SEARCH

SEARCH BY CITATION

Keywords:

  • IgY antibody;
  • cancer biomarker;
  • HER2;
  • telomerase;
  • quantum dots

Abstract

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

In an effort to improve affinity biomarker validation in fixed patient tissue specimens, we have developed a novel quantum dot-based bioimaging system that utilizes chicken IgY antibody for high sensitivity and specificity relative quantitation of cancer proteins. Monospecific, polyclonal IgYs were generated against human HER2 and telomerase, and analytically validated for specificity by western blot and immunohistochemistry on tumor and normal cells and for relative affinity by layered peptide array (LPA). IgYs bound desired targets in cell lines and fixed tissues and showed greater affinity than commercial mammalian antibodies for both HER2 and telomerase proteins. In tissue microarray experiments, HER2 quantitation with IgY antibody and quantum dot imaging correlated well with chromogenic in situ hybridization (CISH), whereas telomerase quantitation suggested a trend toward correlation with prostate cancer Gleason Grade and differentiation. Although patient numbers were small, these findings demonstrate the feasibility of relative quantitation of cancer biomarkers with IgY and quantum dot fluorophores, and show promise for rigorous clinical validation in large patient cohorts. © 2008 Wiley-Liss, Inc.

Few new early cancer biomarkers have surfaced in recent years.1 Among novel proteomic biomarkers for early cancer detection, some have proven controversial.2, 3 Analytical and clinical validation of cancer biomarkers has suffered from bias in the design, conduct and interpretation of such research,4 incompletely validated imaging,5 and lack of affinity standards6 and antibodies that did not, in fact, detect correct targets.7

Here, we describe a novel approach to cell-based bioimaging with relative quantitation for biomarker validation. We report characterization of two new IgY antibodies for quantitation of model cancer biomarker systems, HER2 and telomerase,8 and explore analytical improvements, including low cross-reactivity IgY-isotype chicken polyclonal antibodies raised against recombinant polypeptides; digital quantification of antibody signals with streptavidin-conjugated semiconductor nanocrystals to obviate photobleaching of organic fluorescent dyes; complete z-plane fluorescence image capture using 3D-deconvolution microscopy; high-throughput, automated, robotic slide processing; and quantitative, massively parallel, high-throughput analysis of peptide antigen–antibody interactions by layered peptide array (LPA) technology.9

The data presented here for HER2 and telomerase IgY antibodies in cancer biomarker detection show promise for differentiating patient tumors in tissue microarray (TMA) format from normal control tissues. More generally, these results promise improvements in measuring and standardizing affinity and antibody-based proteomic reagents that may expedite clinical proteomics validation projects for cancer biomarkers.

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

IgY antibodies

Protein sequence domains of human HER2 (human epithelial growth factor family receptor-2/neu, GenBank No. NP_004439.1), and human telomerase (hTERT, GenBank No. NP_003210.1) were selected for design of polypeptide immunogens. Polypeptide antigen sequence and design considerations will be reported elsewhere. The anti-HER2 IgY antibody is termed YXPB-IgY-hHER2/p. The anti-telomerase IgY antibody is named YXPB-IgY-hTel/p (Table I). Each resulted from production experiments that generated a single polyclonal antibody lot. Antigens were expressed as recombinant protein and purified from E. coli cells. Immunogens were injected intramuscularly into one laying hen (leghorn/Rhode Island) for each immunogen. About 200 μg of antigen protein were mixed with Complete Freund's Adjuvant (CFA) for primary immunization. Booster inoculations were performed at 2–3 weeks interval with 100 μg antigen in Incomplete Freund's Adjuvant (IFA) for 3 inoculation cycles. Eggs were collected after a second boost. Antibodies were isolated from egg yolk using the polyethylene glycol (PEG) precipitation method.11 The antibodies were further purified by affinity chromatography with immunogen as affinity ligand. Antibody production was performed on contract by GenWay (San Diego, CA).

Table 1. Primary Antibodies, Immunogens and Sources: Metrology for Affinity Proteomics
AntibodyImmunogenIsotypeReference/Vendor
  1. 1. Ref.8.

  2. 2. Antigen: a synthetic carboxyterminal fragment of the intracytoplasmic domain of human HER2 protein coupled to keyhole limpet hemocyanin.

  3. 3. Antigen: a synthetic peptide to the internal domain of cerbB2 oncoprotein (HER2).

  4. 4. Ref.8.

  5. 5. Antigen: human telomerase amino acids 173–320.

  6. 6. Antigen: full length human telomerase protein expressed in insect cells.10

  7. 7. Antigen: synthetic peptide to amino acids 1–100 of human β-actin sequence.

Anti-human HER2 Antibodies
 YXPB-IgY-hHER2/p1IgY, chicken polyclonalXiao et al., this data
 AO4852IgG, rabbit polyclonalDako (Glostrup, Denmark)
 CB113IgG, mouse monoclonalVentana (Tucson, AZ)
Anti-human Telomerase Antibodies 
 YXPB-IgY-hTel/p4IgY, chicken polyclonalXiao et al., this data
 NCL-hTERT5IgG, murine monoclonalNovocastra (Newcastle upon Tyne, UK)
 NB100-2976IgM, murine monoclonalNovus (Littleton, CO)
Control Antibodies (anti-β-actin)
 ab138227IgY, chicken polyclonalAbcam (Cambridge, MA)

Mammalian antibodies

Commercially available primary mammalian antibodies for comparison with IgY antibodies generated here are summarized in Table I together with their antigens and sources. These were chosen as they are the most widely used antibodies in research studies and therefore considered the best available. Secondary antibodies were anti-IgY biotinylated antibody (GenWay), and anti-IgG biotinylated antibody (Ventana). Each was detected by fluorescence microscopy with streptavidin-Qdot605 (Invitrogen, Carlsbad, CA) as the final affinity detection reagent. As a control, chicken IgY isotype antibody ab14001 specific for human β-actin was obtained from Abcam.

Antibody studies on fixed cultured cells

For HER2, human breast cancer cell lines MCF-7 (low HER2 gene copy number and expression) and SK-BR-3 (high HER2 gene copy number and expression) were studied as a model system. For telomerase, human cell lines A549 and IMR90 were the positive and negative control model system studied,12 with high and low telomerase levels, respectively. Cells were grown on tissue culture chamber slides (Nunc, Rochester, NY) in medium as recommended by American Type Culture Collection (Manassas, VA), at a density of 30,000 cells/cm2. Cell monolayers were fixed in 10% neutral-buffered zinc formalin (Fisher, Pittsburgh, PA) but not paraffin-embedded. Fixed monolayers were preblocked with 5% w/v nonfat dry milk in TBST (50 mM Tris-HCl, 150 mM NaCl, 150 mM Tween 20, pH 7.5), 20°C, for 20 min. For antibody detection, slides were manually incubated with primary antibody, secondary antibody and fluorescent detection reagents. Primary antibody dilutions are: IgY 1:300 (final concentration 3.3 μg/ml), CB11 IgG undiluted as per manufacturer instructions (6.0 μg/ml), NCL-hTERT 1:200 (16.5 μg/ml). Alternatively with high-throughput technology, slides were robotically prepared (reaction with primary and secondary antibodies, and washes) with a Benchmark XT workstation (Ventana).13 IgY-isotype antibodies were detected with rabbit anti-IgY antibody labeled with biotin, followed by semiconductor nanocrystal fluorophore-streptavidin conjugates (Invitrogen). FluoSpheres carboxylate-modified microspheres, 0.2 μm, crimson fluorescent (Invitrogen) were used as calibration standard. Imaging systems for 3D analysis of fluorescence signals from quantum dots, deconvolution of z-plane (depth) images, and integration of the signal with an imaging system were described elsewhere.14

Antibody studies in patient and control tissues with tissue microarrays (TMAs)

Tissue microarrays (fixed in 10% neutral-buffered formalin and embedded in paraffin) obtained from Invitrogen were treated according to the manufacturer's instructions with antibodies and detection systems as above for the cultured cells. TMAs were deparaffininzed with the Benchmark XT system.

Western blotting

Single-dimensional Western blot analysis involved electrophoresis of cell lysates prepared as described,15 on NuPAGE 4–12% Bis-Tris gels (Invitrogen). Gels were equilibrated in buffer and electroblotted onto nitrocellulose membranes. Transfer blots were soaked in blocking buffer containing 5% w/v nonfat dry milk in 1× PBS and 0.1% v/v Tween 20, incubated with primary IgY, IgG or IgM antibodies overnight at 4°C. All primary antibodies were diluted 1:1000. Blots were then washed and incubated with HRP (horse radish peroxidase)-conjugated secondary antibodies, either goat anti-chicken IgY (GenWay) or goat anti-mouse IgG + IgM (Abcam), for 1 hr at room temperature. After washing with Tween-PBS, bound antibody was visualized with 3,3′-diaminobenzidine (DAB; Sigma, St. Louis, MO).

Layered peptide array (LPA) analysis

P-FILM Smart Antibody Affinity membranes (20/20 GeneSystems, Rockville, MD) coated with the peptide or protein of interest, were placed within a vacuum plate (Bio-Rad, Hercules, CA). Antibodies were applied to the 96 wells in the plate in duplicates and incubated for 5 min at room temperature. Vacuum was applied for 5 min followed by washing of the membranes in PBS for 5 min, and then incubation with secondary FITC- or Alexa-conjugated, isotype specific antibodies for 30 min at room temperature with shaking, followed by another wash in PBS. Membranes were dried on a filter paper (Whatman, Florham Park, NJ) and scanned on a Typhoon scanner with 520 BP40 and ALEXA filters (Amersham Biosciences, Piscataway, NJ). Images of the membranes were imported to the ImagePro 4.5 analysis software (MediaCybernetics, Bethesda, MD) for analysis. Each membrane included 96 dots. The optical density was calculated by the program for each well in the membrane by marking a circle containing each dot. The optical density was defined according to the following formula: [OD = −Log10(X/256)] with 256 representing the total number of gray levels in the image and X the individual level of gray of each object (each well of the 96 wells for membrane). The measurements were repeated 2 times, for every duplicate well. Thus, a data set of average optical densities for each antibody in the membranes was generated. Data were imported to Microsoft Excel and mean ± standard deviation values were calculated.

Statistical analysis

For comparative studies, Student's t-test (unpaired) or one-way ANOVA tests (with Bonferroni post test if p < 0.05) were used for statistical analysis. Differences were considered statistically significant if a p value of < 0.05 was achieved.

Results

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

HER2

In fixed cell preparations, chick antibody YXPB-IgY-hHER2/p detected high levels of HER2 protein in HER2-expressing cell line SK-BR-3 compared with low-expressing MCF-7 controls (Fig. 1a). Significant high-level HER2 fluorescence signal was localized to the cell membrane in SK-BR-3 cells as seen in optical sections from deconvolution experiments, whereas only low signal levels were found in the MCF-7 controls.

thumbnail image

Figure 1. Immunohistochemical detection of cancer biomarkers with IgYs and quantum dot fluorophores in positive and negative expressing cell lines. (a) HER2 detection with IgY in high expressing (SK-BR-3, left) and very low expressing (MCF-7, right) cells. (b) HER2 detection, same experiment with mammalian IgG antibody CB11. (c) Human telomerase detection with IgY in high expressing (A549, left) and very low expressing (IMR90, right) cells. (d) Human telomerase detection, same experiment with mammalian IgG NCL-hTERT. The white bar beneath the micrograph represents 10 μm in the left panel of c and 15 μm in the right panel of d; in all other cases it represents 20 μm.

Download figure to PowerPoint

For comparison, similar experiments with commercial mammalian IgG antibody CB11 (Table I) raised against HER2 were also performed. High HER2 expressing cell line SK-BR-3 showed membrane-limited fluorescence when probed with CB11; with this antibody, MCF-7 cells showed low HER2 signal (Fig. 1b). As performance of mammalian IgG antibody AO485 is considered on a par with CB11 in HER2 immunohistochemistry (IHC) testing, the same experiment was not performed with the AO485 antibody.

Telomerase

For chick antibody YXPB-IgY-hTel/p analysis, model telomerase expression cell lines included A549 (high expression) and IMR90 (very low expression control) cells, respectively. The IgY antibody detected epitopes exclusively within the nucleus in A549 cells, but not in nucleus or cytoplasm of low-expressing IMR90 cells (Fig. 1c). Representative images from stacks (n = 20 sections total) of telomerase signals in A549 and IMR90 cells are illustrated in Figure 2. For A549, cell to cell heterogeneity in total telomerase signal was found among nuclei. Detailed data on distribution and heterogeneity of telomerase expression detected with chick YXPB-IgY-hTel/p antibody will be presented elsewhere.

thumbnail image

Figure 2. Representative deconvolution image processing planes (top 4 panels) illustrated in high telomerase expressing cells (A549, left stack) and very low expressing cells (IMR90, right stack). Planes shown for illustration (4) are a subset of the total image planes (20) on which signal quantitation was based. After digital removal of unfocused light by deconvolution software, the total in-focus fluorescence is represented on the bottom grids for high expressors (A549, left bottom) and very low expressors (IMR90, right bottom). The analysis was made on the same cell image as in Figure 1c.

Download figure to PowerPoint

Commercial mammalian IgG antibody NCL-hTERT (Fig. 1d) showed a lower signal level in telomerase-expressing A549 cells, than the chick antibody. Telomerase-specific IgM NB100–297 signals were not detected in either A549 or IMR90 cells with the quantum dot detection methods (data not shown). Measurements of total Qdot605 fluorescence in 10 cells are shown in Figure 3, quantitating the images in Figure 1.

thumbnail image

Figure 3. Quantitation from IHC Imaging with quantum dots detected with IgY and IgG. Blue bars, data from HER2-specific antibodies IgY (dark blue) and IgG CB11 (light blue) in high expressor (SK-BR-3) and very low expressor (MCF-7) cells. Green bars, data from telomerase-specific antibodies IgY (dark green) and IgG NCL-hTERT (light green) in high expressor (A549) and very low expressor (IMR90) cells. FluoSpheres carboxylate-modified microspheres, 0.2 μm, crimson fluorescent (Invitrogen) were used as calibration standard (red bar). HER2-IgY vs. IgG in SK-BR-3 cells, p = 0.04. Telomerase IgY vs. IgG in A549 cells, p = 0.04. Error bars represent standard error of the mean.

Download figure to PowerPoint

In these two model cancer biomarker systems, mammalian IgGs and chick IgYs were detected with the same secondary fluorophores (biotinylated anti-isotype antibodies followed by streptavidin-Qdot605 conjugates). However, because of differences in the antigens, the isotypes of the primary and the second antibodies, comparisons are considered relative, not directly quantitative among isotypes.

Immunodetection with quantum dot fluorophores in cultured cells

Empirically, comparing chicken with mammalian antibody in the HER2 model, IgY antibody YXPB-IgY-hHER2/p showed greater total signal for SK-BR-3 than IgG antibody CB11 (Fig. 3, blue bars, 4,131 vs. 2,428 fluorescence intensity units, p = 0.04). The HER2-related IgY and IgG signals for MCF-7 low-expressing control were not statistically different (p = 0.13).

Comparisons of mammalian with avian antibodies for telomerase as target showed chick YXPB-IgY-hTel/p antibody about 4.6-fold greater signal for A549 cells than mammalian NCL-hTERT antibody (Fig. 3, green bars, 3,269 vs. 712 fluorescence units, p = 0.04). Signal differences for IMR90 low expressing control between telomerase-related IgY and IgG were statistically not significant (p = 0.30). These quantitative differences between signals elicited by avian and mammalian antibodies were proportional, high to low signal, in both HER2 and telomerase model systems.

These data suggest that for the antibodies tested, IgY polyclonal antibodies gave greater signal than mammalian antibodies detected with quantum dot fluorophore-labeled secondary antibodies. Differences may arise from the antibody clonality (mono- vs. poly-clonal), the specific isoforms or processed polypeptide detected by each antibody, and the fact that antibodies were raised to different polypeptide regions of the same target molecules (Table I).

To independently determine whether these antibodies detected the appropriate target proteins by molecular weight and abundance in highly expressing cells, polyacrylamide gel electrophoresis (PAGE) followed by Western blots was performed15 (Fig. 4). On the three vertical panels depicting parallel PAGE gels for HER2 (Fig. 4a), CB11 (top panel) detected a faint band of ∼185 kD in cell extracts from expressor SK-BR-3 cells. No band at this molecular weight was detected in similarly prepared extracts of negative control MCF-7 cells. The AO485 anti-HER2 antibody was not evaluated by Western blotting. Chick antibody YXPB-hHER2/p (middle panel) detected a band of ∼185 kD in SK-BR-3 cells that was absent from control MCF-7 cells. The sample loading control for β-actin (antibody ab13,822, Table I) at 42 kD (lowest panel) shows equal amounts of protein lysate were present in each lane.

thumbnail image

Figure 4. Western blot analysis with HER2 and telomerase antibodies on high and low expressing model cell cultures. Signals detected with horseradish peroxidase/DAB as in text. (a) HER2 expression from Western blots as detected by IgG (top), IgY (middle) and β-actin control (lower) in HER2 high expressor (right lanes) and HER2 very low expressor (left lanes). (b) Telomerase expression from Western blots as detected by IgM (top panel), IgG (second from top), IgY (third from top) and β-actin control (lowest panel).

Download figure to PowerPoint

PAGE/Western blot analysis with anti-telomerase antibodies is shown in the 4 panels of Figure 4b. Neither the commercial IgM antibody NB297–100, nor IgG NCL-hTERT detected signal in cell lines of the telomerase model system (top panel and second from top). Loading controls are shown in the bottom panel. By contrast, the chick telomerase antibody YXPB-IgY-hTel/p detected a visible ∼127 kD band in A549 cells that was absent in IMR90 cells (third panel from top).

Differences in Western blot signal represent different levels of target protein between cell lines. Diaminobenzidine (DAB) detection is semi-quantitative, however the relative protein abundance and approximate size of the targets HER2 and telomerase is consistent with correct detection of the targeted proteins. These data with the immunochemical data above indicate the IgY results are consistent with detection of appropriate target proteins in the cell culture model system, and Western blot bands detected were of appropriate molecular weights. The basis of lack of telomerase Western blot signal in high expression A549 cells detected by commercial antibodies is unknown, although it is well known that antibodies successful in IHC may or may not prove adequate for Western blot experiments.

Quantitative comparisons of IgYs with mammalian antibodies

To further characterize antibody performance, a novel layered peptide array technology (LPA)9 was used. Although the most direct relative performance experiments would be on isotype-matched antibodies generated in response to the identical antigens in various species, a preliminary trend was evaluated by comparing chick IgYs with mammalian antibodies that were commercially available and widely used. Although a comprehensive set of overlapping peptides were not available for these affinity studies,16 a preliminary estimate of relative antibody performance was made with antibodies tested against a dilution series of a limited set of polypeptide antigens by LPA.9 The scope of work presented here precluded re-synthesis of immunogens for all antibodies, although immunogens for the HER2 and telomerase IgYs were available. Therefore, these results are a preliminary comparison among antibodies. The purpose of the LPA work was to collect preliminary data on selected, widely available antibodies with immunogen peptides specifically used to generate the new chick reagents. Immunogen design and sequence will be presented elsewhere.8

LPA data for anti-HER2 reagents (Fig. 5a) include chick IgY antibody (YXPBY-hHER2/p), rabbit IgG antibody (AO485) and mouse IgG isotype antibody (CB11). Antibody dilutions were tested against a polypeptide antigen covalently bound to the proprietary assay membrane.9 All these HER2-specific antibodies tested against the peptide antigen used to raise YXPB-IgG-hHER2/p, were not different from negative control at 0.5–2 μg/ml of antibody (p > 0.05). However, at higher concentrations (8, 16 μg/ml), chick antibody YXPB-IgY-hHER2/h showed greater membrane binding with increase in the amount of antigen peptide bound (at 8, 16 μg/ml, p = 0.01 and 0.004, respectively). The mouse CB11 antibody, which was raised to HER2 (Table I) but not the specific immunogen tested here, also showed activity against this peptide, however, the IgY binding to membrane-bound immunogen was 10–25% higher (p < 0.01). The rabbit antibody AO485 (anti-HER2 IgG) showed no concentration-specific signal increase against this particular peptide. This is not surprising as antibody AO485 was raised against a carboxyterminal intracytoplasmic human HER2 antigen commercially developed for IHC applications (Table I).

thumbnail image

Figure 5. Quantitative LPA Analysis of relative affinities of antibodies for polypeptides corresponding to biomarkers HER2 and telomerase. (a) Comparison of three HER2-related antibodies with increasing antibody concentration to the right (abscissa). HER2 polypeptide (Xiao et al., unpublished data; Table 1) was in constant concentration on the membrane for all antibodies. (b) Comparison among three telomerase-specific antibodies against a human telomerase polypeptide bound to membrane. Statistics: HER2 IgY vs. IgG: 0.5–2 μg/ml, p > 0.05; 4 μg/ml, p = 0.22; 8 μg/ml, p = 0.01; 16 μg/ml, p = 0.004. Telomerase IgY vs. IgG: p-values for 0.5, 1, 2, 4, 8 and 16 μg/ml were respectively, 0.27, 0.02, 0.006, 0.02, 0.94, and 0.005. Error bars represent standard error of the mean.

Download figure to PowerPoint

Results for the antibodies raised against telomerase polypeptides are shown in Figure 5b. For the IgY, binding increased gradually from 0.5 to 16 μg/ml. However, at most antibody concentrations from 1 to 16 μg/ml, YXPB-IgY-hTel/p showed significantly higher binding signal than IgG (1, 2, 4, 16 μg/ml, p = 0.02, 0.006, 0.002, and 0.006, respectively). At 8 μg/ml, binding of the mammalian and avian antibodies is statistically indistinguishable (p > 0.05), and at 0.5 μg/ml, the IgG appeared to have significantly greater binding. Consistent with the immunochemical data, the signal for the mammalian IgM isotype antibody NB100–297 was weak or absent in LPA.

Quantitation with TMAs

With cell culture results indicating correct targeting of reagents and relative quantitation capabilities for both models, formalin fixed, paraffin embedded patient tumor cores were screened with IgY antibodies in commercially available tissue microarrays (TMAs). Representative images are shown from experiments with quantum dot-detected HER2 IgY in breast cancer patient tumor biopsies with HER2 overexpression (Fig. 6a), with normal HER2 expression (Fig. 6b), and in normal colon controls (Fig. 6c); and telomerase IgY detection of telomerase in biopsies from prostate cancer (Fig. 6d), in benign prostate hyperplasia (BPH; Fig. 6e) and in normal colon controls (Fig. 6f).

thumbnail image

Figure 6. Representative figures and image quantitation from tissue microarray analysis of breast and prostate tumors by IgYs for HER2 and telomerase. (a–c) TMA section for which HER2 was detected with IgY and quantum dot fluorophores. (d–f) TMA sections for which telomerase was detected with telomerase-specific IgY and quantum dot fluorophores. Image quantitation (n = 10 cells, randomly chosen) is shown under the images in red histograms for HER2 and in blue histograms for telomerase. The white bar beneath the micrographs represents 20 micrometers except in d, where it represents 10 μm.

Download figure to PowerPoint

Ten cells in each image were randomly selected for quantitation of the levels of IgY-detected signal as above, and these data are represented in the histograms under the tissue section images (Fig. 6). For breast cancer TMAs, overall positive signal intensity was approximately similar to signal detected in cell cultures compared to background and controls. The tumors collectively showed ∼20-fold greater HER2 signal than normal colonic tissue provided in the tissue microarray as detected with IgY and quantum dot imaging. In parallel, prostate tumor tissues collectively showed ∼11-fold overexpression of telomerase nuclear signal compared to BPH.

In the breast cancer TMA, 60 core breast tumor tissues were represented, encompassing both HER2-positive and HER2-negative tumors determined independently by CISH. Double-blind IgY screening here detected a positive signal (above threshold defined by signal in normal colon) in 10 of 60 breast tumors. Independently performed CISH for HER2 gene copy number showed 9 of 60 breast cancer tissue cores had amplified HER2 gene according to TMA supplier. Thus, of the 10 HER2-positive breast tumor cores that expressed high levels of HER2 as detected by IgY HER2 antibody and quantum dot imaging, one was read as negative by CISH. As no clinical data other than CISH result were provided with TMAs, no further analysis of IgY results in these clinical materials was possible.

In a smaller set of prostate tumor TMA cores (n = 20 patients) and parallel TMA cores of BPH, telomerase-positive signals were detected in 20/20 prostate tumor cores. Among these, two prostate tumors had signal levels of 20–60 intensity units and 18 prostate tumors had signal levels of >60 intensity unit levels. Signals in the benign prostatic hyperplasia cores (n = 20 patients) were not above background as defined by the normal colon core included in the TMA.

Clinical data available for prostate tumor patients (patient age, size of tumor, consistency of tumor, cellular differentiation and clinical diagnosis) were compared to IgY imaging results for telomerase. For the 18 prostate tumors for which Gleason Grades were reported, the average and standard deviation of telomerase signal for quantitative telomerase imaging for all tumors (grades 4–10) was 149 ± 40. For those at Gleason Grade 7 and above (n = 14), telomerase signal value was 159 ± 35. Among tumors at Gleason Grades 4, 5 and 6 (n = 4), the telomerase signal value was 96 ± 34.

When telomerase values for prostate tumors (n = 19) were compared with the level of cellular differentiation (well-differentiated, moderate, poor), well differentiated tumors showed an average and standard deviation for telomerase signal value of 145 ± 21 (n = 2). Among moderately differentiated tumors (n = 11), the telomerase value was 133 ± 40. Finally, among poorly differentiated tumors (n = 6), the average telomerase value was 180 ± 8. Like Gleason grading, the trend in the data is lower telomerase index with moderately differentiated tumors, and a higher value among poorly differentiated tumors.

Telomerase quantitation with chick IgY antibodies and quantum dot imaging suggests a trend toward correlation with Gleason Grade and differentiation, however, more work is clearly indicated.

Discussion

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

The ability to quantitate cancer biomarkers depends on the uniqueness of interaction of primary reagents with targets. A major drawback of using mammalian IgG antibodies in clinical diagnostics is reactivity with anti-mammalian IgG antibodies, e.g. human anti-mouse IgG antibodies, rheumatoid factor, and the human complement system, which are often present in clinical samples such as serum or plasma, resulting in false positives, high background, or increased interference.

IgY antibodies have a number of intrinsic biochemical advantages over mammalian IgGs (for a complete review, see ref.17). Because of their structural and amino acid sequence characteristics, IgYs do not activate complement. In addition, IgYs do not bind to proteins A and G, to rheumatoid factors, or to cell surface Fc receptor. Consequently, IgY antibodies have been used successfully in a variety of methods in different areas of research, diagnostics, medical application and biotechnology.17 Using IgY in immunoassays can potentially eliminate false positives, and often results in low background and interference.

In addition to the intrinsic biochemical and immunological advantages, IgY has several other attractive advantages in terms of its production.17 IgY antibodies are usually produced using chickens or ducks as the immunization host and their eggs as the sources for antibody isolation. Thus, IgYs usually have enhanced immunogenicity against conserved mammalian proteins because of the phylogenetic distance between donor and recipient organisms.18 Also, IgY antibodies were found to have high affinity or avidity against human proteins.19–21 Moreover, polyclonal IgY antibodies can be produced more quickly and at lower cost.

In this study, we have generated two novel IgY antibodies for fluorescence quantitation of cancer biomarkers HER2 and telomerase. The results presented here indicate that the IgY antibodies are consistent in analytical validation studies with detection of the appropriate targets. The evidence for this comes from the approximate sizes of the proteins detected, expression in both cell culture systems and fixed patient tumors, and subcellular localizations. The results also indicate that compared with mammalian IgGs or IgMs, detection of the cancer biomarkers with IgY antibodies was more sensitive.

The ability to quantitate cancer biomarkers depends on the degree of quantitation possible with the detection system. The method presented here, comparing high and low expressor cell lines by relative intensity of signals, could be considered as relative quantitation. To be truly quantitative, the use of quantifiable references or calibration standards is needed. Also, sample preparation should be consistent and rigorously controlled. In the present study, the cell lines were fixed in zinc formalin but not paraffin embedded. In addition, the use of commercially acquired TMAs, which were prepared by means unknown, renders the method not fully quantitative.

Biologically, the HER2 IgY appeared to detect overexpression in 1 breast tumor that was not found to have the gene amplified by CISH analysis. Also, the chick IgY for HER2 correlated well with CISH in the other tumors (n = 59 patients).

The fact that telomerase remains a controversial cancer biomarker makes the telomerase IgY data of interest and possible significance. Previous recent studies have found no association between telomerase activity or mRNA, and clinical prostate cancer Gleason grade or stage.22 Some prior IHC studies have used a telomerase antibody that detects nucleolin, not telomerase.7 Although not conclusive, the trends in the present study warrant further investigation of a quantitative relationship between IHC-detected telomerase and clinical status of prostate cancer patients and their tumors.

While these preliminary data suggest a difference in telomerase values between tumors of Gleason Grade 6 and below, and tumors of Gleason Grade 7 and above, and between moderate and poorly differentiated tumors, further studies in larger patient populations are warranted to resolve whether this is a biological difference or due to the small numbers of patients in this preliminary study. While general trends are evident in comparisons of telomerase-IgY elicited fluorescence intensity with clinical parameters in this data set, the number of tumors was felt to be insufficient to evaluate these initial results. While it is tempting to speculate on initial specificity and sensitivity of the IgY IHC results for both the breast cancer data and the prostate cancer data, such comparisons will be deferred until a larger clinical cohort can be characterized in greater detail with automated methods.

We demonstrate here the feasibility of detection and measurement of cancer biomarker protein levels at single-cell resolution in limited cell populations of the size that might be collected as specimens for early cancer detection. Here clinical materials consisted of TMAs. However, in a more useful sampling format, clinical body fluids such as blood, sputum, stool, nipple aspirate, or urine all contain cells that could be analyzed by this method for the presence of HER2- or telomerase-expressing cancer cells. Cells from fine needle aspirates, although not considered an early cancer detection screening modality, would also be amenable to analysis as described here. The combination of IgY isotype-specific antibodies, novel quantum dot detection systems,14, 23 digital collection of in-focus fluorescence signal in all focal planes (3D deconvolution), and robotic, programmed high-throughput slide preparation, set the stage for development of an early cancer diagnostic low-cost, high accuracy test platform may be of clinical value. The TMA results demonstrate that the new IgY reagents can be used to generate relative quantitation data for studying clinical populations in fixed clinical histopathology formats.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Certain commercial equipment or materials are identified in this paper in order to specify adequately the experimental procedures. Such identification does not imply recommendation or endorsement by the National Institute of Standards and Technology, nor does it imply that the materials or equipment identified are necessarily the best available for the purpose. Supported in part by NCI-NIST Interagency Agreement Y1-CN-45016 (PI, PEB). HER2 work supported in part by NIST HER2 Standard Reference Material development funds (PEB) for RM 2373, Standard for HER2 Testing. The authors declare that they have no competing financial interests. Drs. Emmert-Buck and Gannot are inventors of layered peptide arrays, a new technology being co-developed by the NIH and 2020 GeneSystems, Inc., and can receive royalty payments through the NIH technology transfer program.

References

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  • 1
    Anderson NL,Polanski M,Pieper R,Gatlin T,Tirumalai RS,Conrads TP,Veenstra TD,Adkins JN,Pounds JG,Fagan R,Lobley A. The human plasma proteome: a nonredundant list developed by combination of four separate sources. Mol Cell Proteomics 2004; 3: 31126.
  • 2
    Ransohoff DF. Lessons from controversy: ovarian cancer screening and serum proteomics. J Natl Cancer Inst 2005; 97: 31519.
  • 3
    Baggerly KA,Morris JS,Edmonson SR,Coombes KR. Signal in noise: evaluating reported reproducibility of serum proteomic tests for ovarian cancer. J Natl Cancer Inst 2005; 97: 3079.
  • 4
    Ransohoff DF. Bias as a threat to the validity of cancer molecular-marker research. Nat Rev Cancer 2005; 5: 1429.
  • 5
    McCabe A,Dolled-Filhart M,Camp RL,Rimm DL. Automated quantitative analysis (AQUA) of in situ protein expression, antibody concentration, and prognosis. J Natl Cancer Inst 2005; 97: 180815.
  • 6
    Amos MD,Bridges AJ. Standards for autoantibody testing; addressing future needs for autoimmune disease and cancer diagnosis. Cancer Biomark 2005; 1: 2217.
  • 7
    Wu YL,Dudognon C,Nguyen E,Hillion J,Pendino F,Tarkanyi I,Aradi J,Lanotte M,Tong JH,Chen GQ,Segal-Bendirdjian E. Immunodetection of human telomerase reverse-transcriptase (hTERT) re-appraised: nucleolin and telomerase cross paths. J Cell Sci 2006; 119: 2797806.
  • 8
    Xiao Y. IgY antibodies to human telomerase reverse transcriptase. Patent application filed, June 26, 2006, attorney docket number 000479. 00163.
  • 9
    Gannot G,Tangrea MA,Gillespie JW,Erickson HS,Wallis BS,Leakan RA,Knezevic V,Hartmann DP,Chuaqui RF,Emmert-Buck MR. Layered peptide arrays: high-throughput antibody screening of clinical samples. J Mol Diagn 2005; 7: 42736.
  • 10
    Masutomi K,Possemato R,Wong JM,Currier JL,Tothova Z,Manola JB,Ganesan S,Lansdorp PM,Collins K,Hahn WC. The telomerase reverse transcriptase regulates chromatin state and DNA damage responses. Proc Natl Acad Sci USA 2005; 102: 82227.
  • 11
    Polson A,von Wechmar MB,van Regenmortel MH. Isolation of viral IgY antibodies from yolks of immunized hens. Immunol Commun 1980; 9: 47593.
  • 12
    Shay JW,Wright WE. Telomerase therapeutics for cancer: challenges and new directions. Nat Rev Drug Discov 2006; 5: 57784.
  • 13
    Bankfalvi A,Boecker W,Reiner A. Comparison of automated and manual determination of HER2 status in breast cancer for diagnostic use: a comparative methodological study using the Ventana BenchMark automated staining system and manual tests. Int J Oncol 2004; 25: 92935.
  • 14
    Xiao Y,Telford WG,Ball JC,Locascio LE,Barker PE. Semiconductor nanocrystal conjugates. FISH and p. H. Nat Methods 2005; 2: 723.
  • 15
    Xiao Y,Monitto CL,Minhas KM,Sidransky D. Lactoferrin down-regulates G1 cyclin-dependent kinases during growth arrest of head and neck cancer cells. Clin Cancer Res 2004; 10: 86836.
  • 16
    Soldateschi D,Bravaccini S,Berti B,Brogi A,Benicchi T,Soldatini C,Medri L,Fabbri F,De Paola F,Amadori D,Calistri D. Development and characterization of a monoclonal antibody directed against human telomerase reverse transcriptase (hTERT). J Biotechnol 2005; 118: 3708.
  • 17
    Zhang WW. The use of gene-specific IgY antibodies for drug target discovery. Drug Discov Today 2003; 8: 36471.
  • 18
    Gassmann M,Thommes P,Weiser T,Hubscher U. Efficient production of chicken egg yolk antibodies against a conserved mammalian protein. FASEB J 1990; 4: 252832.
  • 19
    Davalos-Pantoja L,Ortega-Vinuesa JL,Bastos-Gonzalez D,Hidalgo-Alvarez R. A comparative study between the adsorption of IgY and IgG on latex particles. J Biomater Sci Polym Ed 2000; 11: 65773.
  • 20
    Lemamy GJ,Roger P,Mani JC,Robert M,Rochefort H,Brouillet JP. High-affinity antibodies from hen's-egg yolks against human mannose-6-phosphate/insulin-like growth-factor-II receptor (M6P/IGFII-R): characterization and potential use in clinical cancer studies. Int J Cancer 1999; 80: 896902.
  • 21
    Stuart CA,Pietrzyk RA,Furlanetto RW,Green A. High-affinity antibody from hens' eggs directed against the human insulin receptor and the human IGF1 receptor. Anal Biochem 1988; 173: 14250.
  • 22
    Bialkowska-Hobrzanska H,Driman DK,Fletcher R,Harry V,Razvi H. Expression of human telomerase reverse transcriptase. Survivin. DD3 and. PCGEM1 messenger. RNA in archival prostate carcinoma tissue. Can J Urol 2006; 13: 296774.
  • 23
    Xiao Y,Barker PE Semiconductor nanocrystal probes for human metaphase chromosomes. Nucleic Acids Res 2004; 32: e28.