1. Patients 4923 bottom 4924 bottom 4924 bottom 2. Normal   4930 top 4928 top 4928 top Profile of cytokine expression in cancer patients and normal subjects using cytokine antibody chips. (Source: Cancer Genomics & Proteomics)

The antibody microarray is gaining popularity among scientists, particularly in the area of cancer research, where the ability to identify patterns at a glance is particularly seductive. New products and services are now available to smooth the transition to this powerful assay format.

The antibody microarray is attracting a great deal of interest because it can measure protein abundance independently of gene expression, unlike its older sibling, the DNA microarray. Changes in gene expression do not always correlate with protein abundance. In cases where there is correlation between messenger RNA (mRNA) and protein expression, that correlation is time-dependent and unique for each protein/mRNA pair.

There are many applications of this technology, but cancer research is one of the largest and most promising. With increasing use of protein biomarkers in the blood for screening and treatment of cancer, it is important to have a good, quantitative method for measuring protein and comparing protein abundances between samples.

There are two main types of antibody arrays: the label-based assay and the sandwich assay. All antibody arrays begin with a number of antibodies bonded to a plate or slide. In the label-based assay, fluorescently labeled target proteins bind to the antibodies on the slide. In a sandwich assay, immobilized antibodies capture unlabeled proteins, and a second antibody is used to detect the bound protein.

Label-based assays have the advantage that they can be highly multiplexed, and commercially available label-based arrays can come standard with as many as 1,000 antibodies. Using label-based arrays, a researcher can also co-incubate test and reference samples, allowing the two proteins to compete for binding sites on the antibodies. Potential disadvantages of the label-based format include interference in the protein-antibody interaction by the fluorescent label.

A label-based approach
Using a label-based approach, Brian Haab, PhD, of the Van Andel Institute, Grand Rapids, Mich., is developing an assay for early detection of pancreatic cancer. By targeting a wide variety of proteins, Haab and his team hope to develop a more accurate assay than one that would be based on a single biomarker alone. Using blood samples from many cancer patients and a cohort of controls from patients with benign disease, Haab's research group quantifies the binding of proteins to each of approximately ninety antibodies and studies the differences between patient groups, with the goal of defining particular groups of proteins that could be used as a diagnostic signature.

Because the same chip containing ninety antibodies is used for each sample, Haab has observed interactions that were unexpected and would not have been found in a method using individual assays. "This is one of the advantages of testing many different antibodies," he says. "You look at antibodies in situations where you wouldn't have otherwise done it. We've found a few things associated with pancreatic cancer we didn't know before. It was surprising."

Pioneering the sandwich array
Sandwich assays improve on specificity but do not work well in highly multiplexed arrays. In order for a sandwich assay to work, there must be a pair of antibodies available for each protein analyte, and there's a risk of cross-reactivity in the detection antibody. However, for groups of 30 to 50 targets, a sandwich-based antibody array can yield good results with low background.

click the image to enlarge   ProteoPlex Microarray. Each slide has 16 wells with an identical array spotted at the bottom of each well. The human array is shown above. Murine arrays are identical but do not have IL-7 or IL-8. Key: AS = alignment spots; PS = positive control spots; IL = interleukin; GM-CSF = granulocyte macrophage-colony stimulating factor; IFN = interferon; TNF = tumor necrosis factor. (Source: Craig Draper, PhD, EMD Biosciences Inc.)

In a review published in Current Proteomics [vol. 1 pp. 199-210 (2004)], Ruo-Pan Huang, PhD, assistant professor at Emory School of Medicine, Atlanta, describes an ELISA-based multiplexed sandwich immunoassay for applications such as measuring cytokine expression levels, detecting multiple bacteria and viruses, diagnosing allergies, detecting protein-protein interactions, and detecting protein modifications. Huang himself pioneered the development of the multiplexed sandwich immunoassay for detection of cytokines, which act as mediators of cellular communication. Cytokines are found in many physiological pathways, including growth, differentiation, apoptosis, healing, and homeostasis. It is now believed that disrupted patterns of cytokine expression in bodily fluids can be used to diagnose, treat, or stage a cancer.

Because cytokines are secreted proteins, the entire assay is simple and fast. Bodily fluids are collected and can be tested with very few intermediary steps. In contrast, a DNA microarray would require a time-consuming DNA purification step. Because of this synchronicity of target molecule and method, many of the first successful antibody microarrays are being developed by cancer researchers. "It also can be used for potential personalized medicine in cancer or autoimmune disease," Huang says. "Maybe we can distinguish the cytokine expression profile that can be used for diagnosis or to guide treatment of certain kinds of disease."

Antibody Arrays Against Bioterrorism Amy Herr, PhD, senior member of the technical staff in the biosystems research department and her colleagues at the Livermore, Calif., facility of Sandia National Laboratories, developed a label-based assay for cholera toxin B-subunit, diphtheria toxin, anthrax lethal factor, and protective antigen that could be used in a toxin detection device for use in the field. A direct assay measured fluorescently labeled toxin bound to capture antibodies spotted onto an epoxide-functionalized slide. A second, competition, assay used labeled toxin molecules as reporters for native toxin in solution. The authors describe these assays as "a facile quantitative route to analyzing solutions for the presence or absence of toxins." In the future, this could provide a quick and safe method of detecting dangerous toxins in a law-enforcement setting.

Jump start with commercial products
For researchers who do not have the time or resources to create their own antibody microarrays, there are a few commercial products available to make the process easier. For example, Gentel Biosurfaces Inc., Madison, Wis., has developed the Path slide, a nitrocellulose slide with a higher signal-to-noise ratio and lower background than traditional nitrocellulose surfaces. The Path slide comes ready to use in 3" by 1" format. The customer can then print it with antibodies using a robotic spotter.
Gentel Biosurfaces plans to launch the same product in 3" by 5" format, coupled with a microtiter plate for compatibility with liquid handlers and microtiter plate readers. "We like microtiter plates because it's a very common platform," says Gentel Biosurfaces chief scientist Bryce Nelson, PhD. "We feel the market is moving to an open platform concept."

 One limitation of antibody arrays is the availability of high-quality antibodies. These antibodies must either be custom-made for the array (an expensive proposition) or purchased from another company that specializes in antibodies. However, the use of products from another company introduces concerns regarding intellectual property. Imagine needing 100 or 1,000 separate licenses for an array containing 100 or 1,000 different antibodies. Additionally, antibody arrays suffer in comparison to DNA microarrays in one significant aspect: it is possible to print an entire genome on a single array. However, an antibody array made with the current technology can only contain a small portion of the proteome (estimated to be about 10⁷ proteins). This is both because proteomics is more complex and because there are so many variables contained in it, such as posttranslational modifications. "We're not there yet," says Nelson. "It's going to take much longer, because proteins are more complicated."

Novagen, a brand of EMD Biosciences, San Diego, offers a prepackaged cytokine array called Proteoplex. This sandwich array comes prespotted with antibodies for each of 12 cytokines, plus controls. The kit includes buffers and fluorescent detection reagents. The cytokines detected by the Proteoplex kit are pro- or anti-inflammatory; thus, this kit is marketed primarily for the study of inflammatory diseases such as arthritis. The customer runs the assay, then returns the plate to Novagen for analysis.

Product manager Craig Draper, PhD, predicts that use of antibody arrays will increase as customer comfort level increases. "It's something in which our company has invested and has taken to its logical conclusion," Draper says. "We have a lot of industrial and academic customers, and they are happy with the discount analysis we offer. People like that they can send slides in, and we get Proteoplex scientists to do the scanning for the customer."

A similar product measures murine cytokines. Raybiotech Inc., Norcross, Ga., offers a number of human and nonhuman cytokine arrays. Researchers can choose from about a dozen standard arrays for cytokines, angiogenesis, inflammation, atherosclerosis, chemo-kines, and metalloproteinases, or they can order custom arrays drawing on a number of choices of antibodies.

Upside-Down Arrays When an antibody array is "backward," such that proteins are immobilized to the solid phase and antibodies are applied in solution, it is sometimes referred to as a "reversed phase" antibody array. A study published in March in BMC Biochemistry [www.biomedcentral.com/1471-2091/6/2] uses this reverse phase approach to study leukocyte membrane protein interactions. Researchers at Sir William Dunn School of Pathology, University of Oxford, UK, and the Hospital for Sick Children, Toronto, Canada, successfully detected the human leukocyte membrane protein CD200 and mapped antigenic epitopes for two monoclonal antibodies in the vicinity of the ligand binding site using this method, and detected CD200 binding to its receptor CD200R. There are a great many leukocyte membrane protein interactions that are as-yet undefined. The characterization of these interactions is complicated by the very low affinities involved, and the fact that denaturization of the protein disrupts the three-dimensional interactions. Also, leukocyte membrane proteins tend to be heavily glycosylated, which could affect protein interactions. It was desirable to design an assay that would maintain the three-dimensional structures and allow for glycosylation. The researchers printed diluted CD200R (ligand) solutions onto an epoxy-coated microscope slide. They then incubated the spotted plate with either antibodies and a detection reagent to map antibody epitopes, or with CD200 and then with antibodies to the hybrid protein. These experiments yielded quantitative results for extremely weak binding interactions between CD200 and CD200R and demonstrated the usefulness of reverse phase antibody arrays for detecting multiple simultaneous protein-protein interactions.

 

 

 

 

 

 

 

 

 

 

 

 

Whatman Schleicher & Schuell, Keene, N.H., offers a full array of products and services from simple, standard cytokine arrays to custom array design with an on-site field specialist to run the samples. The company offers six different arrays for chemokines, cytokines, and angiogenesis in sandwich and label-based ("single-capture") formats on 1" by 3" microscope slides coated with nitrocellulose.

click the image to enlarge   Antibody microarray assay formats. (A) Label-based assays, showing detection in two colors. The left drawing shows a direct-labeling assay, in which two protein pools are respectively labeld with Cy3 or Cy5. The right drawing shows indirect detection, in which two protein pools are respectively labeled with tags such as biotin and digoxigenin. (B) Sandwich assay. A matched pair of antibodies bind the unlabeled analyte, followed by detection by a secondary antibody. (C) Other strategies. Suspensions of cells are incubated on microarrays of antibodies targeting cell surface antigens. Cells bind to the array according to their expression of the targeted antigens, and the level of binding to each antibody spot is quantified by dark field microscopy. The right drawing shows reverse-phase detection. Complex cell lysates are spotted onto surfaces, and the presence of particular proteins in the lysates are qunatified by incubation of antibodies targeting those proteins. (Source: Brian Haab, Van Andel Institute)

Whatman Schleicher & Schuell is one of a number of companies banking on the growth of the market for antibody arrays. Their customers include researchers working in clinical studies, drug discovery, and diagnostics. "The market is maturing slowly," says Michael Harvey, PhD, vice president of research and development at the company. "We need to provide these kinds of services to introduce the technology to the customer." Many of their customers decide to run their antibody microarrays independently after using the company's complete services on the first round. Harvey believes that cancer research has been the first to exploit the antibody array because of its unique applications in pattern recognition. "People in cancer are enamored with the idea of pattern recognition. The patterns give them a synergistic ability to interpret data."

Kinetics of antibody arrays
Bernhard Geierstanger, PhD, of the Genomics Institute of the Novartis Research Foundation, San Diego, is another pioneer in the development of antibody arrays. In a paper published in Clinical Chemistry, Geierstanger reports on the results of an experiment to test the kinetic behavior of a sandwich microarray. The goal was to determine whether the microarray behaved as an ambient analyte or as a mass sensing device. In the former, the concentration of antibodies removed during the course of the reaction is negligible, and does not affect the reaction kinetically. If, however, the microarray removed enough analyte from the solution to affect the overall kinetics, it would become a mass sensing device, and put into question some interpretations of data results.

Geierstanger and colleagues wrote, ". . . A mass sensing device will yield a signal proportional to the total amount of analyte present in the sample solution. Variations in assay volume could be a source of experimental uncertainty in mass-sensitive devices, but should not affect the precision or the detection limit of ambient analyte assays." 

Glimpsing the Future “In the future, the antibody array will play an increasingly important role in cancer research. We want to look at how individual cancers have different protein profiles. These can be used for patient care and management. This kind of approach can help identify a more integrated view of how cancer develops, and help people identify missing factors involved in development of cancer.” — Ruo-Pan Huang, PhD, assistant professor at Emory School of Medicine, Atlanta

Geierstanger captured antibodies onto aminosilane-coated, planar glass slides and incubated the slides with protein solutions (24 mouse serum proteins), then biotinylated detection antibodies. For detection, the group used RLS particles, gold particles coated with an anti-biotin antibody. An instrument was used for detection that measures light scattering. The result of the experiment was that the microarray did indeed behave as an ambient analyte system. This clears the way for developing antibody arrays for clinical applications.

Although most new publications regarding antibody array technology are in the area of cancer research, and especially cytokines, the usefulness of this new technology extends far beyond cancer. Any research in which it would be useful to measure the abundance of a number of different proteins can benefit from the use of antibody arrays.

Patterns of protein expression can be used to diagnose disease or determine appropriate treatment for patients. There are also several appropriate in vitro applications for antibody arrays. When working with tissue and cell cultures, antibody arrays can be used to study protein expression in tumor cells, phosphorylation states of cellular proteins, and posttranslational modifications of proteins. Scientists have also successfully studied enzyme activity using antibody arrays. In the future, the greatest demand for antibody arrays may be in basic research, where scientists looking at systems on the molecular level need to identify large groups of interactions simultaneously, and pick out patterns from these interactions.

About the Author
Shaffer is a freelance writer based in Ann Arbor, Mich.

This article was published in G & P magazine: Vol. 5, No. 4, May, 2005, pp. 26-30.