Research Reports Reuse of Denaturing Polyacrylamide Gels for Short Tandem Repeat Analysis BioTechniques 25:892-897 (November 1998) Allan Tereba, Katherine A. Micka and James W. Schumm Promega Corporation, Madison, WI, USA ABSTRACT Denaturing polyacrylamide gel electrophoretic analysis of amplified polymorphic short tandem repeat (STR) loci using fluorescent markers is a mainstay of forensic and paternity testing. To reduce the drawback of preparing gels or using expensive precast gels, we have developed a simple and rapid method to reuse gels between 2 and 8 times over a period of several days. Following the initial electrophoresis and scan, the original samples are removed from the gel by a 1 1.5-h reverse-electrophoresis step. This step heats the gel for the next set of samples and can be performed several days after the initial electrophoresis. Sample bands remain sharp on subsequent runs, but edge effects (frowning of the outside lanes) become progressively worse and ultimately limit gel reuse. Well distortions and separation of the gel from the plates become problems if the gel is used more than twice. However, degassing the gel solution and bonding the gel to both plates eliminate these problems. Precast gels also can be used multiple times. Using this technique, we have successfully analyzed samples amplified with a nine-locus multiplex system and characterized the separated products using a fluorescent scanner and software. INTRODUCTION Denaturing polyacrylamide gels are an integral part of analyzing relatively small DNA fragments, especially for sequencing (10,14) or size analysis following amplification by the polymerase chain reaction (PCR) (5,15). The discovery of polymorphic short tandem repeats (STR) loci containing dinucleotide tandem repeats (18) and, more recently, trimeric and tetrameric tandem repeats (4,11), provided the resources for using PCR amplification in genotyping DNA samples. The current prominence of this approach in forensic and paternity testing will result in its use for developing national and state databases in both the United States (3) and Europe (7,8). Two methods have been developed to reuse gels for sequence analysis. (i) Taking advantage of the apparently greater stability of ultrathin gels (1), Ansorge et al. (2) demonstrated that 0.1-mm gels could be reused by reloading samples following the first electrophoresis. (ii) Swerdlow et al. (17) used expensive formamide-containing buffer originally developed to stabilize capillary gels (16). In contrast, Robertson et al. (13) were able to use standard sequencing gels twice when analyzing the smaller STR loci. Although these reports showed the feasibility of reusing gels, these approaches have not been widely pursued because of cost or chronic technical difficulties. This report demonstrates the successful multiple reuse of slab gels for STR analysis in conjunction with a fluorescent scanner separate from the electrophoresis apparatus. The shorter electrophoresis time, typically 60 min, and lower wattage compared with previously mentioned methods help to preserve the gel s integrity. Reuse of polyacrylamide gels with this simple procedure will significantly reduce supply costs and time spent in preparing gels and cleaning glass plates. Precast gels also have been shown to be compatible with this technique, resulting in savings of 75% or more in gel costs. MATERIALS AND METHODS Plate Preparation Both plates were treated with γ- methacryloxypropyltrimethoxysilane (9 µl in 3 ml 0.5% acetic acid in 95% ethanol; Reference 9). Excess solution was removed with four washes of ethanol (squirt bottle and Kimwipes ). When only one or two runs were performed, only the top portion of the short plate was treated. Acrylamide Gel Solution Preparation An acrylamide stock solution containing 4% acrylamide:bis (19:1), 7 M urea, 0.5 TBE (44.5 mm Tris base, 1 mm EDTA, 44.5 mm boric acid, ph 8.3) was prepared and stored at 4 C. Before pouring gels, 40 ml of stock acrylamide mixture were warmed to room temperature and then degassed using a water aspirator for 5 min with occasional swirling. Just before pouring the gel, 0.2 ml of 10% ammonium persulfate was added and mixed by gentle swirling. When only two runs were to be performed, the degassing 892 BioTechniques Vol. 25, No. 5 (1998)
Table 1. Relative Bandwidth as a Function of Run Number and Place on Gel step was omitted, and the ammonium persulfate was increased to 0.25 ml. Gels with dimensions of 17 43 0.4 cm were prepared with 34- or 30-well, flat-bottom combs. Shark-tooth combs have also be used successfully. Pre-run The wells were rinsed with 0.5 TBE running buffer, and the gel was pre-run on a Model SA-32 Electrophoresis System (Life Technologies, Gaithersburg, MD, USA) with an EC600 Power Supply (E-C Apparatus, St. Petersburg, FL, USA) at 50 55 W for 30 45 min to attain a plate temperature of 45 50 C. See conditions for precast gels below. Sample Preparation From 0.2 2 ng DNA were amplified on a GeneAmp PCR System 9600 (PE Applied Biosystems, Foster City, CA, USA) using the GenePrint PowerPlex 1.1 System and GenePrint Amelogenin (TMR) Fluorescent Sex Identification System (both from Promega, Madison, WI, USA). Three microliters of Bromophenol Blue Loading Solution (95% formamide, 0.05% bromophenol blue, 10 mm NaOH) and 1 µl of GenePrint Fluorescent Ladder Run Number Bases 1 2 3 4 X a SD b Lane 3 400 2.56 2.29 2.27 2.76 2.47 0.24 300 2.59 2.39 2.18 2.98 2.53 0.34 200 3.30 3.02 2.42 3.00 2.94 0.37 100 2.92 2.55 2.69 3.10 2.81 0.24 Lane 9 400 2.25 2.31 2.30 2.10 2.24 0.10 300 2.17 2.37 2.00 1.96 2.12 0.19 200 2.54 3.01 2.56 2.51 2.66 0.23 100 2.64 2.59 2.77 2.64 2.66 0.08 Lane 17 400 2.28 2.09 2.14 2.37 2.22 0.13 300 2.36 2.32 2.09 2.16 2.23 0.13 200 2.88 3.12 2.93 2.67 2.90 0.19 100 2.92 2.53 2.36 2.39 2.55 0.26 a x - mean or average bandwidths. b SD Standard deviation in bases. (CXR), 60 400 bases (Promega) were mixed with 2 µl of amplified DNA solution, allelic ladder or 1 STR amplification buffer (50 mm KCl, 10 mm Tris-HCl, ph 9.0, 1.5 mm MgCl 2, 0.1% Triton X-100 and 0.2 mm of each dntp), heated to 95 C for 2 min and quick-cooled. Gel Electrophoresis After the pre-run, the wells were rinsed with buffer, and 2.5 3 µl of sample (depending on comb size) were loaded per lane. Electrophoresis continued approximately 60 min. In subsequent runs, migration of the samples was slower, requiring an extra 5 10 min of electrophoresis time to be added after 4 runs. Scanning and Band Analysis Gels were scanned on a FMBIO II Fluorescent Scanner (Hitachi Software Engineering America, South San Francisco, CA, USA) with 505-, 585- and 650-nm wavelength filters for fluorescein, TMR- and CXR-labeled primers, respectively. FMBIO Analysis Software Version 6.0f11 (Hitachi Software Engineering America) was used to analyze the samples along with STaR- Call Allele Calling Software Version 2.0b1 (Hitachi Software Engineering America). Gray-level corrections were optimized for the first run, and the same values were used for analyzing all subsequent runs. Bandwidth was calculated by dividing the calculated volume of each band by its peak fluorescent intensity. Removal of Original Samples From Gel and Second Pre-run and Run The samples from the first run were removed from the gel, and the gel was reheated using reverse-electrophoresis (with electrodes reversed, positive electrode at the top of the gel) at the same wattage for 15 min longer than the first forward run. New running buffer in both reservoirs was used for the reverse electrophoresis. Immediately following the reverse-electrophoresis, (i) the wells were flushed with buffer, (ii) the second set of samples was applied, (iii) the electrodes were set to the original position (with the positive electrode at the bottom) and (iv) the second forward run was performed. Buffer was not replaced at this point. If another run was not planned for the same day, the gel was stored as described below after scanning. The same electrophoresis apparatus was used for both the forward and reverse electrophoresis steps to ensure uniform electrophoretic conditions. Gel Storage Gels were stored by placing 3 5 ml of 0.5 TBE containing 7 M urea in the wells and then wrapping the top and bottom of the gel with paper towels wetted with water. The towels were then wrapped with plastic wrap and stored at room temperature for up to 20 days. Cleaning of Gel-Bonded Plates Plates that had been bonded to the gel on both sides were separated using a spatula and then soaked in 10% NaOH for at least 1 h. The gel could be easily washed off with a stream of water. The plates were washed with 1% Liquinox (Alconox, New York, NY, USA), rinsed with distilled water and dried. Precast Gels R 3 Precast Gels (Hitachi Software Vol. 25, No. 5 (1998) BioTechniques 893
Research Reports Engineering America) prepared with 32 wells and containing 4.5% acrylamide, 7 M urea and 1 TBE buffer were subjected to electrophoresis on a Sequencer Electrophoretic Apparatus (Stratagene, La Jolla, CA, USA) with 1 TBE buffer for 70 min at 40 W giving a plate temperature of 50 C. Samples were prepared and scanned as described. Reverse-electrophoresis was performed for 85 min. RESULTS Reuse of Polyacrylamide Gels A viable procedure for reuse of gels must rapidly and completely remove the previous samples from the gel, keep wells intact, control gel distortions that interfere with data analysis and maintain a predictable sample migration pattern. Electrophoresis with the electrodes reversed (positive electrode at the top) for 15 min longer than the forwardelectrophoresis rapidly removed samples from a used gel. This provides a uniform procedure independent of DNA fragment size. Scans of resulting gels have consistently shown complete sample removal, whether subjected to reverse-electrophoresis immediately after the forward run or stored for 2 weeks (data not shown). Reverse-electrophoresis also serves as a pre-run to heat the gel before loading new samples. Well distortions and separation of the gel from the plates were observed when gels that were made without any special preparation were used more than twice. A simple and effective solution to both of these problems was to degas the acrylamide gel solution for 5 min before gel pouring and to bond the gel to both glass plates. Flat bands, indicating well integrity and lack of bubbles, were observed in the scans of all runs shown in Figure 1. Predictable and preferably uniform migration of samples across the gel allows software programs to characterize individual bands reliably. Sample migration patterns frequently show edge effects on the first run of acrylamide gels, resulting in smiling (see Figure 1, run 1). Heat transfer plates and lower wattage are sometimes used to reduce this effect; however, the easiest solution is not using the outer 2 or 3 lanes. To assess edge effects on reused gels, a gel prepared with outer lanes 9 mm from the side spacers was reused a total of six times. Figure 1 shows the 60 400-base ladder pattern of the samples applied in all 34 wells. Run 2 shows a pattern that is straighter than the first run. This uniform migration is seen through run 4 with the exception of the outer two lanes on each side, which now appear to frown. Based on comparisons of migration distances between runs, this differential migration is actually because of slower migration of samples in the inner lanes. By run 6, the frowning of the outer 2 4 lanes, especially on the right side, is sufficient to interfere with software analysis of the bands. This migration pattern is very predictable from gel to gel (data not shown) and still leaves 26 usable wells for the sixth run. Bandwidth Remains Constant For gel reuse to be useful, the resolution of the gel must be maintained from run to run. As a first approximation of resolution, relative bandwidths were calculated as band volume divid- Figure 1. Gels can be reused multiple times. A 43-cm gel with a 34-well comb was loaded in each well with a 60 400-base fluorescent ladder, subjected to electrophoresis, scanned and subjected to reverse-electrophoresis a total of 6 times over a 4-day period. 894 BioTechniques Vol. 25, No. 5 (1998)
Table 2. SD of Band Size with Run Number SD in Bases Locus Run 1 Run 2 Run 3 Run 4 CSF1PO 0.09 0.11 0.16 0.11 D16S539 0.11 0.06 0.12 0.11 TPOX 0.07 0.08 0.07 0.08 D7S820 0.07 0.08 0.08 0.07 D13S317 0.05 0.05 0.07 0.06 THO1 0.06 0.07 0.07 0.06 vwa 0.03 0.04 0.05 0.05 D5S818 0.03 0.04 0.07 0.03 ed by peak intensity. Sharper bands will have a smaller value than broad bands. Table 1 presents the major band values for lanes 3, 9 and 17, representing lanes across the gel, for runs 1, 2, 4 and 6 of the experiment presented in Figure 1. Lanes 9 and 17, which are not within the significant smiling and frowning" regions, show consistent bandwidth from run to run. Lane 3 appears to have slightly sharper peaks in runs 2 and 4, which might reflect the lack of significant curvature in these runs compared with runs 1 and 6. When this analysis was performed on lanes 1 and 2, significant deviations occurred in runs 4 and 6, respectively. As expected, in all cases, bandwidth increased about 15% between 400 and 100 bases. Analysis of the minor bands of the ladder showed similar results with bandwidth being somewhat sharper but with larger standard deviations (SD). In addition, peak intensity appeared to decrease about 25% between runs 1 and 6. Multiplex STR Analysis To demonstrate that gel reuse can provide reliable data in a practical appli- Figure 2. Reuse of a gel for multiplex STR analysis.during a 3-day period, a single gel was used to analyze two DNA samples amplified under a variety of conditions with a nine-locus multiplex system (runs 1 and 3 and runs 2 and 4, respectively). The blue bands are the fluorescent 60 400-base ladder, and the red and green bands are the nine loci. The allelic ladders, containing each of the known alleles for all nine loci, flank each set of four DNA samples amplified at 2, 1, 0.5 and 0.2 ng. The last set in gels 1 and 3 are negative controls. Vol. 25, No. 5 (1998) BioTechniques 895
Research Reports cation and to compare band resolution with gel run number, an STR system containing nine primer pairs was used to amplify two DNA samples under a variety of conditions. Sample 1 (gel runs 1 and 3) and sample 2 (gel runs 2 and 4) were amplified at various temperatures and concentrations of DNA and Taq DNA polymerase. This resulted in a wide variation in the intensity of the bands upon electrophoresis followed by scanning. The results (Figure 2) confirm and extend the findings already demonstrated with the 60 400-base ladder (Figure 1). Runs 2 and 3 showed a band pattern that was as straight or straighter than run 1, while run 4 began to frown, especially on the right side. There was no evidence of contamination from the previous runs even when some of the lanes were overloaded. When the bands were analyzed using STaRCall Allele Calling Software with range limits of ±1 base (except for THO1 allele 9.3, which had a range limit of ±0.5 bases), all alleles in all gels were correctly identified. Table 2 shows the SD in the calculated base number with all alleles of each locus being grouped and listed according to descending size. The data indicate no significant loss in band resolution with gels run at least 4 times. The deviation is less for the small more rapidly migrating loci. However, even in run 4, the THO1 allele 9.3 could be easily distinguished from the 1-base larger allele 10 (12). These results not only provide practical evidence on the reliability of reusing gels for STR analysis but also indicate that sequencing analysis would not be compromised by reuse of gels. Use of Precast Gels Although the above technique reduces the number of gels that have to be prepared, some laboratories might still prefer to use commercially available precast gels to meet quality-control requirements or to reduce labor-intensive steps. We evaluated individual R 3 Precast Gels over 8 runs spanning 4 days using a 60 400-base ladder. Figure 3 shows runs 2, 4, 6 and 8. The precast gels were somewhat more variable, possibly because of the thinner plates and gel, and frequently slanted on the first run. However, the progressive frowning of the outer lanes observed with the homemade gels was less apparent with the precast gels and occurred primarily on the left side. The precast gels were not as robust in some ways as homemade gels bonded to both plates. First, since they are bonded to only one plate, they showed separation on the edges during the third run, which progressively worsened near the top and bottom on the fourth and subsequent runs (data not shown). This gel separation did not affect the migration of the bands or sample analysis. Second, the wells were less robust than wells from gels bonded to both plates. For example, the material separating wells 14 and 15 of the precast gel broke while flushing before loading the third set of samples. This prevented the use of these wells in subsequent runs. DISCUSSION With the growing worldwide interest in developing databases that contain the genotypes of the criminal population, there is an increased need to streamline and reduce costs of analyses. By demonstrating that gels can be reused with little modification to existing techniques, this study provides one approach to reduce supply and/or personnel costs by over 75% and in addition reduce the danger of working with acrylamide solutions. Although gels can be reused several times, they do progressively degrade on the edges. This degradation is related to Figure 3. Reuse of precast gels. An R 3 Precast, 32-well gel was loaded in every other lane with 60 400-base fluorescent ladder, subjected to electrophoresis, scanned and subjected to reverse-electrophoresis a total of 8 times over a 4-day period. On alternate runs, the samples were shifted one lane to investigate potential contamination. Wells 14 and 15 were not used after run 3 because the spacer between these two wells broke during rinsing before run 3. 896 BioTechniques Vol. 25, No. 5 (1998)
the temperature of the gel during the runs and the overall time of electrophoresis. Reducing running temperatures to 45 from 50 C extends the life of the gel, reduces "smiling" and "frowning" and has not affected gel analysis of the systems used. The shorter running times of the gels and lower wattage used compared with STR loci analyzed on sequencers could also account for the success of this application with fluorescent scanners despite its limited success with sequencers. For samples that contain only large DNA fragments, longer electrophoretic runs can be used to obtain better separation. In these cases, running the samples out of the gel in the same direction is quicker and has been used successfully for at least five runs. However, the bands in the second run are not usually as straight as in the first run and there is the possibility that slowly migrating bands in the first run will be confused with those of the second run. With forensic and paternity samples, it is essential that contamination between sample runs does not occur. Theoretically, when reusing a gel, contamination can occur either by eluted DNA reentering the gel from the upper reservoir or by not completely eluting the original material from the gel. Currently, we do not change buffer following the reverse run because the eluted DNA is diluted at least 100000-fold. For added precaution, the upper-chamber buffer could be changed before sample loading. The more likely source of contamination is the failure of the first set of samples to migrate out of the gel. When this happens, a double image is observed with the first entire set of samples (including complete allelic and base ladders) migrating farther than the second set. Following the precautions stated in the Materials and Methods section (such as reverse runs 15 min longer than the forward run) lanes have always been free of previously run products. We are currently exploring the use of a visible marker to demonstrate that reverse-electrophoresis has eliminated the previous samples. Finally, although we have not looked at reusing gels for DNA sequence analysis, we expect this approach would work well. Scanning sequencing gels that have been loaded with fluorescently labeled sequencing reactions can give excellent results in a short time for DNA lengths up to 300 bases (6). Since the resolution of the gel does not appear to degrade with reuse, this would provide an attractive approach to reducing costs for sequencing short cloned inserts. ACKNOWLEDGMENTS This work was supported by Promega Corporation. We thank Hitachi Software for providing the R 3 Precast Gels. REFERENCES 1.Ansorge, W. and R. Barker. 1984. System for DNA sequencing with resolution of up to 600 base pairs. J. Biochem. Biophys. Methods 9:33-47. 2.Ansorge, W., H. Voss, S. Wiemann, C. Schwager, B. Sproat, J. Zimmermann, J. Stegemann, H. Erfle, N. Hewett and T. Rupp. 1992. High-throughput automated DNA sequencing facility with fluorescent labels at the European Molecular Biology Laboratory. Electrophoresis 13:616-619. 3.Budowle, B., B.W. Koons, K.M. Keys and J.B. Smerick. 1996. Methods for typing the STR triplex CSF1PO, TPOX, and HUMTH01 that enable compatibility among DNA typing laboratories. Adv. Forensic Haemogenet. 6:107-114. 4.Edwards, A., A. Civitello, H.A. Hammond and C.T. Caskey. 1991. DNA typing and genetic mapping with trimeric and tetrameric tandem repeats. Am. J. Hum. Genet. 49:746-756. 5.Frazier, R.R., E.S. Millican, S.K. Watson, N.J. Oldroyd, R.L Sparkles, K.M. Taylor, S. Panchal, L. Bark, C.P. Kimpton and P.D. Gill. 1996. Validation of the Applied Biosystems Prism 377 Automated Sequencer for the forensic short tandem repeat analysis. Electrophoresis 17:1550-1552. 6.Ishino, Y., J. Mineno, T. Inoue, H. Fujimiya, K. Yamamoto, T. Tamura, M. Homma, K. Tanaka and I. Kato. 1992. Practical applications in molecular biology of sensitive fluorescence detection by a laser-excited fluorescence image analyzer. BioTechniques 13:936-943. 7.Kimpton, C.P., P. Gill, E. d Aloja, J.F. Andersen, W. Bar, S. Holgersson, S. Jacobsen, V. Johnsson et al.1995. Report on the second collaborative STR exercise. Forensic Sci. Int. 71:137-152. 8.Kimpton, C.P., N.J. Oldroyd, S.K. Watson, R.R.E. Frazier, P.E. Johnson, E.S. Millican, A. Urquhart, B.L. Sparkes and P. Gill. 1996. Validation of highly discriminating short tandem repeat amplification systems for individual identification. Electrophoresis 17:1283-1293. 9.Kobayashi, Y. 1988. A method to cast thin sequencing gels. Focus 10:73-74. 10.Maniatis, T. and A. Efstratiadis. 1980. Fractionation of low molecular weight DNA or RNA in polyacrylamide gels containing 98% formamide or 7 M urea, p. 299-305. In L. Grossman and K. Moldave (Eds.) Methods in Enzymology, Vol. 65. Academic Press, New York. 11.Polymeropoulos, M., D.S. Rath, H. Xiao and C.R. Merril. 1992. Tetranucleotide repeat polymorphism at the human beta-actin related pseudogene H-beta-Ac-psi-2 (ACTBP2). Nucleic Acids Res. 20:1432. 12.Puers, C., H.A. Hammond, L. Jin, C.T. Caskey and J.W. Schumm. 1993. Identification of repeat sequence heterogeneity at the polymorphic short tandem repeat locus HUMTHO1 [AATG]n and reassignment of alleles in population analysis by using a locusspecific allelic ladder. Am. J. Hum. Genet. 53:953-958. 13.Robertson, J.M., J.B. Sgueglia, C.A. Badger, A.Ç. Juston and J. Ballantyne. 1995. Forensic applications of a rapid, sensitive, and precise multiplex analysis of the four short tandem repeat loci HUMVWF31/A, HUMTH01, HUMF13A, and HUMFES/FPS. Electrophoresis 16:1568-1576. 14.Sambrook, J., E.F. Fritsch and T. Maniatis. 1989. DNA sequencing, p.13.45-13.58. Molecular Cloning: A Laboratory Manual, 2nd ed. CSH Laboratory Press, Cold Spring Harbor, NY. 15.Schumm, J.W., A.M. Lins, C.J. Sprecher and K.A. Micka.1996. High throughput systems for analysis of STR loci, p. 10 19. In Proceedings from the Fifth International Symposium on Human Identification 1995. Promega Corporation, Madison, WI. 16.Swerdlow, H., K.E. Dew-Jager, K. Brady, R. Grey, N.J. Dovichi and R. Gesteland. 1992. Stability of capillary gels for automated sequencing of DNA. Electrophoresis 13:475-483. 17.Swerdlow, H., K. Dew-Jager and R.F. Gesteland. 1994. Reloading and stability of polyacrylamide slab gels for automated DNA sequencing. BioTechniques 16:684-693. 18.Weber, M.L. and P.E. May. 1989. Abundant class of human DNA polymorphisms which can be typed using the polymerase chain reaction. Am. J. Hum. Genet. 44:388-396. Received 15 January 1998; accepted 13 May 1998. Address correspondence to: Dr. Allan Tereba Promega Corporation 2800 Woods Hollow Rd. Madison, WI 53711, USA Internet: atereba@promega.com Vol. 25, No. 5 (1998) BioTechniques 897