The impact of functional genomic study using cDNA microarray

Discrimination of Genetic Polymorphism Using DNA Chips

Ena Wang and Sharon Adams

Department of Transfusion Medicine, Clinical Center, NCI,

National Institutes of Health, Bethesda, MD 20892 (e-mail: Wang_Ena@nih.gov)

HLA class I and class II molecules exhibit extraordinary polymorphism. According to the updated (April 9, 2001) HLA sequence database, 747 HLA class I and 561 HLA-class II alleles sequences have been assigned (http://www.anthonynolan.com/HIG/index.htm). As a consequence, accurate HLA typing for donor and recipient matching in transplantation has become increasingly complex and burdensome. In addition, due to the important role that HLA molecules play in antigen presentation and the stringency of the relationship between epitope and associated HLA allele, high-resolution typing is increasingly requested for appropriate enrollment of patients into immunization protocols aimed at the enhancement of T cell responses (1). Finally, specific HLA alleles may determine disease susceptibility and their identification has been used for diagnostic and/or prognostic purposes (2,3). Therefore, high-resolution HLA typing is increasingly in demand in clinical and experimental settings.

The usefulness of conventional serologic assays for HLA typing has been limited by the availability of allele-specific sera. As antibodies identify structural differences on the surface of HLA molecules, protein structure differences caused by single or limited nucleotide polymorphism particularly within the peptide binding groove of the HLA heavy chain are not detectable by these techniques. However, these differences are of functional significance as they determine the specificity and affinity of peptide binding (4,5) and, therefore, T cell recognition of self as well as allogeneic target cells (6,7). For this reason, functionally significant high-resolution typing of HLA is only achievable through molecular methods.

The application of hybridization-based array techniques to the study of biologic processes has spawned the field of genomics. The completion of the human genome project offers the opportunity to identify genomic variation in cells and tissues quantitatively, simultaneously and automatically under a variety of conditions using high-throughput microarray analysis. Array technology applied to the detection of single nucleotide polymorphisms (SNP) has confirmed most common polymorphisms previously identified by conventional techniques and has identified a large number of new SNP (8). Here, we will review the potential utilization of DNA oligonucleotide-based arrays for the identification of SNP relevant to HLA typing.

The basic principle of microarray technology is the miniaturization of current hybridization systems to allow the simultaneous performance of multiple reactions on a small surface using minimal amounts of reagents (including patient’s samples) and materials (Figure 1). The best example of an application of microarray technology has been the utilization of complementary DNA (cDNA)-based arrays for the study of gene expression profiles. With this method, thousands of individual genes are spotted separately on a solid surface such as a glass slide (those commonly used for routine histo-pathological studies). Test cDNA from various tissues (labeled with red fluorescence) and reference cDNA such as normal tissue or cell lines (labeled with green fluorescence) are then co-hybridized to the immobilized arrayed genes. Genes expressed in common by both test and reference tissues will fluoresce with both colors, represented as yellow. Those present only in the test material fluoresce red, and those present in the reference material fluoresce green. As fluorescence is read by a laser scanner, the fluorescence intensity, reflecting the cDNA expression level from both channels, is recorded separately. The relevant individual gene expression level is presented as the ratio of test (red) fluorescence intensity versus reference (green). Using this method, large studies have been performed to characterize the gene expression profile of various tissues in the context of a variety of physiological or pathological states.

Hybridization to cDNA arrays is facilitated by the length (600-1000 kb) of complementary sequence homology characterizing specific genes and, therefore, it is not suitable for detection of the few and often single nucleotide variations that differentiate human HLA polymorphisms. For this purpose, oligonucleotide-based arrays consisting of 16-25 nucleotide (nt) long oligonucleotides and using low stringency hybridization conditions are required. The first oligonucleotide-based array designed for the detection of allelic variants was reported by Saiki et al 1989 (9). Sequence-specific oligonucleotide probes linked to homopolymer tails were spotted onto nylon membranes and hybridized to biotinylated PCR products. Genotype determination was based on the colorimetric signal intensity produced by specific hybridization. Later, high-density oligonucleotide arrays were developed by Affymatrix (10,11). With this array technology, each oligonucleotide is synthesized on a glass surface in situ. Oligonucleotide chains are protected by a synthetic photolithographic mask that does not allow the addition of further nucleotides unless the mask is removed (de-protection) by a computer-directed beam of ultraviolet light. The solid surface is then flushed separately with individual deoxyribonucleotides. The site(s) activated (de-protected) by the UV light incorporates the nucleotide to its end while the protected sites do not. In this fashion, sequence-specific oligonucleotides can be built directly on a solid surface according to a pre-programmed order. Other fabrication techniques have been described such as the covalent attachment of pre-synthesized oligonucleotides to glass surfaces using a “Bubble Jet” inkjet printing device (12). With this technique, the oligonucleotides are attached to the glass surface by bi-functional crosslinkers that react with the amino group on the substrate and a thiol group on the oligonucleotide probe.

A new strategy known as BeadArray has been recently described. This technology uses fiber bundles attached to micro-sphere beads to form random arrays in which the identity of each bundle/bead combination is known. Theoretically, this method offers a potential 10-fold increase in density of nucleotides, allowing higher throughput at lower cost. The high density is achieved through the use of extremely fine fiberoptic bundles assembled into cassettes designed to fit a standard 96 well or 384 well microtiter plate. Every bundle is composed of 50,000 fibers of 3 m diameter, each with a 2.7 m microsphere imbedded at its tip. These beads are covered with oligonucleotides designed with a hairpin-shaped structure in which the 5’ and 3’ ends are self-complementary while the middle part is specific for the sequence of interest. The self-annealing of the two ends brings a fluorescent label into close proximity to fluorescence-quencher molecules. Hybridization of an unlabeled target containing the specific sequence causes separation of the fluorochrome-labeled end from the quencher end of the probe, thus restoring fluorescence.

The design of high-density oligonucleotide arrays is dependent upon the length of the oligonucleotide probe (N). Any set of probes composed of the four nucleotides can be synthesized in a maximum of 4 x N permutations (13). For detection of unknown SNP, like new HLA alleles, all possible permutations within a sequence of interests could be theoretically designed. The interrogated SNP is designed in the center-most position of the target/probe duplex (14). To interrogate both strands for all possible sequence variations, 2 x 4x x L probes are required (L= length of DNA target, X = number of polymorphic oligonucleotides). For example, to detect each possible individual base polymorphism in the HLA class I region, which occupies approximately a 1000 kb genomic region of DNA, 8 x 106 probes would be needed. Thus, multiple array chips would be needed to cover just the HLA class I loci.

A reverse array strategy for HLA typing has been developed. Instead of using immobilized oligonucleotides specific for individual HLA alleles, PCR-amplified HLA fragments from different individuals are spotted on a solid surface. Labeled sequence-specific oligonucleotide probes are then hybridized to the test samples from different individuals simultaneously. This approach is suitable for screening of large populations for a limited number of major alleles, as in genetic linkage studies and surveys of certain HLA types associated with disease.

If identification of every theoretically possible new HLA allele is not required, sequence information from previously identified alleles allows the design of oligonucleotide arrays for HLA typing with much less complexity. To detect and verify known HLA polymorphisms, sequence-specific oligonucleotides based on the SSOP principle can be synthesized and printed on arrays as probes. In this way, microarray technology would represent a miniaturized form of SSOP in which small amounts of DNA hybridized to a chip would allow detection of individual alleles. The fabrication of large-scale oligonucleotide arrays containing all identified allele-specific probes for HLA typing is conceivable. Single stranded sense and anti-sense consensus sequences for all alleles can be generated by in vitro transcription (15) and used as positive controls for the detection of novel alleles which will not hybridize to any existing oligo-probe. These new alleles can then be verified by sequencing.

PCR amplification-based preparation of test material is commonly used for the detection of genetic variation. Instead of performing each individual PCR reaction separately, multiple primer pairs are reacted together in a single tube (multiplex PCR) using 3’ and 5’ specific primers extended with T3 and T7 promoters respectively. With this method it has been possible to generate pools of sense or antisense single stranded DNA fragments using in vitro transcription with T3 or T7 RNA polymerase (16). This procedure allows direct incorporation of either biotin or a fluorescent label, and the labeled single stranded fragment can then be hybridized to antisense or sense oligonucleotide arrays for SNP discovery (17).

The detection of hybridization can be achieved with any of three different strategies (Figure 2). The first strategy uses 4 nucleotide probes complementary to the same sequence except for the nucleotide in the centermost position. Only one of the four probes will represent a perfect match. With this approach, a perfect match with high-efficiency hybridization of the probe to the target can be distinguished based on the high-signal intensity compared to the intensity of a mismatched signal (gaining signal) (16).

The second strategy utilizes a dual detection system in which the reference (consensus) material is labeled with one fluorescent dye (i.e. green) while the test target is either unlabeled or labeled with another fluorochrome (i.e. red). The two targets are then hybridized to the same array slide. During co-hybridization of the reference and test target, competition of the two targets for the same template binding sites results in decreased consensus (i.e. green) fluorescence if an unlabeled test target is used (loss of signal). If the test target is labeled, then the co-hybridization will result in the presence of dual colors. Signal loss or dual coloring suggests SNP. This approach has shown its effectiveness in discriminating single nucleotide mismatches (13).

The third strategy for the identification of sequence-specific hybridization employs a technique similar to that used for terminal sequencing. An oligonucleotide, typically 18 nt long, encompassing a consensus sequence of interest is spotted on the array. This oligonucleotide is designed to encompass a sequence ending at the 3’end just before the position where a putative SNP is being investigated. The test sample is then hybridized to the oligonucleotide. This complementary test sequence is then used as a template for a single base chain extension (SBCE) reaction using a mixture of four di-deoxynucleotides (ddNTP), each labeled with a different fluorescent dye. Since the SBCE reaction cannot be extended further after incorporation of a ddNTP, the fluorescent color will indicate the identity of the SNP at the 3’ end of the oligonucleotide (18).

Array-based detection of genetic variation has been successfully applied to the identification of SNP and to the mutational analysis of large genes with complex genomic structures. In one study, more than 90,000 oligonucleotide probes were applied to scan a >200 kb coding region of 22 genomic DNA samples and identified 17 out of 18 distinct heterozygous and 8 of 8 distinct homozygous variants (17). On a similar scale, 20 nucleotide-based oligo-arrays have been used to test for the presence of the hereditary of BRCA-1 gene mutation in breast and ovarian cancer, with high accuracy in detecting individual mutations. Examination of 3.2 megabases of human genomic DNA by high-density oligonucleotide arrays in combination with gel-based sequencing has identified a total of 3,241 of candidate SNP (16). Thus, the DNA chip technique, although still burdened by specificity of signal detection, accuracy of data analysis and optimization of experiment conditions, has already proven its potentially important role in complex gene analysis studies. By applying this technique to HLA it is theoretically conceivable that, in the future, high-resolution typing based on simple step hybridization to a small glass slide or fiber-optic bundle will be affordable in most laboratories. Automation of the methodology and interpretation will allow efficient response to the increasing demands for high-resolution HLA typing.

References

1. Bettinotti MP, Marincola FM. 2001. HLA and Cancer: Immunogentics in action. ASHI Quarterly 25:7-10.

2. Rani R, Sood A, Lazaro AM, Stastny P. 1999. Associations of MHC class II alleles with insulin-dependent diabetes mellitus (IDDM) in patients from North India. Hum Immunol 60: 524-531.

3. Perez-luque E, Malacara JM, Olivo-Diaz A, et al. 2000. Contribution of HLA class II genes to end stage renal disease in Mexican patients with Type 2 diabetes mellitus. Hum Immunol 61:1031-1038.

4. Parker KC, Bednarek MA, Hull LK, et al. 1992. Sequence motifs important for peptide binding to the human MHC class I molecule, HLA-A2. J Immunol 149:3580-3587.

5. Rotzschke O, Falk K, Stevanovic S, Jung G, Rammensee HG. 1992. Peptide motifs of closely related HLA class I molecules encompass substantial differences. Eur J Immunol 22:2453-2456.

6. Sette A, Vitiello A, Reherman B, et al. 1994. The relationship between class I binding affinity and immunogenicity of potential cytotoxic T cell epitopes. J Immunol 153:5586-5592.

7. Rivoltini L, Loftus DJ, Barracchini K, et al. 1996. Binding and presentation of peptides derived from melanoma antigens MART-1 and gp100 by HLA-A2 subtypes: implications for peptide-based immunotherapy. J Immunol 156:3882-3891.

8. Hacia JG, Brody LC, Chee MS, Fodor SP, Collins FS. 1996. Detection of heterozygous mutations in BRCA1 using high density oligonucleotide arrays and two-colour fluorescence analysis [see comments]. Nat Genet 14:441-447.

9. Saiki RK, Walsh PS, Levenson CH, Erlich HA. 1989. Genetic analysis of amplified DNA with immobilized sequence-specific oligonucleotide probes. Proc Natl Acad Sci U S A 86:6230-6234.

10. Lockhart DJ, Dong H, Byrne MC, et al. 1996. Expression monitoring of hybridization to high-density oligonucleotide arrays. Nat Biotech 14:1675-1680.

11. Fodor SP, Read JL, Pirrung MC, Stryer L, Lu AT, Solas D. 1991. Light-directed, spatially addressable parallel chemical synthesis. Science 251:767-773.

12. Okamoto T, Suzuki T, Yamamoto N. 2000. Microarray fabrication with covalent attachment of DNA using Bubble Jet technology. Nat Biotech 18:438-441.

13. Chee MC, Yang R, Hubbell E, et al. 1996. Assessing genetic information with high-density DNA arrays. Science 274:610-614.

14. Hacia JG. 1999. Resequencing and mutational analysis using oligonucleotide microarrays. Nat Genet 21:42-47.

15. Wang E, Miller L, Ohnmacht GA, Liu E, Marincola FM. 2000. High fidelity mRNA amplification for gene profiling using cDNA microarrays. Nat Biotech 17:457-459.

16. Wang DG, Fan J-B, Siao C-J, et al. 1998. Large-scale identification, mapping and genotyping of single-nucleotide polymorphisms in the human genome. Science 280:1077-1082.

17. Hacia JG, Sun B, Hunt N, et al. 1998. Strategies for mutational analysis of the large multiexon ATM gene using high-density oligonucleotide arrays. Genome Res 8:1245-1258.

18. Chen J, Iannone MA, Li M-S, et al. 2000. A microsphere-based assay for multiplex single nucleotide polymorphism analysis using single base chain extension. Genome Res 10:549-557.

Close window to return to previous page.