Novel liquid chromatographic determination of cystatin C in human urine

Journal of Chromatography B, 877 (2009) 747–750

Contents lists available at ScienceDirect

Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb

Novel liquid chromatographic determination of cystatin C in human urine Othman Ibrahem Yousef Al-Musaimi a,∗ , Manar Khalid Fayyad b , Adel Khalil Mishal c a

Analytical Research Department, Hikma Pharmaceuticals-Amman, Department of Chemistry, University of Jordan, Amman, Jordan University of Jordan, Amman, Jordan c Hikma Pharmaceuticals, Amman, Jordan b

a r t i c l e

i n f o

Article history: Received 27 November 2008 Received in revised form 28 January 2009 Accepted 2 February 2009 Available online 11 February 2009 Keywords: Cystatin C Protein HPLC Chromatography

a b s t r a c t A rapid and sensitive method for quantitative determination of cystatin C (CC) protein in human urine via HPLC was developed and validated. Acetone has been used as a precipitating agent of CC protein from the urine biological matrix. Separation and detection were accomplished by ion pair liquid chromatography and UV detection. Gradient elution mode was utilized to elute CC with a UV detection of 224 nm. The analysis time was 14 min per sample using Ace C8 (150 × 4.6 mm i.d., 5 ␮m) as a chromatographic column with a flow rate of 1.0 mL/min. Calibration curve with good linearity (r2 = 0.99) within the range from 0.390 to 0.001 mg/mL was obtained. Limits of detection and quantification were 0.001 and 0.002 mg/mL, respectively. Inter-assay and intra-assay variabilities were <15% for all levels and <20% at the limit of quantification level. Major advantages of the method: specific where no false positive results might be obtained and fast where sample pretreatment needs only 1 h. © 2009 Elsevier B.V. All rights reserved.

1. Introduction CC is a cysteine proteinase inhibitor. Large number of pathological processes is controlled by the balance between proteinases and their inhibitors, where an increase of these proteinases has been observed in case of inflammation [1,2]. Concentration of CC protein is considered as an overall index of renal function. CC leaks out from the kidney in case of glomerular filtration rate (GFR) reduction. There are different types of GFR markers, which reflect whether the kidney is functioning properly or not. Some of these markers are exogenous markers, such as: 51 Cr-EDTA, Inulin, (125 I) Iotholamate (99m Tc) TPA. The others are endogenous: microalbuminurea, creatinine, cystatins. . .etc. [3–6]. CC amount in the urine increases by 200 folds at the early stages of kidney failure whereas in the other biological fluids (blood, tears) it increases only by 2 folds [1,7,8]. Several immunoassay methods have been reported for quantification of CC. Particle enhanced nephelometric immunoassay (PENIA), particle enhanced turbidimetric method (PET) and enzyme-amplified single radial immunodiffusion [9–12]. However, the interferences caused by cross reactions (between antigens and antibodies) which might lead to false positive results, long analysis time, in addition, to high coefficient of variation obtained by the mentioned methods were unsatisfactory.

Besides, HPLC-MS analytical method was also applied to quantify CC [13]. However, this method tracked only CC raw material and it is not applicable to CC in biological fluids. In this study, we developed and validated a UV-HPLC method that can be used to separate, detect and quantify CC without complicated sample preparation steps. This method is rapid, simple and sensitive enough to quantify the concentration of CC in unhealthy subjects with good accuracy and precision. 2. Experimental 2.1. Chemicals CC protein was purchased from Scipac (UK), HPLC grade acetonitrile, methanol, 1-hexane sulfonic acid sodium salt, trifluoroacetic acid (TFA) and acetone were purchased from Merck (Germany). 2.2. Urine collection and processing Five patients, 3 males and 2 females (range 43–61 years old) with known history of renal failure were recruited from hospital. Each one gave written informed consent to participate in the study. Urine samples were collected from the patients into polypropylene (acetone-compatible) tubes. Samples were processed within 5 h. 2.3. Sample preparation

∗ Corresponding author. Tel.: +962 78 5782259. E-mail address: [email protected] (O.I.Y. Al-Musaimi). 1570-0232/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jchromb.2009.02.009

CC analyte was precipitated from the urine samples by vortexing them with cold acetone (−10 to −20 ◦ C) for 1 h (keeping the same

748

O.I.Y. Al-Musaimi et al. / J. Chromatogr. B 877 (2009) 747–750 Table 2 CC method precision (ruggedness) (inter-assay).

Table 1 Gradient program. Time/minutes

Mobile phase A%

Mobile phase B%

Preparation #

% CC (Day1) ±15%

% CC (Day2) ±15%

0.01 3.00 8.00 10.00 10.50 14.00

35 20 0 0 35 35

65 80 100 100 65 65

1 2 3 4 5 6 Average RSD % RSD % within 12 preparations

98.9 94.6 97.2 93.5 101.8 90.8 96.1 4.1

104.7 94.7 105.1 91.0 95.5 88.0 96.5 7.3

temperature during vortexing) [14]. Centrifugation at 4400 × g at 4 ◦ C for 30 min took place; the solid precipitate was dissolved in 1 mL of 0.05% TFA (v/v). Finally, solution was transferred to an eppendrof vial and centrifuged at 4400 × g at 4 ◦ C for 15 min. Supernatant was filtered with 0.45 ␮m teflon filter. 2.4. Chromatographic conditions Shimadzu HPLC system (Shimadzu, Japan) with a degasser, low pressure gradient pump, column oven, an autosampler, a UV detector was used. Data acquisition was performed with LC-Solution 1.21 software. A reversed phase Ace C8 (150 × 4.6 mm i.d., 5 ␮m) column was placed in the column oven at 25 ◦ C. Mobile phase A of 0.01 M 1-hexane sulfonic acid sodium salt plus 0.05% TFA, pH 2.4 (filtered through 0.45 ␮m Teflon filter) and mobile phase B (acetonitrile:methanol:mobile phase A) (300:300:225, v/v/v), pH 2.5), were delivered at flow rate of 1.0 mL/min and used for gradient elution of CC. Prior to each analysis, the column was equilibrated with 65% of mobile phase B for 10 min. The gradient was increased to 80% in 3 min and then for 100 for 7 min to complete the separation. The gradient was decreased to return the system to the initial conditions. At 14 min, the HPLC system is ready for the next injections (Table 1). The injection volume was 100 ␮L. CC peak was detected at 224 nm.

5.7 ± 15%

acid and aspirin were tested to check if there is any interference with CC peak. 2.5.5. Stability of analytical solutions CC was spiked into healthy urine at low and high concentration and stored for 3 days at room temperature, in the refrigerator and in the freezer (25, 2–8 ◦ C and −10 to −20 ◦ C), respectively. Samples were processed, prepared and analyzed using the developed method. The concentrations were calculated from the calibration curve and then compared with the actual values. Five replicates were preformed. In order to check the developed method for any potential interference that might be present in the biological matrices such as the endogenous matrix component, metabolites, decomposition products. . .etc., six healthy urine samples from different sources were analyzed. 3. Results and discussion 3.1. Absorbance spectra of CC CC has absorption maximum around 224 nm.

2.5. Assay validation

3.2. Linearity

All of the validation tests were performed according to the ICH Guideline Q2 (R1) [15] and Bioanalytical Method Validation Guidance [16].

Good linearity was obtained for CC over the range 0.390–0.001 mg/mL with r2 = 0.992, (Y (mAU) = 3e+7x + 95,316). LOD was estimated according to the visual evaluation method that involved analysis of samples with known concentrations of analyte and establishing the minimum level at which the analyte can be reliably detected. LOQ was determined by establishing the minimum level at which the analyte can be quantified with acceptable accuracy and precision.

2.5.1. Linearity To establish the calibration curve, eight calibration points were constructed from CC stock solution (0.4 mg/mL CC in healthy urine). Same sample preparation procedure was applied. 2.5.2. Limits of detection (LOD) and quantification (LOQ) Using the calibration curve constructed from the standard solutions, the LOD was defined as the lowest concentration that can be detected but not necessarily quantified and LOQ the lowest concentration that can be determined with acceptable accuracy and precision. 2.5.3. Accuracy and precision Quality control (QC) samples consisting of a high, medium and low concentration of CC analyte, containing (0.400, 0.200 and 0.002) mg/mL of CC in healthy urine were stated at the beginning and the end of assay work to determine the accuracy and the precision. CV or RSD were determined from 6 consecutive preparations (inter-day) and from 10 independent assays from one experimental day (intra-day). 2.5.4. Selectivity Active ingredients of non-prescribed drugs, over the counter drugs (OTC): ibuprofen, caffeine, paracetamol, nicotine, ascorbic

Fig. 1. Chromatograms of recovery preparations, (0.400, 0.200, 0.100 mg/mL CC and blank solution).

O.I.Y. Al-Musaimi et al. / J. Chromatogr. B 877 (2009) 747–750

749

Table 3 Recovery of CC from urine. Concentration (mg/mL)

Preparation

% Recovery

Average

% RSD

0.002

1 2 3

89.9 93.0 87.0

90.0 ± 20%

3.3 ± 20%

0.1

1 2 3

98.8 94.6 93.4

95.6 ± 15%

3.0 ± 15%

0.2

1 2 3

97.2 93.5 103.9

98.2 ± 15%

5.4 ± 15%

0.4

1 2 3

101.8 90.8 101.0

97.9 ± 15%

6.3 ± 15%

Fig. 3. Chromatogram of pathological sample (0.66 mg/mL CC). Table 4 Real samples data (pathological samples). Sample #

Age (years)

Gender (male/female)

CC amount (mg/mL)

1 2 3 4 5

55 43 61 58 49

Male Male Female Female Male

1.09 0.66 0.44 1.11 0.77

Healthy urine samples already contain 0.001 mg/mL CC, accordingly it was used to evaluate the detection limit, where at this concentration the peak due to CC was detected but not necessarily quantified. The lowest concentration that can be determined with acceptable precision and accuracy should be 0.001 mg/mL of CC, but due to the fact that CC is already present in the blank (healthy) urine sample, 0.002 mg/mL of CC will be considered as the quantification limit instead of 0.001 mg/mL CC. Accuracy and precision were determined by analyzing high, medium and low concentrations of QC samples throughout the standard calibration range. Intra-assay and inter-assay precisions were <15% and the mean values of the accuracy were between 90.0 and 98.2% (Fig. 1, Tables 2 and 3 ). 3.3. Selectivity None of the OTC drugs appeared to interfere with the CC peak. Furthermore, six healthy urine samples were analyzed, no interferences were observed with the CC peak. These testify the high

Fig. 2. Chromatogram of pathological sample (0.44 mg/mL CC).

Fig. 4. Chromatogram of pathological sample (1.09 mg/mL CC).

selectivity of the developed method and consequently, no false positive results will take place. Figs. 5 and 6. 3.4. Real samples/application We measured the concentration of CC in unhealthy urine samples. The obtained concentrations were inline to what have been reported in the literature. CC concentration has increased by 200

Fig. 5. Chromatograms of OTC drugs (nicotine, aspirin and ascorbic acid).

750

O.I.Y. Al-Musaimi et al. / J. Chromatogr. B 877 (2009) 747–750

However, the stability decayed for 1 day only when freeze/thaw cycles took place; this phenomenon can be attributed to the surface-induced denaturation during freeze/thaw cycles which affect the proteins [16]. Going to LOQ concentrations of CC, stability of solutions did not exceed 1 day in all storage conditions (Table 5).

4. Conclusion New HPLC analytical method has been developed and validated in order to be used as a diagnostic tool for the susceptible kidney failure subjects where a swift corrective action can be taken.

Acknowledgements Fig. 6. Chromatograms of six healthy urine samples.

Table 5 Stability of analytical solutions. Storage condition

% CC (0.200 mg/mL) ±15%

% CC (0.002 mg/mL) ±20%

After 1 day (25 ◦ C) After 2 days (25 ◦ C) After 3 days (25 ◦ C) After 1 day (2–8 ◦ C) After 2 days (2–8 ◦ C) After 3 days (2–8 ◦ C) 1st thawing cycle after 1 day (−10 to −20 ◦ C) 2nd thawing cycle after 2 days (−10 to −20 ◦ C) 3rd thawing cycle after 3 days (−10 to −20 ◦ C)

95.5 92.6 107.1 111.1 99.0 101.8 87.5

95.6 38.3 BLQ 97.9 59.8 BLQ 94.8

58.9

BLQ

22.9

BLQ

BLQ: below limit of quantification.

folds and more in the urine of the kidney failure patients (Figs. 2–4, Table 4). 3.5. Stability of analytical solutions Stability solutions were prepared by spiking CC into healthy urine at low and high concentration and storing them for 3 days at room temperature, in the refrigerator and in the freezer (25, 2–8 and −10 to −20 ◦ C), respectively. The concentrations were calculated from the calibration curve and then compared with the actual values. Stability of analytical solutions study showed that CC of concentrations higher than the LOQ is stable for 3 days at room temperature and in the fridge.

Financial support has been furnished by Hikma Pharmaceuticals-Amman. I’d like to thank Mr. Ahmad Zalloum and Mr. Eyad hylooz for their full cooperation and support. Special thanks for Dr. Sameer Awad Allah (Hashemite University) and Mr. Tariq Mas’oud (Med LABS) for their kind advice of choosing CC bio marker. Indeed, I cannot express my appreciation for Mr. Ahmad Al-Khaza’eleh, Mr. Mohammad Khair Al-Kurdi and all staff members at Al-Hikma Modern Hospital (Hemodialysis Department) for their continuous supply of pathological samples. References [1] G. Filler, A. Bokenkamp, W. Hofmann, T. Le Bricon, C. Martinez-Bru, A. Grubb, Clin. Biochem. 38 (2005) 1. [2] L.A. Bobek, M.J. Levine, Crit. Rev. Oral. Biol. M 3 (4) (1992) 307. [3] E. Paskalev, L. lambreva, P. Simeonov, N. Koicheva, B. Beleva, M. Genova, R. Marcovska, A. Nashkov, Clin. Chim. Acta 310 (2001) 53. [4] F.J. Hoek, F.A.W. Kemperman, R.T. Krediet, Nephrol. Dial. Transpl. 18 (2003) 2024. [5] Uzun Hafize, Keles Melek Ozmen, Razzan Ataman, Aydin Sevel, Kalender Betul, Uslu Ezel, Gonul Simsek, Halac Metin, Kaya Safiye, Clin. Biochem. 38 (2005) 792. [6] H. Finney, C.J. Bates, Price.F C.P., Arch. Gerontol. Geriat. 29 (1999) 75. [7] Kazuo Uchida, Akiko Gotoh, Clin. Chim. Acta 323 (2002) 121. [8] H. Finney, D.J. Newman, W. Gruber, P. Merle, P.C. Price, Clin. Chem. 43 (6) (1997) 1016. [9] J. Kyhse-Anderson, C. Schmidt, G Nordin, B. Anderson, P. Nilsson Ehle, V. Lindstrom, A. Grubb, Clin. Chem. 40 (10) (1994) 1921. [10] M. Mussap, M. Dalla Vestra, P. Fioretto, A. Saller, M. Varagnolo, R. Nosadini, M. Plebani, Kidney Int. 61 (2002) 1453. [11] O.F. Laterza, P.C. Price, M.G. Scott, Clin. Chem. 48 (5) (2002) 699. [12] M.L. Storme, B.A. Sinnaeve, J.F. Van Bocxlaer, J. Sep. Sci. 28 (2005) 1759. [13] L. Jiang, L. He, M. Fountoulakis, J. Chromatogr. A 1023 (2004) 317. [14] Guidance for Industry: ICH Q2 (R1) Validation of Analytical Procedures: Text and Methodology, 1994, http://www.ich.org/LOB/media/MEDIA417.pdf. [15] Guidance for Industry: Bioanalytical Method Validation, 2001, http://www.fda. gov/cder/guidance/index.htm. [16] E. Cao, Y. Chen, Z. Cui, P.R. Foster, Biotechnol. Bioeng. 82 (2002) 684.