Sequencing Basics

Principles of automated fluorescent DNA sequencing
Instrumentation in our laboratory
What kinds of DNA can we sequence and how much do we need?
Choosing your host strain
Template preparation and purification
Plasmids
PCR products
BAC/PAC clones
Methods for quantitation
Primer design


Principles of automated fluorescent DNA sequencing

Automated fluorescent sequencing utilizes a variation of the Sanger chain-termination protocol developed over 20 years ago. In this method, the DNA to be sequenced acts as a template molecule to which a short, complementary oligonucleotide (primer) will anneal to begin enzymatic extension and amplification of a specific region of double-stranded DNA. The newly created fragments will be complementary to the template DNA. This process takes place during the cycle sequencing reaction, a process that each sample we receive must undergo in order to become amplified and fluorescently labeled for detection on our sequencers.

In this cycle sequencing reaction, template and primer are combined together with a reaction mixture composed of dNTPs, fluorescently labeled ddNTPs, Amplitaq FS polymerase enzyme and buffer. The cycle sequencing reaction is composed of three steps - denaturation, annealing and extension - and takes place in a thermal cycler, an instrument that allows for controlled heating and cooling of our reactions. These steps are repeated for 35 cycles to ensure sufficient amplification of the labeled DNA, and takes about 3 1/2 hours to complete. During the denaturation step, which occurs at 96ºC, the double-stranded template DNA is first separated into single-stranded molecules. At the annealing stage, the temperature is lowered to 50ºC so that the small primer molecules can find their complementary regions on the now single-stranded template DNA and hybridize, or anneal, correctly. The temperature is then raised to 60ºC (extension step) to allow the Taq polymerase enzyme to begin incorporation of nucleotides into growing chains of newly created fragments that are complementary to the single-stranded template DNA. These extension products begin at the end of the primer and extend in the 3’ direction. Chain termination occurs during this extension step. Our reaction mixture contains a mixture of deoxynucleotides (dNTPs) and dideoxynucleotides (ddNTPs), at concentrations that create a statistical probability that a dideoxynucleotide will be incorporated instead of a deoxynucleotide at each nucleotide position in the newly generated fragments. When a dNTP is incorporated, the new fragment will continue to grow. A ddNTP contains a hydrogen atom instead of a hydroxyl group at its 3’ end and cannot participate in further extension. Therefore, when a ddNTP is incorporated, further chain elongation is blocked and this results in a population of truncated products of varying lengths.

When separated by electrophoresis, a "ladder" of these truncated products will form, each differing in size by one nucleotide, with the smallest terminated fragments running fastest on the gel. There are four different ddNTPs that correspond to each of the four DNA nucleotides, and each ddNTP has a different color. Thus, each truncated fragment will contain a fluorescently labeled ddNTP at its 3’ end, and the sequencing ladder will be composed of colored bands. The sequence can then be determined by correlating the color of a band on the gel with its specific ddNTP, and the order in which they ran on the gel. So, the first and smallest band visualized will correspond to the first labeled nucleotide incorporated immediately adjacent to the primer. The second band will be the fragments that consist of 1 unlabeled dNTP and 1 labeled ddNTP that terminated those particular growing chains. The third band will be made up of 2 unlabeled dNTPS followed by 1 labeled ddNTP, and so on up the gel.

Once the sample has been amplified and labeled, it must be electrophoresed for separation of the labeled fragments and their visualization. As mentioned before, the ddNTPs are fluorescently labeled. The attached dyes are energy transfer dyes and consist of a fluorescein energy donor dye linked to an energy acceptor dichlororhodamine dye. This energy transfer system is much more sensitive than a single dye system and allows us to use less DNA for detection and, in addition, allows us to now sequence very large DNA molecules, such as BACs, PACs and even some bacterial genomic DNA, that previously less sensitive methods were unable to manage. These dye-labeled fragments are loaded onto the sequencers and during electrophoresis, migrate either through the polyacrylamide gel or liquid polymer and are separated based on their size. Towards the end of the gel or the capillary, they pass through a region that contains a read window, behind which a laser beam passes back and forth behind the migrating samples. This laser excites the fluorescent dyes attached to the fragments and they then emit light at a wavelength specific for each dye. This emitted light is separated according to wavelength by a spectrograph onto a cooled, charge-coupled device, or CCD camera, so that all four fluorescent emissions can be detected by one laser pass. The data collection software collects these light intensities from the CCD camera at particular wavelength bands, or virtual filters, and stores them onto the sequencer’s computer as digital signals for processing. The analysis software then interprets the fluorescent intensity at each data point and assigns its base call interpretations.

Instrumentation in our laboratory

The RPCI DNA sequencing laboratory currently utilizes two ABI PRISM 3130XL Genetic Analyzers. These instruments are upgraded versions of our original 3100s, with improved hardware and software capabilities. They are automated capillary DNA sequencers manufactured by Applied Biosystems. Each 3130 uses a 16-capillary array, which can run 176 samples in 24 hours using our current configuration or up to 384 samples in 24 hours with a protocol which gives shorter read lengths, but allows for maximum throughput.

With the 3130, sample loading is based on electrokinetic injection. When a plate of samples is placed onto the instrument’s loading deck, the 16-capillary array dips into the first sixteen sample wells, voltage is applied and negatively charged ions, which are primarily the DNA extension products but which may also include interfering substances such as salts, are injected into the capillary array. This is one reason why DNA must be especially clean for runs on the 3130 as smaller negatively charged ions, such as salts, are preferentially injected and can interfere with the migration of the DNA fragments.

A liquid flowing polymer passes through the 16 capillaries in the array to act as the separation matrix. A fresh aliquot of polymer is pushed into the capillaries before each run, flushing out the polymer from the previous run. Previously, our laboratory used the longest capillary array provided for the 3100, 80 cm in length, and the Applied Biosystems POP-4 polymer for the separating medium. This combination gave around 900 bases of useful sequence, and data for sixteen samples could be acquired in 3 hours, 40 minutes with this array and polymer. However, we have since developed modified run protocols that utilize Applied Biosystems newest polymer, POP-7, and a 50 cm array. This polymer was developed specifically for ABI’s high-throughput sequencer (the model 3730) and has been shown to provide longer read-lengths coupled with shorter run times but was not supported for use on the 3100. Prior to upgrading to the 3130 models, we were able to obtain, on average, over 900-950 Phred Q20 bases with our “long-read” protocol and 850-900 Q20 bases with our “standard-read” protocol using our modified protocols. The “long-read” protocol has a run time of 2 hours, 45 minutes. Our “standard-read” protocol is complete in 2 hours, 5 minutes. Some of this work has been presented in Biotechniques. (2004 Jun;36(6):932-3). With the 3130 upgrades, we are able to get similar read lengths and quality scores in a somewhat shorter period of time.

The table below summarizes major differences in run conditions between different arrays and polymers.

Electrophoresis is facilitated by a high-voltage electrical circuit in the 3130. The charge is conducted through the circuit by DNA and ions in the polymer, ions in the buffer used in the instrument, and through electrical wires and electrodes. The 3130 can evenly and efficiently dissipate heat from this voltage because of the large surface area of the fused silica capillaries. It is this attribute that allows for high voltage to be applied during electrophoresis, and as a result, contributes to significantly decreased run times without loss of resolution. Older sequencers such as the Applied Biosystems 377, which employed a slab gel in order to separate DNA fragments, could not dissipate heat as well and had to be run considerably slower and longer to attain read lengths similar to those on the 3130.

Sequencing fragments pass through a detection cell at the end of the array. Shorter fragments migrate more quickly than longer fragments and pass through the detection cell in this order. An argon-ion laser beam causes the electrons on the dyes attached to the DNA fragments to be excited to a higher energy state. When they return to their ground state, they fluoresce. The emitted fluorescence is captured by a charge-coupled device (CCD) camera. This camera is composed of a silica chip with thousands of pixels which store an electrical charge proportional to the intensity of the fluorescence that reaches it. The CCD camera then converts the fluorescence information into electronic data which is transferred to a computer for processing. Software processes this data in order to create an electropherogram where each peak represents one fragment of DNA sequencing product.

What kinds of DNA can we sequence and how much do we need?

The DNA sequencing laboratory can sequence a variety of DNA samples, including plasmids, cosmids, PCR products, single-stranded phage DNA, and BAC or PAC clones. When submitting DNA for sequencing, and if you are providing your own sequencing primers, concentrations for the various types of DNA should adjusted as follows:

* PLEASE ADJUST YOUR DNA AND PRIMER CONCENTRATIONS ACCORDINGLY TO MEET OUR CONCENTRATION REQUIREMENTS - WE PROCESS HUNDREDS OF SAMPLES A WEEK AND DO NOT HAVE TIME TO MAKE YOUR DILUTIONS FOR YOU!

At the concentrations listed above, we will use approximately 3-4 ul per reaction. Also, these measurements do not need to be exactly 200ng/ul or 10 ng/ul or whichever is applicable - these are approximations of how much we’d like and if you are off by a little - say, give or take 30 ng/ul for plasmids, 2-3 ng/ul for PCR products- it will still probably be fine as there is a range of concentrations we can use and still acquire good sequence data. And don’t panic if your yields are much lower than our requirements as you do have a couple of options. First, if you have access to a vacuum centrifuge, aliquot out a volume that contains the TOTAL amount that we would use, dry it down and then resuspend in water to the concentration we would like. For example, if your initial plasmid DNA concentration is only 50 ng/ul, take out 10-15 ul, dry it down and then add 3 ul of water. If you don’t have access to a vacuum centrifuge, then make a note in the comments field of your request form and we will use a greater volume of your sample in our cycle sequencing reaction. However, keep in mind that the maximum volume of DNA we can add to a sequencing reaction is 10 ul, so in order to attain our minimum total amount of DNA your concentration cannot be much less than 50 ng/ul for plasmids or cosmids, 70-90 ng/ul for BAC and PAC samples. If your sample concentration is lower than that, and you don’t have access to a vacuum centrifuge, then let us know and we can dry it down for you in our centrifuge, or we will show you how to do it. Alternatively, you can do an ethanol precipitation and resuspend in a smaller volume to concentrate your samples but remember to remove all traces of ethanol.

Generally, for PCR products, product yield isn’t a problem if your PCR conditions are optimized and robust. However, if you have a large PCR product that needs to be sequenced, that will require more DNA (for example, a 2 kb PCR product would require 100-150 ng of DNA), so you may want to scale up your reaction or pool multiple reactions together. In addition, when providing your own custom primers, please adjust their concentration to 1 uM, or 10 uM for BAC/PAC samples. At these concentrations, we will use 4 ul of primer per reaction. We provide a number of the more common vector primers, free of charge, for sequencing. Please check the link at Primers we provide to see the complete list. And whenever possible, please provide enough DNA and primer for two reactions in case we need to repeat a reaction for any reason.

Choosing your host strain

Choosing your host strain for your template preparation is an important consideration, as some strains will yield significantly better quality DNA for automated sequencing. Certain E.Coli strains may release certain factors, such as endonucleases and large amounts of carbohydrates, during their lysis that can be inhibitory to DNA preparation and sequencing. In addition, certain cellular components, such as positively charged polysaccharides, can compete with the DNA for binding in silica-based DNA preparation methods that are commonly used in many commercial kits. It is always recommended that you check with the manufacturers of specific strains to determine if there are any known difficulties that might arise when using their strains with your applications. Host strains that generally give reliable data include DH5alpha, DH10B, DH1, C600, XL1-Blue, XL10-Gold, TOP-10 and NM294. XL1 Blue grows slower than most strains and can lead to decreased DNA yield, but usually still works well. Strains such as JM101, JM83, HB101, TB1 and TG1 are NOT recommended as they can release large amount of inhibitory factors during lysis that can lead to poor quality DNA that will require extra purification steps.

Other factors to consider when preparing to grow up cultures are media choice and antibiotic selection. Overgrowth of cells should be avoided as cells begin to lyse fairly rapidly after reaching stationary phase, releasing large amounts of degraded chromosomal and plasmid DNA as well as polysaccharides that are difficult to separate away from plasmid DNA. Standard LB media or buffered Richer media is usually recommended, as an overnight culture of 10-12 hours will generally give an appropriate cell density for loading onto columns or membranes, and in addition will have a much lower proportion of degraded cellular materials and contaminating DNA. It’s best to avoid media such as TB or 2xYT as they will reach stationary phase much sooner, and thus an overnight culture will give an inappropriately high cell density as well as a greater percentage of inhibitory substances. If you should have to use media such as these, it’s advisable to monitor culture time and cell number to avoid DNA contamination and column overloading. One trick to try is washing your pelleted cells with 0.5M NaCl just prior to lysing to help remove polysaccharides and media materials that may co-elute with your DNA. Resuspend your pellet in the 0.5 NaCl solution, vortex well and spin again, removing the supernatant when finished.

Antibiotic selection should be maintained throughout to avoid growth of cells that do not contain plasmid. In the absence of effective selection, cells that do not contain plasmid will outgrow cells that do, leading to poor yield of the desired DNA. In addition, some plasmids will metabolize certain antibiotics, such as ampicillin, during cell growth so it is important to provide proper antibiotic concentrations to minimize depletion. Antibiotics should be aliquoted and frozen to maintain optimal efficacy.

Template preparation and purification

Template purity and concentration are the two most important factors in obtaining good quality sequence data. Template purity and concentration are the two most important factors in obtaining good quality sequence data. No, this is not a typo, it’s just that we can’t stress this principle enough when it comes to automated fluorescent sequencing, especially on our 3130s which require even more stringent cleanup methods. Fluorescent sequencing is very sensitive to certain contaminants in the DNA sample and DNA that is considered clean enough for many other molecular biological procedures such as PCR, cloning, or even manual radioactive sequencing, is NOT necessarily clean enough for fluorescent sequencing and can lead to poor quality data or sometimes no data at all. In this section, we list our recommendations on the best ways to clean up and quantitate your DNA, depending on the type of DNA you are trying to sequence.

Those of us working in DNA sequencing core facilities see all kinds of templates, from all sources, prepped and purified in many different fashions. And I think it’s safe to say that Qiagen kits are one of the most universally accepted methods for giving consistently high quality DNA for automated sequencing. This is not to say that there are not other methods that do not work - we’ve listed a few other recommended protocols - but in our experience, Qiagen cleanup methods just have the best track record for reliability and consistency in their final product. You may initially save some money with cheaper kits or homemade preparations, and if you have good bench skills and have successfully used other protocols in the past for sequencing, by all means use what has proved to work for you. But there are times when you may actually end up investing more time and money in re-prepping and resequencing poor quality templates, in the long run. Just my two cents worth...

NEW!  Another option that is available for DNA purification that gives template of high quality for automated DNA sequencing is the Mini-Plasmid DNA Isolation kit available from the Biomedical Research Service Center (BRSC) at SUNY Buffalo. This kit is "a modified and streamlined alkaline lysis mini-prep method from Sambrook et al. (Molecular Cloning, CSH, NY) that eliminates the need for additional RNase treatment and phenol-chloroform extraction. Plasmid DNA can be easily and rapidly purified from an overnight E. coli culture in 15 min with a typical yield of 1 – 5 mg DNA per ml of culture. The DNA is suitable for restriction digestion, probe preparation, and PCR analysis. DNA to be used for sequencing analysis is subjected to an additional 5-min step of PEG precipitation (PS-4) to remove any trace amount of RNA. High-quality DNA sequencing reads can be easily obtained (almost 1,000 bases resolved per sequencing run) if the plasmid DNA is isolated from a suitable E. coli host strain. The kit (sufficient for 300 mini-preps) is stable for at least one year if handled and stored properly." (taken from the BRSC website). At $95 for 300 plasmid minipreps, this kit is significantly cheaper than Qiagen kits (about one-fourth the cost), rapid and very simple to use. We have seen, first-hand, that the quality of DNA obtained from this kit is excellent for sequencing and we would recommend the use of this kit for DNA purification. The BSRC is a local SUNY resource that promotes collaborative translational research, develops and commercializes innovative products, and manufactures quality research reagents for the broad life science community. For more information about the BSRC and it's products, you can check out their website at: http://www.acsu.buffalo.edu/~chunglee/.

A link to the mini-plasmid DNA isolation kit protocol can be found at: Mini-plasmid prep protocol.

Plasmid templates

Considerations when cleaning up plasmid preps

Poor template quality is one of the most common reasons for bad sequence data, as mentioned above, and is a prime consideration when choosing a plasmid cleanup method to give DNA of optimal purity for automated sequencing. Plasmid template quality can be affected by a variety of factors and contaminants including the following:

* Salts or organics left over from template preparation
* Presence of cellular components such as RNA, proteins, polysaccharides or chromosomal DNA
* DNA that has degraded while in storage
* silica fines that carryover from template preparation kits that utilize loose resin or silica solutions

Here is a table of allowable contaminants and their acceptable concentration ranges that will still allow for good sequencing data:

Some comments on the effects of some of these contaminants are listed below:

RNA - As you can see from the value in the chart above, a significant amount of RNA can actually be tolerated in the sequencing reaction. It is a common contaminant in plasmid minipreps, especially when columns are overloaded and the capacity of the lysis buffers is exceeded - overloading can be a problem with Qiagen minipreps. One of the ways RNA can potentially interfere is when quantitating your DNA preps by spectrophotometer- RNA also absorbs at 260 and when there is a large amount present, it can really throw off the accuracy of your concentration. It’s best, then, to treat your template preparations with RNase or high salt precipitation and also to quantitate your samples both by gel as well as spectroscopically. More on that in Methods for Quantitation.

PEG- the presence of residual polyethylene glycol in the template prep can have an inhibitory effect on the cycle sequencing Taq polymerase enzyme and lead to weak signal.

Salts- the processivity of the Taq polymerase used in the cycle sequencing reaction declines in the presence of high amounts of salts. Salt contamination in DNA preps may result from coprecipitation of salts in alcohol precipitations, insufficient removal of supernatant after precipitations or an incomplete wash of the pellet with 70% ethanol. Careful technique should be used when precipitating with alcohol. It has also been demonstrated that acetate ions, as opposed to sodium, potassium or chloride ions, are the most inhibitory in sequencing reactions. When using potassium acetate or sodium acetate, concentrations over 20 mM led to complete failure of the sequencing reactions, while concentrations of 60mM of sodium chloride were required before complete inhibition. Salts can be inhibitory when we sequence samples on our gel-based 377 but are even more problematic when running samples on our 3100 capillary system as these smaller, charged salts are preferentially injected and interfere with the migration of the DNA samples.

Ethanol- ethanol contamination can occur when the sample is insufficiently dried after precipitation or when carried over in an ethanol-containing wash buffer used in some DNA isolation procedures. Contamination with 10% or greater concentrations of ethanol usually leads to failure of the DNA sequencing reaction. Complete drying of the DNA samples is required to remove these traces of ethanol.

Phenol- phenol may be carried over from DNA alkaline lysis methods that utilize phenol and chloroform to remove proteins and other cellular contaminants from cell lysates. Phenol cannot be tolerated in the cycle sequencing reaction as it denatures proteins and will thus degrade the Taq polymerase enzyme used in the cycle sequencing reaction. Chloroform does not have the strong denaturing properties of phenol and doesn’t appear to adversely affect the sequencing reaction.

Cesium chloride - When using a cesium chloride ultracentrifugation density gradient protocol, one can obtain DNA of very high quality suitable for automated sequencing. HOWEVER, it is strongly recommended that you either perform dialysis followed by ethanol precipitation (best method) or do a minimal room temperature isopropanol precipitation to remove all traces of residual cesium chloride as the cesium can inhibit the Taq polymerase used in the cycle sequencing reaction.

EDTA - EDTA can chelate the magnesium required by the Taq polymerase in the cycle sequencing reaction, so when submitting samples, it is best to always have them diluted or resuspended in sterile ddH20 or 1X Tris buffer. Suspension in TE buffer is not recommended, though people have done it and many times there is not a problem. However, providing template DNA in water is an easy thing to do and if there is a problem with your sequence quality, the fact that there is no EDTA in your sample is one potential problem we can eliminate right away.

Qiagen kit principles
As mentioned above, Qiagen makes kits for extracting and purifying all types of DNA, including plasmid DNA. The Qiagen plasmid miniprep kits contain an anion exchange resin that consists of positively charged DEAE groups that preferentially bind the negatively charged DNA backbone. Once the DNA is bound to the resin, other contaminants such as salts, proteins, RNA and carbohydrates are washed away in a low-salt buffer. The DNA is eluted in a second step, using a high-salt buffer to dissociate it away from the resin. This kit is designed to give the highest purity DNA and is somewhat more costly. Another kit offered by Qiagen is the Qiaprep kit which uses a silica gel-based membrane to bind the DNA in the presence of high concentrations of chaotropic salts. Salts are removed with an ethanol wash and the DNA is then eluted with a low ionic strength Tris buffer. This method comes in a spin column format that can be used either in a centrifuge or on a vacuum manifold. This kit gives DNA of a purity that is suitable for sequencing and is also less expensive. However, there are some factors that one should keep in mind when using the Qiagen kits, or any other column-based miniprep kits. First, do not overload the columns - RNA and polysaccharides can compete with the DNA for binding on the column resin and lead to low yields of the desired plasmid DNA. Follow directions for cell growth and use LB media as recommended. Secondly, do not forget to wash the DNA with 70% ethanol following the isopropanol precipitation in order to remove excess salts. Washing two to four times with ethanol can improve DNA quality, and, in addition, extra washes with the binding solution can help remove other contaminants. Thirdly, warming your elution buffer can help improve yield. And lastly, remove all traces of ethanol before resuspending the final product in water.

Alkaline lysis/PEG methods
Bacteria can be lysed to release plasmid DNA by a number of methods including treatment with detergents, organic solvents, heat or alkali. When attempting to isolate large plasmids (>15kb) that are more susceptible to damage during the lysis procedure, a gentler method, such as equilibrium centrifugation or lysis involving a detergent like SDS, should be used. When lysing cells to harvest smaller plasmids, stronger methods, such as alkali treatment, can be utilized. These treatments cause the linear chromosomal DNA of the host to denature and degrade. Closed circular plasmid DNA strands do not separate completely as they are topologically intertwined and when conditions are returned to normal, the strands reanneal precisely in native conformation as superhelical molecules.

Standard alkaline lysis protocols, when coupled with PEG precipitation, can lead to small plasmid DNA of sufficient purity for use in automated fluorescent sequencing. The following is a detailed protocol for alkaline lysis/PEG treatment that is recommended by Applied Biosystems, manufacturers of our DNA sequencers.

Note: To minimize shearing of contaminating chromosomal DNA, do not use a vortex during this procedure.

1. Pellet 1.5 -4.5 ml aliquots of culture for 1 minute in a microcentrifuge at maximum speed.
2. Remove the supernatant by aspiration.
3. Resuspend the bacterial pellet in 200 ul of GET buffer by pipetting up and down.
4. Add 300 ul of freshly prepared 0.2N NaOH/1% SDS. Mix the contents of the tube by inversion. Incubate on ice for 5 minutes.
5. Neutralize the solution by adding 300 ul of 3.0M potassium acetate, pH 4.8. Mix by inverting the tube. Incubate on ice for 5 minutes.
6. Remove cellular debris by spinning in a microcentrifuge at maximum speed for 10 minutes at room temperature. Transfer the supernatant to a clean tube.
7. Add RNase A (DNase free) to a final concentration of 20 ug/ml. Incubate the tube at 37C for 20 minutes.
8. Extract the supernatant twice with chloroform:
a. add 400 ul of chloroform
b. mix the layers by inversion for 30 seconds
c. centrifuge the tube for 1 minute to separate the phases
d. transfer the upper aqueous phase to a clean tube.
9. Add an equal volume of 100% isopropanol. Mix the contents of the tube by inversion.
10. Spin the tube in a microcentrifuge at maximum speed for 10 minutes at room temperature.
11. Remove the isopropanol completely by aspiration.
12. Wash the DNA pellet with 500ul of 70% ethanol. Dry under vacuum for 3 minutes.
13. Dissolve the pellet in 32ul of deionized water.
14. Add 8.0ul of 4M NaCl, then 40ul of autoclaved 13%PEG 8000.
15. Mix thoroughly, then leave the sample on ice for 20 minutes.
16. Pellet the plasmid DNA by spinning in a microcentrifuge at maximum speed for 15 minutes at 2-6C.
17. Carefully remove the supernatant. Rinse the pellet with 500ul of 70% ethanol.
18. Resuspend the pellet in 20ul of deionized water. Store at -15 to -25C.

Cesium chloride purification
Purification of plasmid DNA using a density gradient method such as equilibrium centrifugation in CsCl-ethidium bromide gradients can yield ultrapure DNA and is especially useful for purifying large plasmids or small plasmids that are to be used in more rigorous experiments, such as biophysical measurements. This protocol takes advantage of the fact that ethidium bromide will intercalate in different amounts when it encounters linear or closed circular DNA molecules. Due to the differential binding of this dye, the buoyant densities of linear and closed circular DNA that are formed after centrifugation will be different in CsCl gradients containing saturating amounts of ethidium bromide, thus allowing effective separation of plasmid DNA away from contaminating chromosomal DNA, nicked circular plasmid DNA, RNA and proteins. This method has long been the gold standard for obtaining very pure DNA, but it is rather time-consuming and expensive. Commercial kits or modified alkaline lysis methods provide DNA of adequate quality for automated sequencing, and so eliminate the need for such a stringent method. However, if cesium chloride purification is your method of choice, protocols can be found in the Maniatis Molecular Cloning manuals.


Some final notes on plasmid purifications

-boiling methods of DNA extraction are not acceptable for automated sequencing unless they are phenol/chloroform extracted to remove all proteins. Endonucleases are not always completely inactivated during the boiling procedure and plasmid DNA can be degraded in the presence of Mg++ in subsequent incubation procedures.

And if, in the end, your plasmid prep is still not of optimal purity for good sequence data from our sequencers, here are some recommendations for further cleanup steps that may help:

-purify your DNA by ultrafiltration, using the Centricon-100 Micro-concentrator columns. See protocol at Ultrafiltration/Molecular weight cutoff columns.

-purify by further extraction:
1). Extract the DNA twice with 1 volume of chloroform or chloroform:isoamyl alcohol (24:1 v/v)
2). Add 0.16 volumes of 5M NaCl and 1 total volume of 13% PEG.
3). Incubate on ice for 20 minutes, then centrifuge at maximum speed in a microcentrifuge at 2-6C for 20 minutes.
4). Rinse the pellet twice with 70% ethanol.
5). Dry the pellet in a vacuum centrifuge for 3-5 minutes or to dryness.

-purify using minimal isopropanol precipitations:
1). Add 0.5 volume of 7.5M ammonium acetate or 0.1 volume of 3M sodium acetate to a dilute DNA solution (<100 ng/ul) and mix well.
2). To this total volume, add 0.6 volume of room temperature isopropanol and mix well.
3). Incubate at room temperature for 5 minutes, then centrifuge at maximum speed to pellet the DNA.
4). Fill tube with 70% ethanol and do a complete wash of the pellet.
5). Dry the pellet in a vacuum centrifuge for 3-5 minutes or to dryness.


PCR products

Considerations when cleaning up PCR products

When submitting PCR products for direct sequencing, it is essential to separate the PCR product away from potentially interfering substances that may remain in solution after the PCR reaction. These contaminants can sometimes participate, and thus interfere, in the cycle sequencing reaction and lead to poor quality data or no data at all. Most importantly, primers and dNTPS should be removed. If both forward and reverse PCR primers remain in solution, they will both act as sequencing primers, resulting in multiple peaks from beginning to end as each primer can anneal to complementary strands with different nucleotide composition, and thus lead to overlapping fragments and unreadable data. Excess dNTPs should also be removed as they will upset the specific ratios of dNTPs/ddNTPs required for optimal extension and termination in the cycle sequencing reaction.

Another consideration when choosing your PCR cleanup method is the determination of the presence of a single band or multiple bands that may result from your PCR reaction. Once the PCR reaction is completed, you should run a small aliquot on an agarose gel to assess its quality. If you have a single specific band, then it will be sufficient to remove excess PCR reactants using one of the methods listed below. If your PCR reaction generates multiple bands, an excessive amount of primer-dimers, or low-intensity smearing, it will then be necessary to run your PCR product out on a gel, excise your band of interest, and purify it away from the gel. Alternatively, you may want to spend some time optimizing your PCR reaction to eliminate the presence of these additional bands or artifacts and thus save yourself some downstream time in gel purifying your PCR products for sequencing.

Regardless of which purification method you choose, you should always quantitate your PCR product after purification as no cleanup method will give 100% recovery. Any quantitative estimates you make when initially running your PCR product out on an analytical gel will be inaccurate as you will inevitably lose some product during purification, especially when gel purifying.


-Single PCR band

Qiaquick PCR purification kit
Qiagen makes PCR purification kits that come in either a microcentrifuge or vacuum manifold format, and can process anywhere from one PCR product using spin columns up to 96 samples in a 96 well plate. Fragments ranging in size from 100 bp up to 10 kb can be purified away from primers, dNTPs and salts that can interfere in the sequencing reaction. The protocol is simple and involves binding of DNA to the column or well membrane, washing away of contaminants, and elution of the DNA from the membrane. One nice thing about this kit is there are no extra steps to remove any mineral oil that may be present in the PCR solution.

Exonuclease I/Shrimp Alkaline Phosphatase
ExoSAP-IT is a product manufactured by USB Corporation and is an excellent method for purifying PCR reactions that produce one specific band. The Exonuclease I component functions to hydrolyze any residual single-stranded primers and any single-stranded DNA fragments produced in the PCR reaction. The Shrimp Alkaline Phosphatase degrades any excess dNTPS remaining in the reaction. The ExoSAP mixture is added directly to the PCR reaction, incubated at 37ºC for 15 minutes for hydrolysis of single-stranded DNA and dNTPS, and then incubated at 80ºC for 15 minutes to inactivate the enzymes.

Ultrafiltration/Molecular weight cutoff columns
Amicon Microcon and Centricon devices, which can be purchased from Millipore, are available with 30,000 and 100,000 molecular weight cut-off [MWCO] membranes and can separate primers from the synthesized PCR product. When purifying smaller PCR products (170 bp or less), choose the Centricon 30. For larger PCR products, the Centricon 100 works well. They will concentrate and desalt the DNA simultaneously, typically removing over 90% of the non-incorporated primers and dNTPs. When using either device, a small sample cup, or reservoir, is fitted into a collection tube. The reservoir contains the size-exclusion membrane at its bottom surface. The PCR product is loaded into the sample reservoir along with a suitable volume of distilled water (400 ul when using the Microcon filter, 2 ml for the Centricon), and then spun in a centrifuge. Salts, dNTPs and primers are spun through into the collection tube. To recover purified DNA, remove the sample reservoir from the vial, invert into a new tube and spin once again. The clean PCR product will then be captured in the second collection vial. Due to the increased wash volume and concentration factor, one spin in a Centricon device removes >95% of a 25 bp primer. A second spin removes up to an additional 4%. The Microcon filter will benefit from additional washes - one wash removes approximately 87% of primers, while three washes will filter away between 95%-99% of standard PCR-size primers.

Ethanol precipitation
If performed carefully, ethanol precipitation will yield PCR products of adequate purity for direct sequencing. However, if not done properly, residual primers and dNTPS, as well as any remaining ethanol, will give poor sequence data so be sure to remove supernatant carefully and completely, and keep your eye on the pellet.
Both protocols listed below are for 50 ul PCR reactions, which tend to give better yields after purification, especially for larger products. For PCR reactions greater than 50 ul, proportionally scale up the reagents listed. For reactions less than 50 ul, add water to bring the volume to 50 ul. If mineral oil is present in the PCR reaction, add 100 ul of chloroform, (do not mix with PCR solution), spin at 10,000 RPM for 30 seconds and then aspirate supernatant into a clean tube. The chloroform/mineral oil component will settle to the bottom.

Protocol 1)
- Thoroughly mix 25ul of 7.5M ammonium acetate with 190 ul 95% ethanol. Add purified supernatant (if mineral oil was previously used) or your PCR solution to this mixture. Vortex and place on ice, or at -20ºC, for 30 minutes to precipitate PCR products. Centrifuge at 10,000 RPM for 10 minutes at 4ºC. Completely remove supernatant, and dry the pellet, either by air-drying or in a vacuum centrifuge (do not overdry, a few minutes in a vacuum centrifuge will be fine). Resuspend in 20 ul of water.

Protocol 2)
-Thoroughly mix 5 uL of 3M sodium acetate, pH 4.6, with 100 uL of 95% ethanol. Add purified supernatant or your PCR solution to this mixture. Vortex and place on ice, or at -20ºC, for 30 minutes to precipitate PCR products. Spin tubes in a microcentrifuge at maximum speed for 20 minutes. Completely remove supernatant and discard. Add 300 ul of 70% ethanol to rinse the pellet, vortex briefly, and spin for an additional 5 minutes at maximum speed. Completely remove supernatant and discard, being careful not to lose the pellet. Dry the pellet and resuspend in 50 ul water.

-Multiple PCR bands

Gel purification

When purifying a PCR product by gel filtration, either standard or low melting point agarose with TBE or TAE buffer may be used. However, it is important to use a high quality agarose such as FMC SeaPlaque or SeaPrep to minimize carryover of unwanted contaminants that may inhibit the sequencing reaction. When purifying the PCR band, the gel containing the PCR product of interest is first visualized under a UV light source and the band is then excised out of the agarose gel with a clean, sharp razor - minimize the amount of agarose that is cut out with the band, even if it means losing a bit of DNA. Tiny bits of agarose can have an adverse affect in the sequencing reaction. In addition, it is important to remember that UV light is damaging to your DNA - damage can occur in the time it takes you to cut out your band and it’s effect can become apparent when sequencing your DNA as you may see dramatically shortened read lengths. So whenever possible, reduce the intensity of the UV light source, reduce your exposure times or use a 366 nm UV source if available. Placing a glass plate between the light source and your gel can also help lessen the degrading effect of the UV light on your DNA. Once your band of interest has been isolated from the gel, numerous methods can be used to separate the PCR product away from the agarose slice, including the Qiaquick Gel Purification kit (MOST HIGHLY recommended), Millipore Ultrafree kit with modified TAE buffer, Promega PCR Wizard preps and Bio101 Geneclean kits. Follow the manufacturer’s direction carefully.

PCR products should be examined and quantitated on an agarose gel AFTER gel purification as well, as you will not get 100% recovery and your assumed PCR concentration for sequencing will not be accurate.


BAC/PAC DNA

Considerations when cleaning up BAC/PAC DNA

When attempting to sequence large DNA templates such as bacterial artificial chromosomes (BACs) or P1 derived artificial chromosomes (PACs), the quality of DNA template becomes even more crucial. As with plasmid DNA purification, Qiagen kits give excellent results for purification of BAC DNA for sequencing and they provide a few different methods for cleaning up the template DNA, including Qiafilter cartridges and a Large-Construct kit, the latter being somewhat more involved. One modified protocol of the Qiafilter Plasmid Midi Kit, which has consistently given good quality BAC DNA for sequencing, has been kindly provided to us from Wenjie Wu and we will post his modifications here.

-use 100 ml of BAC culture
-P1 buffer: add 10 ml
-P2 buffer: add 10 ml
-P3 buffer: add 10 ml
- after addition of P3 buffer, incubate the sample on ice for 15 min, then centrifuge at 4000 rpm for 30 minutes at 4ºC (no incubation at room temperature).
-add the supernatant from the centrifuged sample to the cartridge and filter the samples into a clean tube
-apply the filtered supernatant to the Qiagen-tip for further purification and follow remaining Qiagen protocol.
-when eluting the DNA, first warm the QF solution to 65ºC and elute 5 x 1ml to maximize yield.

Other BAC purification methods can be used, including CsCl purification, and some other protocols are available at the following websites:

University of Oklahoma Advanced Center for Genome Technology:
http://www.genome.ou.edu/DblAcetateProcV3.html

Washington University School of Medicine Genome Sequencing Center:
http://genome.wustl.edu/tools/protocols/index.php?lgcdseqprep=1

Note: BAC DNA prepped on the Autogen instrument generally does not work well for sequencing

Methods for Quantitation

There are three methods commonly used for quantitating DNA - spectrophotometry, agarose gel analysis and fluorometry. These methods can also give you a sense of how clean your DNA is, although none of these methods will detect everything. When using the spectrophotometer, you will typically take readings at both A260 and A280, as DNA will absorb at A260 while protein absorbs at A280. These readings will give you an A260/A280 ratio. A good DNA A260/A280 ratio should be between 1.7-1.9. If you find your ratio is smaller, this could indicate contamination of your sample with proteins or organic chemicals. If possible, use a spectrophotometer that is capable of performing a wavelength scan from 220 nm to 330 nm as a scan may be useful in revealing other contaminants such as phenol, which is detectable at 230nm, and particulates in solution that may absorb at 325nm. When calculating your DNA concentration with a spectrophotometer, you will need to know that one A260 O.D. unit (absorbance) of double-stranded DNA contains 50 ng/ul. One O.D. unit is the amount of a substance dissolved in 1 ml that gives an absorbance reading of 1.0 in a spectrophotometer with a 1-cm path length. So to calculate the concentration of DNA you have you will multiply the absorbance value times the conversion factor of 50ng/ul times any dilution factor you may have used and this will give you the concentration of your sample in ng/ul or mg/ml. However, quantitation of your DNA by spec is probably the least accurate of the three methods as both RNA and contaminating DNA will also absorb at 260nm and can potentially lead to inaccurate values. Also, readings greater than 1.0 OD or less than 0.05 are probably not very accurate, limiting the sensitivity and range of this method.

Quantitating by agarose gel electrophoresis tends to be more accurate as you can visualize any contaminating DNA or RNA. Purified DNA should run as a single band on an agarose gel, while uncut plasmid DNA will show three bands: supercoiled, nicked and linear. This method is also useful when quantitating small amounts of DNA such as PCR products. The DNA sample is run on a gel next to a mass ladder composed of fragments of known and varying concentrations and the intensity of your DNA band is visually compared to the intensities of the bands from the mass ladder to estimate DNA concentration. This takes a bit of experience and practice, so you may want to get someone to help you the first few times. And keep in mind that when using an agarose gel, you can’t detect the presence of proteins or chemicals so to get the most information about the amount and quality of your desired DNA, both spectroscopic and gel quantitation methods should be used together.

Perhaps the most accurate method of measurement of the three is the use of a fluorometer. Dyes may be added to the DNA such as Hoechst 33258 Dye or Picogreen which intercalate into AT rich regions of double stranded DNA and are thus quite specific to double stranded DNA, avoiding the problems of contaminating RNA seen when using a spectrophotometer. Common contaminants present in DNA preparations such as salts, urea, ethanol, chloroform, detergents, proteins and agarose do not have significant affect on these assays and thus lead to a much more sensitive and precise method for quanititation. Fluorometric measurements can detect DNA levels as low as 5 ng/ul.


Primer design

If you are already familiar with the recommendations for designing PCR primers, then you’ll find the same apply for sequencing primers, for the most part. If not, then here are some guidelines for sequencing primer design. But keep in mind that even if you follow these guidelines to the letter, some primers just do not work well and the reason is often not clear. On the other hand, if you don’t strictly follow these recommendations and just pick a primer close to your area of interest by eyeballing, oftentimes the primer will work just fine. In addition, there are many programs on the Internet that can help you choose primers, determine potential primer-dimer formation or hairpin structures, and calculate Tm. These programs can be especially helpful when you need to design multiple sets of primers to ensure coverage of a large region of DNA. Look at our section Useful Links for some primer design programs that we have found helpful. That being said, here are some common rules for sequencing primer design:

* primer should be 18-24 nucleotides in length - this helps ensure good and specific hybridization to the template DNA. Long primers (>35-40bp) tend not to work quite as well in sequencing due to the increased chances of secondary structure formation

* G/C content should be approximately 50%, but the range can be 30-80%

* Tm should be between 50º-60ºC - if your Tm is much below that it may not hybridize properly, as the annealing temperature in our cycle sequencing reaction is 50ºC. If necessary, add a few more bases to your primer to increase its Tm. To calculate the Tm of your primer use the following formula:

Tm = 4(G + C) + 2(A + T)º C

* avoid primers with runs of a single nucleotide, especially runs of four or more Gs

* primer should not form primer-dimers with itself (a primer-dimer can occur when nucleotides of one primer molecule anneal to complementary sequences on another primer molecule and form a stable interaction, thus reducing the amount of primer that can participate in the sequencing reaction)

* primers should not form hairpin or secondary structures (a hairpin can occur when one section of the primer is complementary to another section within the same primer molecule and can cause the primer to fold upon itself. It is especially important to avoid secondary structure at the 3’ end of the primer as the primer will not be available to begin synthesis of a new strand by the Taq polymerase).

* avoid palindromes as they are more prone to forming secondary structures.

* whenever possible, make sure that your primer does not have a second hybridization site as the primer will bind in two places of different nucleotide sequence and will show noisy data with many extra peaks

* degenerate primers do not generally work well for sequencing

*It is often suggested that the 3’ end of a primer should have a G or C as the last nucleotide to act as a "clamp" (these nucleotides bind most strongly) when the primer anneals to the template DNA, as the clamp will help avoid "breathing" of the ends. Lately, however, it has been suggested that the use of A/T rich 3’ ends will give greater specificity as the weaker bonds formed by A and T pairing will only stay hybridized if the match is perfect. However, in practical terms, the choice does not really appear to be critical.

As a last note on ordering primers - while purity of a primer is important, it is generally not necessary to have ultrapure primers, such as those purified by HPLC or OPC cartridge, though that is, of course, optimal. As long as the oligo synthesis efficiency was high and the majority of your primer solution contains full-length primers, then simple desalting of your primers should be fine. Primers of poor synthesis quality will not work well for sequencing because, along with your full-length product, there will be a greater proportion of "failure" primers (i.e. truncated primers of less than full length, in varying sizes) that can also act as sequencing primers. Sequence data from these primers will be poor, ranging from unusable to minor "shadow" peaks under the proper peaks. See more in our Troubleshooting section to see how to recognize this problem. When submitting custom primers for sequencing, dilute them in water to the proper concentration.