| Research Article |
Open Access |
|
| Biosensor Based Protein Profiling on Reverse Phase Serum Microarray |
| Ronald Sjöberg1, Lennart Hammarström2 and Peter Nilsson1* |
| 1SciLifeLab Stockholm, School of Biotechnology, KTH – Royal Institute of Technology, Box 1031, SE-17121 Solna, Sweden |
| 2Division of Clinical Immunology, Karolinska Institute, Karolinska University Hospital, Huddinge, SE-14186 Stockholm, Sweden |
| *Corresponding author: |
Peter Nilsson
SciLifeLab Stockholm, School of
Biotechnology
KTH – Royal Institute of Technology
Box 1031, SE-17121 Solna,
Sweden
E-mail: peter.nilsson@scilifelab.se |
|
| |
| Received August 02, 2012; Accepted August 13, 2012; Published August 15,
2012 |
| |
| Citation: Sjöberg R, Hammarström L, Nilsson P (2012) Biosensor Based Protein
Profiling on Reverse Phase Serum Microarray. J Proteomics Bioinform 5: 185-189.
doi:10.4172/jpb.1000233 |
| |
| Copyright: © 2012 Sjöberg R, et al. This is an open-access article distributed under
the terms of the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original author and
source are credited. |
| |
| Abstract |
| |
| The reverse phase serum microarray format enables multi-parallel and simultaneous analysis of literally
thousands of samples, a feature which is of uttermost importance for protein profiling of clinical samples. We have
here screened 2400 serum samples for their potential IgA deficiency by using a fluorescence based reverse phase
serum microarray platform and a biosensor based label-free microarray platform for verification and also compared
our microarray-results to clinical routine ELISA. We have been able to identify possible IgA-deficiencies and to show
the suitability of our microarray-platforms for large-scale screening of clinical serum samples. |
| |
| The two microarray methods show reproducibility and correlation towards each other and low variation
between replicates within each platform. Both of the microarray platforms show less agreement towards ELISA.
The fluorescence based microarray method has been shown to be applicable for large-scale screening of clinically
important serum samples for detection of possibly IgA-deficient patients. Furthermore, it was found that the
microarray based biosensor method could be used for determining the relative differences in concentration of IgA
between the samples. |
| |
| Keywords |
| |
| Reverse phase serum microarray; Protein profiling and
screening; SPR; Biosensor; Label-free |
| |
| Introduction |
| |
| The reverse phase microarray format enables multi-parallel and
simultaneous analysis of literally thousands of samples, a feature which
is of uttermost importance for protein profiling of clinical samples and
which is also often a limiting factor. Other microarray based technology
platforms might have the capacity to profile large numbers of targets
and analytes, but is usually restricted in the multiplexing dimension
of samples. In the reverse phase microarray platform the samples
are immobilised on the substrate in an array configuration and the
affinity reagents are subsequently applied as detection reagents. This
enables large numbers of samples being profiled for one or a few targets
simultaneously under the same experimental conditions while using
low sample volumes. On the other hand, in the forward phase array
where the affinity reagents, antibodies or other binding molecules, are
immobilised on the functionalised surface, the samples containing the
antigens are then applied to the array. This allows for the simultaneous
analysis of a few samples towards a high number of targets. |
| |
| The majority of the described applications of reverse phase
microarrays have so far mainly involved various forms of either cell
lysates [1,2] or tissue lysates [3,4]. We have previously described the
application of serum microarrays. This have been in the context of
analysis of IgA levels in 2000 patients [5], screening for IgA deficiency
in 5000 children [6] and screening for C3 deficiency in newborns with
spotted extracts from paper dried blood spot samples [7]. |
| |
| There is an inherent sensitivity issue with reverse phase serum
microarrays. This limitation arises due to the large dynamic range of
proteins present in blood where many proteins are present in low pg/ml
and thereby fM [8]. Proteins that are present in such low concentrations
in whole blood will be represented by only a few molecules if the
total sample volume is in the sub-nano litre scale spotted in the array
which greatly impedes sensitive measurements. Efforts are needed
and ongoing to develop methodologies and techniques for increased
detectability utilizing various types of signal amplification as for
example reviewed by Nong et al. [9] for DNA-based technologies. It is although still in its current direct setup a platform suitable for detecting
medium to highly abundant proteins while low abundant proteins will
be challenging to detect reliably. Within the IgA profiling the limits
of detection has been found to be in the high ng/ml to the low μg/ml
range [5]. |
| |
| The majority of currently available biosensor instruments does not
allow for multi-parallel analysis in an array-based format, but there are
some platforms that can provide the necessary sample throughput [26]. This makes them suitable as alternative platforms for confirmation
of results obtained from large-scale screenings on reverse phase
fluorescent microarrays. |
| |
| We have previously developed a reverse phase microarray platform
for large-scale simultaneous probing of serum samples [5]. We have
now applied this platform to 2423 serum samples from children in
order investigate the feasibility of using this platform in large-scale
screening of children for detection of IgA-deficiency in combination
with a biosensor based microarray platform for validation. A subset of
those samples with indications of being IgA deficient were transferred
to a SPR-based platform in order to confirm the results, we have also
compared the two microarray platforms to the commonly used ELISA. |
| |
| Materials and Methods |
| |
| Experimental setup |
| |
| 2423 serum samples from children were printed on glass slides and
analysed for their IgA content. 182 of those samples were reprinted on
glass slides and reanalysed as well as printed on a sensor chip for SPRanalysis.
This selected set consisted of 28 samples with a concentration
of IgA of 0.3 mg/ml or less, 100 samples that showed large discrepancies
between ELISA and fluorescence microarrays, and a 54 sample set that
were randomly chosen from the sample pool. A comparison between
replicates and methods were performed as well as a comparison with
ELISA-values for those samples. |
| |
| Fluorescence based microarray |
| |
| The serum samples were diluted 1:10 PBS with 0.5% Tween20,
loaded onto 384 well plates in volumes of 30 μl/well (Genetix) and
printed in duplicate (in 14 identical blocks) on Corning Epoxide
slides (Corning) using a non-contact printing robot (Nano-plotter 2.0,
Gesim). The slides were incubated in a humidity chamber (75%) for 16
hours at 20°C and blocked with Super Block solution (Pierce) using an
air-brush pistol. Polyclonal rabbit anti-human IgA antibodies (DAKO)
were added at a concentration of 46 ng/ml. Alexa Fluor 555 goat-antirabbit
IgG (Molecular Probes) was used as a secondary antibody at a
concentration of 33 ng/ml. The slides were scanned in a G2565BA array
scanner (Agilent) with the photomultiplier tube set to 100% for both
channels and the scan resolution set to 10 μm. The resulting images
were analyzed with GenePix-Pro 5.1 (Molecular Devices) using noncircular
feature alignment. |
| |
| Biosensor based label-free microarray |
| |
| A total of 200 samples consisting of 182 serum samples with known
IgA levels and 18 control samples where diluted 1:10 in 0.5% Tween20
in 1x PBS and printed in duplicates with a non-contact microarray
printer (Nanoplotter2, GeSiM). |
| |
| Blocking of the detection surface was conducted in the FlexChip
instrument (GE Healthcare,Biacore Systems) by filling the flowcell with
0.1% Tween 20 in 1xPBS with 10% Bovine Serum Albumin (BSA Cohn
fraction V, protease free, Saveen Werner) five times for five minutes
each. A baseline was established by flowing the running buffer (0.1%
Tween 20 in 1x PBS) through the flow cell for ten minutes before the
first antibody injection. |
| |
| As a negative control anti-rabbit IgG antibody (Jackson
ImmunoReasearch) was used, anti-human HSA antibody (Jackson
ImmunoReasearch) was used as a positive control, anti-human IgG
antibody (DakoCytomation) was used for verification of normal IgG
levels in the samples and a blank sample consisting of 0.1% Tween 20 in 1xPBS was used to verify that no unspecific binding occurred due to
the dilution buffer. |
| |
| The antibodies where diluted in running buffer to final
concentrations of 50 μg/ml in separate sample tubes and sequentially
recirculated through the flowcell for five minutes each with a five
minute disassociation phase in between every injection. The order of
injection was; anti-rabbit IgG, anti-human IgA, anti-human C3, antihuman
IgG, anti-human HSA, 0.1% Tween 20 in PBS (Figure 3). |
| |
|
Figure 1: Overview of the experimental setup.The sample cohort of 2423
serum samples were initially screened in a high-throughput manner on
fluorescence arrays (1), 182 samples were then re-printed and re-analysed on
fluorescence arrays as well as on the label-free platform (2).These datasets were
then compared with ELISA-values for evaluation purposes (3). |
|
| |
|
Figure 2:A) Quality of spot reproducibility within slides. Plotting the spot
replicates within each slide in a correlation plot with spot replicate one on the
x-axis and spot replicate two on the y-axis showed that the spot replicates have
a good correlation towards each other (R = 0.98) indicating that the spotting
procedure is robust and have good reproducibility and that few spot replicates
are needed in order to get reliable results.
B) Quality of spot reproducibility
between slides. Plotting two slide replicates against each other show good
reproducibility between the replicates with a good correlation (R = 0.98) albeit
with lower correlation for samples with higher contents of target protein.
C)
Quality of spot reproducibility between separate prints. 182 samples were
reprinted and reanalysed on fluorescent arrays and results from the replicate
printings were plotted against each other. This showed comparable results with
good correlation (R = 0.9) between printings which indicates that single printings
would be sufficient for obtaining data. |
|
| |
|
Figure 3:A) The SPR-sensor chip and its flow-cell. The image shows an SPRsensor
chip with attached window gasket. Inlet and outlet is part of the gasket
which is attached after spotting of the array. The area available for spotting the
array is a square of 1 x 1 cm.
B) Schematic illustration of the SPR-sensor chip
flowcell. A schematic illustration of the flowcell and its detection system is shown
where anti-IgA antibodies are circulated through the flowcell and as binding to,
and dissociation from, the immobilised samples occur the change of refraction is
monitored by the detector and recorded.
C: Recorded binding curves from
the first set of three analytes. A total of 400 simultaneously recorded binding
curves from label free analysis with anti-IgG, anti-HSA, and running buffer. Both
anti-IgG and the positive control anti-HSA show strong binding to all spots except
the sample buffer spot. Results from anti-IgG show no correlation to results from
anti-IgA indicating that the IgA-deficient samples contains normal IgG levels.
Anti-HSA functions as a positive control and indicates that all spots are present.
D) Recorded binding corves from the second set of three analytes. Binding
curves from label free analysis with negative control (anti-rabbit), anti-IgA, and
anti-C3 as a second positive control. They show minimal binding of the negative
control to all spots and varying binding from anti-IgA, which is to be expected.
The Spearman correlation coefficient between replicate chips was R = 0.98 which
implies that single chip would be enough in order to validate the results from the
fluorescent arrays. |
|
| |
| The resulting binding curves yields full kinetic data for the binding
by extracting two report points, binding early and binding late, from
the association phase and two report points, stability early and stability
late from the dissociation phase. In this work only the late stability
values were used in the comparison between platforms. This was
repeated on two detection chips to yield a total of four binding curves
for each serum sample. |
| |
| ELISA |
| |
| Total serum IgA levels were determined by sandwich ELISA using
polyclonal rabbit anti-human IgA antibodies (DAKO) and alkaline
phosphatase-conjugated rabbit anti-human serum IgA antibodies
(Jackson ImmunoResearch), added at a concentration of 1.2 μg/ml
and 0.6 μg/ml respectively. Polystyrene plates (Corning) were coated
over night at room temperature with 100 μl per well of the primary
antibody diluted in carbonate bicarbonate buffer (0.05M). The plates
were washed four times with phosphate-buffered saline (PBS) with
0.5% Tween 20 between the incubations. The samples were three-fold
serially diluted in PBS with 0.5% Tween 20. All samples were titrated
against a six-fold serially diluted standard, ranging from 3.1 ng/ml to
100 ng/ml. The samples, the standard dilutions and a blank (PBS with
0.5% Tween 20) were added in duplicate (100 μl/well) and incubated
over night at room temperature. The alkaline phosphatase-conjugated
antibodies were added (100 μl/well) and incubated for 2 hours in room
temperature. p-Nitrophenyl phosphate (Sigma-Aldrich) was added
and the absorbance was read on a Vmax microplate reader (Molecular
Devices). A mean concentration was obtained for each sample using
Deltasoft JV 1.8 (Biometallics). In addition, total serum IgG levels were
determined in the individuals with IgA deficiency using sandwich
ELISA. Polyclonal rabbit anti-human IgG antibodies (DAKO) and
alkaline phosphatase-conjugated polyclonal rabbit anti-human IgG
antibodies (DAKO) were added at a concentration of 0.6 μg/ml and 1.1
μg/ml respectively. The same protocol as for determination of serum
IgA was followed. |
| |
| Data analysis and normalisation |
| |
| For both array platforms a simple global median normalization was
performed in order to correct for experimental artifacts. This was done
by multiplying each data point with the ratio between the median of the
individual slide and the median of all slides. The median intensity of
each spot was averaged based on replicates and the correlation between
the microarrays and ELISA was calculated. A scaling factor was
calculated between the data from the arrays and the data from ELISA
as the ratio between the median of the array data and the median of
the ELISA data and applied to the array data for scaling to mg/ml for
comparison with ELISA values. The coefficient of variation and Pearson
correlation were calculated between replicate ELISA experiments,
replicate microarray printings and between replicate SPR-sensor chips.
All statistical analysis and normalisation were done using R, a language
for data analysis and graphics (www.r-project.com). |
| |
| Results and Discussion |
| |
| We have in a high-throughput fashion screened 2400 serum
samples from children to identify possible IgA-deficiencies with the
goal of utilize a fluorescence-based microarray platform as a tool
for large-scale screening of clinically relevant samples and to use a
biosensor based label-free reverse phase microarray platform for
validation. We have confirmed the results on a subset of 182 samples
consisting of samples that were identified as IgA-deficient on ELISA
and samples that showed large discrepancy between the two platforms.
This was done using a microarray-based biosensor with label-free
surface plasmon resonance detection and the results from both
microarray-based platforms were compared with results from ELISA
in order to compare the results obtained from the microarray platforms
to a clinically common platform and to investigate the suitability of the
SPR-based platform for detecting IgA-deficiencies (Figure 1). |
| |
| Fluorescence microarrays |
| |
| Printing the serum samples yielded spots with uniformly
homogenous morphology. The correlation between replicate slides
and within slides were both r=0.98 and between printings r=0.90. This
high correlation between technical replicates show that the protocol
used enables consistent generation of high quality sample spots which
is necessary in order to ensure that sufficient precision in determining
deficient samples is achieved and to minimise the risk of producing
false negatives (Figures 2A-2C). Separation between deficient samples
and non-deficient samples were achieved to a satisfactory degree
making it possible to identify possible deficiency samples that can be
further analysed in order to validate their lack of IgA (Figure 4D). |
| |
|
Figure 4: A) Scatterplot of the label free SPR-arrays vs. fluorescence arrays. Plotting the two array-based methods show acceptable agreement between
methods with a correlation of r = 0.89. B) Scatterplot of the fluorescence
arrays vs. ELISA. Plotting the fluorescence arrays against ELISA show less
agreement with a lower correlation then between the array-based platforms (r
= 0.66). C) Scatterplot of the SPR-based arrays vs. ELISA. The SPR-based
arrays and ELISA also show less agreement between platforms then the two
array-based platforms (r = 0.76). D) Separation of the deficient samples
from non-deficient. Boxplots of the two groups of samples show satisfactory
separation between groups for the label-free detection method as well. |
|
| |
| This means that it is possible to use this protocol for highthroughput
screening of serum samples for determining which samples
show low IgA-levels that could be a result of IgA-deficiency in the
patient. This would best be achieved by first performing a large initial
screening of samples in order to investigate the relative amount of IgA
and then reanalyse the lowest range to confirm possible deficiencies.
The initial large-scale screening can be performed upon thousands of
samples simultaneously and thus minimise the amount of time and
money needed to perform each sample analysis. This kind of screening
is best performed in a microarray format with fluorescence labelling
due to the possibility to easily print and analyse samples in the tens of
thousands. The subsequent reanalysis of the lowest ranged samples is
more suitable to be performed in a label-free format due to its lack of
complicated and time-consuming blocking procedures and secondary
antibody binding steps. If needed a final analysis of the identified
deficiencies can be analysed on ELISA in order to determine their exact
IgA-concentration. |
| |
| SPR-microarrays |
| |
| We printed 182 serum samples with known IgA levels and 18 control
samples in duplicates forming a total of 400 spots resulting in 400
simultaneous binding curves per injection (Figure 3). Using duplicate
chips and six injections we obtained a total of 4800 binding curves. The
duplicate positive binding curves showed a correlation between chip
replicates of r=0.98as well as sufficient separation between deficient
and non-deficient samples confirming the results from the fluorescence
based analysis. Interrogating the immobilised samples with an anti-IgG
antibody showed that there were no correlation between the IgA-levels
and the general IgG-levels ensuring that the results are specific to the
IgA-levels (Figure 3C). We have chosen to use the stability latepoint
of the dissociation part of the binding curve for measuring the IgAlevels
but the results were similar for the binding early, binding late
and stability early measurements meaning that either of these points on
the binding curve could be used for determining the relative amount
of IgA. |
| |
| Comparison with ELISA |
| |
| When comparing the two microarray-based platforms to ELISA
the results from the microarray-based platforms proved to be more
consistent with each other than with the results from ELISA (Figures
4A– 4C).ELISA show lower correlation between replicates (r=0.57 vs.
r=0.90 for the fluorescence-based platform and r=0.98 for the SPR- based platform) which implicates overall better robustness for the
microarray platforms. This is of high importance in a clinical setting
since it would mean that fewer technical replicates are needed to
minimise the risk of false positives or false negatives. |
| |
| This inconsistency between the two microarray methods and
ELISA becomes more apparent at increasing IgA concentration which
might mean that ELISA lacks accuracy at high concentrations of the
target molecule. Although when using ELISA to identify samples
with IgA deficiency, defined as serum IgA levels below 0.07 mg/ml,
and plotting the two groups for all three platforms they all show good
separation between the deficient samples and the normal samples.
The lower correlation between replicates as well as the low correlation
between the two microarray platforms and ELISA (r=0.66 for ELISA vs.
fluorescence based arrays and r=0.76 for ELISA vs. SPR-based arrays
might be an effect of the low throughput of ELISA that requires the
samples to be analysed in small batches over a long period of time while
the high-throughput array-platforms allow analysis of large sample
cohorts under the same experimental conditions. It could otherwise
have been expected that ELISA and the fluorescence-based microarray
platform would have behaved more similar to each other than to the
SPR-platform due to the similar setup using primary and secondary
labelled binders, while the SPR-platform only need the primary binder
and therefore should be better suited for analysing complex samples. |
| |
| Conclusions |
| |
| We have in this work analysed 2400 serum samples in a large-scale
screening on a microarray platform based on fluorescent labelling
and confirmed the results on a microarray platform based on surface
plasmon resonance. Sufficient separation between deficient and nondeficient
samples was achieved for identification of deficient samples
even though no depletion of the samples was performed before hand.
No correlation between IgA-levels and general IgG-levels could be
found meaning that identified IgA-deficiency samples are not suffering
from general IgG-deficiency. |
| |
| For benchmarking purposes we have compared the two microarray
platforms to ELISA. We found that the fluorescently labelled and
the SPR-based microarray platforms show higher correlation
between replicates than ELISA and confirm each other with better
correlation towards each other than towards ELISA. ELISA show
increasing disagreement with the microarray-based methods at higher
concentrations of the target and show a low reproducibility of results
which imply a lower accuracy for ELISA when analysing for IgA in
serum. |
| |
| This means that utilising a microarray-based platform with
fluorescence-based detection appears highly suitable for screening
large cohorts of samples to determine their relative concentration of
IgA. The set of samples that show the lowest relative concentration can
then be reanalysed on a SPR-based method to screen that smaller subset
of samples in an effective way of validating possible IgA-deficiencies. |
| |
| Acknowledgements |
| |
| We like to thank Magdalena Janzi at Karolinska Institutet for excellent
technical assistance, Björn Persson and Stefan Lövås at GE Healthcare and
Jochen Schwenk at SciLifeLab for fruitful discussions. This study was supported by
the ProNova VINN Excellence Centre for Protein Technology (VINNOVA, Swedish
Governmental Agency for Innovation Systems) and by grants from the Knut and
Alice Wallenberg Foundation. |
| |
|
| References |
| |
- Sevecka M, Wolf-Yadlin A, MacBeath G (2011) Lysate microarrays enable high-throughput, quantitative investigations of cellular signaling. Mol Cell Proteomics 10: M110 005363.
- Nishizuka S, Charboneau L, Young L, Major S, Reinhold WC, et al. (2003) Proteomic profiling of the NCI-60 cancer cell lines using new high-density reverse-phase lysate microarrays. Proc Natl Acad Sci U S A 100: 14229-14234.
- VanMeter AJ, Rodriguez AS, Bowman ED, Jen J, Harris CC, et al. (2008) Laser capture microdissection and protein microarray analysis of human non-small cell lung cancer: differential epidermal growth factor receptor (EGPR) phosphorylation events associated with mutated EGFR compared with wild type. Mol Cell Proteomics 7: 1902-1924.
- Gulmann C, Sheehan KM, Conroy RM, Wulfkuhle JD, Espina V, et al. (2009) Quantitative cell signalling analysis reveals down-regulation of MAPK pathway activation in colorectal cancer. J Pathol 218: 514-519.
- Janzi M, Odling J, Pan-Hammarstrom Q, Sundberg M, Lundeberg J, et al. (2005) Serum microarrays for large scale screening of protein levels. Mol Cell Proteomics 4: 1942-1947.
- Janzi M, Kull I, Sjoberg R, Wan J, Melen E, et al. (2009) Selective IgA deficiency in early life: association to infections and allergic diseases during childhood. Clin Immunol 133: 78-85.
- Janzi M, Sjoberg R, Wan J, Fischler B, von Dobeln U, et al. (2009) Screening for C3 deficiency in newborns using microarrays. PLoS One 4: e5321.
- Jacobs JM, Adkins JN, Qian WJ, Liu T, Shen Y, et al. (2005) Utilizing human blood plasma for proteomic biomarker discovery. J Proteome Res 4: 1073-1085.
- Nong RY, Gu J, Darmanis S, Kamali-Moghaddam M, Landegren U (2012) DNA-assisted protein detection technologies. Expert Rev Proteomics 9: 21-32.
- Klenkar G, Liedberg B (2008) A microarray chip for label-free detection of narcotics. Anal Bioanal Chem 391: 1679-1688.
- Fei YY, Landry JP, Sun YS, Zhu XD, Luo JT, et al. (2008) A novel high-throughput scanning microscope for label-free detection of protein and small-molecule chemical microarrays. Rev Sci Instrum 79: 013708.
- Ozkumur E, Needham JW, Bergstein DA, Gonzalez R, Cabodi M, et al. (2008) Label-free and dynamic detection of biomolecular interactions for high-throughput microarray applications. Proc Natl Acad Sci U S A 105: 7988-7992.
- Wang X, Zhao M, Nolte DD (2008) Area-scaling of interferometric and fluorescent detection of protein on antibody microarrays. Biosens Bioelectron 24: 987-993.
- Mace CR, Striemer CC, Miller BL (2008) Detection of human proteins using arrayed imaging reflectometry. Biosens Bioelectron 24: 334-337.
- Sinensky AK, Belcher AM (2007) Label-free and high-resolution protein//DNA nanoarray analysis using Kelvin probe force microscopy. Nat Nanotechnol 2: 653-659.
- Lynch M, Mosher C, Huff J, Nettikadan S, Johnson J, et al. (2004) Functional protein nanoarrays for biomarker profiling. Proteomics 4: 1695-1702.
- Zheng G, Patolsky F, Cui Y, Wang WU, Lieber CM (2005) Multiplexed electrical detection of cancer markers with nanowire sensor arrays. Nat Biotechnol 23: 1294-1301.
- Torres FE, Kuhn P, De Bruyker D, Bell AG, Wolkin MV, et al. (2004) Enthalpy arrays. Proc Natl Acad Sci U S A 101: 9517-9522.
- Huber F, Lang HP, Hegner M, Despont M, Drechsler U, et al. (2008) Analyzing refractive index changes and differential bending in microcantilever arrays. Rev Sci Instrum 79: 086110.
- Walter JG, Kokpinar O, Friehs K, Stahl F, Scheper T (2008) Systematic investigation of optimal aptamer immobilization for protein-microarray applications. Anal Chem 80: 7372-7378.
- Homola J (2003) Present and future of surface plasmon resonance biosensors. Anal Bioanal Chem 377: 528-539.
- van Vuuren BJ, Read T, Olkhov RV, Shaw AM (2010) Human serum albumin interference on plasmon-based immunokinetic assay for antibody screening in model blood sera. Anal Biochem 405: 114-120.
- de Boer AR, Hokke CH, Deelder AM, Wuhrer M (2008) Serum antibody screening by surface plasmon resonance using a natural glycan microarray. Glycoconj J 25: 75-84.
- Xinglong Y, Dingxin W, Xing W, Xiang D, Wei L, et al. (2005) A surface plasmon resonance imaging interferometry for protein micro-array detection. Sensors and Actuators B: Chemical 108: 765-771.
- Lindquist NC, Lesuffleur A, Im H, Oh SH (2009) Sub-micron resolution surface plasmon resonance imaging enabled by nanohole arrays with surrounding Bragg mirrors for enhanced sensitivity and isolation. Lab Chip 9: 382-387.
- Scarano S, Mascini M, Turner AP, Minunni M (2010) Surface plasmon resonance imaging for affinity-based biosensors. Biosens Bioelectron 25: 957-966.
|
| |
| |
