| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
Departments of1 Laboratory Medicine and Pathology and3 Medicine, Mayo Clinic, Rochester, MN;2 Idaho Technology Inc., Salt Lake City, UT
aaddress correspondence to this author at: Endocrine Laboratory, Hilton 730C, Mayo Clinic, 200 1st St. SW, Rochester, MN 55905; fax 507-284-9758, e-mail grebs{at}mayo.edu
During the last decade, research and clinical use of real-time PCR applications has continued to grow in importance (1). Many laboratories that use real-time PCR with fluorescent probes experience an unexplained loss of probe fluorescence at some stage, in particular with pairs of fluorescence resonance energy transfer (FRET) probes. Photobleaching is often assumed to be the cause. Structural integrity of the oligonucleotides is also a major factor, and its loss has been shown to correlate with repeated freeze–thaw cycles (2). Laboratories guard against these two problems by aliquoting probes and protecting them from light. Despite these precautions, inexplicable FRET probe failures are still observed. In one recent such case, we were able to determine an additional mechanism for FRET probe failure: loss of the phosphate cap from the 3' end of a probe. To our knowledge, this has not been described previously. Our studies revealed that this may be a common and important problem, intrinsic to 3'-phosphate-blocking chemistry. We also found that alternative terminating groups may be a preferable option to 3'-phosphate blocking.
A 3-nmol/L synthesis-scale LightCyclerTM hybridization probe set was purchased from Idaho Technology Inc. Biochem in April 2003. As is common practice, the manufacturer produced a large-scale synthesis and, after shipping our order, archived the remainder for a possible future reorder. The first half of the batch (
-probe set) was sent immediately, whereas the second half (ß-probe set) was stored lyophilized for 6 months at –20 °C and then shipped with our next order.
Oligonucleotides from the first shipment were used in PCR reactions in a LightCycler with satisfactory results. PCR conditions were as follows: 1x LightCycler FastStart DNA Master Hybridization Probe Mix (Roche Diagnostics), MgCl2 (final concentration, 3.5 mM), 0.5 µM each of the forward (5'-GGCCTTTCTGAAGCAAG-3') and reverse (5'-GACGATTTCTTATTTCACAGCTCC-3') primers, 0.2 µM each of the donor (5'-GGACGCAGAGGGGATGG-FITC-3', where FITC is fluorescein isothiocyanate) and acceptor (LCRed640-GTGTATGGGACCCGCCAG-phosphate) probes, and 2 µL of cDNA template mixture. The final reaction volume was 10 µL. The reaction started with an initial melting step at 95 °C for 10 min followed by 45 cycles of 95 °C for 2 s, 57 °C for 10 s, and 72 °C for 5 s.
Loss of fluorescence activity was first observed when we received the ß-probe set. PCR reactions were carried out under the same conditions, but the amplification curves displayed extremely low signal intensities relative to the -probe set that was being replaced (Fig. 1A
). We excluded photobleaching by measuring raw probe fluorescence. We then set up two additional PCR reactions, pairing each -set of oligonucleotides with the corresponding ß-set FRET partners. The loss of fluorescent activity corresponded with the ß-set LCRed640 probe, whereas the -set LCRed640 probe and the fluorescein donor probes from both the - and ß-sets were fully functional (Fig. 1A
).
|
Agarose gel electrophoresis demonstrated a diminished target band at 389 bp as well as a strong additional band between the 50- and 150-bp size markers in the reaction that contained the suspect ß-set LCRed640 probe. The extra band in the ß-probe reaction corresponded to the size of a PCR product that would have been formed if the ß-probe had acted as a primer. Additional PCR reactions were run to explore this possibility, including the use of an unlabeled primer identical in sequence to the LCRed640 probe, which confirmed our hypothesis (Fig. 1B
). Finally, we excised the low-molecular-weight band in the ß-probe reaction from the gel and sequenced it, using the reverse PCR primer as a sequencing primer. The DNA fragment had the sequence predicted if the ß-set LCRed640 probe had acted as a PCR primer, including the full acceptor probe sequence, and terminated at its 5' end (data not shown).
In a typical FRET probe arrangement, it is critical that neither the donor nor the acceptor probe can be elongated during PCR. For the donor probe, the fluorescent dye at the 3' end blocks PCR extension, whereas for the acceptor probe, a nonfluorescent blocking agent, typically a phosphate group, is linked to the 3' hydroxyl of the terminal base. If a blocking group is lost, then a probe can act as a PCR primer. Because of the shorter length of the PCR product formed by this aberrant priming and the higher melting temperature (Tm) of the probes compared with the PCR primers, formation of these aberrant products will be greatly favored over the intended full-length target. If the blocking group is lost from the 3' FRET probe, very few targets will be produced that contain sequences complementary to both the 5' and the 3' FRET probes; hence, little or no detection signal will be observed. Our results demonstrate that this has indeed happened and that extension has occurred from our acceptor probe, implying loss of its 3'-phosphate cap.
We suspect that this is the cause of a significant proportion of FRET probe failures. When we evaluated an unrelated probe set that had suddenly and inexplicably failed and was also purchased in two batches (from Genset Oligos), we found a low-molecular-weight PCR product that was consistent with phosphate-cap loss from the 3' probe (data not shown). We therefore decided to explore whether alternative 3' capping moieties might be less susceptible to degradation. After exploring the available chemistries and their costs, we selected a C3 carbon spacer (cat. no. 20-2913; Glen Research) as a blocker of potentially enhanced stability. A 72-base single-stranded template was designed and synthesized as a complement for probe hybridization (5'-GTCCCTTAAGTAACTAGAATAATGGAATTGGGCTCCTTATAATCAAGCACTCATAACAACATAATCATTGC-3'). Probes were then designed such that the donor fluorescein probe (5'-AATGATTATGTTGTTATGAGTGCTTG-FITC-3') acted as the anchor with a Tm of 62 °C. Three LCRed640 probes were synthesized with identical sequences (5' LCRed640-TATAAGGAGCCCAATT-3') and different 3' terminations. The first had no cap and served as a positive control, the second had a standard phosphate cap, and the third was synthesized with a 3' C3 carbon spacer. Each probe was stored in both water and Tris-EDTA (738 mmol/L Tris-HCl, 0.5 mol/L EDTA, pH 8.3) at 4 and –20 °C. On a weekly basis, the probes were tested in a LightCycler in separate primer extension reactions followed by melting curve analysis, The 10-µL reactions contained (final concentrations) 1x PCR buffer (Idaho Technology Inc.), 4 mM MgCl2, 0.2 mM each of the deoxynucleotide triphosphates, 0.1 MU/L KlenTaq1TM (Ab Peptides, Inc.), and 0.2 µM each of the complementary strand, LCRed640 probe, and fluorescein probe. The melting curve analysis conditions were 3–15 cycles of 94 °C for denaturing and 50 °C for annealing, followed by melting from 35 to 85 °C. In this experiment, any loss of 3' capping of the acceptor FRET probe will lead to probe extension. As the acceptor probe becomes extended, its Tm increases from 51 °C to 
70 °C, making it the anchor for formation of a secondary peak at 62 °C in the negative- derivative melting profiles (Fig. 1C
). This secondary peak corresponds to the dissociation of the fluorescein donor probe. For every secondary peak found, we calculated the peak area and determined the percentage of signal in the secondary peak compared with the total signal. This percentage was used to estimate the amount of extended probe per reaction.
At 4 °C, the 3'-phosphate terminus was unstable in both Tris-EDTA and water. By contrast, the three-carbon cap remained stable throughout all storage conditions tested. Even when stored at 4 °C for 8 weeks, the C3 probe showed no indication of probe extension (Fig. 1C
and Table 1
). None of the oligonucleotides stored at –20 °C showed evidence of degradation during 8 weeks. However, in circumstances in which different buffer systems are used and freeze–thaw cycles are more common, one might see a higher amount of phosphate cap degradation even at lower temperatures.
|
Many factors can induce cleavage of the 3'-phosphate ester bond, which is an intrinsically unstable bond, but most are likely to involve hydrolysis. Therefore, high salt concentrations and acidic or basic pH will probably play a major role (3). Oligonucleotide synthesis involves many chemicals, in particular ammonia compounds, that would fit this profile (4). Various postsynthesis purification methods are aimed at removing these substances, but on occasion, hydrolytic attack on reconstitution of the probe may occur as a result of even minor variations in lyophilization efficiency or pH of the chromatography eluent or as a result of residual salts trapped in the lyophilysate, e.g., as a consequence of an exhausted chromatography column. The fact that the loss of the phosphate cap seems to be slightly accelerated in Tris-EDTA might suggest that microbial contamination could also play a role. The organic matrix and the buffering conditions of Tris-EDTA may allow microorganisms to proliferate and produce catabolic enzymes.
Because of the greater stability of a phosphodiester linkage, as used in the C3 linker, relative to a single phosphoester bond, it appears that all of these concerns can be alleviated by replacing the phosphate cap with an alkyl group via common phosphoramidite chemistries. In our case, we have demonstrated that the C3 linker is indeed an excellent choice for such a blocking agent, but alternatives may include other carbon spacers and 3'-terminal dideoxynucleotides, although the latter would add substantially to synthesis costs. Users interested in such alternatives should discuss availability with their oligonucleotide manufacturers. For probes with conventional 3'-phosphate caps, we would recommend frozen storage, ideally lyophilized, and minimization of any storage time at 4 °C or above, as well as any handling that might allow bacterial contamination.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  C. N. Gundry, S. F. Dobrowolski, Y. R. Martin, T. C. Robbins, L. M. Nay, N. Boyd, T. Coyne, M. D. Wall, C. T. Wittwer, and D. H.-F. Teng Base-pair neutral homozygotes can be discriminated by calibrated high-resolution melting of small amplicons Nucleic Acids Res., June 1, 2008; 36(10): 3401 - 3408. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Dames, R. L. Margraf, D. C. Pattison, C. T. Wittwer, and K. V. Voelkerding Characterization of Aberrant Melting Peaks in Unlabeled Probe Assays J. Mol. Diagn., July 1, 2007; 9(3): 290 - 296. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. T. Seipp, J. D. Durtschi, M. A. Liew, J. Williams, K. Damjanovich, G. Pont-Kingdon, E. Lyon, K. V. Voelkerding, and C. T. Wittwer Unlabeled Oligonucleotides as Internal Temperature Controls for Genotyping by Amplicon Melting J. Mol. Diagn., July 1, 2007; 9(3): 284 - 289. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
