Clinical Chemistry
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Clinical Chemistry 52: 2273-2280, 2006. First published October 19, 2006; 10.1373/clinchem.2006.073569
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
clinchem.2006.073569v1
52/12/2273    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Luo, Y.
Right arrow Articles by Fu, W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Luo, Y.
Right arrow Articles by Fu, W.
Related Collections
Right arrow Automation and Analytical Techniques
(Clinical Chemistry. 2006;52:2273-2280.)
© 2006 American Association for Clinical Chemistry, Inc.

Automation and Analytical Techniques

Rapid and Simultaneous Quantification of 4 Urinary Proteins by Piezoelectric Quartz Crystal Microbalance Immunosensor Array

Yang Luo1, Ming Chen1, Qianjun Wen2, Meng Zhao1, Bo Zhang1, Xiaoyan Li3, Feng Wang1, Qing Huang1, Chunyan Yao1, Tianlun Jiang1, Guoru Cai4 and Weiling Fu1,a

1 Department of Laboratory Medicine, 2 Laboratory Center of Urology, and 3 Endocrinology Department, Southwest Hospital, The Third Military Medical University, Chongqing, People’s Republic of China.
4 The 26th Research Institute, Chinese Electronic Scientific and Technical Group Company, Chongqing, People’s Republic of China.

aAddress correspondence to this author at: Department of Laboratory Medicine, Southwest Hospital, Third Military Medical University, Chongqing 400038, People’s Republic of China. Fax 86-23-65460909; e-mail weilingfu{at}yahoo.com.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results and Discussion
 References
 
Background: Urinary proteins are predictive and prognostic markers for diabetes nephropathy. Conventional methods for the quantification of urinary proteins, however, are time-consuming, and most require radioactive labeling. We designed a label-free piezoelectric quartz crystal microbalance (QCM) immunosensor array to simultaneously quantify 4 urinary proteins.

Methods: We constructed a 2 x 5 model piezoelectric immunosensor array fabricated with disposable quartz crystals for quantification of microalbumin, {alpha}1-microglobulin, ß2-microglobulin, and IgG in urine. We made calibration curves after immobilization of antibodies at an optimal concentration and then evaluated the performance characteristics of the immunosensor with a series of tests. In addition, we measured 124 urine samples with both QCM immunosensor array and immunonephelometry to assess the correlation between the 2 methods.

Results: With the QCM immunosensor array, we were able to quantify 4 urinary proteins within 15 min. This method had an analytical interval of 0.01–60 mg/L. The intraassay and interassay imprecisions (CVs) were <10%, and the relative recovery rates were 90.3%–109.1%. Nonspecificity of the immunosensor was insignificant (frequency shifts <20 Hz). ROC analyses indicated sensitivities were ≥95.8% and, specificities were ≥76.3%. Bland–Altman difference plots showed the immunosensor array to be highly comparable to immunonephelometry.

Conclusions: The QCM system we designed has the advantages of being rapid, label free, and highly sensitive and thus can be a useful supplement to commercial assay methods in clinical chemistry.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results and Discussion
 References
 
Diabetic nephropathy is a common cause of renal failure and is a major complication of both type 1 and type 2 diabetes (1)(2). Early diagnosis and treatment may delay or even prevent the onset of diabetic nephropathy. High– and low–molecular-weight urinary proteins are more sensitive than urea and creatinine clearance for diagnosing initial renal damage (3). Microalbuminuria, defined as protein amounts of 30–300 mg measured during a 24-h urine collection, is a sensitive indicator of histopathologic alteration from nephropathy (confirmed by biopsy) and has been used as a noninvasive method to identify kidney involvement in diabetes mellitus (4)(5)(6). Low–molecular-weight urinary proteins, such as a1-microglobulin (A1M) 1 and ß2-microglobulin (B2M), are the best markers for early detection of renal tubular dysfunction associated with conditions such as heavy metal intoxication, diabetic nephropathy, and pyelonephritis (7)(8), and IgG excretion is closely associated with the extent of tubulointerstitial damage and A1M excretion (9). Thus, the combined analyses of microalbumin (MA), A1M, B2M, and IgG can provide comprehensive information about early renal impairment and subsequent disease progression.

RIA and fluorescence-based immunoassays are conventional approaches to quantification of urinary proteins. Although RIA has many advantages, such as high sensitivity and low cost, the radioactive-labeled material requires special disposal and has stability problems. Fluorescence-based immunoassays also have high sensitivity and have been applied to many platforms (10)(11), but their manipulation procedures are relatively time-consuming and require detection of labeled molecules. Sodium dodecyl sulfate electrophoresis is a useful method, but it has a lower limit of quantification that is inadequate for clinical specimen analysis (12). HPLC cannot absolutely differentiate albumin from other urinary proteins if it is not combined with special detection strategies (13). Immunonephelometry (INM), a newer technique with full automation that has become the major method for protein quantification, requires sophisticated facilities and rigidly matched reagents and offers suboptimal precision, all of which limit its clinical applications (14). Thus a fast, sensitive, label-free, and easy-to-manipulate technique for urinary protein quantification is needed for clinical use.

Piezoelectric quartz crystal microbalance (QCM) immunosensors, because of the inherent specificity of antigen–antibody combinations and high sensitivities of physical transducers, have promising applications to clinical chemistry (15)(16)(17). We have set up a QCM biosensor platform and successfully quantified human chorionic gonadotropin (18) and hepatitis B virus(19). The formula for the QCM mass-frequency effect was first reported by Sauerbrey (20) and later extensively investigated by others (21)(22):

Formula
where {Delta}F is the measured frequency shift (FS) (Hz) of the coated crystal, F is the basic resonance frequency (MHz) of the crystal, A is the area coated, and {Delta}m is the mass deposited.

We constructed and evaluated a rapid, multiplex, label-free immunosensor array for simultaneous and accurate quantification of 4 urinary proteins.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results and Discussion
 References
 
samples
All urine samples were collected from the Southwest Hospital (Chongqing, China) between 2004 and 2005. There were 124 urine samples in this study; 48 were from patients with continuous microproteinuria or macroproteinuria and 76 from healthy individuals with normal urinary protein concentrations. All participants gave signed informed consent and the study was approved by the Ethics Boards of the Third Military Medical University. For all study participants, 24-h urine specimens were collected after the bladder was emptied. Each 24-h collected sample was treated with a protease inhibitor cocktail (set III, Calbiochem) and then, to minimize possible proteolytic degradation of B2M, the pH was adjusted to 7.0 with HCl or NaOH. We then extracted 4-mL aliquots from each collected sample and stored these at –70 °C, with no freeze-thaw cycles before the protein assay. Each aliquot was further divided into 2 equal portions before detection. One portion was analyzed with INM (reference method) and the other with QCM immunosensor (proposed method).

reagents
Staphylococcal protein A (SPA) and monoclonal antibodies against MA, A1M, and B2M were purchased from Sigma-Aldrich. Anti-IgG was purchased from Beckman Coulter. A1M and B2M were obtained from the RIA test reagent set from the Research Institute of Atomic Energy. MA was prepared by diluting the MA powder (Sigma-Aldrich) with doubly distilled water. We prepared phosphate-buffered saline (PBS; 6.4 mmol/L Na2HPO4, 0.9 mmol/L KH2PO4, 137 mmol/L NaCl, 2.7 mmol/L KCl, pH 7.2), and used a blank urine solution containing none of the 4 proteins as negative control. High concentrations of urine quality control reagent (IMMAGE, Beckman) were used as the positive control. Bovine serum albumin (Sigma-Aldrich) was used as the blocking reagent. Other chemicals used were of analytical reagent grade, and doubly distilled water was used throughout.

equipments and apparatus
The piezoelectric crystals (the 26th Research Institute, Chinese Electronic Scientific and Technical Group Company) were 13 x 13 mm2 squares, 0.15 mm thick, AT-cut 10-MHz quartz crystals covered with gold electrodes (4 mm in diameter). The detector (PESA-3000; Fig. 1A ) integrated with logic circuit, thermal controller system, a voltage stabilizer, and 2 x 5 detection wells (selfdesigned), was made by the Jialing Group of China. A personal computer with the self-developed software BIDE (version 2.11) was connected with the detector to monitor the FS and calculate the protein concentrations of the samples. INM was performed with IMMAGE (Beckman) according to the manufacturer’s protocols.


Figure 1
View larger version (90K):
[in this window]
[in a new window]

  		Figure 1. The 2 x 5 immunosensor array and single detection well.

(A), picture of immunosensor array detector. The 2 x 5 detection wells and thermal controller system are demonstrated as follows: (a), 2 x 5 model detection well array. (b), thermal controller system. (B), picture of 6 well-assembled detection wells

assembling 2 x 5 detection wells
The appearance of the 2 x 5 detection wells is shown in Fig. 1BUp . The detection wells were aligned in 2 rows of 5 wells each. All 10 detection wells were designed with the same structure, and each of them was driven by an independent logic circuit so that they worked independently, without mutual interference. The 4 urinary proteins were aligned on the left 4 wells of each row. At the same time, negative and positive controls were assayed in the other 2 wells individually to avoid false-negative and false-positive results.

immobilization of antibodies
First, the quartz crystals were cleaned mechanically with an abrasive cleaner. After rinsing, the crystals were soaked in 1 mol/L NaOH and 1 mol/L HCl for 10 min each. Then the crystals were washed with distilled water and 95% ethanol and dried with a stream of nitrogen gas. Subsequently, 10 µL of SPA solution (5 g/L) in PBS (pH 7.2) was dispersed uniformly on the gold. After overnight incubation, the crystals were rinsed 3 times with distilled water to remove excessive SPA. Then, 5 µL of antibody solution was added, and the crystals were incubated with bovine serum albumin (pH = 7.2) for 1 h at 25 °C to block unoccupied residues.

analysis procedures
Detection wells were assembled with 10 disposable crystals, and the system temperature was set at 37 °C before 90 µL of 0.05 mol/L PBS (pH 7.2) was dripped into each reaction well. The resonance frequency of the crystal was recorded until a steady baseline was obtained (F0) in 4 min during incubation. After 10 µL of protein solution was added, the resonance frequency dropped down gradually and tended to become stable in 10 min, and this stable frequency was read as F1. The whole detection procedure was monitored, and the real-time frequency fluctuation was recorded and displayed via the software. FS (FS = F1 – F0) of each protein was detected, and the final FSs ({Delta}F) {Delta}F = FS1 – FSn were calculated by reading the difference between FSs (sample) and FSn (negative control). The whole detection procedure completed automatically, and the final results were given out by the software in 15 min (Fig. 2B ).


Figure 2
View larger version (26K):
[in this window]
[in a new window]

  		Figure 2. Graph of FSs with various immobilized antibody concentrations.

A high concentration (10 g/L) of each protein was added to determine the optimal immobilized concentrations. Solid line and square, detected FSs induced by series of immobilized MA concentrations; dashed line and circle, detected FSs induced by series of A1M concentration; dotted line and upward-pointing triangle, detected FSs induced by series of B2M concentration; dashed line and downward-pointing triangle, detected FSs induced by series of IgG concentration.

materials methodology comparison and clinical sample detection
First, we generated the calibration curves by plotting FS against various protein concentrations, and the linearities were determined by serially diluting the stock protein solutions. After the analysis procedures, the operational characteristics, including imprecision, accuracy, specificity, analytical sensitivity, and detection limit, were evaluated. Then, the clinical samples were quantified individually with the QCM immunosensor array and INM (Beckman). Results were analyzed with ROC curves for the sensitivity, specificity, and ROC area under the curve (AUC). Bland–Altman analyses were performed to evaluate the residuals plots of the proposed method and reference method. The calibration curves were visualized with Origin (version 7.5). Analyses of the differences between the control group and patient group were carried out by SPSS, version 13. ROC analyses and Bland–Altman analyses were performed with MedCalc, version 8.1 (http://www.medcalc.be).


   Results and Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results and Discussion
References
 
optimization of immobilized antibody concentrations
Methods for immobilizing proteins on gold film, which have been investigated for years, include the SPA method (15)(23), the avidin and biotin method(24)(25), and the self-assembled monolayers method (26)(27). The SPA method was adopted in the present experiment because it can create more reproducible and stable immobilization layers with less added mass for IgG coating (28). Surface density of immobilized antibody is also an important factor for accurate quantification because a low antibody surface density leads to a lower upper limit of detection, whereas redundant antibodies will induce inhomogeneous orientation of the antibodies, which significantly increases steric hindrance on biomolecular interactions (29). Thus, the immobilized antibody concentration must be optimized to obtain high coating efficiency, wide dynamic range, and minimized detection cost.

The FS increases with antibody concentrations of 1– 4 g/L (B2M) or 5 g/L (MA, A1M, and IgG) are shown in Fig. 2Up . FS then reached a plateau and varied little, although the antibody concentration continued to increase. With antibody concentration increases ≥8 g/L, the FS appeared to decrease instead of increasing with the buildup of antibody. The exact reason for this observation is unclear, but taken together with data from a similar setting (29), our findings suggest that at low antibody concentrations the epitopes of the SPA are presented in such a way that the binding pockets of the antibodies can approach them without having to overcome any steric hindrance. With high antibody concentrations with dense packing of the antibodies, the binding pocket of the antibody is sterically repelled by neighboring antibodies when it approaches 1 epitope, a situation that leads to a decrease of FS toward areas of increasing antibody concentrations. Considering immobilization efficiency, detection range, and detection cost, we selected 4 mg/L of B2M and 5 mg/L of A1M, MA, and IgG as the optimal immobilization concentrations in later experiments.

calibration curves and linearity
Calibration curves showed direct proportionality up to 40, 30, 5, and 60 mg/L on MA, A1M, B2M, and IgG, individually and respectively (Fig. 3 ). The regression equations (see Table 1 in the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol52/issue12) had mean (SD) slopes of 75.3 (2.1) and y-intercepts of 63.4 (2.3) mg/L for MA, 76 (2.3) and 113.9 (2.4) mg/L for A1M, 116 (3.0) and 47.7 (7.6) mg/L for B2M, and 121 (2.0) and 80.8 (2.1) mg/L for IgG. The linearity of these curves is shown in Fig. 4A . Within the measuring interval (0.05–40 mg/L in MA, 0.05– 30 mg/L in A1M, 0.01–5 mg/L in B2M, and 0.05–60 mg/L in IgG), the deviations from theoretical values did not exceed 5%, demonstrating good correlation. For protein concentrations >60 mg/L and even up to 100 mg/L, however, the FS tendency increased slowly and deviated from linearity. Thus, these high concentration points should be excluded from the linear range. To obtain accurate results, urine sample concentrations higher than the upper limit of detection should be diluted (10-fold or more) to appropriate concentrations before detection.


Figure 3
View larger version (28K):
[in this window]
[in a new window]

  		Figure 3. Calibration curves of the 4 urinary proteins detected with the QCM biosensor array.

Each point on the calibration curve represents the mean value of 3 independent experiments. Only the first 6 points were taken into account when plotting the linear relationship of FS vs log concentrations. The black line and solid circle represent the concentration of each calibration sample. The gray line and empty circle are the negative control. (A–D), calibration curves of MA, A1M, B2M, and IgG, respectively.


Figure 4
View larger version (20K):
[in this window]
[in a new window]

  		Figure 4. Linearity of the QCM immunosensor array (A) and ROC curves for different urinary proteins in quantifying clinical samples (B).

(A), we assessed assay linearity by use of 3 serially diluted urine samples containing 30 mg/L of A1M, 5 mg/L of B2M, 40 mg/L of MA, and 60 mg/L of IgG individually. In the interval of 500-fold to original concentration, the deviations from theoretical values did not exceed 5%. (B), ROC curves of clinical samples. ROC curves of 4 urinary proteins were plotted from 48 patients and 76 control samples. The AUC of MA (black solid line) is 0.95 (95% confidence interval, 0.89–0.98) with an SE of 0.02. AUC of A1M (dashed line) is 0.99 (0.95–1.0) with an SE of 0.01. AUC of B2M (dotted line) is 0.96 (0.90–0.99) with an SE of 0.02. AUC of IgG (gray solid line) is 0.94 (0.88–0.97) with an SE of 0.03.

imprecision and accuracy
Duplicate tests were performed by detecting very low, low, medium, and high concentrations of quality control (IMMAGE, Beckman). For every sample, tests were repeated 20 times in 1 day for intraassay and repeated on 20 consecutive days in the same manner (mean of 3 duplicates per day) for interassay. The mean intraassay and interassay CVs were 4.35% and 5.22% (MA), 4.98% and 6.30% (A1M), 6.15% and 6.62% (B2M), and 4.81% and 6.49% (IgG) (see Table 2 in the online Data Supplement). The intraassay CV showed a very low SD (mean CV = 5.07%), and the interassay CV had a slightly higher SD (CV = 6.06%), but they were both acceptable for clinical immunoassays (CV <10%) (30).

For the recovery study, which was carried out to evaluate the accuracy of the QCM immunosensor array, all of the recovery rates were within the interval of 90%–110% (see Table 3 in the online Data Supplement). Moreover, the SDs of recovery, including low, medium, and high concentrations of each protein, were sufficiently low (<8) to validate the satisfactory accuracy of the immunosensor array.

specificities of the immunosensor array
Cross-reaction tests were carried out by adding 10 µL of each protein calibrator (10 mg/L) to all 5 detection wells. Results showed mean nonspecific FS values of ~12 Hz (interval, 10.2–15.8 Hz) (see Table 1 in the online Data Supplement), which were significantly lower than the FS induced by a specific antibody–antigen reaction (>220 Hz on 10 mg/L protein; P <0.001). These results indicate that this QCM system has relatively low cross-reactivity and acceptably high specificity.

analytical sensitivities and detection limits
The limit of detection (signal-to-noise ratio, >3) was measured on 20 consecutive aqueous dilutions of the negative control. The limit of quantification was determined by serially diluting stock solutions with PBS (pH = 7.2). The dilutions were measured with 5 replicates, and the last diluted concentration that had a CV <10% between replicates was considered the limit of quantification (10). For MA, A1M, B2M, and IgG, the minimum detectable concentrations were 4 µg/L, 2 µg/L, 3 µg/L, and 12 µg/L, respectively. In addition, the limits of quantification in a urine matrix were 8 µg/L (CV = 6%), 6 µg/L (CV = 8%), 4 µg/L (CV = 7%), and 27 µg/L (CV = 9%), respectively. These quantification limits of the immunosensor array are comparable to those of RIA (31).

reference intervals
The ROC curves for 48 patients and 76 control samples are shown in Fig. 4BUp . The upper reference limit for MA was defined as 18.6 mg/L (97.9% of sensitivity and 82.9% of specificity), and for A1M, B2M, and IgG, upper reference limits were 15.7 mg/L (97.9%, 93.4%), 0.23 mg/L (95.8%, 80.3%), and 16.3 mg/L (95.8%, 76.3%), respectively. Generally, ROC curves definite the upper reference limit of analytic methods based on the highest accuracy (minimal false-negative and false-positive results). This research aims to establish a fast screening method to reveal the early increase of urinary proteins, which requires a cutoff value with the highest sensitivity. In the current research, sensitivities of the 4 proteins were higher than 95%, indicating that the QCM immunosensor can be used as a screening method. Moreover, the sensitivity can be improved to 100% by slightly decreasing the reference value to be extremely careful to avoid missed diagnoses of early renal injury.

clinical sample detection and comparison
We used Bland–Altman difference plot analysis to evaluate the correlation of 2 diagnostic methods, QCM immunosensor array and INM, for clinical sample detection (Fig. 5 ). A Bland–Altman difference plot analysis for MA showed a mean (SD) difference (QCM minus INM) of 0.68 (1.7) mg/L, and the limits of agreement (d – 1.96 S to d + 1.96 S, –2.7 to 4.1 mg/L) were sufficiently narrow (see Table 4 in the online Data Supplement). Similarly, A1M, B2M, and IgG had narrow ranges of agreement limit, suggesting good consistency and clinical comparability between these 2 methods.


Figure 5
View larger version (29K):
[in this window]
[in a new window]

  		Figure 5. Bland–Altman difference plot for urinary protein results by QCM immunosensor array and INM.

(A), Bland–Altman difference plot comparing urinary MA concentrations obtained with QCM immunosensor array against INM (IMMAGE; Beckman). (B–D), Bland–Altman difference plot comparing urinary proteins concentrations of A1M, B2M, and IgG, respectively. The solid line represents the mean difference in quantitative measurement of urinary proteins between the methods, and the dashed lines are mean ± 1.96 SD.

According to the criteria above, significant differences existed between the control and patient groups. Protein concentrations in the patient group were significantly higher than those of the control group (P <0.01): mean (SD) 29.3 (5.0) mg/L for the patient group vs 13.4 (4.6 mg/L) for the control group for MA, 22.8 (3.8) mg/L vs 9.0 (3.5) mg/L for A1M, 0.47 (0.14) mg/L vs 0.16 (0.07) mg/L for B2M, and 31.3 (9.6) mg/L vs 11.2 (4.1) mg/L for IgG. Although samples from a larger population are needed to verify the accuracy of this method, the significant differences between these 2 groups validated the clinical application of this method for renal injury diagnosis. It should be mentioned, however, that the specificity of IgG was only 76.3%, and the specificities of MA and B2M were also <85% (Fig. 4BUp ). Thus, a positive diagnosis of diabetes nephropathy should to be made on the basis of proteins concentrations in combination with other physical signs.

Quantification of urinary B2M has traditionally been problematic because of its instability in acidic conditions, especially at pH <5.5. This situation is complicated by the considerable variation of urinary pH during a 24-h period. To avoid degradation of B2M, we added a protease inhibitor cocktail and adjusted the pH to 7 after urine collection. Storage at –70 °C also contributed to the conservation of B2M in the urine. Furthermore, for urinary protein detection analysis, 24-h urine collection extraction is more accurate than spot urine collection, which, for the latter, is susceptible to unstable urine pH and transient proteinuria (30).

In addition to high accuracy and sensitivity, the QCM immunosensor array also allows rapid detection and easy handling. Sample preparation without the need for centrifugation greatly shortens analysis time. The total analysis time, including detection well assembly, incubation, and real-time detection was within 15 min, which is much shorter than conventional methods such as RIA and INM. During QCM immunosensor analysis, assembling detection wells by mounting a new disposable crystal took only 1 min. Incubation took <4 min with a stable oscillation frequency in PBS, and then the real-time detection procedure could be completed in 10 min. The endpoint judgment time is based on our preliminary experiments. We had continuously recorded the frequencies of 30 individual samples for 20 min and found out that the FSs of all the samples become stable within 10 min (see Fig. 1Up in the online Data Supplement).

The label-free technique, the principal feature of the QCM biosensor array, greatly simplifies reagent preparation procedures and eliminates safety concerns associated with radioactive materials. A label-free method also avoids some essential interference induced by the fluorescence scanning procedure. In summary, the label-free strategy not only guarantees the accuracy of the test, but also shortens the detection time and decreases the detection cost. Given that QCM immunosensor analyses are fairly rapid and relatively easy to operate, this approach lends itself particularly well to the rapid development of an analytical method and the analyses of large numbers of samples in clinical laboratories. More importantly, this method could provide convenient monitoring of the status of kidney function, especially in patients with a history of diabetes mellitus.

The rate of analysis, however, was essentially limited by instrumentation—only 2 samples were analyzed each time. Nevertheless, 4 urinary proteins could be simultaneously quantified within 15 min. The low cross-reactivity indicates that each detection well works independently, making it feasible to increase the rate of analysis to hundreds per hour by increasing the array format. At the same time, we are continuing our investigations of high integration and high throughput to explore wider applications. Given the ability to increase sample throughput, QCM immunosensor arrays will find increasing clinical applications.

In conclusion, the QCM immunosensor described in this study allows for simultaneous quantification of 4 nephropathy-related urinary proteins (MA, A1M, B2M, and IgG) with a method that is accurate, fast, label free, and easy to perform.


View this table:
[in this window]
[in a new window]

  		Table 1. Frequency shifts of cross-reactions on protein immunosensor array.


   Acknowledgments
 
This study was supported, in part, by grants from the Chinese National Natural Science Foundation (3040010 and 30270388), the Major Project Chinese National Programs for High Technology Research and Development (863 Program, 2002AARZ2023) and the International Cooperation Program for Science and Technology Development (2004DFA00600).


   Footnotes
 
1 Nonstandard abbreviations: A1M, a1-microglobulin; B2M, ß2-microglobulin; MA, microalbumin; INM, immunonephelometry; QCM, quartz crystal microbalance; FS, frequency shift; SPA, staphylococcal protein A; PBS, phosphate-buffered saline; AUC, area under the curve.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results and Discussion
 References
 

  1. Bloomgarden ZT. Diabetic nephropathy. Diabetes Care 2005;28:745-751.[Free Full Text]
  2. Stehouwer CDA, Gall M-A, Twisk JWR, Knudsen E, Emeis JJ, Parving H-H. Increased urinary albumin excretion, endothelial dysfunction, and chronic low-grade inflammation in type 2 diabetes: progressive, interrelated, and independently associated with risk of death. Diabetes 2002;51:1157-1165.[Abstract/Free Full Text]
  3. Corso A, Zappasodi P, Pascutto C, Bosoni T, Mangiacavalli S, Lorenzi A, et al. Urinary proteins in multiple myeloma: correlation with clinical parameters and diagnostic implications. Ann Hematol 2003;82:487-491.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  4. Pegoraro A, Singh A, Bakir AA, Arruda JA, Dunea G. Simplified screening for microalbuminuria. Ann Intern Med 1997;127:817-819.[Abstract/Free Full Text]
  5. Viswanathan V, Nair MB, Suresh S, Chamukuttan S, Ambady R. An inexpensive method to diagnose incipient diabetic nephropathy in developing countries. Diabetes Care 2005;28:1259-1260.[Free Full Text]
  6. Mogensen CE. Microalbuminuria predicts clinical proteinuria and early mortality in maturity-onset diabetes. N Engl J Med 1984;310:356-360.[Abstract]
  7. Bakoush O, Grubb A, Rippe B, Tencer J. Urine excretion of protein HC in proteinuric glomerular diseases correlates to urine IgG but not to albuminuria. Kidney Int 2001;60:1904-1909.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  8. Jung K, Pergande M, Graubaum HJ, Fels LM, Endl U, Stolte H. Urinary proteins and enzymes as early indicators of renal dysfunction in chronic exposure to cadmium. Clin Chem 1993;39:757-765.[Abstract/Free Full Text]
  9. Bazzi C, D’Amico G. The urinary excretion of IgG and alpha1-microglobulin predicts renal outcome and identifies patients deserving treatment in membranous nephropathy. Kidney Int 2002;61:2276.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  10. Choi S, Choi EY, Kim HS, Oh SW. On-site quantification of human urinary albumin by a fluorescence immunoassay. Clin Chem 2004;50:1052-1055.[Free Full Text]
  11. Lakowicz JR, Malicka J, Matveeva E, Gryczynski I, Gryczynski Z. Plasmonic technology: novel approach to ultrasensitive immunoassays. Clin Chem 2005;51:1914-1922.[Abstract/Free Full Text]
  12. Maachi M, Fellahi S, Regeniter A, Diop M-E, Capeau J, Rossert J, et al. Patterns of proteinuria: urinary sodium dodecyl sulfate electrophoresis versus immunonephelometric protein marker measurement followed by interpretation with the knowledge-based system MDI-LabLink. Clin Chem 2004;50:1834-1837.[Free Full Text]
  13. Sviridov D, Meilinger B, Drake SK, Hoehn GT, Hortin GL. Coelution of other proteins with albumin during size-exclusion HPLC: implications for analysis of urinary albumin. Clin Chem 2006;52:389-397.[Abstract/Free Full Text]
  14. McElderry LA, Tarbit IF, Cassells-Smith AJ. Six methods for urinary protein compared. Clin Chem 1982;28:356-360.[Abstract/Free Full Text]
  15. Chou S-F, Hsu W-L, Hwang J-M, Chen C-Y. Determination of {alpha}-fetoprotein in human serum by a quartz crystal microbalance-based immunosensor. Clin Chem 2002;48:913-918.[Abstract/Free Full Text]
  16. Godber B, Thompson KSJ, Rehak M, Uludag Y, Kelling S, Sleptsov A, et al. Direct quantification of analyte concentration by resonant acoustic profiling. Clin Chem 2005;51:1962-1972.[Abstract/Free Full Text]
  17. He F, Zhang L. Rapid diagnosis of M. tuberculosis using a piezoelectric immunosensor. Anal Sci 2002;18:397-401.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  18. Zhang B, Mao Q, Zhang X, Jiang T, Chen M, Yu F, et al. A novel piezoelectric quartz micro-array immunosensor based on self-assembled monolayer for determination of human chorionic gonadotropin. Biosens Bioelectron 2004;19:711-720.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  19. Chen M, Liu M, Yu L, Cai G, Chen Q, Wu R, et al. Construction of a novel peptide nucleic acid piezoelectric gene sensor microarray detection system. J Nanosci Nanotechnol 2005;5:1266-1272.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  20. Sauerbrey GZ. Use of quartz crystal vibrator for weighting thin films on a microbalance. Z Phys 1959;155:206-222.[CrossRef]
  21. Kanazawa KK, Gordon JG. Frequency of a quartz microbalance in contact with liquid. Anal Chem 1985;57:1770-1771.[CrossRef]
  22. Theisen LA, Martin SJ, Hillman AR. A model for the quartz crystal microbalance frequency response to wetting characteristics of corrugated surfaces. Anal Chem 2004;76:796-804.[Medline] [Order article via Infotrieve]
  23. Wang H, Liu Y, Yang Y, Deng T, Shen G, Yu R. A protein A-based orientation-controlled immobilization strategy for antibodies using nanometer-sized gold particles and plasma-polymerized film. Anal Biochem 2004;324:219-226.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  24. Ouerghi O, Touhami A, Jaffrezic-Renault N, Martelet C, Ouada HB, Cosnier S. Impedimetric immunosensor using avidin-biotin for antibody immobilization. Bioelectrochemistry 2002;56:131-133.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  25. Lue RY, Chen GY, Hu Y, Zhu Q, Yao SQ. Versatile protein biotinylation strategies for potential high-throughput proteomics. J Am Chem Soc 2004;126:1055-1062.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  26. Bronson RT, Michaelis DJ, Lamb RD, Husseini GA, Farnsworth PB, Linford MR, et al. Efficient immobilization of a cadmium chemosensor in a thin film: generation of a cadmium sensor prototype. Org Lett 2005;7:1105-1108.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  27. Zhou XC, Huang LQ, Li SF. Microgravimetric DNA sensor based on quartz crystal microbalance: comparison of oligonucleotide immobilization methods and the application in genetic diagnosis. Biosens Bioelectron 2001;16:85-95.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  28. Babacan S, Pivarnik P, Letcher S, Rand AG. Evaluation of antibody immobilization methods for piezoelectric biosensor application. Biosens Bioelectron 2000;15:615-621.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  29. Duschl C, Sevin-Landais AF, Vogel H. Surface engineering: optimization of antigen presentation in self-assembled monolayers. Biophys J 1996;70:1985-1995.
  30. European Urinalysis Guidelines: Summary. Scand J Clin Lab Invest Suppl 2000;231:1-86.[CrossRef][Medline] [Order article via Infotrieve]
  31. Ekins RP. New perspectives in radioimmunoassay. Nucl Med Commun 1993;14:721-735.[Web of Science][Medline] [Order article via Infotrieve]



The following articles in journals at HighWire Press have cited this article:


Home page
 Clin. Chem.Home page
A. A. Ellington, I. J. Kullo, K. R. Bailey, and G. G. Klee
Antibody-Based Protein Multiplex Platforms: Technical and Operational Challenges
Clin. Chem., February 1, 2010; 56(2): 186 - 193.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
clinchem.2006.073569v1
52/12/2273    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Luo, Y.
Right arrow Articles by Fu, W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Luo, Y.
Right arrow Articles by Fu, W.
Related Collections
Right arrow Automation and Analytical Techniques

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS