| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Biology and Genetics |
Departments of Pathology [A. S., S. V., T. R. B., A. M. C.], Urology [A. M. C.], Radiation Oncology [M. K. N., T. S. L.], and Biostatistics [D. G.], Comprehensive Cancer Center [T. S. L., A. M. C.], University of Michigan Medical School, Ann Arbor, Michigan 48109
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, and uncoupling protein 2, among others. Consistent with this observation, an increased percentage of apoptosis was observed in irradiated LoVo cells. Interestingly, we also observed radiation-induced down-regulation of carcinoembryonic antigen, a prototypic cancer biomarker. Selected proteins assessed by microarray were validated by traditional immunoblotting. Taken together, our work suggests that protein/antibody microarrays will facilitate high-throughput proteomic studies of human cancer and carcinogenesis. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Current methods for detecting protein-protein interactions include the yeast two-hybrid system, coimmunoprecipitation and phage display (6) . Similarly, detection and quantitation of proteins has relied on basic techniques such as one/two-dimensional gel electrophoresis, immunoblotting, ELISA, and RIA. However, a large proportion of proteins expressed by a cell cannot be detected using these methodologies (6) . The dynamic range of these techniques makes it difficult to visualize minor changes in protein expression. Additionally, most of these well-established methods for protein detection and quantitation are not amenable to high-throughput applications and have thus far not been effectively applied to a biochip format. Hence, it becomes essential to formulate a technique analogous to DNA microarrays for high throughput analysis of expressed proteins.
One way to detect the expression of proteins in a complex mixture is to use immobilized antibodies as probes. The feasibility of developing such an array-based platform to study protein expression has been indicated by several reports. These include the use of membrane-based arrays to screen for DNA, RNA, or protein-binding targets in bacterial expression libraries (7) and filter-based ELISAs to characterize single-chain antibody libraries (8) . Array formats have also been used to monitor protein-protein interactions (9 , 10) , quantitate auto-antibodies (11) , detect enzyme targets, and analyze interactions between proteins and small molecule drugs (12) . Recently, Haab et al. (13) used a protein/antigen microarray to study interactions between 115 antigen-antibody pairs in the background of nonspecific proteins. In the study presented here, we implement an analogous protein microarray system to monitor endogenous protein levels in cancer cells. More specifically, we used antibody microarrays to monitor proteomic profiles of LoVo colon carcinoma cells exposed to ionizing radiation. This approach yielded both known radiation-regulated proteins such as p53 and DR5,3 and several novel associations such as DFF40/CAD and CEA.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1 million cells were treated with 4 Gy radiation and subsequently cultured. The radiation was administered at 1–2 Gy/min at room temperature using an Atomic Energy of Canada Limited Theratron (60Co) calibrated with an ionization chamber directly traceable to National Institute for Standard and Technology. At the end of postirradiation culture, the cells were washed and scrapped into PBS containing EDTA-free protease inhibitor mixture (Roche Diagnostic Corporation, Indianapolis, IN). These were then used for protein extraction.
Protein Microarrays.
Antibodies (146) against proteins involved in stress response, cell cycle progression, and apoptosis were used to create antibody arrays. Antibodies were either provided by Stressgen Biotechnology, British Columbia, Canada or purchased from Neomarkers, Fremont, CA (See Supplementary Table 1). The antibodies were diluted in PBS and spotted on poly-L-lysine coated or superaldehyde-modified glass slides (Telechem International Inc., Sunnyvale, CA) at 100 µg/ml concentration in a volume of 1 nl using a high precision robotic arrayer (Gene Machines, San Carlos, CA). In arrays used to profile cell lysates each antibody was replicated 
10 times on the chip. Similarly 100 µg/ml of antibodies to luciferase (Chemicon International Inc., Temecula, CA), transferrin (ICN Biomedicals, Costa Mesa, CA), thrombin (Sigma Chemical Co., St. Louis, MO), guinea pig IgG (Sigma Chemical Co.), MIP2, and MIP1 (R & D Systems, Minneapolis, MN) were used for validation of the antibody chip. The antigen chip consisted of transferrin (Sigma Chemical Co.), insulin (Sigma Chemical Co.), recombinant human TRAIL (14)
, TWEAK or rabbit, goat, guinea pig, and human immunoglobulins (Sigma Chemical Co.) each at 100 µg/ml. In experiments carried out to detect the sensitivity of the assay, rabbit IgG (Sigma Chemical Co.) was used at concentrations ranging from 3 µg/ml to 100 µg/ml The humidity was maintained between 40 to 45% during all of the print runs. The printed slides were either used immediately or stored at -20°C until later use.
Protein Extraction and Delipidation.
The cells were lysed in PBS containing 1% NP40 (Sigma Chemical Co.) and protease inhibitors for 15 min at 4°C. Cellular DNA was sheared by brief sonication, and the solubilized proteins were delipidated by using either Triton-X114 (Sigma Chemical Co.; Ref. 15
) or ExtriGel beads (Pierce, Rockford, IL) according to manufacturer’s instructions. For detergent extraction, the cell lysates were treated with 1% Triton-X114, and phase separation was allowed to occur at 37°C for 5 min (15)
. The aqueous phase was separated by centrifugation and recovered. Alternatively, delipidation was carried out using ExtriGel column. The flow-through, which contains the delipidated cell lysates, was collected and subsequently concentrated using a Biomax Ultra Filter (Millipore Co., Bedford, MA). Protein estimation was carried out according to the Bradford method (Bio-Rad, Hercules, CA; Ref. 16
). The delipidated cell lysates were then used for labeling and incubated with protein microarrays.
Labeling of Purified Proteins or Cell Lysates with Fluorescent Dyes.
In experiments carried out to test the sensitivity of an antigen chip, antirabbit IgG (Sigma Chemical Co.) was used at 1 mg/ml concentration. Experiments to test the specificity of antigen chip were carried out by labeling antitransferrin, antiguinea pig IgG, and anti-insulin, each at 100 µg/ml, or antibodies to IgGs (Sigma Chemical Co.) from goat (120 µg/ml), rabbit (72 µg/ml), or guinea pig (135 µg/ml) in the background of 1% BSA. Validation of specificity of an antibody chip was done using purified antigens labeled at 1 µg/ml concentration in presence of 1% BSA. The dynamic range of the antibody chip was measured in the background of 1% BSA using 5–5000 ng/ml of purified transferrin. Delipidated cell lysates were also labeled at a concentration of 100 µg/ml. All of the labelings were done in freshly prepared 0.1 M of sodium bicarbonate buffer (pH adjusted to 9.3 with IN NAOH). We used a dual fluorescent-labeling assay analogous to the one used with DNA microarrays (17)
. The reference samples and experimentals were labeled separately using either Cy5 or Cy3 dyes. (Amersham Pharmacia Biotech Inc., Piscataway, NJ). The labeling reaction was allowed to proceed for 1 h at room temperature in the dark. The reaction was quenched by addition of hydroxylamine (Sigma Chemical Co.) to a final concentration of 1 M. The labeled proteins were pooled and separated from the free dye using a Sephadex G-25 (Amersham Pharmacia Biotech Inc.) gel filtration column (10-ml bed volume) pre-equilibrated in PBS containing 0.1% BSA. The labeled proteins, which migrated ahead of the unbound dye were monitored visually and collected. These were then passed through a 20-µ filter (Millipore Co.) and reduced to a final volume of 40 µl using a Biomax 5K concentrator (Millipore Co.). Tween 20 (Sigma Chemical Co.) was added to the probe at a final concentration of 0.01%, and the probe mixture was incubated with the spotted proteins on the glass slide microarrays (see below).
Detection of Protein Interactions on Microarrays.
Superaldehyde slides containing spotted proteins (stored at -20°C) were thawed and blocked in PBS containing 1% BSA (Sigma Chemical Co.) in 0.5% Tween 20 for 1 h at room temperature. The slides were then incubated with labeled protein mixture in a humidified chamber for 1 h at room temperature. They were washed three times with high salt PBS-T (PBS containing 1 M of sodium chloride and 0.1% Tween 20) and subsequently with low salt PBS (PBS containing 10 mM of sodium chloride), each for 10 min. The slides were dried by centrifugation at 500 x g for 5 min and analyzed using a microarray scanner (Axon Instruments Inc., Foster City, CA). The ratios of Cy5 to Cy3 were calculated for each of the spots using the manufacturer’s software package (Genepix 3.05).
Protein Microarray Data Analysis.
Primary analysis was done using the Genepix software package. Images of scanned microarrays were gridded and linked to a protein print list. Initially, data were viewed as a scatter plot of Cy3 versus Cy5 intensities. Cy3:Cy5 ratios are determined for the individual proteins along with various other quality control parameters (e.g., intensity over local background). The Genepix software analysis package flags spots as absent based on spot characteristics.4
Furthermore, bad spots or areas of the array with obvious defects were manually flagged. These criteria included spots exhibiting any of the following: signal only in one channel, nonuniform distribution of Cy5 or Cy3 signal intensities, regions of high background, significant deviations from the average intensity ratio, or small diameters (<50 µ). Flagged spots were not included in subsequent analyses. Data were scaled such that the average median ratio value for all of the spots was normalized to 1 (done separately for each array), with the premise that the average spot on the chip would represent unchanged protein expression.
Statistical Analysis of Differential Protein Expression.
In this analysis, a data-based criterion was determined, above or below which proteins were found to be differentially expressed. To determine such a cutoff level, a hierarchical model approach was used. Let Yij be the jth expression ratio on the ith protein, i = 1,... ,146; j = 1,... ,ni. We assumed the following statistical model:
 |
2 and 
2 were assumed to be known. Using such a model, the variance of the posterior distribution of µi for given protein expression measurements, d, can be shown to be:
 |
1 - a
, 1 + a
, where a is a constant greater than zero, was used to identify a region that represents expression that is not significantly different from 1. A useful aspect of this approach is that information from different proteins is pooled in a sensible way. A conservative "cutoff" rule would be to take a = 1.5, which corresponds to 86% confidence value. Using such a criterion, proteins having their mean expression levels outside this interval could be considered differentially expressed.
The parameters in the model were estimated using a two-stage approach. The µi and 2 in the first level of the model were estimated using ANOVA. Then, 2 was computed as the variance of the estimated mean expression ratios for the proteins and weighted by the number of measurements. The estimates of 2 and 2 were combined to give an estimate of d, which was calculated to be 0.07. Thus NMRAT values outside the interval of 0.74–1.26 would be considered differentially expressed with 68% statistical confidence, whereas those outside the interval of 0.6–1.4 would be considered differentially expressed with 86% statistical confidence.
Immunoblot Analysis.
Proteins from LoVo cell lysates were separated on SDS-PAGE. Separated proteins were transferred onto polyvinylidene difluoride or nitrocellulose membranes and blocked for 2 h in PBS containing 5% nonfat milk and 0.1% Tween 20. The membranes were then incubated overnight at 4°C with antibodies directed against BAX, bcl-2, Calnexin, ERK1/2, FADD, FLIP, DR5, STAT1, DFF40/CAD, DFF45/ICAD (Stressgen Biotechnology, British Columbia, Canada), TF, CEA (Neomarkers and Oncogene Research Products, San Diego, CA), UCP2, p53 (Calbiochem, La Jolla, CA), actin (Santa Cruz Biotechnology, Santa Cruz, CA), and antiphosphotyrosine (Upstate Biotechnology, Lake Placid, NY). Membranes were washed with PBS-T and incubated with horseradish peroxidase-conjugated secondary antibody for 2 h at room temperature. Detection was achieved using the enhanced chemiluminescence system following the manufacturer’s instructions (Amersham Pharmacia Biotech).
Measurement of Percentage of Apoptosis.
LoVo cells (control or irradiated with either 4 or 8 Gy radiation; n = 3) were harvested after 4 h in culture and fixed in 70% ethanol for 30 min. Cells were stained with 1% propidium iodide in PBS and were analyzed under fluorescent microscope. Cells with completely fragmented nuclei were considered apoptotic (18)
. Cells (200) were counted in random for each treatment, and the percentage of apoptotic cells (percentage ± SE) was plotted for each treatment.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The array was sensitive enough to detect 6.25 pg of spotted antigen (Fig. 1, A and B)
. A linear increase in spot intensity from 4 ± 1.8 to 3126.4 ± 157 (arbitrary intensity units) was observed with increasing amounts of the antigen (R2 = 0.99; Fig. 1B
). Importantly, a mixture of Cy3-labeled antiguinea pig IgG and Cy5-labeled antitransferrin detected only their cognate antigens on a microarray comprised of purified transferrin, guinea pig IgG, and TRAIL (Fig. 1C)
. Recombinant human TRAIL (14)
, which was used as a control, did not react with the probe mixture (Fig. 1C)
demonstrating the specificity of the assay. Furthermore, to emulate a cell lysate or serum milieu our probe labelings were carried out in a background of excess BSA (1% BSA). Similarly, an antigen microarray consisting of transferrin, guinea pig IgG, insulin, and recombinant TWEAK (which served as a nonspecific control) was probed with a 1% BSA mixture containing antitransferrin Ab labeled with Cy5, antiguinea pig IgG labeled with Cy3, and anti-insulin Ab labeled separately with both the fluorescent dyes (Fig. 1D)
. As predicted, spots corresponding to transferrin and guinea pig IgG fluoresced at their respective wavelengths with little or no cross-reactivity (Fig. 1D)
. Spots corresponding to insulin fluoresced equally in both channels and are represented as yellow spots (Fig. 1D)
. Negligible immunoreactivity was observed with TWEAK, which served as a nonspecific control protein (Fig. 1D)
. To increase the scale of our antigen chip, we spotted multiple, different antigens in replicate onto glass slides (Fig. 1E)
. Arrays containing 96 spotted elements corresponding to immunoglobulins of goat, rabbit, guinea pig, and human origin were probed with antibodies directed against guinea pig IgG (Fig. 1E
, top panel), rabbit IgG (Fig. 1E
, middle panel), and goat IgG (Fig. 1E
, bottom panel) labeled separately with Cy3 dye. Each of these antibodies distinctly picked up their cognate antigens in the background of BSA.
|
. Antibodies to thrombin, luciferase, MIP2, and MIP1
served as negative controls and were not immunoreactive (Fig. 2A)
.
|
). Using this "two-color approach," variability attributable to spot size and amount of antibody deposited is minimized, making it more valid to make comparisons across microarrays.
Profiling of Endogenous Protein Levels in LoVo Cells Treated with Ionizing Radiation Using Microarrays.
One of the primary goals of this study was to monitor the levels of endogenous proteins in neoplastic cells. We used the protocol depicted in Fig. 3
to interrogate protein expression in LoVo colon carcinoma cells treated with 4 Gy ionizing radiation and cultured for 4 h after irradiation. The 4-h time point for analysis of radiation-induced protein expression by microarray was chosen based on an earlier report showing maximum elevation of protein levels in human malignant myeloma cells in response to X-ray-induced stress (19)
. Experiments were done using 1920 element protein microarrays consisting of 146 antibodies spotted in replicate (Fig. 4A)
. Microarray analysis of untreated LoVo cells compared with itself showed considerably fewer differences in protein expression than compared with radiation-treated LoVo cells (Fig. 4B)
. A scatter plot of protein profiles of untreated LoVo cells compared against itself showed minimal divergence (Fig. 4C
, left) with an R2 of 0.83. This is remarkably different from the scatter plot of irradiated LoVo cells compared against untreated controls, which showed a higher degree of divergence (Fig. 4C
, right) with an R2 of 0.4. Thus, this suggests differential expression of proteins in LoVo cells after radiation treatment.
|
|
. Interestingly, among the differentially expressed proteins was a cohort of apoptotic regulators of which the level appeared to have increased on radiation treatment (Fig. 5)
. Data from five antibody microarrays comparing radiation-treated cells with untreated cells were used (see "Materials and Methods" for details). In this focused analysis, 1327 expression level measurements from 20 proteins were considered, and measurements were normalized to adjust for between-chip variability. To determine differential expression for each protein, a one-sample t test was performed. The null hypothesis was that the mean expression level for the jth protein is 1.0, j = 1,... ,20. No adjustments for multiple comparisons were made. In addition, the protein expression measurements from the chips were treated as independent observations. Normalized median of Cy5:Cy3 ratios was calculated for each protein and graphically represented (Fig. 5B)
. The expression patterns obtained from the graph were confirmed by immunoblot analysis for the selected proteins. The data from these two independent analyses were used to formulate an empirical interval for unchanged protein expression. Thus, proteins of which the levels were unchanged by radiation treatment had an NMRAT in the range of 0.85–1.15 (Fig. 5B
, gray shaded region), whereas the up-regulated proteins showed NMRAT values >1.15 and down-regulated proteins had NMRAT values <0.85 (Fig. 5B)
. Independent statistical analysis using a hierarchical model approach determined the "cutoff ratios" for unchanged protein expression to range between 0.74 and 1.26, corresponding to 1 SD. This was similar to the interval obtained from analysis of microarray data. Importantly, use of a more stringent interval of 0.6–1.4 corresponding to 1.5 SDs (see statistical analysis section of the "Materials and Methods") still resulted in DFF40/CAD, DFF45/ICAD, DR5, FLIP, STAT1, TF, UCP2, and CEA to be classified as differentially regulated. Furthermore, microarray results were also validated by performing traditional immunoblot analysis on selected proteins of interest (Fig. 5C)
.
|
). Likewise, microarrays revealed increased tyrosine phosphorylation of cellular proteins as reported previously (24)
. We also identified up-regulation of several novel radiation-responsive proteins including DFF40/CAD, DFF45/ICAD, UCP2, DcR2, and STAT1 (Fig. 5, B and C)
. Interestingly, levels of the cancer biomarker CEA were significantly decreased on radiation treatment. Levels of actin, caspase 3, FADD, BAK, Bcl2, BAX, ERK1/2, and Calnexin were found to be unchanged by both microarray and immunoblot analysis (as were most other proteins; Fig. 5, B and C
). The specificity of the microarray assay was additionally supported by a fluorescent dye reversal experiment wherein the irradiated sample was labeled with Cy3 instead of Cy5 (Fig. 5C)
. Similar profiles of differentially expressed proteins were obtained irrespective of the dye used to label the lysates.
Analysis of Levels of CEA on Irradiation of LoVo Cells.
Having observed radiation-induced down-regulation of CEA by protein microarrays (Fig. 5C)
, we investigated the expression of this protein in more detail. We confirmed the down-regulation of CEA in radiation-treated LoVo cells by performing immunoblot analysis with two distinct antibodies directed against the CEA protein, namely CEA-Ab1, raised against native CEA protein purified from human colonic adenocarcinoma and CEA/CD66e Ab-3 raised against human colon carcinoma extract. Both antibodies demonstrated a quantitative decrease in CEA levels 4 h after irradiation (Fig. 6, A and B)
. This down-regulation of CEA was seen as early as 2 h and with higher levels of radiation (8 Gy; Fig. 6C
). Importantly, CEA levels began to recover 24 h after irradiation (Fig. 6C)
. Blots were stripped and reprobed with anticaspase3 antibody to confirm equal loading (Fig. 6C)
.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The protein chip described in this study is based on immobilization of "bait" proteins in an array format onto poly-L-lysine- or aldehyde-coated glass surfaces. The microarrays are then probed with a mixture of fluorescent-labeled antigens or cell lysates. Such a system offers the flexibility to spot peptides, protein fragments, and expression libraries, and is amenable to high-throughput screening of protein-protein, protein-RNA, and protein-DNA interactions. It can also be used to study translational or post-translational modifications such as phosphorylation, dephosphorylation, proteolytic cleavage, and dimerization, which are important signaling responses of a cell to external stimuli (6) . For example, we observed increased tyrosine phosphorylation of cellular proteins in radiation-treated LoVo cells using both microarray and Western blotting techniques. Our results with purified proteins demonstrated high sensitivity and specificity of the protein microarray system. The range of detection was 6.25–50 pg of spotted proteins. Thus, similar to the system described by Haab et al. (13) , the protein microarray described in this study has the potential to detect low amounts of marker proteins in samples of interest.
In addition to the technical advances made by this study, we also characterized a number of radiation-responsive proteins in LoVo cells, some of which may contribute to radiation-induced apoptosis and cell cycle arrest (Fig. 7)
. Earlier reports using two-dimensional gel electrophoresis have shown up-regulation of a group of radiation-responsive proteins termed X-ray-induced proteins, in human malignant myeloma cells (19)
. These X-ray-induced proteins were reported to have a molecular mass in the range of 126–275 kDa and are induced 3 h after irradiation with maximum expression levels at the 4 h time point (19)
. Hence, in this study we used 4-h postirradiated LoVo cells for microarray analysis of protein expression. Gene products demonstrated previously to be up-regulated by radiation were identified in this screen including p53 (20)
, FLIP (23)
, UCP2 (25)
, TRAIL (22)
, and DR5 (which is a receptor for TRAIL; Ref. 21
). Our study identified additional proteins dysregulated by radiation, which to our knowledge has not been reported previously, including DFF40/CAD, DFF45/ICAD, DcR2, STAT1, and CEA. Interestingly, among the proteins identified to be differentially regulated by protein microarray in response to radiation, DFF40/CAD, DFF45/ICAD, FLIP, DR5, STAT1, UCP2, and CEA had NMRAT values outside the 0.6–1.4 interval (1.5 SD). This range was designated independently by statistical analysis to represent the zone of unchanged protein expression with 86% confidence value.
|
In addition to detecting elevated levels of proapoptotic molecules such as p53, DFF40/CAD, DR5, and TRAIL, we also detected increased levels of FLIP, which has been shown to inhibit apoptosis by a mechanism that interrupts FADD-caspase 8 signaling (29) . Similar to FLIP, we also detected elevated levels of another antiapoptotic protein, DcR2. Ionizing radiation has been shown previously to up-regulate DcR3, which has been implicated in attenuating radiation-induced apoptosis of colon cancer cells by interfering with TRAIL signaling (30) . Taken together, up-regulation of apoptotic inhibitors FLIP and DcR2 may represent a protective mechanism induced by ionizing radiation to delay programmed cell death.
Apoptotic signals also affect the transcriptional program of the cell. One such transcriptional factor shown previously to be involved in radiation-induced apoptosis is STAT1 (31)
, which is phosphorylated at Ser727 in response to radiation (31)
. Interestingly, we detected elevated levels of STAT1 protein in irradiated LoVo cells. Additional studies using activation-specific antibodies will be required to confirm whether radiation-induced STAT1 activation in LoVo increases both protein levels and STAT phosphorylation.
Antibody microarrays also detected up-regulated levels of the mitochondrial uncoupling protein, UCP2, in radiation-treated cells. The mRNA transcript levels for this protein have been shown by DNA microarray analysis to be increased by radiation in a murine B lymphoma model (25) . Voehringer et al. (25) hypothesized a role for UCP2 in mitochondrial uncoupling and loss of membrane potential leading to release of mitochondrial apoptogenic factors. Also, Nicolo et al. (23) and Taneja et al. (32) have independently reported radiation-induced decreases in mitochondrial membrane potential as a major contributing factor in the activation of the cell death pathway by radiation. In this context, the detection of elevated levels of UCP2 protein provides additional evidence for its role in mediating radiation-induced loss of mitochondrial membrane potential (25) and consequent activation of the cell death machinery. Early studies from our laboratory indicate a concomitant decrease in membrane potential of irradiated LoVo cells with an increase in percentage of apoptosis.
In addition to detecting the radiation-induced up-regulation of various proteins by microarray, we also detected down-regulation in levels of CEA after radiation treatment. Immunoblot analysis showed that in response to both 4 Gy and 8 Gy radiation doses, CEA levels decreased within 4 h and begin to recover to control levels within 24 h. To our knowledge, such profound alterations in CEA levels after radiation-treatment have not been documented. Because CEA is a secreted biomarker used to monitor colon cancer recurrence in patients undergoing radiation therapy (33) , radiation-mediated CEA down-regulation may have important clinical implications.
Such high-throughput profiling of minute amounts of analyte would require the use of protein chips containing functionally active antibodies. Thus, the antibodies that have been tested for ELISA or immunoprecipitation would be considered ideal for such an application. Also, it is important to keep slides containing the spotted proteins in a humidified atmosphere to maintain the antibodies in their native and hydrated state. We have noticed that after 4 weeks of storage, signal:noise ratio of the chips decreased resulting in a slight decrease in sensitivity. This may also be attributable to possible dehydration and denaturation of spotted antibodies. Thus, the printed chips yield better hybridizations when used at the earliest time after printing. It is also important to note that the freeze thawing of the antibodies or antigens should be avoided to preserve the immunoreactivity of these molecules. Thus, antibodies, which lose their ability to bind the cognate antigens on hybridization, would appear as yellow spots. These spots would not be considered during the course of analysis. In addition, the nonspecific cross-reactivity of antibodies becomes less of an issue in a ratio-based detection system as ours. Also a high signal:noise ratio is essential for optimum monitoring of changes in protein expression. Some of the factors found to contribute to a high signal:noise ratio include labeling efficiency and removal of lipids. Optimum labeling is achieved under alkaline conditions using dyes freshly resuspended in DMSO. However, the extent of labeling might vary for different proteins. Finally, the abundance of antigens in the extract could also be a factor determining the effective monitoring of protein expression alterations in cells. Thus, changes in low abundance proteins might be detected to a lesser extent compared with their highly expressed counterparts.
In summary, our study describes the development of an array-based approach for high throughput analysis of proteomic profiles in cancer cells. A modified version of our system should be applicable to clinical tumor specimens and is currently under development. Using a protein microarray approach we identified several potential regulatory sites for radiation-induced apoptosis signaling in LoVo cells. Furthermore, we also report a remarkable radiation-induced down-regulation of the cancer biomarker CEA. Taken together, protein microarray technology should complement gene expression microarrays in the large-scale national and international efforts to molecularly profile human cancer.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported in part by a Pilot Grant from the University of Michigan Bioinformatics Program (to A. M. C.). 
2 These authors contributed equally.
3 To whom requests for reprints should be addressed, at Department of Pathology, The University of Michigan Medical School, 1301 Catherine Road, MSI Room 4237, Ann Arbor, Michigan 48109-0602. Phone: (734) 936-1887; Fax: (734) 763-6476; E-mail: arul{at}umich.edu
4 The abbreviations used are: DR, death receptor; FADD, Fas-associated death domain; TRAIL, tumor necrosis factor-related apoptosis inducing ligand; CEA, carcinoembryonic antigen; ERK, extracellular signal-regulated kinase; MIP, macrophage inflammatory protein; TRF, transferrin protein; FLIP, FLICE-like inhibitory protein; UCP, uncoupling protein; DcR, decoy receptor; DFF, DNA fragmentation factor; CAD, caspase-activated DNase; ICAD, inhibitor of caspase-activated DNase; NMRAT, normalized median of ratios; IgG, immunoglobulin; TWEAK, TNF-like weak inducer of apoptosis; MyD, myeloid differentiation antigen; TF, tissue factor; NHS, N-hydroxysuccinimidyl.
5 Internet address: www.microarrays.org.
Received 6/ 1/01. Accepted 8/16/01.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
uses a different signaling pathway. Proc. Natl. Acad. Sci. USA, 96: 13956-13961, 1999.This article has been cited by other articles:
|
|
 |
|
  X. C. Song, G. Fu, X. Yang, Z. Jiang, Y. Wang, and G. W. Zhou Protein Expression Profiling of Breast Cancer Cells by Dissociable Antibody Microarray (DAMA) Staining Mol. Cell. Proteomics, January 1, 2008; 7(1): 163 - 169. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Song, I.-m. Shih, R. Salani, D. W. Chan, and Z. Zhang Annexin XI Is Associated with Cisplatin Resistance and Related to Tumor Recurrence in Ovarian Cancer Patients Clin. Cancer Res., November 15, 2007; 13(22): 6842 - 6849. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Goldin-Lang, K. Pels, Q.-V. Tran, B. Szotowski, F. Wittchen, S. Antoniak, T. Willich, H. Witt, M. Hummel, D. Lenze, et al. Effect of ionizing radiation on cellular procoagulability and co-ordinated gene alterations Haematologica, August 1, 2007; 92(8): 1091 - 1098. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Pollack A Perspective on DNA Microarrays in Pathology Research and Practice Am. J. Pathol., August 1, 2007; 171(2): 375 - 385. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M.-H. Tsai, J. A. Cook, G. V.R. Chandramouli, W. DeGraff, H. Yan, S. Zhao, C. N. Coleman, J. B. Mitchell, and E. Y. Chuang Gene Expression Profiling of Breast, Prostate, and Glioma Cells following Single versus Fractionated Doses of Radiation Cancer Res., April 15, 2007; 67(8): 3845 - 3852. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Fan, K. M. Ahmed, M. C. Coleman, D. R. Spitz, and J. J. Li Nuclear Factor-{kappa}B and Manganese Superoxide Dismutase Mediate Adaptive Radioresistance in Low-Dose Irradiated Mouse Skin Epithelial Cells Cancer Res., April 1, 2007; 67(7): 3220 - 3228. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. M. Ahmed, S. Dong, M. Fan, and J. J. Li Nuclear Factor-{kappa}B p65 Inhibits Mitogen-Activated Protein Kinase Signaling Pathway in Radioresistant Breast Cancer Cells Mol. Cancer Res., December 1, 2006; 4(12): 945 - 955. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Ellmark, J. Ingvarsson, A. Carlsson, B. S. Lundin, C. Wingren, and C. A. K. Borrebaeck Identification of Protein Expression Signatures Associated with Helicobacter pylori Infection and Gastric Adenocarcinoma Using Recombinant Antibody Microarrays Mol. Cell. Proteomics, September 1, 2006; 5(9): 1638 - 1646. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Sanchez-Carbayo Antibody Arrays: Technical Considerations and Clinical Applications in Cancer Clin. Chem., September 1, 2006; 52(9): 1651 - 1659. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Smith, M. B. Watson, S. L. O'Kane, P. J. Drew, M. J. Lind, and L. Cawkwell The analysis of doxorubicin resistance in human breast cancer cells using antibody microarrays. Mol. Cancer Ther., August 1, 2006; 5(8): 2115 - 2120. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Nettikadan, K. Radke, J. Johnson, J. Xu, M. Lynch, C. Mosher, and E. Henderson Detection and Quantification of Protein Biomarkers from Fewer than 10 Cells Mol. Cell. Proteomics, May 1, 2006; 5(5): 895 - 901. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. B. Haab Using Array-Based Competitive and Noncompetitive Immunoassays Am. Assoc. Cancer Res. Educ. Book, April 1, 2006; 2006(1): 148 - 154. [Full Text] [PDF]
|
|
|
|
|
|
A. Gelbard, C. T. Garnett, S. I. Abrams, V. Patel, J. S. Gutkind, C. Palena, K.-Y. Tsang, J. Schlom, and J. W. Hodge Combination Chemotherapy and Radiation of Human Squamous Cell Carcinoma of the Head and Neck Augments CTL-Mediated Lysis. Clin. Cancer Res., March 15, 2006; 12(6): 1897 - 1905. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. B. Travis, C. S. Rabkin, L. M. Brown, J. M. Allan, B. P. Alter, C. B. Ambrosone, C. B. Begg, N. Caporaso, S. Chanock, A. DeMichele, et al. Cancer Survivorship--Genetic Susceptibility and Second Primary Cancers: Research Strategies and Recommendations J Natl Cancer Inst, January 4, 2006; 98(1): 15 - 25. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Orchekowski, D. Hamelinck, L. Li, E. Gliwa, M. VanBrocklin, J. A. Marrero, G. F. Vande Woude, Z. Feng, R. Brand, and B. B. Haab Antibody Microarray Profiling Reveals Individual and Combined Serum Proteins Associated with Pancreatic Cancer Cancer Res., December 1, 2005; 65(23): 11193 - 11202. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. W. Olle, A. Sreekumar, R. L. Warner, S. D. McClintock, A. M. Chinnaiyan, M. R. Bleavins, T. D. Anderson, and K. J. Johnson Development of an Internally Controlled Antibody Microarray Mol. Cell. Proteomics, November 1, 2005; 4(11): 1664 - 1672. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Hamelinck, H. Zhou, L. Li, C. Verweij, D. Dillon, Z. Feng, J. Costa, and B. B. Haab Optimized Normalization for Antibody Microarrays and Application to Serum-Protein Profiling Mol. Cell. Proteomics, June 1, 2005; 4(6): 773 - 784. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Yeretssian, M. Lecocq, G. Lebon, H. C. Hurst, and V. Sakanyan Competition on Nitrocellulose-immobilized Antibody Arrays: From Bacterial Protein Binding Assay to Protein Profiling in Breast Cancer Cells Mol. Cell. Proteomics, May 1, 2005; 4(5): 605 - 617. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. B. Haab Antibody Arrays in Cancer Research Mol. Cell. Proteomics, April 1, 2005; 4(4): 377 - 383. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Sheehan, V. S. Calvert, E. W. Kay, Y. Lu, D. Fishman, V. Espina, J. Aquino, R. Speer, R. Araujo, G. B. Mills, et al. Use of Reverse Phase Protein Microarrays and Reference Standard Development for Molecular Network Analysis of Metastatic Ovarian Carcinoma Mol. Cell. Proteomics, April 1, 2005; 4(4): 346 - 355. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Wang, Y.-C. Hu, S. Dong, M. Fan, D. Tamae, M. Ozeki, Q. Gao, D. Gius, and J. J. Li Co-activation of ERK, NF-{kappa}B, and GADD45{beta} in Response to Ionizing Radiation J. Biol. Chem., April 1, 2005; 280(13): 12593 - 12601. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N Dainiak, S K Schreyer, and J Albanese The search for mRNA biomarkers: global quantification of transcriptional and translational responses to ionising radiation Br. J. Radiol., January 1, 2005; Supplement_27(1): 114 - 122. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. T. Garnett, C. Palena, M. Chakarborty, K.-Y. Tsang, J. Schlom, and J. W. Hodge Sublethal Irradiation of Human Tumor Cells Modulates Phenotype Resulting in Enhanced Killing by Cytotoxic T Lymphocytes Cancer Res., November 1, 2004; 64(21): 7985 - 7994. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Horimoto, M. B. Resnick, T. A. Konkin, J. Routhier, J. R. Wands, and G. Baffy Expression of Uncoupling Protein-2 in Human Colon Cancer Clin. Cancer Res., September 15, 2004; 10(18): 6203 - 6207. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Sreekumar, B. Laxman, D. R. Rhodes, S. Bhagavathula, J. Harwood, D. Giacherio, D. Ghosh, M. G. Sanda, M. A. Rubin, and A. M. Chinnaiyan Humoral Immune Response to {alpha}-Methylacyl-CoA Racemase and Prostate Cancer J Natl Cancer Inst, June 2, 2004; 96(11): 834 - 843. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. B. Nielsen, M. H. Cardone, A. J. Sinskey, G. MacBeath, and P. K. Sorger Profiling receptor tyrosine kinase activation by using Ab microarrays PNAS, August 5, 2003; 100(16): 9330 - 9335. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. F. Petricoin and L. A. Liotta Clinical Applications of Proteomics J. Nutr., July 1, 2003; 133(7): 2476S - 2484. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. A. Giuliano High-Content Profiling of Drug-Drug Interactions: Cellular Targets Involved in the Modulation of Microtubule Drug Action by the Antifungal Ketoconazole J Biomol Screen, April 1, 2003; 8(2): 125 - 135. [Abstract] [PDF]
|
|
|
|
 |
|
 C. N. Otis, P. A. Krebs, A. Albuquerque, M. M. Quezado, X. S. Juan, M. E. Sobel, and M. J. Merino Loss of Heterozygosity of p53, BRCA1, VHL, and Estrogen Receptor Genes in Breast Carcinoma: Correlation with Related Protein Products and Morphologic Features International Journal of Surgical Pathology, October 1, 2002; 10(4): 237 - 245. [Abstract] [PDF]
|
|
|
|
 |
|
 M. A. Knepper Proteomics and the Kidney J. Am. Soc. Nephrol., May 1, 2002; 13(5): 1398 - 1408. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Rubin, M. Zhou, S. M. Dhanasekaran, S. Varambally, T. R. Barrette, M. G. Sanda, K. J. Pienta, D. Ghosh, and A. M. Chinnaiyan {alpha}-Methylacyl Coenzyme A Racemase as a Tissue Biomarker for Prostate Cancer JAMA, April 3, 2002; 287(13): 1662 - 1670. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
