| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Department of Biotechnology, University of Turku, Tykistökatu 6, FIN-20520 Turku, Finland.
a Author for correspondence. Fax 358-2-3338050; e-mail
harri.harma{at}utu.fi.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
Methods: Streptavidin was covalently coated on 107-nm nanoparticles containing >30 000 europium molecules entrapped with ß-diketones. In a model assay system, these nanoparticles were used to trace biotinylated prostate-specific antigen (PSA) in a microtiter plate format.
Results: The detection limit (mean + 3 SD of the zero calibrator) of biotinylated PSA was 0.38 ng/L, corresponding to 10 fmol/L or 60 zeptomoles (60 x 10-21 moles) of PSA. Moreover, single nanoparticles, representing individual PSA molecules, were visualized in the same microtiter wells with a time-resolved fluorescence microscope using a x10 objective. Single nanoparticles, possessing high specific activity, were also detected in solution by a standard time-resolved plate fluorometer.
Conclusions: The universal streptavidin-coated europium(III) nanoparticle label is suitable for detection of any biotinylated molecule either in solution or on a solid phase. The europium(III) nanoparticle labeling technology is applicable to many areas of modern biochemical analysis, such as immunochemical and multianalyte DNA-chip assays as well as histo- and cytochemistry to improve detection sensitivities.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
Ultra-high-throughput screening, chip-based technologies, multitarget detection systems, and diagnostic screening are rapidly growing areas, and an ongoing trend is to miniaturize these assay formats. In microarray and microspot techniques, spatial resolution of individual reactive sites on a chip becomes extremely important. At the same time, improved labeling and detection technologies are required to analyze smaller sample volumes and to measure sample on a limited solid-phase area. The use of intense fluorescence labels with high specific activity and low nonspecific signal is a prerequisite before optimal miniaturization can be accomplished. Obviously, bioimaging also benefits from such a labeling technology.
Time-resolved fluoroimmunoassays were introduced in the early 1980s (7). The high specific activity and very low background signals provided by the temporal separation of the long-lived emission signals have enabled detection of analytes with superior sensitivity (7)(8). The most successful time-resolved fluorescence technology, the commercial DELFIA concept (Perkin-Elmer Life Sciences), is based on the enhanced signal produced by lanthanide ions as they are dissociated and successively entrapped by ß-diketones, detergents, and synergistic agents (9). We previously have shown that 10 000 intrinsically fluorescent europium chelates (nondissociative) can be monitored on a single microparticle by a confocal time-resolved fluorometer (10). To further improve the sensitivity provided by nondissociative labels, we describe here a europium-label detection technology applicable to various bioanalyses. These nanosized polymer labels contain >30 000 europium molecules entrapped by ß-diketones, which possess one of the highest quantum yields of the known lanthanide chelators. The polymer shell efficiently removes fluorescence-quenching water from the vicinity of the chelate by producing a hydrophobic environment. Furthermore, carboxylic groups on the surface of the label enable simple and convenient conjugation of virtually any biomolecule. With this technology, we demonstrate detection limits of zeptomoles for biotinylated prostate-specific antigen (PSA)1 in a conventional microtiter format and visualization of PSA molecules in the same microtiter wells by a time-resolved microscope.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
The number of europium(III) ions in a single nanoparticle was determined by adding the nanoparticles to a 1 mL/L Triton X-100 solution or to the enhancement solution (DELFIA; Perkin-Elmer Life Sciences). The signals obtained by the Victor 1420 multilabel counter were compared with a europium calibrator in the enhancement solution.
The time-resolved fluorescence excitation and emission spectra of nanoparticles (3 pmol/L) in 1 mL/L Triton X-100 and the spectra of europium ions (1 nmol/L) in the enhancement solution were measured with the LS5 spectrofluorometer (Perkin-Elmer Instruments).
coating of nanoparticles with streptavidin
Nanoparticles were prewashed with 10 mmol/L phosphate buffer, pH
7.0, on a Nanosep microporous centrifugal filter (300 kDa cutoff; Pall
Filtron). Phosphate buffer was added to the particles, and the solution
was sonicated with a tip sonicator (Labsonic U; B. Braun) at 80 W for
5 s. Carboxyl groups on the surface of nanoparticles were
activated with 10 mmol/L
N-(3-dimethylaminopropyl)-N'-ethylcarbodiimide
and N-hydroxysulfosuccinimide (Fluka) for 30 min. The
activated particles were washed once with 10 mmol/L carbonate buffer,
pH 9.0, and 15 µmol/L streptavidin was added. After 2 h of
incubation, the streptavidin-coated particles were washed five times
with a 2 mmol/L Tris-HCl solution, pH 7.0, and stored at 4 °C.
detection limit of nanoparticles in solution
The detection limits of the different-sized particles were
determined with a Victor 1420 multilabel counter (Perkin-Elmer Life
Sciences). Nanoparticles were diluted into the 1 mL/L Triton X-100
solution or the enhancement solution, and a 200-µL aliquot of the
nanoparticle solution was monitored. The lower limit of
detection was calculated from the signal of the dilutions containing no
particles (plus 3 SD).
psa immunoassay
PSA (100 µg; provided by Dr. H. Lilja, Lund University Hospital,
Malmö, Sweden) was biotinylated with 60 µmol/L biotin
isothiocyanate (Perkin-Elmer Life Sciences) in 500 µL of 50
mmol/L carbonate buffer, pH 9.8, for 2.5 h. The biotinylated
PSA was separated from unbound biotin on NAP-5 and NAP-10 columns
(eluted with Tris buffer). The PSA assay was carried out on an
anti-PSA microtiter plate [monoclonal antibody (mAb)-H117;
Perkin-Elmer Life Sciences]. In the two-step assay, 5 µL of
biotinylated PSA was first incubated in 30 µL of a commercial assay
buffer (Perkin-Elmer Life Sciences) for 40 min. After a washing step,
2 x 108 streptavidin-coated
nanoparticles were added to the well in 40 µL (4-h incubation at room
temperature). After the final wash, the quantity of the analyte was
detected directly on the surface of the microtiter well with the Victor
1420 fluorometer in time-resolved mode. The detection limit of the
assay was calculated based on the signal for the zero calibrator plus 3
SD.
imaging of nanoparticles
After the PSA dilution curves were established, the same
microtiter wells (no additional sample preparation) were studied with a
time-resolved fluorescence microscope to visualize PSA molecules. A
Nikon Eclipse E600 light microscope (Nikon) equipped for time-resolved
fluorescence imaging (flashlamp, chopper, CCD camera) was used. A
rotating chopper was used to close the light path to the camera during
the excitation pulse (10 µs; wavelength, 340 nm; Xenon flashlamp;
Perkin-Elmer Life Sciences) and the delay time (between excitation and
light collection). Thereafter, the rotating chopper opened the light
path to the camera for 600 µs (emission, 615 nm) and then closed the
light path. The measurement cycle was repeated at 50 Hz for 30–45 s. A
cooled scientific CCD camera (UltraPix; Perkin-Elmer Optoelectronics)
was used to detect nanoparticles. The objectives used were x10 and
x60 with numerical apertures of 0.3 and 0.85, respectively. The
time-resolved images were analyzed with the Image-Pro Plus, Ver. 4.0,
image analysis software (Media Cybernetics).
Scanning electron microscope (SEM) images were taken with a S360 electron microscope (LEO Electron Microscopy). For the SEM images, the particles were dried on an etched-track polycarbonate filter (50 nm; Isopore; Whatman) and coated with gold.
|
Results and Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
shows that no significant difference was observed in the
excitation and emission spectra of europium(III) ions in the
enhancement solution or the nanoparticles in the Triton X-100 solution
because the chelating ligand was the same in both solutions. The
difference in the far-ultraviolet region of the excitation spectra can
be attributed to the absorbance of the polystyrene polymer matrix of
the nanoparticles. In addition to the measured spectra, decay times
were determined for the 107-nm nanoparticle suspension in the 1 mL/L
Triton X-100 solution as well as for europium(III) in the enhancement
solution. Decay times of 
720 µs were measured for both solutions,
indicating identical coordination site occupancy of europium(III) ions
in the two systems.
|
The signals of lanthanide chelates are known to be low, as low as
1000-fold lower than signals of rapidly decaying fluorophores, mainly
because the emission light is temporally resolved and only a fraction
of the emitted light is collected. The concentration of chelated
europium(III) ions in the nanoparticles was 80 mmol/L, which would
apparently cause an inner-filter effect when applied to rapidly
decaying fluorophores (11). It is not uncommon that rapidly
decaying fluorophores self-quench in micromolar concentrations.
However, when lanthanide chelates are used, no such effect has been
found even in high (millimolar) concentrations. Therefore, the loss in
signal attributable to the temporally resolved emission can be
effectively compensated by use of nanoparticles containing a high
concentration of lanthanide chelates.
The detection limits of the nanoparticles were measured in 200 µL of
the 1 mL/L Triton X-100 solution with the standard time-resolved plate
fluorometer. As shown in Fig. 2
, one single 408-nm nanoparticle was detected in solution, and
the detection limits for the 210- and 107-nm particles were 8 and 75,
respectively.
|
imaging of nanoparticles
In Fig. 3A
is shown an image of nanoparticles (noncoated, 107 nm) dried
on an etched-tracked membrane when a Nikon Eclipse E600 light
microscope equipped for time-resolved fluorescence measurements was
used for the imaging (x60 objective and 30-s exposure time). To ensure
that each measured individual dot was a single particle, we dried
1000 nanoparticles (calculated from particle concentration, 1%
solids; diameter, 107 nm) in a small area on the membrane and took
images using a x10 objective and a 45-s exposure time. Image analysis
software was used to calculate the number of individual measured dots.
The number of dots corresponded accurately to the assumed particle
number, 1018 ± 104 (eight individual dried areas). However, the
image analysis could not solely ensure that each dot was a single
particle molecule. Therefore, we compared images of nanoparticles in
the time-resolved fluorescence microscope (Fig. 3, A and B
) and in the
SEM (Fig. 3C
). These images clearly demonstrated that each measured dot
in the microscope image was, indeed, a single nanoparticle.
|
Nanosized particles are commonly used in agglutination tests. In these
tests, it is desired that the particles aggregate when exposed to
analyte. In many assay concepts, aggregation is not, however, a desired
property. Although nanosized labels have been used in several
single-molecule imaging, DNA, and immunoassay studies, very little
information, if any, has been given about the obvious aggregation
problem. Because the time-resolved fluorescence microscope allowed
quantitative particle counting, we had strict control over the
nanoparticle suspensions used, as exemplified in Fig. 4
: noncoated (Fig. 4A
), aggregated streptavidin-coated (Fig. 4B
),
and nonaggregated streptavidin-coated (Fig. 4C
) nanoparticles measured
with a time-resolved fluorescence microscope and SEM. If the assays
were performed with an aggregated particle suspension, it would be
obvious that the number of single nanoparticles was not accurate, and
no reliable results could be obtained. To demonstrate this effect, the
noncoated nanoparticles were diluted first with
H2O or 5 mol/L NaCl solution and subsequently
10 000-fold with the 1 mL/L Triton X-100 solution. Because a high
sodium concentration is known to aggregate nanosized carboxyl
particles, the final dilution of the NaCl-treated solution was assumed
to be highly aggregated. The non-NaCl-treated and NaCl-treated
solutions measured in the standard time-resolved plate fluorometer
yielded the same signals (data not shown). However, the CVs varied
significantly. For the non-NaCl-treated and NaCl-treated nanoparticle
solutions, the CVs were 0.4% and 21%, respectively, when six
individual measurements were compared.
|
psa immunoassay
A model PSA assay was performed on a microtiter plate coated with
a PSA-specific mAb [mAb-H117 (12)]. Biotinylated PSA was
incubated in the first assay step, followed by a second incubation with
streptavidin-coated 107-nm nanoparticles. The analyte reaction volume
was small (30 µL) because we wanted PSA to react on the bottom of the
plate for direct surface detection. In this concept, no additional
enhancement step was required. Fig. 5
shows the dilution curve of the PSA assay. The response was
linear over three orders of magnitude. The lower limit of detection was
0.38 ng/L, corresponding to 10 fmol/L or 60 zeptomoles (60 x
10-21 moles) of PSA molecules. This is more than
10-fold lower than the previously reported molecule quantity in a
microtiter plate-based PSA assay (0.38 ng/L for a 5-µL sample vs 1
ng/L for a 50-µL sample) (13). The present PSA
assay setup cannot be considered a true assay because the analyte was
biotinylated.
|
To perform a true noncompetitive sandwich PSA assay, we biotinylated a second PSA-specific mAb (mAb-5A10). This work (14) was a continuation of the work presented here. The detection limit was 1.6 ng/L, which is not significantly lower than that of the present assay, although this assay concept suffered from increased nonspecific binding produced by the additional tracer molecule, biotinylated mAb-5A10. The incubation time of streptavidin-coated nanoparticles with the surface-bound biotinylated mAb corresponding to the analyte concentration was short, 5 min. We have also theoretically and experimentally studied the kinetics of the nanoparticles in a sandwich-type assay format and found that confining the number of surface-coupled antibodies or other biospecific binding reactants, such as streptavidin, to the nanoparticles produced enhanced association rate constants when compared with a soluble biospecific binding reactant (Soukka T, Härmä H, Paukkunen J, Lövgren T. Immunoassays based on multivalent nanoparticle-antibody bioconjugates utilize kinetically enhanced monovalent binding affinity. Submitted for publication).
In Fig. 6
, images of the same microtiter wells measured for the
biotinylated PSA dilution curve (above) were taken with the
time-resolved fluorescent microscope using a x10 objective and a 45-s
exposure time. The calculated number of individual nanoparticles
(Image-Pro analysis software) followed the concentration of PSA as
anticipated, producing a dilution curve with a lower limit of
detection, 0.8 ng/L, that was in a good agreement with the result
measured with the standard time-resolved plate fluorometer. With this
technology, imaging of individual PSA molecules was possible, although
the practical limit of the assay was defined by nonspecific binding and
single molecules were detected only in principle.
|
At the lower limit of detection, 35 000 PSA molecules were detected, and hence, in the probe area of the microscope (x10 objective; 1.3 mm2), 650 nanoparticles representing PSA molecules should have been found (estimated area for 30-µL volume was 70 mm2). However, the number of nanoparticles found in the microscope image was <5% of the estimated amount at the lower limit of detection. Obviously, the large size of the nanoparticles restricted somewhat the access of nanoparticles to every PSA molecule on the surface, and the random coupling of biotin through the primary amines of PSA was also likely to cause steric hindrance. Furthermore, very small amounts of streptavidin or streptavidin fragments were apparently present in solution. It is not difficult to imagine that some tetrameric streptavidin molecules or streptavidin fragments dissociated from the nanoparticle surface when the total number of streptavidin in one assay was in the order of 1011. The soluble streptavidin molecules bound fast and without restricted accessibility to biotinylated analyte. In conventional assays, the dissociation of streptavidin would not be detectable because the sensitivities of these assays would not allow distinguishing of very small quantities of analyte. If these factors influencing the reduced efficiency of the PSA detection can be controlled, a more than 10-fold improvement in the lower limit of detection could possibly be reached.
PSA molecules with nanosized time-resolved fluorescent labels in a microtiter plate format were imaged with a x10 objective. With this objective, no specialized reaction well was required for the detection of PSA because the working distance of the objective was large, 16 mm. We used a 45-s exposure time to image the nanoparticles, but the individual nanoparticles were visible after a 1-s exposure. The longer exposure time was used only for demonstration purposes to improve the signal-to-noise ratio (>100 with a x10 objective and 45-s exposure time). Because a shorter exposure time can be applied, a more than 2-fold (10-fold reduced europium content on one nanoparticle) smaller particle size could be used to detected analyte molecules, which in turn makes the nanoparticle approach more suitable for other imaging technologies, such as cyto- and histochemistry.
In conclusion, we have introduced a detection technology based on nanosized europium(III) particles. Biotinylated PSA was detected with a detection limit of 60 zeptomoles in a microtiter plate format using a standard time-resolved plate fluorometer, and individual PSA molecules were visualized with a time-resolved microscope. Clearly, microarray and microspot technologies could benefit from the use of these intense time-resolved fluorescent labels in biochemical analyses, e.g., detecting subtle variations in the expression of mRNA, as well as imaging technologies such as cyto- and histochemistry. Multiplex measurements could also be performed with the current labeling technology because terbium(III), samarium(III), or dysprosium(III), all well known and characterized in solution with chelating ligands, can be used instead of europium(III) on nanoparticles. As the time-resolved labeling technology has improved and become more familiar over the years, the detection technology has also gained more interest among instrument manufacturers. Consequently, a high-sensitivity time-resolved labeling technology such as the current nanoparticle approach could be readily implemented in any standard laboratory.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  J. O. Koskinen, R. Vainionpaa, N. J. Meltola, J. Soukka, P. E. Hanninen, and A. E. Soini Rapid Method for Detection of Influenza A and B Virus Antigens by Use of a Two-Photon Excitation Assay Technique and Dry-Chemistry Reagents J. Clin. Microbiol., November 1, 2007; 45(11): 3581 - 3588. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Xu and Q. Li Multiple Fluorescent Labeling of Silica Nanoparticles with Lanthanide Chelates for Highly Sensitive Time-Resolved Immunofluorometric Assays Clin. Chem., August 1, 2007; 53(8): 1503 - 1510. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Rundstrom, A. Jonsson, O. Martensson, I. Mendel-Hartvig, and P. Venge Lateral Flow Immunoassay Using Europium (III) Chelate Microparticles and Time-Resolved Fluorescence for Eosinophils and Neutrophils in Whole Blood Clin. Chem., February 1, 2007; 53(2): 342 - 348. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. M.E. Azzazy, M. M.H. Mansour, and S. C. Kazmierczak Nanodiagnostics: A New Frontier for Clinical Laboratory Medicine Clin. Chem., July 1, 2006; 52(7): 1238 - 1246. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Kokko, N. Johansson, T. Lovgren, and T. Soukka Enzyme Inhibitor Screening Using a Homogeneous Proximity-Based Immunoassay for Estradiol J Biomol Screen, June 1, 2005; 10(4): 348 - 354. [Abstract] [PDF]
|
|
|
|
|
|
P. Huhtinen, A.-M. Pelkkikangas, S. Jaakohuhta, T. Lovgren, and H. Harma Quantitative, Rapid Europium(III) Nanoparticle-Label-Based All-in-One Dry-Reagent Immunoassay for Thyroid-Stimulating Hormone Clin. Chem., October 1, 2004; 50(10): 1935 - 1936. [Full Text] [PDF]
|
|
|
|
|
|
D. S. Fran and M. W. Sundberg RE: Europium Nanoparticles and Time-resolved Fluorescence for Ultrasensitive Detection of Prostate-specific Antigen Clin. Chem., September 1, 2001; 47(9): 1743 - 1744. [Full Text] [PDF]
|
|
|
|
|
|
T. Soukka, J. Paukkunen, H. Harma, S. Lonnberg, H. Lindroos, and T. Lovgren Supersensitive Time-resolved Immunofluorometric Assay of Free Prostate-specific Antigen with Nanoparticle Label Technology Clin. Chem., July 1, 2001; 47(7): 1269 - 1278. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. P. Diamandis Signal Amplification in Time-resolved Fluorometry Clin. Chem., March 1, 2001; 47(3): 380 - 381. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), 210-nm (
), and 408-nm
(•) particles in the 1 mL/L Triton X-100 solution (200 µL) measured
with a standard time-resolved plate fluorometer.
