 1987;20:329-332. [ISI][Medline] [Order article via Infotrieve]
16. Beyer C, Alting IH. Enzymatic measurement of creatine in erythrocytes. Clin Chem 1996;42:313-318. [Abstract/Free Full Text]
17. Jiao Y, Okumiya T, Saibara T, Tsubosaki E, Matsumura H, Park K, et al. An enzymatic assay for erythrocyte creatine as an index of the erythrocyte life time. Clin Biochem 1998;31:59-65. [ISI][Medline] [Order article via Infotrieve]
18. Yagi K, Kiuchi K, Saito Y, Miike A, Kayahara N, Tatano T, et al. Use of a new methylene blue derivative for determination of lipid peroxides in foods. Biochem Int 1986;12:367-371. [ISI][Medline] [Order article via Infotrieve]
19. Hiraga Y, Kinoshita T. Post-column derivatization of guanidino compounds in high-performance liquid chromatography using ninhydrin. J Chromatogr 1981;226:43-51. [ISI][Medline] [Order article via Infotrieve]
20. Okumiya T, Sasaki M, Park K, Sumiyoshi K. High sensitive analysis of guanidino compounds using linear gradient elution HPLC (in Japanese). Rinsho Byori 1988;36:313-317. [Medline] [Order article via Infotrieve]
21. Romero PJ, Romero EA, Winkler MD. Ionic calcium content of light dense human red cells separated by percoll density gradients. Biochim Biophys Acta 1997;1323:23-28. [Medline] [Order article via Infotrieve]
22. Brok F, Ramot B, Zwang E, Danon D. Enzyme activities in human red blood cells of different age groups. Isr J Med Sci 1966;2:291-296. [ISI][Medline] [Order article via Infotrieve]
23. Sass MD, Vorsanger E, Spear PW. Enzyme activity as an indicator of red cell age. Clin Chim Acta 1964;10:21-26.
24. Sabine JC. Erythrocyte cholinesterase titers in hematologic disease states. Am J Med 1959;27:81-96.
25. Nakashima K, Nishizaki O, Andoh Y, Takei H, Itai A, Yoshida Y. Glycated hemoglobin in fractionated erythrocytes. Clin Chem 1989;35:958-962. [Abstract/Free Full Text]
26. Morse EE. Mechanisms of hemolysis in liver disease. Ann Clin Lab Sci 1990;20:169-174. [Abstract]