| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Abstracts of Oak Ridge Posters |
(1 Centre de Recherche en Infectiologie de l’Université Laval, Centre Hospitalier Universitaire de Québec (Pavillon CHUL), Quebec City, Québec, Canada; 2 Division de Microbiologie, Faculté de Médecine, Université Laval, Quebec City, Québec, Canada;
aaddress correspondence to this author at: Centre de Recherche en Infectiologie de l’Université Laval, Centre Hospitalier Universitaire de Québec, Pavillon CHUL, 2705 Laurier Blvd., Quebec City, Québec, G1V 4G2 Canada; fax 418-654-2715, e-mail Michel.G.Bergeron{at}crchul.ulaval.ca)
Hybridization of double-stranded DNA with nucleic acid probes is hampered by competition between the complementary nontarget strand and the probe (1). This competition is stronger with surface-bound probes that expose a target strand with a long dangling end toward the media(2)(3). This situation is more problematic in infectious disease diagnostics, which requires high sensitivity, sometimes as low as 1 genome copy. Microarray integration into microfluidic systems results in increased speed and sensitivity while allowing automation of the whole hybridization process(4) and could lead to the development of point-of-care diagnostic devices. Actual limitations of diagnostic microarrays include the design of ultrasensitive capture probes that are highly specific and discriminant. Some techniques have been developed to produce single-stranded DNA targets for sequencing and hybridization, but these methods may necessitate additional steps or modification of the target strand or have poor sensitivities(5)(6)(7)(8). Degradation of the complementary strand can alleviate competition problems and allow more flexible probe design for the target strand. Therefore, we developed a simple 5-min method for rapid single-step selective digestion of the complementary strand with lambda exonuclease, which leads to increased hybridization signals and improved differentiation of single-nucleotide polymorphisms (SNPs) on DNA microarrays.
Cy3-labeled primers were used to generate the target strands, and phosphorylated primers were used to generate the complementary strands. Nonphosphorylated primers were also used to verify the protection provided by the Cy3-labeled primers. All oligonucleotides were purchased from Integrated DNA Technologies. Purified genomic DNA (1 ng) from Neisseria meningitidis (ATCC-13077), Listeria monocytogenes (CCRI-4862), and Candida krusei (ATCC-28870) were PCR-amplified with a PTC-200 thermocycler (Bio-Rad Laboratories; 1 min at 94 °C, then 40 cycles of 1 s at 95 °C for the denaturation step, 10 s at 60 °C for the annealing step, and 20 s at 72 °C for the extension step) in a PCR reaction mixture containing 50 mmol/L Tris-HCl at pH 9.1, 16 mmol/L (NH4)2SO4, 4.5 mmol/L MgCl2, 2.15 g/L BSA, 0.2 mmol/L dNTPs, primers at 1.0 µmol/L each targetting the TUF or TEF-1 gene (see Table 1 in the online Data Supplement that accompanies the online version of this Abstracts of Oak Ridge Posters at http://www.clinchem.org/content/vol53/issue11), and 0.05 kU/L KlenTaq (AB peptides).
Amplicons from N. meningitidis (42.1 mg/L), L. monocytogenes (37.1 mg/L), and C. krusei (42.5 mg/L) were digested with 10 units of lambda exonuclease (New England Biolabs) at 37 °C in the PCR buffer described above for 5 min unless otherwise noted. Nondigested control amplicons were denatured at 95 °C for 5 min and then cooled on ice for 3 min, whereas digested products were directly used for microarray hybridization without prior heat treatment. Digestions were monitored with capillary gel electrophoresis performed on the Bioanalyzer 2100 (Agilent Technologies). For the comparison between C. krusei phosphorylated (7.9 mg/L) and nonphosphorylated (7.6 mg/L) complementary strand digestion, amplicons were purified with the QIAquick PCR Purification Kit (QIAGEN) and then digested in lambda exonuclease buffer for 30 min. Microfluidic hybridization experiments were conducted with nonpurified amplicons from L. monocytogenes (23.1 mg/L) digested with the exonuclease for 5 min. Oligonucleotide probes (see Table 2 in the online Data Supplement) with a 5' amino-linker were suspended in Micro Spotting Solution Plus (TeleChem International) and spotted at 30 µmol/L on Super Aldehyde slides (Genetix) with a VIRTEK SDDC-2 Arrayer (Bio-Rad Laboratories).
All amplicons were subjected to hybridization at room temperature (22–25 °C) in 6x saline-sodium phosphate-EDTA (SSPE; EMD Biosciences), 0.03% polyvinyl-pyrrolydone (Sigma), and 30% formamide (Sigma). Passive hybridizations were conducted with 20-µL volumes containing 1.75–10 mg/L of target amplicons. Passive hybridization (1 h) was performed with a glass lifterslip (Erie Scientific) apposed to the microarray slide. Active hybridization (5 min) was achieved with a compact disc-based polydimethylsiloxane microfluidic device, as previously described (4). Washing was performed in 0.2x SSPE + 0.1% sodium dodecyl-sulfate, followed by rinsing in 0.2x SSPE. Slides were scanned with a ScanArray 4000XL (PerkinElmer), and the hybridization signals were quantified with GenePix 6 (MDS Analytical Technologies). A Cy3-labeled control oligonucleotide (see Table 3 in the online Data Supplement) was used to validate the microarray subarrays. Subarrays with a median signal of 5000 fluorescence units or more for the control oligonucleotide were considered valid, and those with lower signals were discarded. All hybridization signals were corrected for background (the median background was 100 fluorescence units or less) and were then expressed as a percentage of the control oligonucleotide signal.
Hybridization of amplicons from N. meningitidis, L. monocytogenes, and C. krusei digested directly in PCR buffer with the lambda exonuclease all yielded signal enhancements (up to 14.5-fold; Table 1
). Microfluidic electrophoresis has shown that around 90% of double-stranded DNA with phosphorylated complementary strands and Cy3-labeled target strands was digested compared to an undigested control (data not shown).
|
To monitor the influence of the Cy3 label on lambda exonuclease digestion, C. krusei amplicons produced with Cy3/phosphorylated or Cy3/nonphosphorylated primers were purified and then digested with lambda exonuclease. Passive 1-h hybridization of the digested C. krusei amplicons with nonphosphorylated complementary strands yielded a significant hybridization signal increase of 5-fold or more (Fig. 1A
). A higher signal increase (15-fold) was obtained by digestion of the phosphorylated complementary strand with lambda exonuclease. Selective digestion of the complementary strand yielded a strong hybridization signal with the C. krusei–specific capture probe ECkruH715, whereas no signal was measured with nondigested control amplicons (Fig. 1A
). Digested amplicons yielded higher hybridization signals, facilitating the differentiation of SNPs, as exemplified by the signal with the perfectly matched C. krusei probe ECkruH703, which is at least twice as strong as the signal with the single-mismatched Candida albicans probe ECalbH596. Nondigested amplicons could not distinguish this SNP.
|
For active hybridization, amplicons were introduced into the microfluidic unit juxtaposed to a microarray slide (4). Microfluidic hybridization of the digested product showed a 6- to 9-fold increase in the measured hybridization signals compared to nondigested control amplicons (Fig. 1B
).
Selective direct digestion of complementary strands with lambda exonuclease in PCR buffer yielded up to a 14.5-fold increase in microarray hybridization signals compared to nondigested control amplicons. Because the complementary strand is eliminated from the hybridization material, little or no interaction with the dangling end occurs, thus preventing displacement of the target strand hybridized to the microarray capture probe (2).
Several investigators have used exonuclease digestion of the complementary strand to produce single-stranded DNA targets for hybridizations (8)(9)(10)(11)(12)(13)(14)(15)(16). However, these methods are more cumbersome than our 5-min method because they require additional steps, longer digestion times (15–45 min), and in some cases, more complex modifications of the target strand. Although most investigators have suggested that single-stranded DNA targets can yield better hybridization signals, only a few(9)(10) directly compared digested products with nondigested controls. Under optimal conditions, investigators observed at most a 2-fold increase in hybridization signals with purified DNA targets compared to a 14.5-fold increase with our simple 5-min method.
We have investigated how the presence of a Cy3 label at the 5' end of the target strand influences lambda exonuclease digestion. Because the lambda exonuclease has 20x more affinity for a phosphorylated 5' end than a hydroxylated 5' end (17), we compared the digestion of both of these substrates paired with a 5' Cy3-labeled target strand. We conducted the digestion in lambda exonuclease buffer at 37 °C for 30 min for optimal enzymatic activity. The combination of a 5' Cy3-labeled target primer with a phosphorylated complementary strand yielded the best signal increase. Amplicons produced by a Cy3-nonphosphorylated primer combination also yielded an increased hybridization signal, suggesting that lambda exonuclease has less affinity for the Cy3-labeled 5' end than for the hydroxylated 5' end, thereby somehow protecting the target strand.
Competition for the target strands between the complementary strands and the capture probes during passive hybridization has been previously reported by our group (2). Although hybridization time is shortened with the microfluidic platform(4) and the hybridization kinetics is increased, competition between complementary strands and capture probes also occurs. Digestion of the complementary strands by lambda exonuclease resulted in a 6- to 9-fold increase of the hybridization signals obtained with the microfluidic platform compared with the nondigested control amplicons. This finding confirms that the generation of single-stranded DNA targets results in increased fluorescence signals both with passive and active microfluidic microarray hybridizations.
The 1-step digestion procedure presented here increases hybridization signals and facilitates SNP differentiation without any further modification for protection of the target strand, such as incorporation of phosphorothioate nucleotide analogs (8). Protection of the target strand likely comes from the Cy3 label used for microarray detection as suggested above. Increased hybridization signal allows the use of more stringent washing conditions along with the use of shorter microarray probes for better SNP differentiation. This digestion step eliminates the need for high-temperature denaturation and can be performed directly in a standard PCR buffer. The simplicity of our digestion and hybridization procedures make them amenable to integration into fully automated compact diagnostic devices.
Acknowledgments
Grant/funding support: This research was supported by Génome Québec and Genome Canada.
Financial disclosures: None declared.
Acknowledgments: We thank Dominique Boudreau, Richard Giroux, and Isabelle Martineau for their help in capture probe design.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
