Comparison of fast protein liquid chromatography (FPLC ...icmr.nic.in/ijmr/2009/march/0305.pdf ·...

7
Haemoglobinopathies are a group of autosomal- recessive inherited human disorders prevalent in many parts of the world. The prevalence of β-thalassaemia in the lower portion of northern Thailand was found to be 1.4 per cent and 24 per cent haemoglobin (Hb)E 1 . Heterozygote screening and genetic counselling are essential for the prevention and control of severe β- thalassaemia disease, homozygous β-thalassaemia and β-thalassaemia with HbE (β-thal/HbE). Diagnosis of β-thalassaemia can be made by observing increase in Comparison of fast protein liquid chromatography (FPLC) with HPLC, electrophoresis & microcolumn chromatography techniques for the diagnosis of β-thalassaemia S. Tangvarasittichai, O. Tangvarasittichai & N. Jermnim * Department of Medical Technology, Faculty of Allied Health Sciences & * Clinical Laboratory of Naresuan University Hospital, Naresuan University, Phitsanulok, Thailand Received May 31, 2007 Background & objectives: β-thalassaemia is a genetic disorder and an important health problem around the world. Quantitative haemoglobin A 2 (HbA 2 ) levels are used for the diagnosis of β-thalassaemia. The conventional methods are high performance liquid chromatography (HPLC), electrophoresis, and microcolumn chromatography techniques. We established a fast protein liquid chromatography (FPLC) method, to measure quantitatively of HbA 2 levels, and compared its efficacy with conventional methods. Methods: The FPLC method, using a DEAE Sepharose, Hi Trap anion-exchange column chromatography technique was set up for HbA 2 measurement. In this study, 220 blood samples were screened for haemoglobin type by FPLC technique and also using HPLC, microcolumn chromatography and electrophoresis. Results: The FPLC results were highly correlated (r = 0.985, P<0.001) with those of HPLC for quantification of HbA 2 as well as cellulose acetate electrophoresis (r = 0.977) and microcolumn chromatography (r = 0.980). The FPLC method showed 100 per cent sensitivity and specificity, positive and negative predictive value for β-thalassaemia diagnosis. In addition, the FPLC method was simple, rapid, low cost and reproducible. The HbA 2 /E range of FPLC for β-thalassaemia was 6-10 per cent, HbE trait was 10-40 per cent, β-thalassaemia/HbE was 40-60 per cent and homozygous HbE was more than 60 per cent. Interpretation & conclusions: Our findings suggested that FPLC method could be used as a cost-effective method for routine β-thalassaemia diagnosis. Key words Anion-exchange chromatography - β-thalassaemia - FPLC - haemoglobin A 2 /E 242 Indian J Med Res 129, March 2009, pp 242-248

Transcript of Comparison of fast protein liquid chromatography (FPLC ...icmr.nic.in/ijmr/2009/march/0305.pdf ·...

Haemoglobinopathies are a group of autosomal-recessive inherited human disorders prevalent in many parts of the world. The prevalence of β-thalassaemia in the lower portion of northern Thailand was found to be 1.4 per cent and 24 per cent haemoglobin (Hb)E1.

Heterozygote screening and genetic counselling are essential for the prevention and control of severe β-thalassaemia disease, homozygous β-thalassaemia and β-thalassaemia with HbE (β-thal/HbE). Diagnosis of β-thalassaemia can be made by observing increase in

Comparison of fast protein liquid chromatography (FPLC) with HPLC, electrophoresis & microcolumn chromatography techniques for the diagnosis of β-thalassaemia

S. Tangvarasittichai, O. Tangvarasittichai & N. Jermnim*

Department of Medical Technology, Faculty of Allied Health Sciences & *Clinical Laboratory of Naresuan University Hospital, Naresuan University, Phitsanulok, Thailand

Received May 31, 2007

Background & objectives: β-thalassaemia is a genetic disorder and an important health problem around the world. Quantitative haemoglobin A

2 (HbA

2) levels are used for the diagnosis of β-thalassaemia.

The conventional methods are high performance liquid chromatography (HPLC), electrophoresis, and microcolumn chromatography techniques. We established a fast protein liquid chromatography (FPLC) method, to measure quantitatively of HbA

2 levels, and compared its efficacy with conventional methods.

Methods: The FPLC method, using a DEAE Sepharose, Hi Trap anion-exchange column chromatography technique was set up for HbA

2 measurement. In this study, 220 blood samples were

screened for haemoglobin type by FPLC technique and also using HPLC, microcolumn chromatography and electrophoresis.

Results: The FPLC results were highly correlated (r = 0.985, P<0.001) with those of HPLC for quantification of HbA

2 as well as cellulose acetate electrophoresis (r = 0.977) and microcolumn

chromatography (r = 0.980). The FPLC method showed 100 per cent sensitivity and specificity, positive and negative predictive value for β-thalassaemia diagnosis. In addition, the FPLC method was simple, rapid, low cost and reproducible. The HbA

2/E range of FPLC for β-thalassaemia was 6-10 per cent, HbE

trait was 10-40 per cent, β-thalassaemia/HbE was 40-60 per cent and homozygous HbE was more than 60 per cent.

Interpretation & conclusions: Our findings suggested that FPLC method could be used as a cost-effective method for routine β-thalassaemia diagnosis.

Key words Anion-exchange chromatography - β-thalassaemia - FPLC - haemoglobin A2/E

242

Indian J Med Res 129, March 2009, pp 242-248

HbA2

value by the conventional methods, including

high performance liquid chromatography (HPLC), electrophoresis, microcolumn technique, erythrocyte indices and morphology2-4. Although measurements of HbA

2 by cellulose acetate electrophoresis5 and

microcolumn chromatography6 are reproducible and accurate, these methods are labour-intensive and time consuming, and automation HPLC method, with high-throughput screening was often used in routine screening of haemoglobin7,8.

We modified the fast protein liquid chromatography (FPLC) method from microcolumn chromatographic techniques for HbA

2 determination6. FPLC, a general

purpose system for protein separation and purification, was not specifically designed for clinical analysis. However, earlier studies using FPLC for the separation of urinary protein9, several plasma proteins10, and for diagnosis of β-thalassaemia suggested that this form of chromatography might be used to identify protein profiles or variability within a single protein, which could be of clinical significance11-13.

Therefore we established the method for FPLC application in the separation of HbA

2 in our

laboratory and compared the efficacy of this method of HbA

2 measurement with routinely used methods

including HPLC, electrophoresis and microcolumn chromatography techniques.

Material & Methods

Clinical samples: A total 220 adult blood samples (2 ml) were obtained from routine thalassaemia screening unit at Naresuan University Hospital at Phitsanulok, Thailand during March to September 2007. These blood samples were identified by cellulose acetate electrophoresis at alkaline pH (Helena Labs, Beaumont, Texas, USA), DEAE sepharose microcolumn14, and HPLC (Variant, Bio-Rad Laboratories, California, USA) and these were 125 cases with normal haemoglobin (HbA

2 < 3.5%) and 95

carriers [35 β-thalassaemia trait (HbA2 = 3.5-10%), 48

HbE trait (HbE = 10-40%), 7 homozygous HbE (HbE > 60%) and 5 β-thal/Hb E (HbE = 40-60%)]. This

study protocol was approved by the Ethics Committee of Naresuan University, Phitsanulok, Thailand.

Methods: For FPLC, haemolysate was prepared by mixing 50 µl of EDTA-blood sample with 10.0 ml of Tris buffer A. Then, 0.5 ml of haemolysate was passed through a 5 x 0.5 cm (1 ml) column of diethylaminoethyl (DEAE) sepharose, Hi Trap (GE Healthcare, Sweden) connected to the FPLC (ÄKTA prime, Amersham Biosciences, USA) with flow rate 2 ml/min. The effluent from column was monitored by a single path ultraviolet monitor at 280 nm in a 10 mm path-length high resolution flow cell, and the chromatogram was saved in the computer, fractions were then separated to HbA

2 or HbE (in the same

fraction, HbA2/E), HbA and HbF. The reagents for

FPLC were as linear gradient of buffer A [50 ml of stock Tris buffer (as Tris 60.57 g mixed with 500 ml distilled water, adjusted pH 9.0 with 4 M HCl) was diluted in distilled water 1,000 ml, added 0.1 g KCN (adjusted pH 8.1 with 4 M HCl)], and buffer B (diluted 500 ml of buffer A with 500 ml 1M NaCl). The gradient profile to achieve the separation was given in Table I.

Data analysis: The correlations between variables were performed using SPSS computer programme version 11.0 (SPSS, Chicago, IL).

Results

In FPLC, the retention times of HbA2, HbA and

HbF chromatograms of normal subjects were 0.36 ± 0.026, 3.71 ± 0.33 and 6.815 ± 0.07 min, respectively.

Table I. Linear gradient profile of buffer A and B

Buffer A, volume (ml) Buffer B, %

10.011.012.013.014.015.015.520.0

01030507090

1000

Table II. Values of HbA2 and HbE determined by FPLC, HPLC microcolumn and electrophoresis techniques

Sample n FPLC HPLC Microcolumn Electrophoresis

Normalβ-thalassaemiaHbE, β-thal/HbE

1253560

3.52 ± 0.727.97 ± 1.52

35.00 ± 23.63

2.56 ± 0.445.02 ± 0.72

40.97 ± 26.93

2.01 ± 0.62 5.65 ± 1.05

42.31 ± 28.01

2.71 ± 0.66 7.01 ± 1.56

38.36 ± 28.44

Values are mean ± SD

TANGVARASITTICHAI et al: COMPARISON OF FPLC TECHNIQUES FOR β-THALASSAEMIA DIAGNOSIS 243

(A2) (B2)

Fig. 1. HbA2 chromatogram of β-thalassaemia blood sample analysis by FPLC (A2) and HPLC (B2).

This method was capable of separating haemoglobins within 10 min. The HbA

2 chromatogram of normal and

β-thalassaemia trait sample were the same retention time, but the concentration of HbA

2 in β-thalassaemia

trait was increase, as in Fig. 1 showed the HbA2

chromatogram of β-thalassaemia trait in the same blood sample analysis by FPLC (A1) and HPLC (B1). While the HbE with HbE trait and homozygous HbE had the same retention times but its concentration was much higher than the HbA

2, (Figs 2, 3) because HbA

2 and HbE

could not be separated by FPLC and HPLC techniques. The mean and standard deviation of HbA

2 and HbE

from 220 samples, normal were 3.52 per cent ± 0.75 (mean ± SD) by FPLC, 2.56 per cent ± 0.44 by HPLC, 2.01 per cent ±0.62 by microcolumn, and 2.71 per cent ± 0.66 by electrophoresis, β-thalassaemia were 7.97 per cent ± 1.52 (mean ± SD) by FPLC, 5.02 per cent ± 0.72 by HPLC, 5.65 per cent ± 1.05 by microcolumn, and 7.01 per cent ±1.56 by electrophoresis, HbE, β-thal/HbE were 35.0 per cent 23.63 (mean±SD) by

244 INDIAN J MED RES, MARCH 2009

FPLC, 40.97 per cent ± 26.93 by HPLC, 42.31 per cent ± 28.01 by microcolumn, and 38.36 per cent ± 28.44 by electrophoresis, were shown in Table II. A positive correlation was seen between HbA

2/E levels measured

by FPLC method and those with the HPLC (r = 0.985 P<0.001), electrophoresis, (r = 0.977, P<0.001), and microcolumn chromatography technique, (r = 0.980 P<0.001). The within-run and between-run precision of the methods were assessed by deriving the standard deviation of all measurements obtained during the

10 wk period. The data were then expressed as percentage coefficient of variation (CV). Additional within-run precision and between-run precision data were obtained using samples from two subjects having different haemoglobin type (Table III) which was included at the beginning and end of each of the 10 sample runs. From the 220 adult blood specimens screened for haemoglobin types, 35 were identified as β-thalassaemia, 48 samples of HbE traits, 7 samples of homozygous HbE, 5 samples of β-thal/HbE and 125 samples as normal. The diagnostic

(A3) (B3)

Fig. 2. HbA2/E chromatogram of from HbE trait blood sample analysis by FPLC (A3) and HPLC (B3).

TANGVARASITTICHAI et al: COMPARISON OF FPLC TECHNIQUES FOR β-THALASSAEMIA DIAGNOSIS 245

accuracy of FPLC was all 100 per cent sensitivity and specificity, positive predictive and negative predictive value (Table IV). The FPLC cut-off value of normal samples determined by 2.5th percentile to 97.5th percentile of HbA

2 was 3.08-5.93 per cent, which

was the only higher value than HPLC. FPLC cut-off values for HbA

2/E diagnosis were as follows: Normal

(<6%), β-thalassaemia (7-10%), HbE-trait (>10-40%), homozygous HbE (>60%) and β-thal/HbE (40-60 %).

Discussion

The FPLC, a general purpose system for protein separation and purification was not specifically

designed for clinical analysis. This method was adapted from the microcolumn chromatography technique with some modification in the buffer systems as used in this technique. Buffer with pH 8.1 gave the optimal HbA

2/E separation. The most interesting was the good

precision, accuracy and reliability of this method. The FPLC method carried out as the same procedure as that of HPLC method. The Hi Trap Sepharose, anion-exchange column does not shrink or swell, or leak functional groups. The large numbers of haemoglobin samples that could be run on a Hi Trap Sepharose column and the self-preparation buffer kept the costs very low. In FPLC, cost/test was about 0.14 US$, very

(A4)

Fig. 3. HbA2/E chromatogram of homozygous HbE blood sample analysis by FPLC (A4) and HPLC (B4).

246 INDIAN J MED RES, MARCH 2009

(B4)

Table III. Within-run and between-run precision of FPLC method

Hb type Within-run assay (n=20)

Between-run assay(n=20)

Normal β-thalassaemia

Normal β-thalassaemia

Hb A2, (%)

Mean 4.52 8.92 4.54 8.96 SD 0.168 0.448 0.44 0.584 CV, % 3.72 5.02 5.37 6.52Hb A, (%) Mean 94.82 91.42 95.14 91.86 SD 4.342 4.422 4.56 4.62 CV, % 4.58 4.84 4.79 5.03Hb F, (%) Mean 0.82 1.02 0.86 1.07 SD 0.061 0.084 0.082 0.088

CV, % 7.44 8.24 9.53 8.22

Table IV. β-thalassaemia diagnosis accuracy of FPLC method

HbA2

β-thalassaemia Normal

4-6 %< 4 %

350

0125

Total 35 125

Sensitivity = (35/35 x 100) = 100 per centSpecificity = (125/125 x 100) = 100 per centPositive predictive value = (35/35 x 100) = 100 per centNegative predictive value = (125/125 x100) = 100 per cent

cheap when compared with HPLC (about 4.20 US$/test). Also the cost of a Hi Trap sepharose column was lesser than HPLC column by about 10 times.

The haemoglobin chromatograms eluted from this anion-exchange column were converted with the conventional HPLC cation-exchange chromatography pattern. HbA

2 was eluted in the first chromatogram.

This pattern was very helpful in the screening of HbA

2/E fraction. The quantitative comparative

analysis demonstrated that the FPLC systematically produced HbA

2 value higher than those from HPLC

method. This may be due to the Hi Trap Sepharose, a weak anion exchange column chromatography. The speed of separation and the reproducibility of the HbA

2/E fraction obtained on the anion-exchange

column by this FPLC method were slightly lower than HPLC. After Hb separation, the column was washed by the buffer B to remove all Hb and protein residue out of column. Usually, in the sample with β-thalassaemia trait, the chromatogram was similar to that of the normal sample, but the concentration of HbA

2 was slightly increased. In HbE trait sample,

HbE was eluted at the same retention times as the HbA

2 found in normal samples or samples with the

β-thalassaemia trait, but the concentration of HbE was more than HbA

2 chromatogram. The baselines of the

chromatograms by FPLC method were quite irregular, and the processing of the calculation for the percentage of each chromatogram was done by manual operation. This may be a disadvantage of FPLC method.

In conclussion, the FPLC method appeared to be an appropriate method for identification and quantification of the β-thalassaemia, HbA

2/E, and could be a useful

screening and diagnostic tool in laboratories equipped with a FPLC analyzer.

Acknowledgment

Authors acknowledge the medical staff at the thalassaemia research unit and laboratory of the Naresuan University Hospital, Thailand, for blood sample collection and technical assistance, and the Faculty of Allied Health Sciences, Naresuan University for financial support. Authors thank Associate Professor Dr Mary Sarawit at the Naresuan International College and also Associate Professor Dr Timothy E O’Brien at Loyola University Chicago for their critical reading and correcting the manuscript.

References

1. Tangvarasittichai O, Sitthiworanan C, Dechgitvigrom W, Sanguansermsri T, Jeenapongsa R. Prevalence and heamatological parameters of thalassaemia in lower Northern Thailand. Haema 2005; 8 : 241-4.

2. Schmidt RM, Rucknagel DL, Necheles TF. Comparison of methodologies for thalassaemia screening by Hb A2 quantitation. J Lab Clin Med 1975; 86 : 873-82.

3. Efremov GD. An evaluation of the methods for quantitation of haemoglobin A

2: results from a survey of 10,663 cases.

Hemoglobin 1977; 1 : 845-60.

4. [No authors listed]. Recommendations for selected methods for quantitative estimation of Hb A

2 and for Hb A

2 reference

preparation. International Committee for Standardization in Haematology. Br J Haematol 1978; 38 : 573-8.

5. Marengo-Rowe AJ. Rapid electrophoresis and quantitation of haemoglobins on cellulose acetate. J Clin Pathol 1965; 18 : 790-2.

6. Chamrasratanakorn T, Sanguansermsri T, Punyakeaw A. A modified microcolumn chromatography for HbA2 determination. Bull Chiang Mai Assoc Med Sci 1998; 31 : 32-5.

7. van der Dijs FP, van den Berg GA, Schermer JG, Muskiet FD, Landman H, Muskiet FA. Screening cord blood for Haemoglobinopathies and thalassaemia by HPLC. Clin Chem 1992; 38 : 1864-9.

8. Riou J, Godart C, Hurtrel D, Mathis M, Bimet C, Bardakdjian-Michau J, et al. Cation-exchange HPLC evaluated for presumptive identification of Haemoglobin variants. Clin Chem 1997; 43 : 34 - 9.

TANGVARASITTICHAI et al: COMPARISON OF FPLC TECHNIQUES FOR β-THALASSAEMIA DIAGNOSIS 247

9. Shibasaki T, Gomi H, Ishimoto F, Miyahara T. Urinary N-acetyl-beta-D-glucosaminidase isoenzyme activity as measured by fast protein liquid chromatography in patients with nephrotic syndrome. Clin Chem 1990; 36 : 102-3.

10. Cooper EH, Turner R, Johns EA, Lindblom H, Britton VJ. Applications of fast protein liquid chromatography TM in the separation of plasma proteins in urine and cerebrospinal fluid. Clin Chem 1983; 29 : 1635-40.

11. Desai SN, Colah RB, Mohanty D. Comparison of FPLC with cellulose acetate electrophoresis for the diagnosis of beta-thalassaemia trait. Indian J Med Res 1998; 108 : 145-8.

12. Jeppsson JO, Jerntorp P, Sundkvist G, Englund H, Nylund V. Measurement of Haemoglobin A1c by a new liquid-chromatographic assay: methodology, clinical utility, and relation to glucose tolerance evaluated. Clin Chem 1986; 32 : 1867-72.

13. Marz W, Siekmeier R, Scharnagl H, Seiffert UB, Gross W. Fast lipoprotein chromatography: new method of analysis for plasma lipoproteins. Clin Chem1993; 39 : 2276-81.

14. Tangvarasittichai O, Tangvarasittichai S. An effective method for hemoglobin E detection: DEAE sepharose microcolumn. Lab Hematol 2009; 15 : 10-2.

Reprint requests: Dr Surapon Tangvarasittichai, Department of Medical Technology, Faculty of Allied Health SciencesNaresuan University, Phitsanulok 65000, Thailand

e-mail: [email protected]

248 INDIAN J MED RES, MARCH 2009