The Chipping News

Recent months have seen a surge of activity in the field of protein microarrays. No wonder: Gene expression-profiling is faster and more powerful thanks to improvements in DNA microarray technology. Now researchers want to apply these benefits to boost the speed of proteomics research. But developing these tools is not easy, as protein arrays present technical challenges not faced by DNA microarray manufacturers. "You can attach any two pieces of DNA the same way and expect [them] to behave the same way," explains Nick Naclerio, chief business officer of Zyomyx (www.zyomyx.com), a Hayward, Calif.-based company offering protein arrays and microfluidics-based chip production. But proteins are heterogeneous, he adds, thus making it difficult to develop methods to attach them to biochips and have them remain functional. Proteins are also more difficult to synthesize than DNA and are more likely to lose structural or functional properties in different environments or when modified. Unlike DNA, where the sequence is all that matters, a protein's three-dimensional structure must be preserved. Despite these problems, researchers and biotech companies have successfully developed protein arrays, most based on conventional DNA microarray technology. For example, Brian Haab, working in the laboratory of DNA microarray pioneer Patrick Brown of the Stanford University School of Medicine, designed protein microarrays for the detection and quantitation of 115 proteins in solution. Haab spotted the proteins onto poly-l-lysine-coated microscope slides using a robotic arrayer, and visualized protein binding using a standard DNA microarray scanner.1 Similarly, Harvard University researchers Gavin MacBeath and Stuart Schreiber recently described the design and use of high-density protein microarrays for high-throughput screening of protein function using an automated contact printer to spot 150-200-µm spots onto glass slides coated with an aldehyde-containing silane reagent.2 Protein arrays offer distinct advantages over conventional techniques for quantitating proteins in a complex mixture. Haab, now at the Van Andel Research Institute in Grand Rapids, Mich., explains that antibody assays using protein arrays are much quicker than those using two-dimensional gel electrophoresis or liquid chromatography. He adds that protein arrays facilitate data analysis because each spot is a separate entity. "With 2-D gels, matching up spots is quite difficult from gel to gel," Haab says. Protein arrays also improve upon standard protein-protein interaction techniques such as yeast two-hybrid, says MacBeath. Arrays give "direct interactions, whereas yeast two-hybrid can often be compounded by bridging interaction from other proteins expressed in the cell," he says. In addition, once a protein is expressed and purified, it can be arrayed, whereas with yeast two-hybrid experiments, "you have to worry about whether the protein is expressing and folding correctly, whether it is able to get into the nucleus, and you have to worry to some extent about orientation—whether the interaction will actually lead to transcriptional activation." MacBeath noted that his group's unpublished data shows that protein arrays give fewer false positives and negatives than yeast two-hybrid experiments performed with the same proteins. Several companies have recently entered the protein microarray market. BD Biosciences-Clontech (www.clontech.com) of Palo Alto, Calif., produces the Ab Microarray and Hypromatrix of Millbury, Mass., (www.hypromatrix.com) offers AntibodyArrays.3 Some companies, such as TeleChem International (www.arrayit.com) of Sunnyvale, Calif., and PerkinElmer Life Sciences of Boston (lifesciences.perkinelmer.com), sell DNA microarray equipment for use in protein array production. Other companies offer special attachment chemistries or even complete packages specifically designed for the protein array industry. Complete Array Systems Former Baylor College of Medicine professor T. William Hutchens, developed one of the first protein array applications, surface-enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF MS), in the late 1980s. The technique's patents were licensed exclusively to Molecular Analytical Systems (MAS), founded by Hutchens and colleagues, and as the parent SELDI company, MAS licensed certain rights under the technology to other companies. These include Fremont, Calif.-based companies Ciphergen Biosystems Inc. (www.ciphergen.com) and LumiCyte Inc. (www.lumicyte.com). According to LumiCyte CEO Hutchens, there is a fundamental distinction between the rights to SELDI held by these two companies. Whereas Ciphergen sells SELDI instruments and consumables for use by customers, it is not a service company. LumiCyte, on the other hand, is a service company, catering mostly to pharmaceutical and biotechnology companies. Ciphergen's ProteinChip® System encompasses arrays, readers, and software. The system is primarily a biomarker discovery and analysis tool that allows the user to "pan and fish" a sample for differences in protein expression without having to pre-identify the proteins in the sample. The arrays are aluminum chips that contain eight or 16 chemically treated spots for affinity capture of proteins from a few microliters of sample. The ProteinChip Reader uses SELDI-TOF MS to detect and analyze femtomolar quantities of proteins up to 300 kDa in size.4 Owing to differences in sensitivity, Ciphergen's ProteinChip System can be seen as a complementary tool, rather than a replacement to existing proteomics technologies such as 2-D gel electrophoresis. According to Dick Rubin, director of marketing, Ciphergen's system tends to be able to detect lower molecular weight proteins than 2-D gel electrophoresis, and is better able to detect proteins at extreme isoelectric point ranges. The system is less adept at detecting larger, or heavily glycosylated proteins. Zyomyx also offers a complete protein array and detection platform. The company's initial product, expected later this year, is the Protein Profiling Biochip™ system for simultaneous measurement of proteins from a very small liquid sample. The system will consist of a chip, fluorescent reader, and assay reagents. Zyomyx is also developing chips for functional assays, such as protein-protein interactions. Naclerio explains that Zyomyx's unique chip-manufacturing system offers a high degree of control over the arrays' uniformity. The molecules organize themselves on the chip's surface, which is then fully saturated with antibodies that are oriented such that their binding sites are exposed to solution. In contrast, with conventional chips, proteins can orient in different ways. "In many cases a large fraction of antibodies are not even available to the solution because their recognition sites are facing down in the membrane, gel, or slide," Naclerio explains. All About Attachment NextGen Sciences (www.nextgensciences.com) of Cambridgeshire, UK, has developed novel attachment and manufacturing methods for protein chips. Rather than attaching proteins directly to chips via charged coatings, such as an aminosilane-based substrate, NextGen Sciences expresses proteins in vectors encoding a 15-amino acid-long sequence called a BioTag™, which is specifically biotinylated by the enzyme, biotin ligase. Proteins expressed with a terminal BioTag will adsorb to the surface of streptavidin-coated plastic slides. "In about two minutes, [the protein] is nearly covalently attached to the surface of the slide, and it does not come off unless you pull the streptavidin off as well," explains Kevin Auton, NextGen CEO. Because the tag is small, says Auton, it does not affect protein solubility or activity. And since the biotinylation occurs on the same residue of the tag for every protein spotted onto the array, all proteins are oriented in the same way on the chip's surface and are free to "float" in solution while tethered to the slide's surface. "This gives us at least an order of magnitude increase in linearity, compared with other attachment methods," says Auton. NextGen's attachment system solves problems associated with nonspecific attachment of proteins to charged surfaces, says Auton. For example, random attachment can cause distortion of the protein's native structure or inactivation of functional groups. Auton points out that some groups have been successful with other non-covalent attachment methods, such as Heng Zhu and colleagues in Michael Snyder's laboratory at Yale University, who developed yeast proteome chips using nickel-coated glass slides and polyhistidine-tagged proteins.5 But this method can be ineffective if the interaction between the protein and the slide surface is too weak. "If you start to wash the chip very hard, you can displace the proteins from the chips," Auton says. NextGen has also developed a microfluidics system for chip processing that researchers can use to manufacture chips using under 100 µl of sample. This high-throughput system will be available through PerkinElmer Life Sciences, one of NextGen Sciences' corporate partners. Bothell, Wash.-based Prolinx Inc.'s Versalinx™ protein-immobilization technology has been commercially available since October 2001, and is based on the company's Versalinx chemical affinity system (www.prolinxinc.com).6 The system relies on the interaction between two synthetic, low-molecular-weight molecules, phenyldiboronic acid (PDBA) and salicylhydroxamic acid (SHA), which form a stable complex under a variety of conditions.6 Karin Hughes, vice president, R&D, explains that the SHA substrate used on the slides is presented in a three-dimensional, "polymer brush" format that offers qualities optimal for protein array work. According to Hughes, the Versalinx array substrates have very low nonspecific protein binding, and, unlike other three-dimensional substrates, do not require a blocking step to prevent nonspecific binding. Also, conjugation of protein to PDBA occurs separately from the immobilization of the protein to the slide, allowing more uniform orientation of the proteins on the surface of the array and resulting in an increase in signal. Prolinx offers customers SHA-coated glass slides and a series of PDBA-based reagents that scientists can use to modify proteins at different sites, depending on the availability of functional groups on a protein. For example, reagents exist for modification of proteins through lysine and cysteine residues, or through carbohydrate groups on glycosylated proteins. "You can optimize the activity of your protein by selecting the best reagent for modification. And, even though all of those proteins are modified differently, they can all be spotted on the same slide," says Hughes. The Versalinx system thus offers spot-to-spot optimization, a feature not available with conventional microarrays. Additionally, according to Hughes, because no additional blocking or purification steps are required, all steps in array production, from conjugation to spotting, can be automated. Courtesy of Prolinx Inc.That's Not a Broken Traffic Light: Protein arrays open up a variety of detection schemes, including direct labeling of arrayed proteins, labeling of detection antibodies, and dual-labeling methods. PerkinElmer Life Sciences offers a three-dimensional polyacrylamide-based substrate, called HydroGel, which is specifically developed for immobilizing proteins to a glass microarray slide. The thin HydroGel layer on the slide surface swells when proteins are spotted onto it, providing a solution-like environment in which the proteins retain their native structure and activity. The proteins are immobilized on the gel through van der Waals interactions and require no chemical modifications for adsorption to the slide. Unlike nitrocellulose-based substrates, HydroGel has low intrinsic fluorescence, facilitating detection of low abundance proteins. Researchers can use HydroGel arrays for protein-expression profiling and functional assays. PerkinElmer does not currently sell premade arrays, but rather offers tools for researchers to manufacture them on their own. Keene, NH-based Schleicher & Schuell's (S&S) ready-to-use FAST™ slides (www.arraying.com) are designed for protein microarray construction. These glass slides are coated with a nitrocellulose polymer that binds proteins non-covalently, and, according to company literature, maintains the arrayed proteins' activity. S&S recently announced the opening of a protein array service facility, providing researchers with a way to produce large volumes of customized, FAST-based arrays in as little as two months. Increasing Detection Sensitivity New Haven, Conn.-based Molecular Staging (www.molecularstaging.com) has developed protein arrays that incorporate the company's Rolling Circle Amplification (RCA) technology to improve the detection sensitivity of what is essentially a standard sandwich immunoassay. The company's arrays are composed of more than 100 capture antibodies. After treatment with an experimental sample, the array is exposed to a mixture of biotinylated target antibodies, followed by an anti-biotin antibody conjugated to a short oligonucleotide primer. A DNA circle complementary to the primer is hybridized to the primer; in the presence of DNA polymerase and nucleotides, a long single strand of DNA is generated that remains attached to the analyte. Using a conventional microarray scanner, binding events can then be detected by hybridization with fluorescently labeled oligonucleotide complementary to this DNA sequence.7 According to chief operating officer Stephen Kingsmore, this approach increases assay sensitivity 1,000-fold. Molecular Staging has used RCA-based chips to identify biomarkers associated with diseases such as cerebral palsy, severe combined immunodeficiency, and autism. According to Kingsmore, the company has been able to identify biomarker sets to diagnose these diseases at the molecular level. He explains that the identification of such biomarker sets would be impossible if the assays were to be performed one at a time, using an ELISA, for example. "When we looked at the cerebral palsy study, one protein at a time as an ELISA would, we didn't find any biomarkers. However, when we took the data set as a whole, it turned out that there were a number of markers in there [that] ... additively produced a very powerful diagnostic set," he says. Parallel Expression Another challenge to overcome in protein array production is the expression of many proteins simultaneously with similar modifications so that diverse proteins with varying stabilities remain fully functional and correctly folded on the array. Sense Proteomic Ltd. (www.senseproteomic.com) of Cambridge, UK, uses its proprietary COVET® technology to clone an entire cDNA library and to modify each protein in the same manner without knowledge of the individual genes' sequence. Every protein in the array is expressed with a fusion tag that facilitates purification and subsequent immobilization on the array. COVET technology is compatible with many recombinant expression systems, and, according to product manager Rhian Hayward, generates thousands of correctly folded, fully functional proteins. Sense Proteomic launched its first functional protein array product, the Expressarray™ p53, in January; the array contains wild-type and 49 mutant p53 human tumor-suppressor protein variants. The company followed that release with an array of functional potassium channel regulatory proteins, called the "Kv subunit protein array," in March. According to Hayward, Sense Proteomic, Ltd., will launch several additional array products this year. Protein array and biochip technology is a rapidly expanding field and companies are constantly releasing new products and developing new protocols. Companies that have jumped on the protein array bandwagon emphasize that, even though the field is still developing, it has much to offer researchers today. "One misconception in some customers' minds is that protein arrays are some kind of futuristic topic that they might be able to work on or work with someday," says Rich Fisler, business manager of microarray systems at PerkinElmer. "[But] the products that we sell right now will allow just about any lab to get started developing or experimenting with protein arrays." Aileen Constans (aconstans@the-scientist.com) is a contributing editor. References 1. B. Haab et al., "Protein microarrays for highly parallel detection and quantitation of specific proteins and antibodies in complex solutions," Genome Biology, 2:research4.1-13, Jan. 22, 2001. 2. G. MacBeath, S.L. Schreiber, "Printing proteins as microarrays for high-throughput function determination, Science, 289:1760-3, 2000. 3. A. Constans, "An alternative to two-hybrid," The Scientist, 15[21]:21, Oct. 29, 2001. 4. D. Fitzgerald, "Pursuing proteomes," The Scientist, 14:12[28], June 12, 2000. 5. H. Zhu et al., "Global analysis of protein activities using proteome chips," Science, 293:2101-5, Sept. 14, 2001. 6. A. Constans, "Linking up," The Scientist, 14[18]:23, Sept. 18, 2000. 7. B. Schweitzer et al., "Multiplexed protein profiling on microarrays by rolling-circle amplification," Nature Biotechnology, 20:359-65, April 2002. Protein Nanoarrays IN FOCUS | By Aileen Constans Protein microarray technology is still in its infancy, yet protein nanoarrays are already on the horizon. A research team led by Chad Mirkin, director, Northwestern University's Institute for Nanotechnology, recently demonstrated the use of dip-pen nanolithography (DPN)—a technique in which an atomic force microscope (AFM) is used as a nanoscale "pen" to deliver a molecular "ink" to a surface—to generate protein arrays with 100-350-nm features.1 These prototypical arrays contained only one type of protein, but Mirkin says more complicated nanoscale arrays are coming. Nanoarrays offer several advantages over microarrays, according to Mirkin, including higher feature density, smaller sample volumes, and the potential for faster readers. With a micrometer-scale receptor pad, molecular binding events can only be detected using a label, says Mirkin. As a result, scientists generally use fluorescent labels and fluorescence readers to detect binding events. But this approach is problematic, as chemical modification can change protein surface characteristics, potentially preventing them from binding to ligands immobilized on the array.2 If the receptor pad is reduced to the nanometer scale, however, the binding event will change every physical property of that pad, says Mirkin. AFMs could be used to quickly detect such binding. Nanoarrays can probe biological recognition events that cannot be studied with conventional biochips; thus, they extend the microarray's application range. "If we want to develop a detection system for something larger than a protein or DNA," —such as an organism,— "how can we do it?," asks Mirkin. One way, he says, is to build recognition sites for the biological structure to be detected on a solid surface using DPN. "Those structures can be made of small molecules, proteins, oligonucleotides—anything that we want to use as an ink for the DPN process. We could draw thousands of patterns, and ask not 'how do the biomolecules interact with the components of these patterns,' but 'how does the larger biological structure interact with this pattern itself?'" Once a specific macromolecular pattern is found to act as a receptor for a larger biomolecule, it can then be used in a detection system for structures that have in the past been very difficult to detect, such as the anthrax spore, Mirkin explains. Several companies are developing protein nanoarrays and nanoarray readers, including Ames, Iowa-based BioForce Laboratories Inc. (www.bioforcelab.com), and Chicago-based NanoInk Inc. (www.nanoink.net). BioForce is currently testing its NanoArray™ platform with antibody-capture chips, virus- and pathogen-detection chips, and cancer-screening chips, according to CSO Eric Henderson. NanoInk, is developing a commercial software package that enables researchers to use AFMs for DPN. According to acting CEO Chris Anzalone, NanoInk is also developing instrumentation for multi-pen systems that would allow patterning of hundreds or thousands of different substrates simultaneously. Henderson notes that interest in nanoarrays has gone from "the curiosity of a 'fringe technology' to [researchers asking] 'how can we use this?'" In particular, sample-intensive microarray applications, such as screening small biopsies or quantities of blood, can benefit from further assay miniaturization. Whether these nanoscopic products can supplant microscale protein arrays remains to be seen. References 1. K.B. Lee et al., "Protein nanoarrays generated by dip-pen nanolithography," Science, 295:1702-5, March 1, 2002. 2. T. Kodadek, "Protein microarrays: Prospects and problems," Chemistry and Biology, 8:105-15, 2001.

The Chipping News

Interested in reading more?

Become a Member of