| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Molecular Diagnostics and Genetics |
1 Centre de Recherche en Infectiologie de l’Université Laval, Centre Hospitalier Universitaire de Québec (Pavillon CHUL), Québec, Canada.
aAddress correspondence to this author at: Centre de Recherche en Infectiologie de l’Université Laval, Centre Hospitalier Universitaire de Québec (Pavillon CHUL), Sainte-Foy, Québec, Canada, G1V 4G2. Fax 1-418-654-2715; e-mail Michel.G.Bergeron{at}crchul.ulaval.ca.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: Superparamagnetic beads were functionalized with oligonucleotide capture probes of different lengths and used to capture GBS genomic DNA (gDNA). A rapid extraction procedure was used to provide DNA from GBS cultures or vaginal/anal samples with added GBS. Hybridization reactions consisting of functionalized beads and target DNA in 30 µL of hybridization buffer were performed for 1 h at room temperature, followed by washing and resuspension in water. Captured DNA was then detected using quantitative PCR.
Results: A 25-mer capture probe allowed detection of 1000 genome copies of purified GBS DNA. The ability to detect GBS was improved by use of a 50-mer (100 copies) and a 70-mer capture probe (10 copies). Detection of approximately 1250 CFU/mL was achieved for diluted GBS broth culture and for vaginal/anal swab samples with added GBS.
Conclusion: Oligonucleotide-functionalized superparamagnetic microbeads efficiently capture GBS gDNA from both bacterial cultures and vaginal/anal samples with added GBS. Efficiency of gDNA capture increases with oligonucleotide length. This technology could be combined with sample preparation and detection technologies in a microfluidic system to allow point-of-care testing for GBS.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Magnetic separation technology, an increasingly popular approach for purification and concentration, has been successfully applied to many different analytes and sample types (5)(6). The application of oligonucleotide-functionalized magnetic microbeads for isolation of specific nucleic acid sequences has improved reaction kinetics compared to filter-based DNA hybridization (7). Moreover, target nucleic acids can be concentrated from diverse organic samples and purified to remove PCR inhibitors (8)(9)(10). Hence, sequence-specific hybridization capture on magnetic microbeads, when combined with target amplification approaches such as PCR or with signal amplification technologies such as polymeric biosensors (11)(12), could provide a rapid and sensitive method for nucleic acid detection from diverse organic samples.
Group B streptococcus (GBS) 1 infections remain the leading bacterial cause of neonatal illness and death in Western countries (13). Infants who have GBS disease can require prolonged hospitalization and expensive support therapy. Survivors may suffer permanent disability, such as developmental delay or loss of vision (14). During pregnancy, GBS screening from vaginal/anal swabs is included in most prevention programs. However, culture-based screening detects only 87% of women infected with GBS at delivery (15). To address this problem, we previously developed a rapid (<1 h) quantitative PCR (qPCR) assay for GBS detection from vaginal/anal specimens collected from pregnant women at delivery (1)(16). This assay is currently being adapted for a point-of-care test (POCT) that will allow rapid, simple, and cost-effective GBS diagnosis at the bedside.
We aim to specifically capture a GBS genomic DNA (gDNA) target with oligonucleotide probes linked to magnetic microbeads. This approach can purify and concentrate the target, allowing sensitive and specific detection of GBS from clinical samples. Moreover, it is amenable to automation and use in POCT systems. In the present study, we report the development of magnetic capture hybridization (MCH) for the isolation of GBS DNA from vaginal/anal swab samples taken from pregnant women at delivery.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) that had a primary amino group attached by a linker (IDT®) to its 5' end. Approximately 2.85 x 108 beads were washed (twice) for 10 min in 25 mmol/L 2-(N-morpholino)-ethane sulfonic acid (MES; Sigma-Aldrich) buffer (pH 6.0). Oligonucleotides were added and incubated for 30 min at room temperature with slow (8 rpm), end-over-end rotation (SER) with a tube rotator to prevent bead sedimentation. N-(3-dimethylaminopropyl)-N'-ethylcarbodiimide (EDC, Sigma-Aldrich) solution in cold MES buffer was added to the mixture and incubated overnight at 4 °C with SER. Bead functionalization was tested for different concentrations (0–0.469 mol/L) of EDC. Beads were then washed (4 times) in 0.05 mol/L PBS (8 g/L NaCl, 0.2 g/L KCl, 0.12 g/L KH2PO4, 0.91 g/L Na2HPO4, pH 7.4), blocked with 0.05 mol/L Tris-HCl (pH 7.4) for 15 min at room temperature, and resuspended in MES buffer. Functionalization efficiency was qualitatively assessed by fluorescence microscopy.
|
hybridization capture of target nucleic acids
Oligonucleotides for specific GBS gDNA capture, targeting the gene encoding the Christie-Atkins-Munch-Petersen (CAMP) factor (cfb), were designed using Oligo® primer analysis software (version 6.70, Molecular Biology Insights). Oligonucleotides not complementary to GBS sequences were designed using gene tuf (bacterial gene encoding the elongation factor Tu) sequences from Staphylococcus aureus and atpD (bacterial gene encoding the beta subunit of ATP synthase) sequences from Bacillus globigii (Table 1
). Functionalization reactions were performed using different concentrations of a nonfluorescent 50-mer oligonucleotide capture probe. Beads were then mixed with a complementary (Cy3) or a noncomplementary (Cy5) fluorescently labeled oligonucleotide target in hybridization reactions consisting of 1.43 x 109 beads/mL and 2 µmol/L fluorescent oligonucleotide target in 20 µL of buffer [6x saline-sodium phosphate-EDTA ((6x SSPE); 0.9 mol/L NaCl, 60 mmol/L NaH2PO4·H2O, 6 mmol/L EDTA, pH 7.4), 0.03% polyvinylpyrrolidone (PVP), 30% formamide]. Reactions were incubated overnight at room temperature with SER and protected from light. Beads were then washed with 2x SSPE/0.1% sodium dodecyl sulfate and then with 2x SSPE before resuspension in 2x SSPE. Fluorescence microscopy was then used to qualitatively assess the extent of hybridization.
Beads functionalized with different concentrations of nonlabeled capture probes were then used in hybridization experiments with ultrapure GBS gDNA, which was detected by qPCR as previously described (16) after capture on the beads, washing, and release. gDNA capture experiments were then performed with beads functionalized with oligonucleotide capture probes of different lengths (25-, 50-, and 70-mer).
Hybridization reactions consisted of 10 to 1 x 105 GBS genome copies and 3.17 x 108 beads/mL in 30 µL of hybridization buffer. Reactions were incubated for 1 or 3 h at room temperature with SER. Beads were washed (3 times) with 0.1x SSPE/0.1% sodium dodecyl sulfate, once with 0.1x SSPE, and then resuspended in ultra-pure H2O. Captured DNA was released from the beads by heating at 95 °C for 10 min and then cooled on ice. Tubes were centrifuged briefly to recover condensate, and beads were held at the tube wall by a magnet (MPC®-S; Invitrogen). Finally, the supernatant was removed and used in PCR reactions.
Ultrapure GBS gDNA was detected by MCH-qPCR with or without the presence of purified Escherichia coli gDNA. Smaller hybridization-reaction volumes of 9 or 18 µL were tested, and a 10-fold increased bead number was used in 30-µL reactions.
determination of detection limit
A rapid cell lysis procedure (BD GeneOhmTM Lysis Kit; BD Diagnostics-GeneOhm) was used to extract crude gDNA from pure GBS cultures or GBS cultures added into previously frozen, nonidentified GBS-negative vaginal/anal samples (provided by BD Diagnostics-GeneOhm). These clinical samples were collected during a previous study (approved by the research ethics committees of participating hospitals) and stored at –80 °C.
Enumeration of bacteria in broth cultures was performed by colony-forming unit (CFU) counting, and their DNA content was analyzed by GBS-specific qPCR (16). Dilutions of these broth cultures were used to determine the limit of GBS detection by the optimized MCH-qPCR protocol. The limit of detection (LOD) was defined as the lowest concentration in a dilution series that was detected by the method. Comparisons of the LOD for pure gDNA (genome copies) and crude lysates (CFU) were made, based on the ratio of genome copies to CFU counts obtained from numerous experiments. Finally, dilutions of pure GBS gDNA or whole cells in TE buffer (10 mmol/L Tris-HCl, 1 mmol/L EDTA, pH 8.0) or GBS-negative vaginal/anal samples were tested.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Increasing EDC concentration from 0.078 mol/L up to 0.469 mol/L allowed more efficient formation of amide bonds between COOH groups and amines. More intense fluorescence on beads indicated concomitant increasing amounts of bead functionalization with the probe (Fig. 1
, B–D). Higher concentrations of EDC (up to 1.565 mol/L) did not significantly enhance the visible amount of bead functionalization (data not shown). As such, 0.469 mol/L EDC was used in all subsequent experiments.
|
Beads functionalized with a nonlabeled 50-mer oligonucleotide capture probe were used in hybridization reactions with a fluorescently labeled, complementary oligonucleotide target. This procedure showed that capture probes attached to the beads were available for hybridization and enabled assessment of the effect of probe density on the bead surface. Unhybridized beads and those mixed with a noncomplementary fluorescent oligonucleotide target showed no fluorescence, indicating the specificity of target capture (data not shown). Unfunctionalized beads also showed no fluorescence after incubation with a fluorescent oligonucleotide target (Fig. 1E
). Beads hybridized with the complementary target showed increasing fluorescence with an increasing extent of functionalization (reactions performed using 5, 10, and 20 µmol/L capture probe; Fig. 1
, F–H). Images of beads functionalized with 30 or 40 µmol/L oligonucleotide capture probe were not visibly different from those of beads functionalized with 20 µmol/L of probe (data not shown).
gbs GDNA capture
To determine the optimal bead-surface capture probe density for gDNA capture, the same batches of beads previously used for capture of fluorescently labeled oligonucleotide were used in capture experiments with pure GBS gDNA.
Beads functionalized in reactions with 20 µmol/L oligonucleotide more reproducibly detected 100 GBS genome copies in 30-µL hybridization reactions than beads functionalized at the other capture probe concentrations tested. Hence, this amount of functionalization was used for all further experiments.
Reactions of smaller volumes (9 and 18 µL) with the same number of beads were tested and found to improve detection, probably because the concentration of reactants increased the rate of probe-target interactions. However, because smaller reaction volumes equate to more concentrated target DNA (which is difficult to achieve) or fewer targets per reaction, a 30-µL hybridization reaction was chosen for all further experiments.
Oligonucleotide capture probes of different lengths (25-, 50-, and 70-mer; Table 1
) were used to functionalize beads to test the effect of probe length on the LOD of GBS gDNA. Use of 25-mer capture probes allowed detection of 1000 genome copies of purified GBS gDNA. The LOD was improved by use of 50-mer (100 copies) and 70-mer capture probes (10 copies). Also, reducing hybridization time from 3 to 1 h did not affect the LOD (Table 2
). Thus 1-h hybridization with the 70-mer capture probe was selected for further testing of specific GBS gDNA recovery. In this format, the presence of up to 8.6 x 107 genome copies of purified E. coli gDNA did not significantly affect the LOD of ultrapure GBS gDNA. This result gave an initial indication of the specificity of hybridization capture and the effect of nontarget DNA on GBS gDNA detection. However, we more fully tested capture specificity using GBS gDNA or whole cells added to clinical samples (see below).
|
To increase the probability/rate of bead-target interactions, we increased 10-fold the number of beads per 30-µL hybridization reaction. This increase improved the reproducibility of gDNA detection at low copy numbers. With this optimized MCH-qPCR method, we found that the efficiency of GBS gDNA capture and detection diminished with increasing copy number (Table 3
). This result may be attributable to bead saturation with GBS gDNA causing steric hindrance and electrostatic repulsion to further gDNA capture.
|
A rapid cell lysis procedure (BD GeneOhm Lysis Kit) was used to provide crude gDNA preparations from pure GBS broth culture dilutions. The optimized MCH-qPCR method enabled reproducible detection of GBS at a concentration of 1250 CFU/mL (Table 4
), demonstrating the compatibility of these 2 methods. Notably, in 1 duplicated experiment, GBS were detected at 1250 CFU/mL in both replicates, whereas 625 CFU/mL were detected in only one replicate, roughly indicating the LOD for this procedure. Our observations from repeated experiments comparing qPCR results and CFU counts show that 1 CFU equates approximately to 3 genome copies. As such, the observed LOD represents approximately 19–38 genome copies per 30-µL MCH-qPCR reaction.
|
To investigate the effect of clinical samples on this procedure, the same GBS broth dilutions were added to 11 previously frozen, GBS-negative vaginal/anal swab resuspensions. Crude gDNA preparations were made as before and tested by the optimized MCH-qPCR method. The LOD of GBS detection from clinical samples did not differ from that of detection from pure culture (Table 4
). Because clinical samples contain high numbers of background microorganisms and PCR inhibitors, this finding indicated the specificity of GBS gDNA capture and the efficiency of its purification.
To further test the applicability of crude gDNA extraction, detection was compared to that of pure GBS gDNA added to TE buffer and GBS-negative clinical sample crude lysates. Pure GBS gDNA was detectable at 60–125 genome copies per MCH-qPCR reaction in TE buffer (3 replicate experiments) and at 30–63 genome copies per MCH-qPCR reaction in crude lysates (2 replicates). As such, clinical samples did not inhibit MCH-qPCR. Moreover, this LOD was equivalent to that achieved with added cells (Table 4
). These results suggest that the rapid cell lysis protocol provides crude gDNA preparations that do not interfere with specific MCH-qPCR detection of GBS.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Initially, microbeads were reacted with a Cy3-labeled oligonucleotide to demonstrate functionalization. Results were unsatisfactory, however, because little fluorescence was observed on the beads when examined with fluorescence microscopy. The concentration of EDC initially used was probably insufficient, because other investigators have used much higher concentrations (17)(18)(19). Hence, all variables in the protocol were analyzed, and EDC concentration was found to be the crucial factor for successful bead functionalization with aminated oligonucleotides.
The optimal bead-surface capture probe density for gDNA hybridization is likely to be lower than for oligonucleotide capture, owing to steric hindrance. Indeed, capture of fluorescently labeled oligonucleotide targets continued to increase with higher functionalization amounts, whereas gDNA capture was optimal at lower amounts. Similarly, the length of the capture probe may affect the efficiency of gDNA capture (20)(21). Our results showed that the use of longer oligonucleotide capture probes does indeed lead to a better LOD, on the basis of subsequent qPCR amplification of captured GBS gDNA. This result might be explained by a stronger interaction between gDNA and longer capture probes due to the formation of more hydrogen bonds (20)(21). An additional factor explaining better capture could be a reduction of steric hindrance to hybridization by longer capture probes (21). In the latter case, it might be envisioned that a region of the capture probe close to the bead surface is less available for hybridization owing to physical constraints of long gDNA molecules interacting with microbeads. As such, longer capture probes provide an available hybridization area further from the bead surface for easier interaction with target gDNA. Indeed, polyethylene glycol (22) and polyT spacers(23) have previously been used to aid hybridization on microarrays via orientation of oligonucleotide probes further from the solid phase. PolyA spacers have also been used in a similar fashion on microbeads (24).
With the use of optimally functionalized beads, we were able to detect as few as 10 GBS genome copies by MCH-qPCR. Therefore, the detection limit was 1000 genome copies/mL, or approximately 300 CFU/mL. This estimation is based on our observation that 1 CFU equates to approximately 3 genome copies (data not shown), which is due, in part, to GBS cell clustering during replica plating enumeration experiments.
MCH-PCR has previously been used for the detection of pathogens in clinical, food, and environmental samples. For example, a 10-fold increase in Mycobacterium avium subspecies paratuberculosis DNA detection sensitivity was achieved by a 2-step MCH-PCR method compared to PCR alone (8). Moreover, more positive results were achieved from intestinal tissue extracts from patients with Crohn disease, and only MCH-PCR detected this bacterium in feces of bovines (containing approximately 10–100 CFU/g) with Johne disease. Marsh et al. (25) developed a 2-day, closed-system test, to prevent sample cross-contamination, with a detection sensitivity of 2.5 x 104 CFU/g of feces. This test could be used to detect the organism in pooled ovine fecal samples (1 naturally infected and 99 noninfected sheep). Later, a simpler, direct PCR test was found to be more sensitive than this MCH-PCR system (26). The authors attributed the poor performance of MCH-PCR to "inefficiencies in liquid- phase hybridization and solid-phase magnetic bead capture". They also referred to the work of Arnal et al. (27), which showed that a MCH-PCR method capable of detecting enteroviruses (9)(28) could not be adapted for hepatitis A virus detection from feces and shellfish samples. The conclusion of Arnal et al. (27) was that MCH-PCR is applicable to simple or dilute samples, such as culture supernatants, but not to complex samples such as feces or shellfish extracts. For the latter, organic or silica-based total RNA extraction methods were superior. However, in these enterovirus detection experiments, a 1-step capture method was used as opposed to separate liquid-phase probe hybridization and solid-phase capture steps.
For 1-step capture, the density of capture oligonucleotide probes on the bead surface is critical for optimal target nucleic acid hybridization, as is the length of the capture probe and its positioning away from the bead surface via nontarget-hybridizing linkers (20)(21). Moreover, in the experiments of Arnal et al. (27), beads, nucleic acid samples, and captured target were all diluted before the final PCR detection, thus reducing the sensitivity of detection. The poor recovery of magnetic beads from mussel tissue and fecal samples might also be rectified by the incorporation of sodium dodecyl sulfate in the hybridization buffer (29). Nevertheless, MCH-PCR was faster and simpler to perform than organic extraction and did not use toxic chemicals, although final detection sensitivities were similar for both methods. Furthermore, others have demonstrated improved sensitivity, speed, and reliability of detection by MCH-PCR (10)(30). Hence, MCH can be used, when optimized, to aid the detection of specific nucleic acid sequences.
Many variables must be considered in optimizing a MCH-PCR protocol for any target/sample type. For nucleic acid capture, these variables mainly relate to probe-target hybridization and solid-phase capture of the hybrid. Buffer compositions and concentrations of probes and beads are vital for efficient recovery of the target. Multiple capture probes may also aid recovery of target sequences for which shearing may separate capture sequences from PCR/detection-probe sequences in genomic nucleic acids. Moreover, for PCR detection of gDNA, capture of both strands may further decrease the LOD. We selected a 1-step over a 2-step MCH protocol because rapidity and simplicity are important considerations in the intended POCT use of this method.
A rapid cell lysis procedure was used to provide crude GBS gDNA for MCH-qPCR experiments. This method was validated by use of both pure and mixed GBS broth cultures and GBS-negative vaginal/anal samples to which different numbers of GBS cells were added. Consequently, we were able to reproducibly detect as few as 1250 CFU/mL from supplemented GBS-negative vaginal/anal samples.
Little precise information is available regarding the extent of maternal GBS colonization, with most studies being qualitative or semiquantitative. Generally, heavy colonization is scored as 3+ or 4+ bacterial growth on agar-plate culture, which is estimated to represent 104 to 105 CFU/mL of vaginal material or per swab (31). Heavy maternal colonization is thought to confer a greater chance of neonatal colonization and infection (32)(33)(34). However, Yancey et al. (35) noted that there have been reports of GBS early-onset disease and fatalities among infants born to lightly colonized mothers. Light vaginal colonization may be 10 1 CFU or less per swab (36) or per milliliter of vaginal material (37). Notably, our data show that a GBS load of approximately 103 CFU/swab can be detected by our current MCH-qPCR protocol. Moreover, we are developing a system to concentrate captured target DNA in a microfluidic system (unpublished data); this method promises to further improve detection sensitivity.
Lim et al. (33) showed that heavily colonized mothers are twice as likely as lightly colonized mothers to give birth to colonized infants. However, several reports have shown that the majority of women with detectable GBS have light colonization at <104 to 105 CFU/mL (31)(33)(37)(38). As such, this group constitutes a considerable number of women delivering infants at risk for GBS infection, for which GBS detection requires lengthy enrichment culture using standard techniques. Our method, however, provides sensitive GBS detection (approximately 100-fold more sensitive than by standard agglutination tests) in <4 h, which is consistent with other rapid detection methods (3).
Adaptation of cell lysis and DNA capture to a microfluidic system will facilitate automation. Prototype versions for cell lysis have already been demonstrated to be efficient (39)(40). Ultimately, transfer to a microfluidic micro total analysis system will further decrease test time to <1 h, which would be compatible with POCT. Furthermore, the use of polythiophene biosensors promises to allow sensitive detection of target DNA without prior amplification (11)(12). Hence, we will adapt and combine all of these systems for the capture and detection of GBS DNA from vaginal/anal swab samples.
Concentration of targets from typically milliliter-sized, real-life clinical samples to micro or nanoliter volumes for microfluidic processing poses an important technical challenge. This challenge has previously been addressed for nucleic acids by nonspecific capture on magnetic beads (3). However, to increase detection specificity (especially where PCR is not used), specific target hybridization capture provides a distinct advantage over nonspecific methods. Magnetic bead-based purification is also readily automated, which is conducive to use in micro total analysis systems.
The results we report here show a proof of concept for specific GBS gDNA capture from clinical samples. With careful selection of specific capture probes, however, this technology can be easily applied to any nucleic acid target for detection and diagnostic purposes. Moreover, this method can be readily automated and combined with microfluidics for the development of POCT or portable testing devices.
|
Acknowledgments |
|---|
Financial disclosures: None declared.
Acknowledgments: We thank BD Diagnostics-GeneOhm (Quebec City, QC, Canada) for providing the vaginal/anal samples.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  L. G. Henry, L. Sandberg, K. Zhang, and H. M. Fletcher DNA Repair of 8-Oxo-7,8-Dihydroguanine Lesions in Porphyromonas gingivalis J. Bacteriol., December 15, 2008; 190(24): 7985 - 7993. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
excitationmax = 550 nm,