Find answers below containing general information, troubleshooting tips, and recommended controls on T7-based DNA linear amplification (TLAD).

Last updated 3/29/2006 by Chih Long Liu

[1]: General Questions about the method

• [1.1] What is DNA linear amplification?
• [1.2] What is TLAD?
• [1.3] How does TLAD work?
• [1.4] What advantage does TLAD have over LM-PCR, R-PCR and other exponential-based amplification methods?
• [1.5] What advantages do LM-PCR, R-PCR and other exponential-based amplification methods have over TLAD?
• [1.6] What applications can benefit from TLAD?
• [1.7] What updates have been made to the TLAD protocol since 2003?

[2]: Getting Started

• [2.1] What is a suitable DNA template for TLAD?
• [2.2] How should my T7 primer be synthesized?
• [2.3] Why do I need to use a cacodylate-containing buffer for the TdT step?
• [2.4] Where can I find ddCTP?
• [2.5] Why is there an extra Buffer RPE wash in the post-IVT cleanup?
• [2.6] How long will it take to amplify a sample?
• [2.7] How much will it cost per sample?
• [2.8] What reverse transcription protocol should I use?

[3]: Troubleshooting

[3.1]: Low yields

• [3.1.1] Why am I getting low yields?
• [3.1.2] What are the suggested positive controls?
• [3.1.3] Why is my positive IVT control not working?
• [3.1.4] How can I eliminate RNase contamination?
• [3.1.5] Why is my positive amplification control not working?
• [3.1.6] I think the Qiagen MinElute columns may be having problems. How can I fix this?

[3.2]: Other problems

• [3.2.1] I'm getting template-independent product (TIP). What can I do to reduce or eliminate this?
• [3.2.2] Will dsRNA form in my amplified product, and if so, will it negatively impact reverse transcription and microarray hybridization?
• [3.2.3] Will the 5' poly-A tracts in the aRNA interfere with downstream applications?
• [3.2.4] I'm getting low cDNA yield in my reverse transcription reaction. Why?

A: Amplification of DNA is a method that takes template DNA and makes many copies of it. PCR (polymerase chain reaction) is one prime example. However, methods such as PCR are exponential amplification methods, meaning that the number of copies made of the template DNA increases at an exponential rate.

A: TLAD is T7-based Linear Amplification of DNA. This amplification method was first published in BMC Genomics 4:19 on April 2003, with an extensive update in October 2005, appearing in Chapter 7 (PDF) of the Whole Genome Amplification Methods Express Series books from Scion Publishing Ltd. A number of studies (Genome Biology 5(9) R62, Cell 120(2) 169-81, PLoS Biology 3(10) e328, Cell 123(2) 233-48) have been published that employ this method. It is one of the first amplification methods developed to amplify complex mixtures of DNA in a linear fashion. It was developed using an in vitro transcription strategy used commonly in RNA amplification methods.

A: Figure 1 from BMC Genomics 4:19 is a diagram of the method. TLAD begins with a terminal transferase (TdT) tailing of the template DNA, followed by annealing with a primer adapter containing the T7 promoter sequence. After second strand synthesis with DNA Polymerase I Klenow fragment, in vitro transcription of the template takes place, generating aRNA amplification product. For further details, refer to the BMC Genomics publication.

A: The largest shortcoming of exponential amplification approaches is their diminished fidelity when amplifying complex mixtures of DNA. In an exponential amplification, variances or fluctuations in the amplification efficiency of a given amplicon early in the amplification will themselves amplify to considerable and detectable differences at the end of the amplification, because of the exponential kinetics. In linear amplification, however, the linear amplification kinetics minimizes such variances in outcome relative to exponential amplification. The differences in fidelity is described in detail in the BMC Genomics publication.

A: The main advantage that LM-PCR and R-PCR have over TLAD is that they can produce microgram amounts of amplified product from subnanogram quantities of starting dsDNA template. This is particularly important for methods such as ChIP-chip (reviewed here) if the epitope being immunoprecipitated is rare and/or the amount of chromatin limiting.

A: TLAD was primarily designed as a protocol to replace R-PCR in ChIP-chip experiments. The ChIP-chip method is essentially Chromatin ImmunoPrecipitation followed by analysis using DNA microarrays, or "chips". Thus, the ChIP-chip method stands to benefit the most from TLAD. However, other applications that require large amounts of amplified material and/or require the improved amplification fidelity offered by a linear amplification method can stand to benefit from TLAD.

A: A number of minor and not-so-minor updates have been made to the TLAD protocol since 2003, when the BMC Genomics publication was first published. Here is a brief list of the more notable changes:

• Reaction volumes for TdT and Second Strand Synthesis halved. This cuts enzyme usage by 50% for those two steps.
• NEB switched the buffer provided with the DNA Polymerase I Klenow fragment (5000 U/ml) from EcoPol Buffer to NEB Buffer 2.
• A reaction volume table is now provided for the Second Strand Synthesis step, to minimize template independent product formation.
• Qiagen implemented a new requirement that the MinElute columns be stored at 4C when not in use.
• A suggested T7 RNA polmerase enzyme boost step is now provided for small dsDNA templates (<300 bp)

A: The criteria defining a suitable dsDNA template consist of the following:

• Size between 100-1000 bp. Anything smaller than 100 bp won't be efficiently recovered by the Qiagen columns used in the TLAD method. Sizes between 1-5kb are probably permissible, though TLAD has not been tested extensively with that size range. Sizes >5kb should probably be amplified with an alternate method, such as strand displacement.
• Absence of 3' phosphate groups. Terminal transferase requires free 3' hydroxyl (-OH) groups in order to add the homopolymer tail. Some DNA fragmentation methods (sonication, nuclease digestion, and yes, certain restriction digests) leave behind 3' phosphate groups. Use the optional CIP method to remove these groups.
• Low incidence of 3' recessed ends. Terminal transferase tails best with dsDNA with either blunt or protruding 3' ends. Sonication and certain restriction digests can generate dsDNA with a significant proportion of 3' recessed ends. If yields are low, fill-in of the 3' recessed ends with Klenow fragment DNA Polymerase I is suggested.

A: The "B" at the end of the T7 primer sequence means that the last base is degenerate for C, G, or T. In other words, when you have the primer synthesized, the end result should be a mixture of primers containing identical sequences except for the last base, where a third will end in C, a third end in G, and a third end in T. The "B" serves as an anchor. Also, be sure to obtain HPLC or PAGE purification for the primer, since desalting the primer leaves behind shortened termination products that may lower the yield and/or result in generation of template independent product.

A: During the development of the TLAD method, the Roche TdT enzyme (from calf thymus) was originally used. However, in our hands the TdT enzyme did not perform consistently and varied lot to lot, which we speculated was due to this enzyme having been derived from a natural source (this seems to be tacitly acknowledged (our speculation) by the fact that they no longer offer the calf thymus-derived TdT and that they made this announcement (pdf)). The switch was made to NEB recombinant TdT (Roche hadn't offered a recombinant TdT at that time; they do now). Curiously, however, the NEB TdT enzyme wouldn't tail when prepared with the provided reaction buffer (NEB Buffer 4), and it was determined that this was due to the dTT in NEB Buffer 4 precipitating the cobalt chloride in the reaction solution. When the cacodylate-containing TdT buffer (provided with the Roche enzyme) was tried, however, the NEB TdT performed satisfactorily and consistently. Thus, we strongly recommend using the cacodylate-containing buffer, despite the inconvenience of having to work with something containing arsenic. We recently learned that the Roche TdT buffer is now available (1 ml, $6, catalog # 11 243 276 103). However, it may be necessary to contact them by phone in order to order this item, as it currently does not appear on their website. (In the past, one would have to purchase the Roche TdT enzyme in order to obtain pre-made TdT buffer).

A: ddCTP, or dideoxycytidine triphosphate, is the dideoxy form of CTP and is commonly used in the Sanger method for DNA sequencing. It is used as a tail terminator in the TdT tailing reaction, allowing for a relatively tight tail size distribution in the population of dsDNA template molecules being tailed. It was once available (pdf) from Invitrogen (our originally recommended supplier), who has since stopped carrying the product. The following sources carry ddCTP, although we have not tested them: VWR, Fisher Scientific, and Sigma. Amersham also carries this product, and we have tested it successfully. We currently recommend Amersham's product over the others, since VWR and Fisher are reselling the ddCTP from other lesser known sources, and the Sigma product is in a dry lyophilized salt form.

A: In the post-IVT cleanup, the Qiagen RNeasy kit is used to purify RNA from the IVT reaction mixture. The protocol followed is similar to the RNA reaction cleanup protocol provided by the kit handbook, except for an extra Buffer RPE wash. We have found this wash to be necessary if the RNA is to be used for subsequent microarray applications involving fluorescence, since it is necessary to remove any trace GITC salts that may otherwise remain and produce undesirable background fluorescence on the microarrays.

A: It takes approximately an afternoon and the following morning. For 1-10 samples, it takes about 3.5-4.5 hours during the afternoon and about 1-1.5 hours during the following morning. Note that this is the time it takes to go from the tailing reaction to aRNA product. The optional CIP treatment adds another 1.5 hours, and the optional reverse transcription reaction will take an additional 4-6 hours (depending on the protocol used). FYI, it is possible to amplify up to 100 samples in two 12-hour days without robotic assitance, though this requires that all necessary tubes be labeled in advance.

A: A single sample will cost approximately $16.54 to amplify (May 2005 prices). This includes the cost of the optional CIP treatment. However, this does not include consumables or common reagents, such as tips, tubes, nuclease-free water, or nucleotides. Furthermore, the cost per sample is calculated based on bulk-size purchases of the appropriate enzymes and kits. Smaller-size purchases may increase the cost by 20-30%.

A: Any reverse transcription (RT) protocol you have experience with should be acceptable; the only requirement being that the RT reaction be primed with degenerate primers (e.g. pdN6 random hexamers). Typically 5 ug of pdN₆ is used for 2-4 ug of aRNA. The protocol we typically use is here (pdf). This is also downloadable from the Downloads section.

A: There are a whole myriad of factors that may contribute to low amplification yields. The easiest way to determine the cause is to run the two suggested positive controls, since this will quickly identify where the problem is occuring in the TLAD method.

A: One is the IVT control. Here, the IVT starting material is 250 ng of the pTRI-Xef linearized plasmid provided with the Ambion IVT kit. If not using the kit, an appropriate amount of a dsDNA template that contains the pT7 promoter, with prior history of use as a successful T7 RNA polymerase template, should be used. Yields should typically range from 100-140 ug, limited by exhaustion of the nucleotides in the reaction mixture and the rated 100 ug binding capacity of the Qiagen RNeasy column. Visualization on a non-denaturing agarose gel should yield two intense bands ~0.9 kb and ~1.5 kb..

A: This is likely due to suboptimal IVT conditions, which can be any of one of the following:

• RNase contamination. For getting rid of RNase contamination, see [3.1.4].
• NTPs went through too many freeze-thaw cycles. NTPs are very sensitive to freeze-thaw cycles and should go through no more than three. Try aliquoting the NTPs into as many aliquots as necessary.
• Excessive evaporation during the long incubation time. Use smaller-volume tubes.

A: RNase contamination can be the bane of labs that routinely work with RNA. RNases are everywhere - on benchtops, glassware, non-certified tips and tubes, and on the surface of your skin (to name a few locations). RNase A, the main culprit in undesired RNA degradation, survives the autoclaving process, so autoclaving tips, tubes, and glassware is of no use in decontaminating RNases. A more detailed article on this subject can be found here.

A: This can be due to a myriad of factors, though the most common reason is if an unsuitable or suboptimal dsDNA template is used. See [2.1] for criteria defining a suitable dsDNA template.

It's also possible that the Qiagen columns (particularly the MinElute columns) may have issues; see [3.1.6].

Another possibility is that the Terminal Transferase tailing reaction is not working properly, resulting in the lack of poly-T tails at the 3' ends of the template DNA, resulting in a lack of annealing sites for the T7 primer in the subsequent second strand synthesis reaction. Below is an image of a 2% agarose in 1X TAE gel taken during development of the TLAD method.

Reaction conditions: ~100 ng 190 bp PCR product, 20 ul total reaction volume, 20 min. incubation at 37°C. Amounts are end concentrations of NEB TdT enzyme used in each reaction. 2.0 U/ul is the end concentration used in the official protocol. NEB TdT is provided at 20 U/ul.
Note: this reaction was done before ddCTP-facilitated tail termination was optimized in the protocol, hence the long tail lengths.

If you suspect that your TdT tailing reaction is the culprit, you should conduct a TdT tailing reaction, scaled up to a 20 ul volume, with PCR product between 150-1000 bp. Purify your product with the Qiagen MinElute Column, then run it on a 2% agarose gel. You should get a smear approximately 20-40 bp longer than your original product. If you get an incomplete smear (such as what you would find in the 0.5-1.5 U/ul lanes), you may want to try increasing the amount of enzyme used, or use a fresh lot of enzyme. Any untailed DNA will not produce IVT-valid template after second strand synthesis, and your final yield will drop accordingly!

back to top

A: The Qiagen MinElute columns usually perform consistently well, but they do have a number of attributes that require proper attention for maximum performance:

• The suggested elution volume is 20 ul, twice the manufacturer's recommended volume of 10 ul. This is because proper and complete wetting of the column matrix is essential in maximizing yields. We have found that using the 10 ul volume may reduce yields by up to 50%, depending on the efficiency in which the column matrix was wetted.
• Qiagen began recommending in 2004 that the columns be stored at 4^oC while not in use. If you have some columns that have been sitting at room temperature for months, we suggest using a fresh batch.
• The MinElute column should be carefully washed during the washing step to ensure that there are no traces of Buffer ERC. We haven't found major problems with trace amounts of Buffer ERC in the performance of the TLAD method, but we do notice that it sometimes interferes with quantification by A₂₆₀. In those cases, contaminating Buffer ERC may produce a high linear spectrum with a slope of -1 from 220-400 nm, obscuring the nucleic acid spectrum and peak at A₂₆₀. To avoid this, ensure that when pipetting Buffer PE into the column that the inside lip of the column (which is recessed and flush with the bottom edge of the inside cap when closed) is irrigated with Buffer PE during the pipetting step. Buffer ERC can occasionally be trapped there and provide a source of trace salts that can interfere with subsequent UV absorbance measurements.

A: TIP most frequently appears as a distinct 100 bp band. You will find an example of this in the WGA Methods Chapter 7 (PDF), Figure 3, Lane 4. We speculate that it is caused by IVT-valid dsDNA template formed from T7 primer-dimers. TIP usually occurs if the primer to template dsDNA mass ratio significantly exceeds 5:1. You can do two things to mitigate this:

• Ensure that the starting amount of dsDNA used is accurately quantified. Most UV spectrophotometers take cuvettes with a minimum of 50-100 ul and have a lower detection OD of 0.01 absorbance units. If your spec sample's concentration is less than 0.5 ng/ul in the cuvette, then your UV spec will not accurately quantify your sample. Use a spec that takes small cuvettes (1-10 ul), use a NanoDrop, or use fluorescence such as PicoGreen.
• Use the correct amount of primer and appropriate reaction volume for Second Strand Synthesis. Refer to the enclosed table in the protocol available in the Download section.

A: It's certainly possible, though we have not tested this directly. In our experience, our reverse transcription typically yields around 80-85% for reverse transcrition of aRNA amplified from sonicated dsDNA having an average size of 600 bp. We did test and determine, in our hands, that a low-complexity mixture (~300 unique DNA species) may have a reduced dynamic range of 60-70% of a corresponding mixture where each species was single-stranded and missing its complementary strand. However, most genomic DNA mixtures are likely to be more complex and consequently less likely to suffer from this problem. Furthermore, R-PCR and LM-PCR (described in [1.4]) would also have the same problem since they also amplify both strands of the dsDNA template.

A: The 5' poly-A tracts do not appear to interfere, at least as it applies to the ChIP-chip method. As far as microarray hybridization is concerned, as long as the microarray probes (i.e. the DNA probe adhered to the microarray surface) do not contain significantly large poly-A tracts, these poly-A tracts should not cause any problems. Furthermore, one can optionally follow the lead of gene expression profiling hybridization protocols by adding in poly(A) or poly(dA), which is used to block the poly-T tracts in the cDNA probe. Both poly(A) and poly(dA) are available from Sigma.

A: This may occur due to the following reasons:

• Insufficient random primer used. We use a reverse transcription protocol where 5 ug random hexamer is used for every 2-4 ug of aRNA.
• aRNA is insufficently denatured prior to primer annealing due to dsRNA formation and/or presence of large GC-rich tracts. Try increasing the denaturing temperature and/or lengthening the denaturing incubation step.
• aRNA is short (<300 bp). aRNA is random primed, so most of the reverse transcripts will be shorter than the aRNA template. For small templates, and in most common reverse transcription protocols that use spin columns as a post-RT cleanup method, this may mean that a significant proportion of the reverse transcripts may be lost due to their being smaller than the molecular weight cutoff of the spin column being used. There isn't really any good way around this except to either use more starting aRNA or to use a spin column with a lower molecular weight cutoff. (Note that an anchored oligo-dT primer will be of no use here because the poly-A tract on the aRNA is at the 5' end of the molecule!)

back to top

For those of you not intimately familiar with TLAD, we recommend that you read Chapter 7 of the Whole Genome Amplification Methods Express lab manual, which has just been published (Oct. 2005).

Last updated 3/29/2006 by Chih Long Liu