| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Department of Chemistry, University of Pittsburgh, Pittsburgh, PA 15260.
2
Division of Laboratory Genetics, Mayo Clinic, Rochester,
MN 55905.
3
Department of Pathology, University of Pittsburgh
Medical Center, Pittsburgh, PA 15213.
4
University of Pittsburgh Cancer Institute, Pittsburgh,
PA 15260.
a Address correspondence to this author at: Department of Chemistry, University of Virginia, Charlottesville, VA 22901. Fax 804-243-8852; e-mail jpl5e{at}virginia.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: Capillary electrophoresis was explored as an alternative to slab gel electrophoresis for the analysis of PCR-amplified products indicative of T- and B-cell malignancies as a means of defining the elements for silica microchip-based diagnosis. Capillary-based separations were replicated on electrophoretic microchips.
Results: The microchip-based electrophoretic separation
effectively resolved PCR-amplified fragments from the variable region
of the T-cell receptor-
gene (150–250 bp range) and the
immunoglobulin heavy chain gene (80–140 bp range), yielding
diagnostically relevant information regarding the presence of clonal
DNA populations. Although hydroxyethylcellulose provided adequate
separation power, the need for a coated microchannel for effective
resolution necessitated additional preparative steps. In addition,
preliminary data are shown indicating that polyvinylpyrrolidone
may provide an adequate matrix without the need for microchannel
coating.
Conclusions: Separation of B- and T-cell gene rearrangement PCR products on microchips provides diagnostic information in dramatically reduced time (160 s vs 2.5 h) with no loss of diagnostic capacity when compared with current methodologies. As illustrated, this technology and methodology holds great potential for extrapolation to the abundance of similar molecular biology-based techniques.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A miniaturized variant of CE exploits extant microfabricated chip technology to create an electrophoretic chip device that provides even greater advantages for rapid and cost-effective clinical analysis. Perhaps the greatest advantage of the electrophoretic chip platform over the slab gel and even the capillary formats is the potential for integrating sample processing steps directly (5)(6)(7)(8). In combination with parallel processing, this has the potential to create the rapid, high-throughput platform demanded by current clinical diagnostic applications. The potential impact on the clinical sector for electrophoresis using microfabricated devices has been discussed (9) with supporting preliminary data from applications such as serum protein separations (10) and immunoassays (11)(12)(13). Clinically relevant DNA separations on microfabricated devices have also been shown, including the separation of PCR products of the dystrophin gene (14) and DNA-sequencing products (15).
The development of electrophoretic microchip technology is in its infancy, although an increased number of research groups are reporting their production and use (5)(6)(8)(11)(12)(13)(14)(15)(16)(17)(18)(19). Despite the fact that few reports have demonstrated the application of electrophoretic microchips to real-world analysis, it is clear that this technology is ideally suited to the rapid analysis of PCR-amplified DNA.
A multitude of diagnostic assays currently in use incorporate PCR-based amplification for detection of aberrant DNA sequences (mutations, rearrangements, deletions, and translocations) correlative with the onset of disease (1). Such is the case for the diagnosis of T- or B-cell lymphoma, where PCR results can be available within 1 day whereas Southern blot studies require 1 week (20). Clinical diagnosis of T- or B-cell lymphoproliferative disorders takes advantage of the maturation process of the T and B cells by focusing on the analysis of the genes encoding for the T-cell receptor (TCR) and immunoglobulin heavy chain (IgH). The majority of oncogenic events begin after rearrangement (21); therefore, the offspring of a transformed cell will possess the same specific gene sequence produced when the parent cell underwent rearrangement. A predominance of a single sequence will, therefore, be a mark of clonality, an indication of malignancy. However, as a result of the combinatorial nature of the rearrangement process, a normal cell population will possess a polyclonal population, a diverse variety of low abundance sequences. PCR-based assays have been developed to analyze the lengths of the DNA fragments produced when a variety of primers for the variable and joining regions are used (multiplexed PCR) (22). When multiplexed PCR and subsequent fragment separation are used, a normal cell population displays a large variety of DNA fragment sizes, whereas a malignant population displays a predominance of one DNA fragment size.
As with other PCR assays, established protocols for analyzing B- and T-cell amplified products have relied on slab gel electrophoresis to identify products and provide semiquantitative data. CE provides a more rapid, automated platform for high-resolution PCR fragment analysis with excellent detection sensitivity provided by laser-induced fluorescence (LIF) detection (23). In an earlier report, Oda et al. (24) took the first steps toward utilizing CE for T-cell malignancy diagnosis by defining initial conditions for separation of the PCR-amplified fragments. In this report, this work is extended to include the analysis of B-cell malignancy assay samples and the methodology translated to the microfabricated chip, where separations can be achieved over an order of magnitude faster. Comparisons are made between slab gel, capillary, and microchip electrophoretic results with respect to analysis time, resolution, and diagnostic capabilities. In addition, with a vision toward a simple and robust microchip platform for diagnostic DNA fragment detection, the results of separations that were performed in a microchip containing uncoated channels are presented; these results, although preliminary in nature, illustrate the feasibility for this methodology.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
instrumentation
CE instrumentation.
A Beckman P/ACE System 5510 (Beckman
Instruments) was used for CE analysis. Fluorescence detection used a
P/ACE System Laser Module 488 with a P/ACE LIF detector, which excites
at 488 nm and collects emission at 520 nm ± 10 nm. A laser power
of 0.25 mW was presented at the capillary. Instrument control
and data collection were performed with an IBM 486 ValuePoint computer
utilizing System Gold software (Ver. 8.1).
Electrophoretic microchip instrumentation.
The separation
voltages were supplied by an in-house manufactured high-voltage power
supply controlled by a program written in Labview. For fluorescence
detection, the 488-nm line of an argon ion laser (532R-BS-AO4; Melles
Griot) was expanded to fill the back aperture of an objective. The
expanded beam was reflected off a beam splitter (505DRLP02; Omega
Optical) set at 45° to the incident beam and into the channel of an
electrophoretic microchip by an objective (20x/numerical aperture,
0.5). Fluorescence emitted by the sample was collected by the
objective and focused by a 200-mm lens onto a 400-µm pinhole
(25). The fluorescence wavelength was spectrally filtered by
a 530 nm bandpass filter (Omega Optical 530DF30, FWHM 30 nm), collected
by a photomultiplier tube (PMT, Hamamatsu R38960), and processed via a
program written in Labview.
Electrophoretic microchip microfabrication.
Electrophoretic
chips were microfabricated by the Alberta Microelectronics Center.
Channels were etched in one glass plate to which a glass top plate was
bonded. Electrophoretic chips consisted of an eight-channel
arrangement, with each channel composed of an injection cross with a
sample-to-sample waste distance of 1.45 cm, inlet-to-outlet distance of
6.65 cm, and the junction 0.5 cm from the inlet (Fig. 3
). Detection
occurred 4.2 cm from the injection cross. The sample channel was 100
µm wide and 10 µm deep, and the separation channel was 50 µm wide
and 10 µm deep. One-millimeter holes were drilled in the top glass
plate to allow access to the channels. Cut pipette tips were epoxied
onto the top plate to form larger wells for solutions and placement of
gold-coated electrodes.
|
pcr amplification of tcrg rearrangements
Rearranged T-cell receptor- (TCRG) gene sequences
were amplified using multiplexed PCR as described previously
(20). For the TCRG variable segments, five
primers (TCRG V2, 3, 4, 8, and 9) were used along with three
primers for the TCRG joining segments (JGT12 with consensus
sequences for J1.3 and J2.3, JGT3 with consensus sequences for J1.1 and
J1.2, and JGT4 with consensus sequences for J1.2). The PCR reaction
mixture included 0.75 U of Taq polymerase (Taq Gold-P; Perkin-Elmer),
200 µmol/L each of dATP, dCTP, dGTP, and dTTP, 1.5 mmol/L
MgCl2, and 1.0 µmol/L of each primer in
standard PCR buffer (Perkin-Elmer) for a total volume of 25 µL. PCR
cycling conditions included a 10-min hold at 95 °C, followed by 40
cycles of PCR (94 °C for 30 s, 55 °C for 1 min, 72 °C for
1 min) and a 10-min extension at 72 °C.
pcr amplification of igh rearrangements
IGH PCR was performed in 25-µL reactions using 1 U of
Taq polymerase (PE Biosystems) and a consensus 5' framework III
variable region primer (AGG TGC AGC TGG TGC AGT CTG G) with a mixture
of three consensus primers directed at heavy chain joining
regions 1, 2, 4, and 5 (ACC TGA GGA GAC GGT GAC CAG GGT), 3 (TAC CTG
AAG AGA CGG TGA CCA TTG T), and 6 (ACC TGA GGA GAC GGT GAC CGT GGT).
PCR thermocycling conditions included an initial denaturation for
5 min at 95 °C, followed by 35 cycles of PCR (95 °C for 15
s, 52 °C for 20 s, 72 °C for 30 s) and a 5-min
extension at 72 °C.
analysis via slab gel electrophoresis with sybr green i staining
For the TCRG gene rearrangement assay, 3.5 µL of the
amplified PCR mixtures was loaded onto an 8% polyacrylamide (19:1, by
weight) minigel. The DNA marker was pGEM (Promega). Electrophoresis was
performed using a 1x TBE buffer for 75 min at 200 V. Gels were stained
with SYBR Green 1 (1 µg/L; Molecular Probes) with gentle agitation
for 30 min at room temperature. Stained gels were illuminated with a
300 nm ultraviolet transilluminator and photographed with Polaroid 667
black and white print film through a SYBR Green I gel stain
photographic filter (1- to 2-s exposure; f-stop, 4.5).
For IGH gene rearrangement PCR assays, 30 µL of PCR reaction product mixed with 3 µL of blue sucrose was loaded into lanes of an 8% polyacrylamide gel (16 cm x 20 cm x 1.5 mm) and electrophoresed in 1x TBE buffer for 2.5 h at 180 V. The gel was stained with SYBR Green I and photographed with type 57 Kodak film. MspI digests of pBR322 (New England Biolabs) were used as molecular weight markers.
pcr product analysis via lif-ce
For CE separations using HEC (described above), a 37 cm x 50
µm (effective length, 30 cm) fluorocarbon (FC)-coated µ-SIL
capillary (J & W Scientific) was used. YO-PRO-1 (1 µmol/L) was used
as a fluorescent intercalator. The DNA marker, HaeIII
digest of pBR322 (Boehringer Mannheim Biochemicals), was diluted 1:100
with 10 mmol/L Tris, 1 mmol/L EDTA for a final concentration of 4.20
mg/L. The PCR-amplified samples were diluted 1:10 in 10 mmol/L Tris, 1
mmol/L EDTA. FC-coated capillaries were first rinsed with 20 column
volumes of water. HEC separations used the following method: 3-s
electrokinetic injection of water at 3.5 kV, 10-s injection of sample
at 8.5 kV (230 V/cm), 3-s injection of water at 3.5 kV. Separation
followed at 8 kV (216 V/cm), using reversed polarity (inlet cathode,
outlet anode) and maintaining the capillary at 20 °C. The capillary
was rinsed with 5 column volumes of water between runs and prerinsed
with 10 column volumes of fresh (unelectrophoresed) buffer before each
run.
For CE separations using PVP (described above), 37 cm x 50 µm bare silica capillaries (Polymicro Technology) were first conditioned by rinsing with 20 column volumes of water. Injection and run settings were as stated for the HEC/FC-capillary system.
pcr product analysis via microchip electrophoresis
Microchip channels were coated with linear polyacrylamide (PA) by
a modified Hjertén method (26). Samples were desalted
by placing 10 µL of undiluted sample in a
Microcon® YM-10 Centrifugal Filter Device
(Amicon) with 90 µL of water and then centrifuged at 7g
for 10 min. One hundred microliters of water was then added,
vortex-mixed, and centrifuged for 10 min at 7g; this
procedure was repeated five times, and the eluent was discarded. To
reconstitute the sample, 10 µL of water was added, vortex-mixed,
centrifuged for 30 s at 2g; the cartridge was then
inverted and centrifuged at 7g for 2 min. The same HEC and
PVP sieving matrices were used as described for the CE analyses.
Channel preparation and between-run rinses were performed as described
for the capillary system. Microchip sample injection was performed by
applying a 400 V (275 V/cm) potential across the sample and sample
waste reservoirs, with the sample at ground. For separation, the sample
and sample waste were grounded and -200 V was applied to the inlet and
900 V to the outlet (165 V/cm). Fluorescence was collected at 10 Hz.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(TCRG) and immunoglobulin heavy chain
(IGH) gene rearrangement PCR products on an acrylamide gel.
Fig. 1
shows SYBR Green-stained polyacrylamide gels for these assays.
In these analyses, a normal polyclonal population of lymphoid cells
will yield a series of DNA fragments that appear as a "smear" or a
series of discrete bands, depending on the resolution. Conversely, a
clonal lymphocyte population, indicative of a lymphoid neoplasm, will
yield one or more dominant bands. The PCR primers used for multiplexed
amplification of the TCRG variable region(s) yield fragments
in the 150- to 250-bp range (Fig. 1A
). The polyclonal fragments typical
of a normal sample (negative for a T-cell lymphoproliferative disorder)
are shown in lane 9 (negative control) and in lane 6 (patient sample).
Both cases display the smear of DNA fragments typical of
polyclonal populations. In contrast, the presence of a dominating DNA
fragment indicative of clonality is illustrated by the patient samples
in lanes 2, 3, and 4, which exhibit one or more densely stained bands
in the 150- to 250-bp region and compare favorably with the positive
controls in lanes 7 and 8. Samples in lanes 1 and 5 were
negative/equivocal for clonality.
|
Interrogation of B-cell populations via the same molecular approach is
shown in Fig. 1B
. Here, multiplex PCR amplification of the variable
region(s) of the IGH gene yields fragments in the 80- to
140-bp region (Fig. 1B
). As with the T-cells, normal samples produce a
polyclonal population of DNA fragments that are observed as
low-abundance DNA bands spanning the 80- to 140-bp range (lanes 2–4
and 6). The lower molecular weight fragments yielded by the
IGH gene amplification are resolved better than the
corresponding T-cell amplification products and, therefore, are not
observed as a smear but rather as a series of discrete bands. However,
with B-cell samples that display clonality, dominant band(s), which
signal the presence of the monoclonal population, are clearly observed
and indicative of a B-cell lymphoproliferative disorder (lanes 1, 5, 7,
and 8).
The results from CE analysis of selected samples analyzed via slab gel
electrophoresis in Fig. 1
are shown in Fig. 2
. CE conditions included a 37 cm x 50 µm FC-coated
capillary containing 10 g/L HEC solubilized in 1x TBE buffer at
pH 8.6. A fluorescent intercalator, YO-PRO-1, was added to the buffer
for LIF detection. Fig. 2A
shows CE separations representative of
TCRG gene rearrangement products and compares these
electropherograms with a separation of the lower molecular weight
fragments in a pBR322 HaeIII digest (bottom panel). The top
panel (T1) shows a DNA fragment pattern indicative of a polyclonal
population, with the stained acrylamide gel (inset) displaying an
equivalent banding pattern. Because of the band (gel) and peak
(capillary) indicated by the arrows, this patient was diagnosed as
negative/equivocal. A monoclonal fragment pattern is illustrated in the
middle panel (T2, with acrylamide gel inset). For this sample, the
diagnosis is obvious, based on the presence of a dominant band on the
gel and the corroborative peak in the electropherogram. The
corresponding PCR-amplified IGH rearrangement products were
also analyzed by CE, and the results are shown in Fig. 2B
along with
the pBR322 HaeIII digest marker (bottom panel). The top
panel (B1) shows a positive sample, with a broad band on the gel being
partially resolved into two peaks by CE; this is unequivocal
identification of a monoclonal population. Conversely, the middle panel
(B2) illustrates a characteristic profile for a sample negative for a
B-cell malignancy, with the widespread polyclonal population of
fragments displayed as a sawtooth pattern produced by the
high-resolution separation.
|
microchip-based electrophoretic analysis of tcrg and
igh gene rearrangements
Efficient microchip-based electrophoretic analysis requires a
multichannel electrophoretic microchip and an optical apparatus capable
of sensitive fluorescence detection (Fig. 3
). The microchip illustrated in Fig. 3A
contains eight
independent microchannels with the corresponding reservoirs for each.
Although parallel analysis can be performed, only single-channel
analysis was required in this initial study. Injection of DNA was
accomplished electrokinetically at the intersection between the
separation channel (50 µm wide and 10 µm deep) connecting the inlet
and outlet reservoirs and the cross channel (100 µm wide and 10 µm
deep) that connects the sample and sample waste reservoirs. The sample
reservoir-to-sample waste distance was minimized to negate any
electrophoretic artifacts that may result from separation in the cross
channel itself during injection. With a separation distance (cross
channel intersection-to-detection point) of 4.2 cm, the effective
length of the separation channel was sevenfold shorter than the
capillary used to generate the separations shown in Fig. 2
. Detection
was accomplished by focusing the laser beam on the center of the
microchannel at the detection window of the microchip (shown in Fig. 3A
). Fig. 3B
shows the configuration of the home-built confocal
epifluorescence microscopic detection system. The system uses an argon
ion laser that is filtered spectrally and expanded to fill the back
aperture of a microscope objective lens to focus the beam tightly
inside the channel. Fluorescence is collected via the same objective
lens and, after passage through the optical components displayed in
Fig. 3B
, is detected by a photomultiplier tube. A single
software program controls both the detection system and the
electrophoresis high-voltage system, synchronizing the separation with
the detection. This configuration is ideal for microchip-based
electrophoresis because it allows for sensitive detection while leaving
the microchip surface completely accessible for manipulation of sample,
solutions, and electrophoresis hardware.
Microchip-based electrophoretic separation of DNA markers.
After the microchannels were coated with PA, the microchannels were
filled and equilibrated with 10 g/L HEC in 1x TBE containing
micromolar concentrations of YO-PRO-1 as an intercalating dye for
double-stranded DNA detection. A typical separation of DNA markers,
HaeIII digest of pBR322, produced the microchip
electropherogram shown in Fig. 4
A. The high resolution observed is comparable to separation of
the same DNA markers in the 37 cm x 50 µm capillary (Fig. 4B
).
|
Analysis of TCRG and IGH gene rearrangements.
The samples
evaluated by gel electrophoresis (results shown in Fig. 1
) were also
analyzed by microchip electrophoresis with optimized conditions for
sample preparation (sample desalting and a 1:10 dilution in 10 mmol/L
Tris, 1 mmol/L EDTA). Fig. 5
shows a comparison of the capillary and electrophoretic
microchip formats for separation of the TCRG gene
rearrangement products for four samples. The profile for sample T1
(Fig. 5B
) displays the same series of low-abundance fragments (fragment
size range, 150–250 bp) as observed for gel and capillary separations
and also allows for detection of the suspicious single band (fragment
size, ~140 bp on the gel; indicated by the arrow in Fig. 5
) that led
this sample to be classified as negative/equivocal. A typical negative
sample profile is displayed for sample T4, which contrasts the profiles
for positive samples T2 and T3 that clearly demonstrate clonality.
|
Analysis of the IGH gene rearrangement products via
microchip electrophoresis also yielded profiles similar to those seen
for the CE analysis (Fig. 6
). Electropherograms B2 and B3 are representative of negative
samples, whereas electropherograms B1 and B4 are characteristic of
positive samples. A slight reduction in resolution can be seen in the
electropherograms for sample B1 with the microchip (Fig. 6B
) compared
with the capillary (Fig. 6A
) separations. However, it is improbable
that this slight reduction will influence the diagnostic capacity.
|
Evaluation of PVP as a sieving matrix for DNA separations in bare
silica channels.
PVP, a polymer that has been shown to function as
a sieving matrix for DNA and also a dynamic coating for bare silica
capillaries (27), was evaluated for DNA separations
in bare silica channels. Dynamic coatings can be added to the
separation buffer to temporarily adhere to, and therefore deactivate,
the silica surface. This is in contrast to traditional capillary
coatings, such as the FC and PA coatings, that are covalently bound to
the capillary surface and require time-consuming coating processes.
Initially, this dynamic coating/sieving matrix polymer was tested in a
37 cm x 50 µm bare silica capillary for comparison with the
standard HEC/FC-coated capillary approach. Fig. 7
A shows that the PVP/bare silica approach for separating the DNA
fragments (HaeIII digest of pBR322) provides effective
resolution of DNA in less than 15 min (Fig. 7A
) in addition to
achieving single-base resolution as evidenced by the separation of the
123- and 124-bp fragments.
|
Because of the potential advantages of a dynamic coating sieving matrix
and the positive results obtained with the capillary format, PVP was
tested as a DNA sieving polymer for microchip electrophoresis. Fig. 7B
shows the separation obtained in a 4.2-cm channel. The resolution
obtained with the microchip was clearly inferior to that obtained with
HEC on the microchip with the same channel length (Fig. 4A
). Despite
the lower resolution, the IGH gene rearrangement products
were analyzed with the PVP system to test applicability to a clinically
relevant analysis (Fig. 8
). As shown, those samples that were positive (B1 and B4) and
negative (B2 and B3) for B-cell clonality are clearly distinguishable.
Similar profiles are observed for the negative samples (B2 and B3) when
compared with the capillary and microchip separations in HEC (Fig. 6
, B2 and B3). However, the positive sample (B1 and B4) profiles display
multiple peaks that were not observed in the HEC systems (Fig. 6
, B1
and B4).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although slab gel electrophoresis is clearly the established method for interrogation of PCR-amplified DNA fragments, capillary-based electrophoresis has been gaining momentum since its introduction at the beginning of this decade as an alternative format (3). Although not performed in this work, sensitivity comparisons have been performed and have shown that sensitivity is equivalent between the CE and slab gel techniques (24). Both electrophoretic techniques (capillary and slab gel) can detect a positive control diluted 1:100 in thymus-derived DNA but failed to detect a 1:1000 dilution. Microchip electrophoresis, a miniaturized format of CE, has begun to attract attention as a higher efficiency embodiment of CE (14)(15) where the potential for parallel processing and integration of other chemistries can be realized (9)(23).
Oda et al. (24) achieved TCRG gene rearrangement
product separation by CE in 17 min using 10 g/L HEC in a 47 cm x
50 µm DB-17-coated capillary at 260 V/cm. The results given in Fig. 2A
parallel these results and show that the amplified products of
TCRG genes can be separated in less than 15 min via CE in
FC-coated µ-SIL-FC capillaries, which were found to provide better
reproducibility and longer lifetimes for DNA analysis than DB-17-coated
capillaries. The resolution obtainable with the conditions described in
the original study (24), although adequate for the T-cell
analysis, was not readily applicable to B-cell analysis. The smaller
fragments (80–140 bp) amplified with B-cell gene rearrangements (see
Fig. 1
) require higher resolution conditions in this region than
conditions that provide adequate separation of the primers and product.
Perhaps most important is the demonstration that the optimized
conditions developed with CE could be translated directly to the
microchip for DNA analysis. In the capillary system, FC-coated
µ-SIL-FC capillaries were used because of their adequate stability
and reproducibility; however, microchips with FC-coated channels are
not commercially available. The method of Hjertén
(26), which provides a relatively stable PA coating of the
silica surface, is one of the most common methods utilized for silica
deactivation. Using this methodology, comparison of the microchip-based
electrophoretic separation with separation in the capillary clearly
demonstrates a substantial decrease in analysis time with a negligible
loss of resolving power. In fact, a standard approach for calculating
resolution (29) showed that resolution on the microchip was
higher for the larger fragments (increased ~28% for 267- to 587-bp
fragments) but lower for the shorter fragments (decreased ~24% for
184- to 234-bp fragments; decreased ~55% for 51- to 104-bp
fragments). Despite these minor changes in resolving power (which can
be accounted for by differences in the silica surfaces, the surface
coatings, and/or discrepancies in the applied voltages), it is clear
that the same diagnostic information can be extracted from both
systems. This is exemplified by the ability to detect the suspicious
band/peak (fragment size ~140 bp on the gel; indicated by the arrow
in Fig. 5
) in sample T1, which led this sample to be classified as
negative/equivocal. This is of seminal importance when envisioning how
multiplex microchip electrophoresis (parallel analysis of many samples)
could impact throughput and turnaround time in molecular diagnostics.
The benefits (e.g., in this particular case, of reducing analysis time
from 2.5 h with slab gel electrophoresis to 160 s with
microchip electrophoresis with no apparent compromise in the quality of
the information provided) are obvious.
Although the performance (resolution) of the microchip and capillary systems were comparable in terms of extracting diagnostic information, the robustness of the two platforms differ at this point. Effective analysis of the PCR samples via CE required only a simple 1:10 dilution of the PCR product with Tris/EDTA buffer. This is consistent with the previous observations that the salt concentrations typically associated with standard PCR mixtures do not to cause extensive problems with capillary-based separations (30)(31)(32). In fact, studies have shown that PCR product injected directly into the capillary for CE analysis does not affect the quality of the separation (32). This contrasts results with the PA-coated microchip, which appeared to be extremely sensitive to the high salt concentrations of the PCR mixture. This is in agreement with previous studies in which either desalting (33)(34) or extensive dilutions (14) were necessary sample pretreatment steps. The loss of electrophoretic functionality is probably the result of deterioration of the channel coating because exposure to salt led to loss of efficiency (broader peaks) and increased migration times. Although desalting the samples before microchip analysis circumvented this problem, this presents a disadvantage because it burdens the test protocol with additional steps that eventually impact the cost of the test. However, it is noteworthy that this problem may be resolved by using different surface passivation approaches. Hofgärtner et al. (35) have recently shown that use of chemistries that provide a more stable surface coating can obliterate the salt-sensitive nature of microchip electrophoresis. In that study, the successive injection of several hundred PCR samples that had not been desalted or extensively diluted was achieved without impacting the quality of the separation.
Under the conditions described in this report, replicate analysis of
the same sample or of many different samples showed reproducible
profiles, provided that the PCR samples were desalted before analysis.
However, an extensive statistical analysis of reproducibility needs to
be performed to evaluate the robustness of this electrophoretic
platform. It is clear that the reproducibility of DNA electrophoresis
and robustness of the microchip will be directly related to how well
the microchannel surface is passivated. As far as covalent modification
of the surface is concerned, the PA coating appears adequate but salt
sensitive, and improved deactivation chemistries (35) are
likely to improve on this. Perhaps the most convenient approach will be
the use of a polymer that not only provides sieving of the DNA but also
deactivates the surface in a dynamic fashion. This is attractive
because most covalent coatings that deactivate silica surfaces are
difficult to produce uniformly and reliably and have limited lifetimes.
The virtues of PVP (27)(36) and poly(ethylene
oxide) (37) as polymers that can dynamically
deactivate silica surfaces and provide a sieving matrix for capillary
DNA separations have been proclaimed. We show, for the first time, that
the microchannel can be dynamically deactivated with PVP
(Mr 360 000) and PCR-amplified DNA
resolved with the same polymer. The resolution was poorer than that
observed with HEC, possibly because the surface of the microchannel was
rougher than in a capillary (38) and, hence, not as
effectively deactivated. The multiple peaks observed in the
IGH gene rearrangement products (Fig. 8
) were not seen in
the HEC/FC capillary system (Fig. 6A
), but similar multiple peak
profiles were obtained for the same samples separated in bare silica
capillaries using PVP. Despite the multiple peak profile, which is
currently under investigation, the use of PVP as a dynamic
coating/sieving matrix for DNA separations in uncoated electrophoretic
microchips under unoptimized conditions allows diagnostic information
about IGH gene rearrangement to be obtained. Although there
are idiosyncratic issues to be resolved, these results illustrate the
potential of dynamic coatings that also function as sieving matrices
for simplifying the use of uncoated electrophoretic microchips for
molecular diagnostics.
In conclusion, slab gel, capillary, and microfabricated chip electrophoresis have been used to analyze TCRG and IGH gene rearrangements, molecular diagnostic assays used to detect T- and B-cell lymphoproliferative disorders. The same information obtained from the slab gel and CE was extracted by the electrophoretic microchip, except with an analysis time that was dramatically reduced. Decreasing the time needed for electrophoresis from as long as 2.5 h on the slab gel to 15 min on the capillary and finally to 160 s on the electrophoretic microchip has obvious implications for diagnostic testing. Although the microchip separations primarily used a PA channel coating and HEC for separation of the DNA fragments, the use of PVP as a dynamic coating/sieving matrix for DNA separations was shown to be feasible. This approach may eliminate the time-consuming channel coating process and irreproducibility problems accepted with silica coatings, allowing for minimal preparation and rapid analysis. Inherent in considering electrophoretic microchip technology as an analytical tool in clinical diagnostics will be its comparison with conventional methods. In this report, microchips are shown to reduce electrophoretic analysis time by 60-fold while maintaining full diagnostic capacity. This is the first step toward accepting this technology as a new paradigm in diagnostics.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  O. A. Gra, J. V. Sidorova, E. A. Nikitin, A. Y. Turygin, S. A. Surzhikov, A. L. Melikyan, A. B. Sudarikov, A. S. Zasedatelev, and T. V. Nasedkina Analysis of T-Cell Receptor-{gamma} Gene Rearrangements Using Oligonucleotide Microchip: A Novel Approach for the Determination of T-Cell Clonality J. Mol. Diagn., April 1, 2007; 9(2): 249 - 257. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. F.Y. Li and L. J. Kricka Clinical Analysis by Microchip Capillary Electrophoresis Clin. Chem., January 1, 2006; 52(1): 37 - 45. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Ferrance, K. Snow, and J. P. Landers Evaluation of Microchip Electrophoresis as a Molecular Diagnostic Method for Duchenne Muscular Dystrophy Clin. Chem., February 1, 2002; 48(2): 380 - 383. [Full Text] [PDF]
|
|
|
|
|
|
I. L. Medintz, L. Berti, C. A. Emrich, J. Tom, J. R. Scherer, and R. A. Mathies Genotyping Energy-Transfer-Cassette-labeled Short-Tandem-Repeat Amplicons with Capillary Array Electrophoresis Microchannel Plates Clin. Chem., September 1, 2001; 47(9): 1614 - 1621. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Medintz, W. W. Wong, L. Berti, L. Shiow, J. Tom, J. Scherer, G. Sensabaugh, and R. A. Mathies High-Performance Multiplex SNP Analysis of Three Hemochromatosis-Related Mutations With Capillary Array Electrophoresis Microplates Genome Res., March 1, 2001; 11(3): 413 - 421. [Abstract] [Full Text]
|
|
|
|
|
|
H. Tian, L. C. Brody, S. Fan, Z. Huang, and J. P. Landers Capillary and Microchip Electrophoresis for Rapid Detection of Known Mutations by Combining Allele-specific DNA Amplification with Heteroduplex Analysis Clin. Chem., February 1, 2001; 47(2): 173 - 185. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Tian, L. C. Brody, and J. P. Landers Rapid Detection of Deletion, Insertion, and Substitution Mutations via Heteroduplex Analysis Using Capillary- and Microchip-Based Electrophoresis Genome Res., September 1, 2000; 10(9): 1403 - 1413. [Abstract] [Full Text]
|
|
|
|
|
|
W. T. Hofgartner, A. F.R. Huhmer, J. P. Landers, and J. A. Kant Rapid Diagnosis of Herpes Simplex Encephalitis Using Microchip Electrophoresis of PCR Products Clin. Chem., December 1, 1999; 45(12): 2120 - 2128. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||
