Apolipoprotein E genotyping by capillary electrophoretic analysis of restriction fragments.
Apolipoprotein (apo) E is an important constituent of several plasma lipoproteins, mainly VLDL and HDL and chylomycrons? It is involved in the redistribution of lipids in the liver and is implicated in growth and repair of injured neurons in the nervous system . Apo E has been associated with the risk of developing cardiovascular diseases and in familial type III hyperlipoproteinemia . More recently a strong association between apo E and Alzheimer disease has been demonstrated . Human apo E exists in three main isoforms ([epsilon]2, [epsilon]3, and [epsilon]4) related to two polymorphic sites at codons 112 and 158 of the gene located on chromosome 19 (Table 1). These isoforms arise from three alleles (E2,E3, and E4 respectively) combined in six different genotypes. The [epsilon]4 isoform has been associated with Alzheimer disease as a major risk factor. In particular in late-onset Alzheimer subjects with known familial occurrence, the E4 allele frequency is much higher than in age-matched individuals, whereas E2 is much lower . Furthermore the age of onset of the disease is related to the [epsilon] allele dose . Over 40 independent studies discussing apo E/Alzheimer association in different populations have been published so far with comparable results .
Alzheimer disease has a severe impact in the elderly population worldwide. Large studies to follow the elderly population with respect to this pathology are currently under way, as are clinical trials to test drugs with benefical effects on affected individuals. This makes the typing of apo E isoforms very important in Alzheimer studies. An interesting recent review focused on apo E relevance in laboratory medicine and reported the complete range of analytical techniques devoted to the apo E polymorphism determination . Phenotyping is usually performed by means of isoelectrofocusing techniques, but these present several drawbacks in their application to this specific analytical problem. Apo E genotyping has been developed to avoid such problems. Several recent papers dealt with the analysis of the apo E genotype by means of PCR amplification from genomic DNA . Most of the cited studies involved restriction fragment analysis of the amplified region encompassing codons 112 and 158 of the apo E gene [7, 8]. This procedure has been widely adopted, although it has some drawbacks particularly related to complex electrophoresic pattern due to partial enzymatic digestion and to the requirement of large quantities of amplified DNA . Modified procedures have been proposed to overcome such problems [9,10].
Here we present the identification of the apo E genotype by means of capillary electrophoresis analysis. A relatively viscous sieving media is required to achieve a good separation of the relatively small DNA fragments (48-91 bp) to be analyzed. High-performance capillary electrophoresis (HPCE) with cross-linked polyacrylamide gel offers the best performance in terms of resolution, and recently two groups [11,12] proposed such a technique for apo E genotyping. However, such a sieving matrix cannot be replaced. Several non-gel-sieving media, prepared from polymers, have been proposed to separate DNA molecules. These have a key feature in that they are replaceable after each separation, thus performing very reproducibly from run to run. By using methyl cellulose as a replaceable sieving matrix, the digestion fragments are sized in a very short time (10 min) with a commercial automated apparatus for capillary electrophoresis with UV on-line detection. We present results demonstrating the speed, sensitivity, and reliability of this analytical procedure.
Materials and Methods
The preparation of the sample for the determination of the apo E genotype followed a well-estabilished procedure [7,8] with minor modifications. Samples (178) were processed at different times. DNA extraction from whole blood was performed with a commercial kit based on chromatographic columns. This material was amplified, giving the expected DNA fragment. The amplified samples were checked for purity and quantity on a standard agarose gel (10 g/L) and digested with Cfol, an isoschizomer of Hhal.
DNA EXTRACTION AND APO E AMPLIFICATION
DNA was extracted and purified from 200 [micro]tL of whole blood, drawn from consenting individuals, with the blood DNA extraction kit (Qiagen, Chatsworth, CA). DNA was collected in 200 [micro]L of water (25-150 ng/L) and stored at 4'C. Oligonucleotides P1 and P2  used for the amplification of the polymorphic region were purchased from Genset (Paris, France), diluted to 20 [micro]mol/L, and stored at -20[degrees]C. The DNA amplification was performed in a thermal cycler (Perkin-Elmer, Norwalk, CT) in a total volume of 100 [micro]L. The amplification mixture contained Taq buffer diluted 1:10 as suggested by the manufacturer, deoxynucleotides (200 [micro]mol/L each), 40 pmol of each primer, and 8 [micro]L of purified genomic DNA. The mixture was overlaid with mineral oil and heated at 95 [degrees]C for 10 min. Taq polymerase (2.5 U; Finnzyme, Espoo, Finland) was then added and the samples were subjected to five cycles as follows: 95 [degrees]C for 1 min and 72 [degrees]C for 3 min and then to 30 cycles as follows: 95 [degrees]C for 1 min, 60 [degrees]C for 1 min, and 72 [degrees]C for 1 min. Samples were then incubated at 72 [degrees]C for 10 min and stored at 4 [degrees]C.
DNA DIGESTION AND PURIFICATION
After the amplification, 40 U of Cfol (Boehringer Mannheim, Mannheim, Germany) were added and the samples were digested overnight at 37 [degrees]C. Before digestion, the sample can be purified by means of Microspin columns (S200; Pharmacia, Uppsala, Sweden) to remove nucleotides, primers, and PCR buffer. The digested sample was then isopropanol-precipitated or ultrafiltered (we recommend this procedure) by means of MC 5000 columns (Millipore, Bedford, MA), washing two times with 400 [micro]L of water. The resulting solution was adjusted to 20 [micro]L with water.
CAPILLARY ELECTROPHORESIS ANALYSIS
Analysis was performed on a BioFocus 3000 capillary electrophoresis system (Bio-Rad, Hercules, CA) by means of a standard cartridge containing a 50 cm X 50 [micro]m coated capillary tube thermostated at 20 [degrees]C. The capillary was washed with TBE 1 x (89 mmol/L Tris base, 89 mmol/L boric acid, and 1 mmol/L EDTA disodium salt) and filled with a 12 g/L solution of Methocel (Methylcellulose high density; Fluka, Buchs, Switzerland) in TBE 1 x. This solution was prepared as follows: 1 g of Methocel was slowly added to water (50 mL) at 90 [degrees]C. The solution was allowed to cool at room temperature under constant stirring for 30 min and then cooled at 4 [degrees]C in a refrigerator, continuing the stirring. After 2 h it was centrifuged, filtered through a 0.45-[micro]m filter, and stored at 4 [degrees]C. The solution was mixed with TBE 10x, and adjusted with water to a final concentration of 12 g/L in TBE 1x.
Samples were electroinjected at 10 kV for 5 s. Restriction fragments were separated during a 10-min run at 15 kV and detected at 260 nm.
The digestion of the amplified region encompassing the two polymorphic sites of interest generates a number of fragments. Fig. 1A shows the restriction map for this region with respect to the sequence of the three most common alleles. The expected electrophoretic pattern for the six resulting genotypes is presented in Fig. 1B. Only the DNA fragments ranging from 48 to 91 by are diagnostic for the genotyping of this locus.
After some trials we selected methylcellulose (high viscosity) as the replaceable sieving media to be used in capillary electrophoresis analysis of these relatively short DNA fragments. We optimized the separation variables by means of a DNA size marker prepared by Cfol digestion of a commercial preparation of pUC18 (fragments ranging from 393 to 21 bp). This DNA ladder was analyzed by changing several variables (applied voltage, temperature, polymer concentration, sample injection conditions, capillary tube length, diameter, and coating). A standard 50 cm x 50 [micro]m coated capillary column (Bio-Rad) was found to be appropriate for high resolution and speed of analysis. Alternatively, the analysis was performed on a 50 cm X 50 [micro]Lm DB-1 coated capillary column (J&W Scientific, Folsom, CA). As shown in Fig. 2, the diagnostic 48-,63-,72-,82-, and 91-bp bands are well resolved in <10 min. All six common genotypes can be easily discriminated. The run-to-run reproducibility was tested by injecting the same sample 30 times, refilling the capillary tube each time. The CV for the retention time is 2.2%. CV for peak area is 21.7%. With an external calibrafor (bromophenol blue), the CV for normalized peak area is 5.6%. The filling of the capillary takes 5 min (50 cm x 50 [micro]m column) and therefore the reuse of the sieving matrix would be desirable to speed up the analysis. We tested this possibility by injecting the same sample several times and performing the analysis as explained without sieving matrix reneval. The first six injections gave a CV for the retention time within that reported, indicating that multiple use of the matrix is feasible.
[FIGURE 1 OMITTED]
The reliability of apo E genotyping by restriction digestion and polyacrylamide gel electrophoresis (PAGE) has been recently discussed. Some important drawbacks can affect the analysis. Because of partial digestion, "ghost" bands appear in PAGE, biasing the correct genotyping. Bands having different intensities have been indicated as a possible source of problems during ethidium bromide/ PAGE analysis . In addition, relevant quantities of amplified DNA should be loaded on the gel because of the intrinsic low sensitivity of ethidium bromide staining with small DNA fragments. Appel et al.  recently proposed an alternative procedure requiring radioactive labeling. In an attempt to overcome such problems we used HPCE for the analysis of the restriction fragments generated by HhaI (CfoI) digestion. This technique was recently proposed by Baba et al.  and Schlenck et al. X12] for apo E genotyping. Both used a cross-linked polyacrylamide gel as separation medium, thus gaining very high resolution . UV-  or laser-induced fluorescence  was used as detection system.
The resolution we obtained with methylcellulose was lower than that gained by polyacrylamide , although it is sufficient to discriminate all six common apo E genotypes. However, such a sieving matrix yielded very reproducible results because of the possibility of its renewal after runs, whereas polyacrylamide capillary columns decrease their performance during their lifetime.
We used UV detection, gaining good sensitivity. PCR usually gave 400 ng of amplified DNA that was precipitated, digested, desalted, and electroinjected several (1030) times, yielding reproducible electropherograms similar to those shown in Fig. 2. This was achieved without the need for an expensive laser apparatus for fluorescence detection, although this grants a much higher sensitivity X12]. We observed a large difference comparing the signal intensity we obtained with that shown in ref. 11. This is probably due to the accurate desalting of digested samples performed before electroinjection. This allows an on-line concentration, particularly for smaller fragments.
HPCE, giving a quantitative estimate of each band by UV on-line detection, helps in preventing genotyping errors due to partial digestion. In fact, peak integration allows the comparison of the relative intensities of the bands on a numerical basis. Although the peak intensity ratio can vary because of the inherent variability of electrokinetic injection, the presence of "ghost" bands, like that in Fig. 2F (the shoulder close to the 63-bp fragment), can be easily detected and correctly classified. Samples that are difficult to discriminate, because of partial digestion, can be directly redigested and analyzed because of the extremely small sample quantities required by the capillary electrophoresis analysis. Sample loading by electroinjection has proven to be very efficient, although it requires a preliminary optimization with respect to the actual concentration of the sample. The precipitation performed after digestion yields very concentrated samples that must be injected for a very short time to avoid overload and loss of resolution. Despite this fact, it is very simple to determine the genotype in very faint samples by means of longer injection time, thus overcoming the sensitivity problem. The injection of the same sample 30 times demonstrates the sensitivity of this approach and its reliability (low CV for retention time). The speed of the analysis can be further increased by shortening the capillary tube. In a 20-cm x 50-[micro]m tube, acceptable separation was gained in <3 min (data not shown). Preparation of the sieving matrix has proven to be critical to the final result. The reported procedure must be followed carefully as explained to yield consistent results. The results achieved with this method in our laboratory, during a large-scale apo E genotype study, are extremely positive. On some occasions, in partially digested samples, genotyping by PAGE and by HPCE were in disagreement, but after further digestion, the genotyping by HPCE was confirmed.
[FIGURE 2 OMITTED]
In conclusion, this procedure could be of interest to clinical laboratories involved in large-scale apo E genotype analysis , thanks to the high degree of automation attainable with HPCE combined with the speed, sensitivity, and reliability of the proposed analytical protocol.
We gratefully acknowledge Telethon (grant E213) for partial financial support.
Received (revised) March 11, 1997; accepted March 17, 1997.
[1.] Mahley RW, Innerarity TL, Rail SC, Weisgraber KH, Taylor. Apolipoprotein E: cholesterol transport protein with expanding role in cell biology. Science 1988;240:622-30.
[2.] Breslow JL, McPherson J, Sanciacomo TR, Third JLHC, Tracy T, Glueck CJ. Studies of familial type III hyperlipoproteinemia using as a genetic marker the apoE phenotype E2/2. J Lipid Res 1982;23:1224-35.
[3.] Corder EH, Saunders AM, Strittmatter WJ, Schemechel DE, Gaskell PC, Small GW, et al. Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer's disease in late onset families. Science 1993;261:921-3.
[4.] Corder EH, Saunders AM, Risch NJ, Strittmatter WJ, Schemechel DE, Gaskell PC, et al. Protective effect of apolipoprotein E type 2 allele decreases risk of late onset of Alzheimer's disease. Nature 1994;7:180-4.
[5.] Strittmatter WJ, Roses AD. Apolipoprotein E and Alzheimer disease. Proc Natl Acad Sci U S A 1995;92:4725-7.
[6.] Siest G, Pillot T, Regis-Bailly A, Leininger-Muller B, Steinmetz J, Galteau MM, Visvikis S. Apolipoprotein E: an important gene and protein to follow in laboratory medicine [Review]. Clin Chem 1995;41:1068-86.
[7.] Kontula K, Aalto-Setala K, Kuusi T, Hamalainen L, Syvanen A. Apolipoprotein E polymorphism determined by restriction enzyme analysis of DNA amplified by polymerase chain reaction: convenient alternative to phenotyping by isoelectric focusing. Clin Chem 1990;36:2087-92.
[8.] Hixson JE, Vernier DT. Restriction isotyping of human apolipoprotein E by gene amplification and cleavage with Hhal. J Lipid Res 1990;31:545-8.
[9.] Appel E, Eisenberg S, Roitelman J. Improved PCR amplification/ Hhal restriction for unambiguous determination of apolipoprotein E alleles. Clin Chem 1995;41:187-90.
[10.] Reymer WA, Gronemeyer BE, van de Burg R, Kastelein JJP. Apolipoprotein E genotyping on agarose gels [Tech Brief]. Clin Chem 1995;41:1046-7.
[11.] Baba Y, Tomisaki R, Sumita C, Tsuhako M, Miki T, Ogihara T. High resolution separation of PCR product and gene diagnosis by capillary electrophoresis. Biomed Chromatogr 1994;8:291-3.
[12.] Schlenck A, Visvikis S, O'Kane M, Siest G. High resolution separation of PCR product and gene diagnosis by capillary electrophoresis. Biomed Chromatogr 1996;10:48-50.
(1) Consiglio Nazionale delle Ricerche, Istituto di Tecnologie Biomediche Avanzate, Via Fratelli Cervi 93, 20090 Segrate, Italy.
(2) Universita degli Studi di Milano, Diparfimento di Scienze e Tecnologie Biomediche, Via Fratelli Cervi 93, 20090 Segrate, Italy.
*Author for correspondence. Fax 39-2-26422770; e-mail debellis@ itba.rni.cnr.it.
(3) Nonstandard abbreviations: ape, apolipoprotein; HPCE, high-performance capillary electrophoresis; TBE, Tris-boric acid-EDTA buffer; and PAGE, polyacrylamide gel electrophoresis.
GIANLUCA DE BELLIS, (1) * GIULIANA SALANI, (1) SILVIA PANIGONE, (1) FERRUCCIO BETTI, (1) LUIGIA INVERNIZZI, (1) and MASSIMO LUZZANA (2)
Table 1. Correspondence among apo E protein isoforms, amino acid and codon composition at polymorphic sites, and genotypes. Isoform Position 112 Position 158 Allele E2 Cys (TGC) Cys (TGC) [epsilon] 2 E3 Cys (TGC) Arg (CGC) [epsilon] 3 E4 Arg (CGC) Arg (CGC) [epsilon] 4
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Molecular Pathology|
|Author:||de Bellis, Gianluca; Salani, Giuliana; Panigone, Silvia; Betti, Ferruccio; Invernizzi, Luigia; Luzza|
|Date:||Aug 1, 1997|
|Previous Article:||Extraction of glyceric and glycolic acids from urine with tetrahydrofuran: utility in detection of primary hyperoxaluria.|
|Next Article:||Diagnosis of lysosomal storage disorders: evaluation of lysosome-associated membrane protein LAMP-1 as a diagnostic marker.|