| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Abstracts of Oak Ridge Posters |
1
Advanced Diagnostics Group, Dade Behring Inc., PO Box 49013, San Jose, CA 95161-9013
a author for correspondence: fax 408-239-2707, e-mail alan_dafforn{at}dadebehring.com
Many of the emerging technologies in clinical chemistry and research require the ability to perform hundreds or thousands of measurements on a single sample such as amplified DNA, typically by contacting the sample with an array of different probes or other reagents. If these array approaches are to be practical, the underlying technology must be simple, robust, inexpensive, and amenable to automation. The Luminescent Oxygen Channeling Immunoassay (LOCITM) is a recently developed homogeneous assay method that should be suitable for arrays because of its simplicity. However, to perform large numbers of measurements on reasonable sample sizes (e.g., 500 different measurements on aliquots of a 50-µL volume), it must be possible to detect LOCI signals from very small volumes. Miniaturization has also become a central theme in other areas of clinical chemistry (1). Accordingly, we have constructed a LOCI microscope and used it to demonstrate sensitive detection of analytes in small volumes for three types of assays of potential interest in arrays: detection of a single-stranded DNA fragment, detection of a double-stranded DNA amplicon, and an immunoassay for a protein.
LOCI is a sensitive (femtomolar) detection method that uses chemiluminescence to quantify latex agglutination (2). This technique utilizes one latex particle dyed with a photosensitizer and a second dyed with a chemiluminescent dye, both having binding ligands on their surfaces. Particle suspensions are mixed with the sample, and cross-linking by any analyte present leads to formation of bead pairs or higher aggregates. When the suspension is then illuminated, singlet oxygen is generated by the sensitizer particle, migrates to the chemiluminescent particle, and generates light. Nonspecific signals are low because singlet oxygen decays before it can reach unpaired particles.
A small-volume LOCI reader was constructed by modifying a fluorescence
microscope to allow sample illumination with a 678 nm laser and
monitoring of chemiluminescence with a photomultiplier tube.
Illumination was accomplished either through the objective, using a
beam splitter, or from below the stage, using a shutter to protect the
photomultiplier; similar results were obtained in either mode. Samples
were read by adding a 10-µL aliquot to a hemocytometer cell and
imaging a single field. The volume imaged was defined by the depth of
the cell (100 µm) and the diameter of the field (365 µm) as 
10
nL.
Three different assays were compared by performing incubations as described previously (2), and then removing aliquots of the final mixture and reading either using the microscope or using our regular LOCI readers. These prototype readers illuminate and read 20- or 80-µL aliquots through standard optics. Reagents and procedures for two of the analytes, an oligonucleotide linker and thyroid-stimulating hormone (TSH), were essentially as described previously (2). Briefly, in one assay an oligonucleotide linker included the sequences 5'-(TACT)5 and T20 separated by a short spacer. This linker forms bead pairs between sensitizers with conjugated A24 and chemiluminescent particles with conjugated 5'-(AGTA)6. In the other assay, TSH binds first to a chemiluminescent bead conjugated with one monoclonal anti-TSH antibody and to a second biotinylated monoclonal anti-TSH antibody. Streptavidin beads are then added to form bead pairs if TSH is present. Assays were read by three cycles (five for TSH) of illuminating for 1 s at 678 nm followed by 1 s of light collection.
Detection of double-stranded DNA was demonstrated using an amplicon
from Chlamydia trachomatis. A 518-bp sequence from cryptic
plasmid pLGV440 was amplified by conventional PCR using the primers
5'-GGA CAA ATC GTA TCT CGG GTT ATT-3' and 5'-GGA AAC CAA CTC TAC GCT
GTT-3'. The final concentration of the amplified DNA was estimated as
40 nmol/L by comparison to an earlier LOCI calibration curve.
Various dilutions of the amplicon were mixed with 2 µg of each LOCI
bead (same as for single-stranded DNA detection above). Solutions also
contained each of the following probes at a concentration of 50 nmol/L
in a total volume of 40 µL of PCR buffer (70 mmol/L KCl, 10 mmol/L
Tris-HCl, 4 mmol/L MgCl2, 0.2 g/L
acetylated bovine serum albumin, pH 8.2):
Target-specific binding regions are underlined. X represents OCH2CH((CH2)4NH2)CH2OH, a group introduced to block the 3' end against possible enzymatic extension. Solutions were covered with 20 µL of mineral oil in 200-µL MicroAmpTM tubes (PE BioSystems), and then heated to 95 °C for 2 min to denature the amplicon, cooled to 50 °C for 15 min to allow probes to bind, then cooled to 37 °C for 70 min to allow formation of the complex with beads. Aliquots were then read as above.
The results for all three assays are summarized in Fig. 1
. The ratio of specific signal (signal - background) to
background is presented as a function of analyte concentration in each
case to allow comparisons over many orders of magnitude. In general,
the assays lost only 14- to 64-fold in signal/background over a 2000-
to 8000-fold decrease in volume. As expected, 1000-fold decreases in
signal were partially compensated by decreased background. (Raw
background counts for large volume and 10 nL are as follows: DNA
Linker, 6506 and 60; Chlamydia DNA, 4777 and 68; and TSH,
14 934 and 183.) As they should be, the curve shapes in Fig. 1
are in
general parallel and log-linear except when very little signal is
present. The decreasing signal seen at high concentrations of
Chlamydia amplicon is frequently observed when LOCI is
used to detect DNA amplification in a closed tube. It is believed to
result from saturation of bead surfaces, but it can be avoided by
changes in protocol (3).
|
Detection limits for each assay were defined as the concentrations at which observed signals exceeded background by 3 SD and were estimated from linear plots of low concentration points. The change in the lower limit of detection as a function of volume was larger than that in signal/background because the statistical CV of the background increased at small volumes. The limits for large volumes and for 10 nL were as follows:
3 pmol/L) Because most array applications are likely to involve detection and quantification of undiluted amplicons or of expressed proteins, LOCI has ample limits of detection for most applications even in very small volumes.
The imprecision of the specific signal (signal - background) was determined in the TSH assay as 82 ± 10 counts (n = 5; CV = 12%) at 3.57 pmol/L TSH and 731 ± 68 counts (n = 5; CV = 9%) at 35.7 pmol/L. The imprecision was also estimated using a single bead dyed with both sensitizer and chemiluminescent dye according to the basic procedures described previously (2). Ten replicate aliquots of a bead suspension gave 2450 ± 328 counts (CV = 13%). The most likely source of the observed variability is manual positioning of each well of the slide under the microscope objective in this prototype instrument.
Realization of practical LOCI arrays will also require a sample cassette. This could be as simple as a matrix of closed wells, each containing LOCI reagents for a specific measurement and interconnected by channels to distribute the sample. Fortunately, the dimensions required are easily obtainable by inexpensive molding techniques (4). For example, a well with a 50-nL volume could be 125 µm deep and 714 µm in diameter. If 0.5 cm of channel on the average was required to connect each well and the channel had a cross-sectional area of 104 µm2, then another 50 nL per well would be required to fill the channel. Thus, a 50-µL sample volume would be sufficient to fill 500 wells and the necessary connecting channels.
In summary, LOCI offers several advantages for signal detection
from arrays and other miniaturized devices: The assay retains ample
sensitivity for analytes of likely interest in such devices. An
oligonucleotide could be detected at 1 pmol/L (6000 molecules), the
protein TSH could be detected at 2 pmol/L, and a DNA amplicon could be
detected even at a 1:10 000 dilution. In addition, arrays large enough
for clinical diagnostic purposes should be feasible (500 or more
measurements/sample). Homogeneous assay arrays should also be much
simpler to manufacture than many types of arrays because no surface
chemistry must be performed on a chip. The absence of surface
chemistry or absorption should also give greater reproducibility
compared with spotting technologies and simplify quality control. The
use of generic reagents also simplifies preparation of large arrays.
Finally, homogeneous assays offer relatively fast kinetics and
simplicity of protocol.
We thank Neal DeChene, John Pease, Sharat Singh, and Raj Singh for supplying many of the reagents used in this work and Sam Rose for many helpful insights.
Footnotes
1 present address: PE
Biosystems, 850 Lincoln Centre Dr., Foster City, CA 94404 
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  J. F. Glickman, X. Wu, R. Mercuri, C. Illy, B. R. Bowen, Y. He, and M. Sills A Comparison of ALPHAScreen, TR-FRET, and TRF as Assay Methods for FXR Nuclear Receptors J Biomol Screen, February 1, 2002; 7(1): 3 - 10. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or a large volume (
; 80 µL in A and
B, 20 µL in C).