Hybridization Array Technologies


Key Points

  • Array technology provides a unique and powerful approach to screen a sample for dozens to thousands of genes.

  • Current array fabrications allow for diverse platforms with better detection and reduced cost for high-density applications.

  • Positive signals are generated when a tagged nucleic acid moiety hybridizes with its complementary probe localized on a solid support (i.e., “chip”).

  • Care must be taken when performing and interpreting microarray experiments because numerous factors must be considered in analysis.

  • Numerous applications of array technology have been proposed in the past few years, ranging from molecular staging of tumors to the identification and characterization of microbial agents.

  • Improvement of quality and yield of nucleic acid extraction from fixed and frozen tissues combined with continued miniaturization and cost reduction of platforms are making array technology more commonplace and user friendly.

In only a few short years, hybridization array technology, which enables the performance of thousands of simultaneous hybridization reactions on a solid substrate within a single analytical procedure, has gone from theoretical construct to practical reality. Such massively parallel determinations offer previously unimaginable opportunities for diagnostic application, ranging from gene sequencing and detection of genetic polymorphisms to measurement of gene expression profiles in cancer cells. Microarray-based systems offer a platform for quantifying the expression of thousands of genes at the same time. Therefore, an unprecedented advantage of this technology is that it affords analysis of whole cassettes of genes, or patterns of gene expression, rather than analysis of one, two, or three genes at a time. The general schema for microarray processing includes the generation of nucleic acid (i.e., deoxyribonucleic acid [DNA]) complementary to genes of interest that is laid out in microscopic quantities on solid surfaces at defined positions (the “probe”) nucleic acid (genomic DNA or complementary DNA [cDNA]) from samples is added over the surface—cDNA binds and the presence of bound DNA is detected by fluorescence after laser excitation (for a general review, see ).

It is clear that hybridization array technology has revolutionized the way we approach the study of disease. The one gene–one protein–one function approach is no longer the mainstay of scientific design. Array technology introduced a complicated mirror image of this approach—in other words, multiple functions, multiple proteins, multiple genes. The power of this technology lies in the “screening” capability that it provides. Under-defined experimental conditions, one now has the ability to determine changes in the patterns of genes, identifying multiple genes acting in concert. These patterns can then be “mined” through various computer algorithms, grouped into functional categories (e.g., respiratory genes, inflammatory genes), and used to derive signatures diagnostic of disease or to generate novel hypotheses about pathogenesis and cell function. The relationship of these genes to one another (increased versus decreased expression) are determined and generally expressed as fold-change. Microarray findings of differential gene expression require confirmation by other techniques such as quantitative reverse transcriptase polymerase chain reaction (RT-PCR) or additional independent microarray studies.

After candidate genes or groups of genes are identified, classic approaches to gene function (i.e., gene knockout or gene knockdown) can then be employed to elucidate or validate gene function. More important, hybridization array technology, with subsequent data mining, allows genes not previously known to be associated with the disease to be considered, generating new paradigms for future study. Similar platforms can be used to sequence large genes and to scan the genome for single nucleotide polymorphisms and methylation sites. This chapter reviews the theoretical basis for the various matrix hybridization platforms and provides a perspective on their clinical application, both currently and in the future.

Array Technologies

A hybridization array is the molecular equivalent of a spreadsheet, on which each cell or address reveals a specific piece of data, usually inferred from the binding of a ligand to its specific target. The first array-based methods were exploited in immunoassays ( ; ), and arrays also have been proposed for parallel studies of diverse targets such as proteins, lipids, carbohydrates, and small molecules ( ; ; ; ; ). Similar to antigen–antibody interactions on immunoarrays, the fundamental principle of nucleic acid arrays is detection of specific hybridization between complementary strands. The principles of nucleic acid hybridization and an overview of various platforms are covered in Chapter 68 . The effectiveness of solid-support hybridization formats was pioneered with the use of nitrocellulose membranes ( ) for dot blots ( ), line probes, and Southern blots ( ). Changing the labeled entity from the probe to the sample was first referred to as reverse line-blot (or dot blot ) hybridization and led to the beginning of interrogating one sample for multiple potential targets. Moving from dozens to thousands of targets has come about rapidly over the last few years, driven in part by the unique convergence of microfabrication, robotics, and bioinformatics technologies and the demand for high-throughput genetic analysis resulting from the Human Genome Project.

The field is bursting with potential, and many investigators foresee diagnostic and prognostic applications in tumor classification, chemotherapeutic responsiveness, pathogen detection, monitoring antibiotic resistance, and characterizing inflammatory responses. Array technology may make the dream of personalized molecular medicine, in which the genes of an individual patient may be assessed for risk factors and disease susceptibility and to predict response to therapy, a reality. Studies using gene expression profiling to predict diseases include identifying which trauma and burn patients will develop sepsis and multisystem organ failure ( www.gluegrant.org ) genes involved in the pathogenesis of chronic obstructive pulmonary disease ( ), and which patients would likely have recurrence of cancer ( ), among others. However, challenges of reproducibility and laboratory quality control remain to be addressed before this enormous potential is realized.

Macroarrays

The term macroarray has been applied to formats in which areas of probe localization, often called features , are large enough to be visualized without magnification. These have been manufactured on nylon or nitrocellulose membranes or as plastic strips with linear arrays of bound target. Because of the size of the features, the density of macroarrays is much lower than microarrays, ranging from a few dozen to hundreds or even thousands of probes. They are usually deposited onto the membrane by printing or dot blotting and then dried and stored for future use.

Nylon, plastic, and glass (silicon wafer) are standard supports to make macroarrays. Nylon suffers several drawbacks compared with silicon, which is a preferred matrix for hybridization arrays. Besides its porous nature, nylon displays high autofluorescence, limiting the sensitivity of fluorescence-based detection because of high background values. The latter also precludes the possibility to develop ratiometric (or two-color) assays. This approach uses two targets, each labeled with a different fluor. Most commonly, one is a constant reference control used in direct comparison with the unknown sample, with results expressed as a ratio of the two emitted fluorescent signals ( ).

Macroarrays are currently used for targeted applications. For example, some commercial arrays target all the known cytokine genes or those within specific signal transduction pathways that are activated during infectious and inflammatory processes. Other arrays, targeting the genes altered during oncogenesis, have been used in cancer research (e.g., Atlas Human Arrays, Clontech ; GeneFilters, Research Genetics). Other commercial applications, usually in kit format, are used to detect and type specific PCR products, such as in the cystic fibrosis mutation screen, human leukocyte antigen typing, or human papillomavirus (HPV) typing. Furthermore, nylon-based arrays have been used to detect bacterial colony–based gene expression ( ) and thyroid cancer detection ( ). The advantages of macroarrays include the ability to use standard hybridization equipment, simplified reading or scanning, and affordability.

Some groups have been successful in producing high-density nylon arrays on a custom basis for gene discovery purposes ( ; ; ; ; ; ). Even in high-density configurations, the size of the nylon arrays requires a relatively large volume of hybridization fluid, a practical limitation with respect to generating a probe from small amounts of diagnostic material. Because microarrays, rather than macroarrays, represent the high-density, high-throughput platform of choice, the bulk of this chapter focuses on applications of microarrays.

Microarrays

Assay miniaturization saves time while cutting costs in biomedical diagnostic applications and research. Working with smaller volumes reduces reagent consumption, increases sample concentration, and improves reaction kinetics. These improvements allow the investigator to determine hundreds or thousands of results in the time formerly required for a single experiment. Microarrays are now available from several commercial suppliers, along with reading and fabrication equipment for custom manufacture of dedicated arrays for specific research or diagnostic applications.

Microarray Substrates

A major distinction between microarrays and macroarrays consists of the choice of a nonporous solid support. By preventing diffusion of the target nucleic acids, nonporous surfaces (plastic, glass, or silicon substrates) allow faster hybridization kinetics and easier washing steps. To allow spatial discrimination of numerous reactions performed simultaneously, the molecular probes are bound to a solid surface in defined arrays. Deposition of probes on a solid substrate is also more amenable to automation and enables higher array densities with optimal image definition. These features apply to any type of microarray, from synthetic oligonucleotides to cloned cDNAs or PCR products ( ). Most microarrays are now developed on glass. As opposed to plastic derivatives, glass transparency and lack of autofluorescence allows low-background, fluorescence-based detection. However, as materials science becomes more involved in array technologies, alternative substrates, including coated glass and plastic substrates, may become available.

Microarray Fabrication

3′-End functionalization (i.e., chemical modification) of nucleic acids—oligonucleotides, PCR products, cDNAs, or peptide nucleic acid oligomers—is required for covalent immobilization on either glass or polypropylene surfaces ( ; ). For example, treatment of glass slides with silane allows amino-covered glass to bind amino-linked probes, using bifunctional molecules such as a dialdehyde or a diisothiocyanate ( ; ). Alternatively, glass coating with a polycation (e.g., polylysine) allows direct charge-coupled binding of polyanionic DNA probes ( ); an ultraviolet photo-cross-linking step adds covalent bonds to the ionic interaction. Covalent binding is essential to permit stringent washes and therefore accurate discrimination of hybridized species. Several protocols have been published to address surface activation ( ; ; ; ). An interesting alternative consists of deposition of small patches of activated polyacrylamide to which presynthesized oligonucleotides are attached by microinjection ( ; ; ).

The packing density of probes attached to solid surfaces influences greatly the performance of hybridization matrices. Poor coupling efficiency, which is often encountered on glass matrices, can result in low probe density and low signal-to-noise ratios. On the other hand, overly dense packing of oligonucleotides onto a solid surface causes spatial hindrance. This problem may be even more formidable with longer biomolecules such as cDNAs. Hybridization yields can be increased by up to two orders of magnitude by introducing spacers between the surface and the oligonucleotides ( ). The longer the spacer, the better the hybridization, but, interestingly, there is an optimal spacer length beyond which hybridization yield declines ( ; ). For example, a 40-carbon atom spacer provides a reported 150-fold hybridization enhancement ( ). The density of oligonucleotides is approximately 0.1 pmol/mm 2 on a glass surface, two orders of magnitude less than on aminated polypropylene. Thus, a potential advantage of glass over plastic matrices at present is an optimal oligonucleotide density associated with lower spatial hindrance ( ). Improved surface chemistries may one day allow production of arrays on plastic sheets or films, which may dramatically reduce production costs. To this end, film-based chip arrays have recently been described ( ).

The terminal composition (5′ end) of the oligonucleotides influences their duplex yield as measured by relative intensity. As expected, G:C-rich 5′ extremities lead to a better yield than their homologues (same composition but different sequences) ( ). Approaches to modify the 5′ end of oligonucleotides (e.g., covalent addition of a degenerated linker) can therefore be considered to minimize this contribution. Alternatively, knowledge that mismatches at the ends are less destabilizing and therefore more difficult to discriminate by hybridization ( ) has led to clever improvements for detection of hybridization events. Because polymerases and ligases are most sensitive to terminal (as opposed to internal) mismatches, enzymatic methods have been developed with improved detection stringency (solid-phase minisequencing [ ; ; ], genetic-bit analysis [ ], or ligation-assays [ ; ]).

Commercially available arrays have been previously reviewed by . Fabrication of microarrays falls into two broad categories: those made by direct probe delivery onto the solid surface and those made by in situ synthesis. A third method, which was developed by Nanogen (San Diego), consists of prefabricated oligonucleotides captured onto electroactive spots on silicon wafers ( ; ; ; ; ). Modification of the electric field through independent, spatially addressable electrodes can speed up hybridization and, when the polarity is reversed, provide stringent washing ( ). Recently, large single-stranded, circular-sense molecules were used as probes for DNA microarrays and showed stronger binding signals than those of PCR-amplified cDNA probes ( ).

Delivery Technologies

Deposition of presynthesized biomolecules was initiated by Pat Brown and colleagues ( ; ). Biochemical substances (e.g., proteins, peptides, oligonucleotides, cDNAs) are prepared, purified, and stored in microtiter plates. Small quantities of molecules are mechanically delivered on precisely defined locations on a solid surface using different precision robotic systems. This method is very flexible in terms of biochemical composition, microarray topology (e.g., size and density of the spots, replicate spots for each molecule), and ease of prototyping, allowing for design alteration for different applications. Mechanical methods allow the synthesis of moderate- to high-density microarrays (≤∼16,000 probes/cm 2 [ ]). It is the method of choice when large numbers of microarrays are needed with the same composition, as well as for long sequences using PCR products. Several companies are now manufacturing and selling premanufactured arrays (e.g., MicroMax, NEN Life Science) or arrayers based on deposition ( ) as an alternative to building an arrayer ( ).

Delivery procedures consist of different microprinting systems. Most arrayers use fountain pen–like microdispensing tips (e.g., Cartesian Technologies), but some original systems such as the pin-and-ring technique have been developed (Genetic MicroSystems). In general, because of surface tension, static, and other effects, the reproducibility of printed arrays is lower than in situ synthesis and must be corrected by separate experiments (repeats), replicate spots (on the same slide), multicolor fluorescence detection, or other specific detection algorithms ( ; ). Because of its flexibility and availability, printed array technology will probably have the largest impact on research and clinical laboratories over the next few years.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here