It is now clear that DNA sequence variation will play an important role in human health and disease prevention for the foreseeable future, as recently discussed in The New England Journal of Medicine . Our DNA sequence is also a record of our ancestry; that is, who are we related to and where our ancestors came from. This information, and possibly, our talents and behavioral traits as well are encoded in our DNA. Therefore, it will be important for students to learn about how information is encoded in DNA, and how the information in a region of DNA of interest can be read out. First, we would like to briefly explain exactly what the term “DNA sequence variation” means. The human DNA sequence is very similar among all individuals. On average, there is only one difference every approximately 800 nucleotides when the sequences from two randomly selected individuals are compared. Usually an individual has one of two possibilities for the four letters at a particular place in the DNA, and these differences occur at the same places in the DNA for everyone. Moreover, we now know where nearly all of these differences are located. These single nucleotide differences among individuals make up approximately 80% of DNA sequence variation, and they are referred to as single nucleotide polymorphisms (SNPs) . Additionally, approximately 20% of human variation is due to sequence insertions or deletions (indels) of various lengths . In this work we will be concerned with SNPs.
As a typical example of a SNP, a small portion of human chromosome 16 containing 1,200 nucleotides is shown in Fig. 1. This DNA sequence is the same in practically everyone, except for the single nucleotide “c” encircled and colored green. In individuals of European or African descent there is usually a “c,” as shown, whereas in individuals of Asian descent there is usually a “t” at this position. However, we have two copies of each of the non-sex chromosomes, and many individuals are heterozygotes, having inherited a “c” form one parent and a “t” from the other. More than ten million such SNPs have now been found. Some of these variants are known to be associated with particular human traits or conditions. Information about SNPs is freely available from SNPedia  and from the National Center for Biotechnology Information (NCBI)  over the internet.
Powerful technologies are available to determine which SNP alleles an individual has. Obtaining this information is what is meant by the term genotyping. Although these technologies were very important for determining exactly how the information encoded in our DNA influences our health, the equipment required is very expensive for its initial purchase and for its continued operation, generally precluding the purchasing of DNA sequencing machines and microarray technologies for student laboratories. Of course, student DNA samples can be sent out for analysis at a moderate cost, but then the students do not get hands-on experience. Here, we provide a novel inexpensive method for rapidly sequencing and genotyping DNA, thereby providing students with a hands-on understanding of DNA analysis and of the sequence variation among individuals, which is the basis of future medicine.
The 5-week laboratory course given at Purdue University teaches students first how to sequence a provided DNA sample, how to use NCBI bioinformatics software to compare the DNA sequence that they determined with the actual DNA sequence, and finally how to purify their own DNA from a cheek swab, amplify a region of interest containing an informative SNP, and to genotype themselves by DNA sequencing. The students perform nearly all of the protocols required. Students also learn to use SNPedia and NCBI databases to learn about the DNA sequence variant genotyped and others of medical interest. Additionally, students learn about the relative abundance of the SNP allele that they possess in different world populations, and some medical implications of having particular SNP alleles.
For the course at Purdue University we used the SNP described above. This SNP is in the ABCC11 gene, and is officially designated rs17822931 . The “C” allele causes the formation of wet sticky brownish earwax (cerumen), whereas the “T” causes dry flaky whitish earwax. The wet earwax phenotype is dominant. Thus, genotypes CC and CT cause the wet sticky phenotype, whereas TT causes the dry phenotype. Genotype frequencies at the rs17822931 site among different ethnic populations of the world have been determined and have been reported . The class (fifteen students plus TA and instructor) included: Korean, Chinese, Vietnamese, Indian, African American, European American, South American, and Jewish ethnicities.
Materials and Methods
Operation of the Teaching Laboratory
The 5-week lab was one of several lab modules available to Biology undergraduates (usually seniors). The lab runs smoothly when students are organized into groups of two. Student groups work at their own pace until they achieve success with each protocol. At Purdue University, a 4 hr lab is given two days a week. The teaching assistant (co-author K.S.) prepared the chemically treated gel plates for the class as needed, as well as the stock solutions for gel membrane development. The teaching assistant also cut properly sized Whatman 3MM paper wicks, blotting paper, and nylon membrane strips for the class. Acrylamide containing gel solutions and 5× TBE (Tris, Borate, Ethylelediaminetetracetic acid) gel buffer concentrate can either be prepared by the TA or can be purchased. The students perform all of the other operations. The instructor gave a brief lecture on the protocols to be used that day, and the TA demonstrated proper technique.
Reagents and Supplies
Gel plates (1.5 × 4 in × 1.0–1.2 mm) were custom made. Alternatively, if long reads are not needed, standard 1.5 × 3 in microscope slides can be used. Gel loading devices were laser machined from standard single edge razor blades (VWR Industrial Razor Blades, Surgical Carbon Steel, Single Edged No. 9). Polymerase chain reaction (PCR) primers were from IDT. γ-Methacryloxypropletrimethoxy silane reagent was purchased from Sigma. Surfasil siliconizing reagent was from Pierce. M13mp18 DNA, biotin, and streptavidin were purchased from New England BioLabs. NBT: 4-Nitro blue tetrazolium chloride, 300 mg in 3 mL dimethylformamide and BCIP: X-phosphate/5-bromo-4-chloro-3-indolyl-phosphate, 150 mg in 3 mL dimethylformamide were purchased from Boehringer Mannheim. Hybond-N+ nylon membrane was from Amersham; urea (ultrapure) was from Invitrogen; Triton X-100 detergent was from BioRad. Sequenase Sequencing and Thermo-Sequenase Cycle Sequencing Kits were from USB. Glass weight plates 2 × 4 × 0.25 inch (74 g) were obtained from Lafayette Glass, Lafayette, IN. Other commercial suppliers are listed below in the descriptions of the protocols.
Sanger Sequencing and Preparing Gels
Sanger sequencing with Sequenase was performed as recommended by USB. M13mp18 (for the students) and any convenient double-stranded DNA vector or construct (for the instructors) were used along with a 5′-biotinylated universal sequencing primer (HPLC-purified) from IDT. G, A, T, C samples were heated 5 min at 95 °C and quenched in ice just before use. Gel plates were siliconized or silane treated as recommended by the manufactures. Plates were cleaned scrupulously just before use by students using 70% ethanol and dried streak-free using Kim Wipes. Strips (1/16 in) of Scotch Crystal Clear Tape were used as spacers on the silane treated plate. The inner surfaces of glass plates were made lint/dust-free using “canned air.” Sandwiches were made using a taped silane plate and a siliconized plate. Standard 6% sequencing gel solution containing 19:1 acrylamide:bis acrylamide, 1× TBE, 8 M urea, and TEMED (Tetramethylethylenediamine) is used to “pour” the gels. The sequencing gel solution is good for 2 months, stored refrigerated. Just before use, initiate slow polymerization with 11 µL of freshly prepared 10% APS (Ammonium persulfate) for 1.5 mL sequencing gel solution; mix well. Stagger the plates of a sandwich by a one half plate length, tilt the plates downward by about 30–45 degrees, and pipette in enough gel solution for it to flow to the end (∼100 µL). Then continuously reduce the stagger of the plates, while pipetting in more gel solution as needed, until the stagger is less than about 1/8 inch. Each gel should take only about 1 min to fill, and it should require 200–300 µL of 6% sequencing solution, depending upon how much solution drips out from the bottom. With well-cleaned plates, the gel solution front flows uniformly without bubble formation. Care should be taken to not leave excess gel solution on the pipetted end. After pipetting, immediately lay the gel horizontally on a piece of Parafilm, and stack two glass weights on it in a staggered fashion so that the uncovered end of the sandwich (the pipetted end, which faces up) does not contact the glass weight. Contact causes liquid to be drawn out of the sandwich, resulting in a poorly formed gel. Allow the gels to polymerize for 1.5 hr. The polymerized gel recedes slightly from the spacers, and should appear as rectangular in form with sharp edges when the sandwich is viewed against a backlight. Polymerized gels should be submerged in 1× TBE, 8 M urea, and are good for two weeks when stored refrigerated. The thickness of these gels will be 50 µm. Such ultra-thin gels allow high voltage gradients to be applied across them with low current (less than 2 mA), and run rapidly.
Gels are run horizontally in a standard mini-agarose gel apparatus at 900 V. A high voltage power supply was used. Alternatively, two or three lower voltage power supplies can be connected in series to achieve 900 V. Run times were approximately 8 min. Gel apparatus chambers are filled with appropriate volumes of 1× TBE electrode buffer. Care must be taken to insure that electrode buffer does not splash under the platform, which will result in a current path under the gel instead of through the gel, giving a high current. A gel from the storage solution is rinsed with deionized water, and gently dried. The gel sandwich should be shaken briskly to remove liquid from the channels that formed between the spacers and the gel. The plates are separated (using an ordinary single edge razor blade inserted between the plates at a corner and twisted), being careful not to splash any liquid on the gel. The gel loading area must be dry. The gel-bonded plate is laid horizontally in the apparatus. Thoroughly wet 2 cm × 5 cm wicks made from Whatman 3MM paper in 1× TBE electrode buffer, and blot dry the end that will contact the gel. Drying allows the wick to stick to the gel. Place the wicks over about 0.5 cm of the gel at each end. Use a 3/8 in section of a gel plate as a top wick holder. Dry and replace the siliconized plate so that it covers the bottom wick and leaves a sample loading area of about 0.5 cm (see Fig. 2a). The wicks, which extend into the electrode buffer in each gel apparatus chamber provide electrical contact and keep the gel from drying out. A current check should give 1.5–2.0 mA for 900 V (constant voltage setting).
Using a P-20 micropipette set at 1 µL with a standard non-tapered yellow tip, pick up a very small amount of sample by capillary action, without depressing the piston. Then, depress the piston very slightly so that loading buffer bulges slightly from the tip, and carefully place a small bead of sample at the very end of each of tooth of the loading device to be loaded (Fig. 2b). Without delay (because 20 nanoliters will evaporate significantly after a few minutes), stamp the samples into the gel in the center of the loading area. The loading device teeth should cut through the gel and evenly contact the underlying glass, using an exclusively vertical motion. We find it very helpful in loading to use a “backstop” support fashioned from the supplied plastic gel comb. The backstop support can be made by gluing an appropriately sized piece of a glass microscope slide to a standard agarose gel comb that fits the apparatus. This backstop support is firmly clamped to the apparatus in its normal position, making sure that it aligns with the gel loading area. The loading device with samples is pressed against the support, making it easier to position the samples in the center of the gel and to deliver the samples without any lateral motion, resulting from jittery hands, which can rip the gel during loading. Better results are obtained if the same amount of sample is placed on each tooth of the loading device. After stamping the samples into the gel, the gel should be run for 2.5 min at 900 V. Then the top wick and wick holder should be repositioned to cover the previously exposed loading area. The wick should be positioned close to, but not touching the upper plate; whereas the manufacturer's straight edge of the wick holder should abut the top plate. Unless the loading area is covered, the gel will dry out in the exposed region and the current will fall to zero in a few minutes. The current should remain constant at 1.5–2.0 mA throughout the approximately 8 min run. The faster moving bromophenyl blue dye of the DNA sample buffer runs approximately with 20 nucleotide long chains, whereas, the slower moving xylene cyanol FF dye runs approximately with 80 nucleotide long chains. At the end of a typical run the bromophenyl blue dye should have just entered the bottom wick, becoming visible, and the xylene cyanol FF should be approximately half way down the gel.
Label a gel-sized piece of Hybond N+ nylon membrane in the lower left-hand corner with pencil (ink will wash off) to identify the gel and the DNA side. Wet the membrane in 1× TBE, then blot it dry with a paper towel. The membrane should be handled using forceps. After the gel has run for the desired time, remove the top plate and wicks. Place the bottom plate with the gel on the bench-top in an undisturbed area. Without delay, blot your membrane dry, and very carefully position the bottom of the membrane at the bottom of the gel without touching the gel; then roll it down letting it stick to the gel from the bottom to the top. Place 2 pieces of gel plate-sized Whatman paper on the membrane and then two glass weights. Leave the blot undisturbed for 10 min. Remove the weights and the Whatman paper, and peel the membrane off the gel. The DNA should now be transferred to the membrane, and one should easily see the XCFF dye band on the membrane. UV crosslink the DNA to the membrane. We used a Stratagene UV crosslinker, default conditions, which takes approximately 1 min. Sandwich the membrane between two pieces of Whatman paper for storage at room temperature, if it is not to be processed for DNA detection immediately.
The sequencing primer was labeled with biotin. Therefore, all of the extended DNA chains making up the sequencing ladders will be biotin labeled. The biotin label is used to couple the DNA to alkaline phosphatase, which causes BCIP, in the presence of NBT, to form a fine purple-blue precipitate that specifically coats the membrane-bound DNA, giving purple-blue bands on a white background. This is a very sensitive method, capable of easily detecting femta-gram amounts of DNA in bands. To detect the DNA, first shake the membrane vigorously in blocking solution (5.0 g SDS (Sodium dodecylsulphate), 0.73 g NaCl, 0.24 g Na2HPO4, 0.11 g NaH2PO4.H2O, d-H2O to 100 mL) for 5 min in a capped 50 mL tube containing 25 mL of blocking solution using a rotary shaker. The same tube can be used throughout. Discard the blocking solution, add 5 mL of blocking solution containing 5 µL (1 mg/mL) streptavidin, and rock the membrane gently for 10 min with the DNA side facing up. Next, wash the membrane two times for 5 min each in 25 mL of 0.1× blocking solution using vigorous shaking. Then, gently rock the membrane 10 min using 5 mL of blocking solution plus 5 µL (0.5 mg/mL) biotinylated alkaline phosphatase (BAP) reagent. Then, wash twice for 5 min each using vigorous shaking with 25 mL Triton X-100 buffer (5.0 mL 1 M Tris-HCl, pH 8.0, 5.0 mL 20% Triton X-100, 0.20 mL 0.5 M Na2EDTA, 0.73 g NaCl, H2O to 100 mL). Next, wash twice for 5 min each using vigorous shaking with 25 mL pH 9.5 buffer (0.88 g NaCl, 1.87 g Tris base, 1.0 mL 1 M MgCl2, d-H2O to a little less than 200 mL, titrate with 1 N HCl to pH 9.5, d-H2O to 200 mL). Finally, place the membrane on a clean glass weight plate, with the DNA-side facing up. Add 2.5 µL NBT to 1.0 mL of pH 9.5 buffer and quickly mix by vortexing. Then add 3.3 µL BCIP and vortex again. The solution should be clear yellow. Pipette the NBT/BCIP solution on top of the membrane and cover the membrane with a clean untreated glass plate so that a thin film of NBT/BCIP solution covers the entire membrane. Avoid forming bubbles over the membrane. Then, lay a piece of aluminum foil loosely on top to keep the photosensitive solution out of light. Leave the membrane in the solution for at least 1 hr for the color to develop. The purple-blue color will continue to darken for several hours. Membranes can be left to develop overnight. After the bands are dark enough, briefly (5 min) wash the membranes with Triton X-100 solution for 5 min, then, rinse with d-H2O. Set the membrane down on a paper towel, hold on to an end with forceps, and dry it, both front and back, with a hair dryer. The developed membrane can be stored indefinitely at room temperature in the dark. Developed membranes are scanned at 300 dpi and the images digitally enlarged for analysis of the sequencing ladders.
DNA Isolation, Amplification, and Cycle Sequencing
Human DNA was prepared from a cheek swab (using a Q-tip cotton swab) by the spin basket method described by Kephart . The DNA was eluted from the spin basket with sterile 100 µL of 0.2 mM Na2EDTA. For PCR amplification, samples were prepared in a designated clean area using filtered pipette tips, and contained: 36 µL sterile deionized water, 5 µL 10× Taq Buffer advanced (self-adjusting Mg++) (5 PRIME), 1 µL dNTP Mix (10 mM each), 1 µL Forward primer, 8.8 pmol/µL, 1 µL Reverse primer, 8.8 pmol/µL, 5 µL DNA (prepared as described above). The mixture was overlaid with 30 µL mineral oil (PCR Reagent, Sigma), and incubated in a heating block at 75 °C for 5 min. Then 1 µL (5 U/microliter) Taq Polymerase (5 PRIME) was added and samples were transferred to a Perkin–Elmer Thermal Cycler after the PCR block temperature reached 75 °C. PCR cycling conditions were: 94 °C, 30 sec; 62 °C, 1 min; 68 °C, 1 min for 35 cycles. The forward PCR primer was 5′-TTGGGCTGAGGAACTGGAGAATGA-3′ and the reverse PCR primer was 5′-CAAGGCTTCACCGCCTTTGGGA-3′. The products of PCR amplification were checked by agarose gel electrophoresis: 10 percent of the sample was run directly along with a 100 bp DNA ladder (BioRad). Typical results are shown in Fig. 3, where the expected 701 bp product is the predominant product. The 701 bp band varied in intensity from sample to sample, as did the background signal which likely arises from bacterial contamination. We found that a low degree of background on the agarose gel as in lane 3 could be tolerated, whereas for more severe contamination the DNA prep was repeated. The PCR-amplified DNA was cleaned-up using the QIAquick PCR Purification Kit (QIAGEN), including a spin-column wash with 35% (wt/vol) guanidine hydrochloride after the sample binding step to disrupt primer-dimers. DNA was eluted in 50 µL 10 mM Tris-HCl, pH 8.5. Samples were micro-dialyzed (Millipore 0.025 micrometer VSWP) for 15 min against 40 mL 0.1 M NaCl, 10 mM Tris-HCl, pH 8.0, 1 mM Na2EDTA, then for 15 min against 40 mL 10 mM Tris-HCl, pH 8.0, 0.2 mM Na2EDTA, using the same dialysis disc. Cycle sequencing was performed using the USB Thermo-Sequenase Kit. A mixture containing: 3 µL DNA, 10 µL sterile deionized water, 1 µL (1.5 picomole) 5′-biotinylated-sequencing primer, 2 µL Reaction Buffer, and 2 µL Theormo-Sequenase was prepared. To each of the four tubes (G, A, T, C) containing 4 µL each ddNTP termination mixes, 4 µL of the above mix was added, followed by 10 µL mineral oil. Samples were placed into the thermal cycler, and the 50 cycle program: 95 °C, 30 sec; 55 °C, 30 sec; 72 °C, 1 min was initiated. The sequencing primer was 5′-TCGCTAAACCTCTGAAGCCT-3′. After cycling was completed, 4 µL stop solution was added to each tube, the mixture was vortexed, microfuged, and the oil was removed. G, A, T, C samples were kept frozen until needed. G, A, T, C samples were heated to 95 °C for 5 min, then quenched in ice before loading on the sequencing gel.
Results and Discussion
In preparation for the first day of class, the TA prepared about 30 Mini-Seq gels, which can be stored in the refrigerator for up to 2 wk as described in Materials and Methods. Students used the first class to prepare gels (about five per group), and to become familiar with the DNA sequencing system. Familiarization with the system involved practicing applying about 20 nanoliters of DNA sample buffer to each tooth of the loading device (Fig. 2b), and then practicing stamping the samples into a sacrificed gel from which the top plate had been removed from the sandwich. The exposed gel was just placed on the bench top, and stamped repeatedly. After successfully accomplishing these tasks, students practiced setting up a gel for running (Fig. 2a), and then they loaded the gel with sample buffer and ran the gel for several minutes so that the two dye fronts separated by several centimeters, and appeared as two parallel lines. The students also practiced how to properly cover up the gel loading area, as described in Materials and Methods.
On the second day, students shared four tubes containing G, A, T, C sequencing reactions, which were prepared by the instructor using a convenient vector or construct and universal primer. Student groups loaded these samples in duplicate (GATCGATC) on their gels, ran the gels, blotted them onto nylon membranes, and developed their membranes as described in Materials and Methods. Because the development reaction takes at least 1 hr, students generally left their membranes in development solution overnight, and the next morning the instructor washed and dried the membranes for the students.
In the third lab, students who got good looking sequencing ladders went on to enlarge the membrane images, to read the sequence, and to use NCBI basic local alignment search tool 2 (BLAST2) to align their sequence read with the actual sequence. Some groups were able to read more than 100 nucleotides with few errors. See ref. [  for sequencing ladders obtained using an older version of this system. Common errors were calling doublets as singlets, triplets as doublets, or interchanging two nucleotides from closely spaced bands. Student groups who were not successful on their first attempt for one reason or another repeated the experiment. All groups were able to read at least 40 nucleotides of sequence with greater than 95% accuracy after two or three attempts.
In the fourth and fifth labs, student groups preformed their own sequencing reactions using M13 DNA, and sequenced the DNA. Student groups also made more gels as needed.
For the remainder of the 5 weeks, each student prepared a sample of their own DNA from a cheek swab, PCR amplified an approximately 700 bp fragment encompassing the SNP, cleaned up the PCR product, and sequenced the region of the DNA from a sequencing primer that was approximately 40 nucleotides away from the primer 3′ end (see Fig. 1). Ideal gel results from the three possible genotypes are shown schematically in Fig. 4. Only four nucleotides before the SNP (arrow) and five nucleotides after the SNP are shown. The nucleotide sequence flanking the SNP is given at the right of the third gel drawing. For the CC genotype, the sequence reads GGCCC (from bottom to top), whereas the TT genotype gel reads GGCCT. All bands ideally have the same intensities. For each of these two genotypes (CC or TT), both the maternal and the paternal DNA sequences were the same, and the sequence ladders had only a single nucleotide at each position, as would be found for a pure DNA sample. For the CT genotype (center gel), there are two bands at the SNP position (one in the C lane and one in the T lane), and these bands have half the intensities of the others. This pattern arises because one parent contributed a C at that position and the other a T. So the DNA sequenced was a mixture containing equal amounts of two different sequences. The SNP regions of actual gels from the class for the three possible genotypes are shown in Fig. 5. Whereas the gels were not perfect, they look similar to those in the simulation (Fig. 4), and the genotypes are easy to deduce.
For extra credit, students who complete the above experiments early can request cheek swabs from their parents and/or siblings to construct a family pedigree. An example is shown in Fig. 6a (S.T.) for student S.T. who found that she was a CT heterozygote. There are seven different pedigree structures that could give rise to a heterozygous daughter: CC for one parent, TT for the other; CC for one parent, CT for the other; TT for one parent, CT for the other (two ways for each of these three possibilities); or CT for each parent. Figure 6a (Mom) shows that mom was CT, and Fig. 6a (Dad) shows that dad was also CT, consistent with their phenotypes of having yellow sticky earwax. The daughter (S.T.) thus had a 50 percent chance of being CT. Interestingly, she had a 25 percent chance of being TT and thereby having a phenotype different from her parents. The family pedigree is shown in Fig. 6b.
Students learned a great deal about DNA analysis, and felt a sense of accomplishment after mastering the protocols in this hands-on laboratory. In addition, students found it exciting to sequence their own DNA to see how their genotype compared with their phenotype, with the genotype expected from their ancestry, and to see which of their particular alleles were contributed by each of their parents. Student privacy issues were not a concern because of the innocuous nature of the SNP chosen.
The authors would like to thank the students of the 2011 BIOL 44208 DNA Sequencing Lab class for generating some of the data presented in this paper. The authors have no conflicts of interest.