Methods and devices for measuring differential gene expression

Abstract

This invention includes methods for identifying nucleic acids in a sample of nucleic acids by observing sequence sets present in the nucleic acids of the sample and then identifying those sequences in a nucleic acid sequence database having the sequence sets observed. In a preferred embodiment, a sequence set consists of two primary subsequences and an additional subsequence having determined mutual relationships. The methods include those for observing the sequence sets and those for performing sequence database searches. This invention also includes devices for recognizing in parallel the additional subsequences in a sample of as well as methods for the use of these devices. In a preferred embodiment, the devices include probes bound to a planar surface that recognize additional subsequence by hybridization, and the methods of use include features to improve the specificity and reproducibility of this hybridization.

Description

[0001] This application claims the benefit of provisional application Serial No. 60 / 105,305, filed Dec. 3, 1997, which is hereby incorporated by reference in its entirety.1 FIELD OF THE INVENTION[0003] The field of the invention relates to methods and devices for qualitatively and quantitatively observing nucleic acids in a sample of nucleic acids, and more particularly to methods and devices that recognize the presence of a set of subsequences in each nucleic acid in the sample and identify the nucleic acid from a set of subsequences by reference to a database of sequences likely to be present in the sample.2 BACKGROUND[0004] Modem biology teaches the importance of genes and gene expression to processes of health and disease. New individual genes causing or predisposing to conditions or diseases are now reported almost daily. Additionally, it is commonly understood that observing and measuring the spatial and temporal patterns of gene expression in health and disease will contribut...

AI Technical Summary

Benefits of technology

[0020] It is a broad object of this invention to provide methods and devices for observing and measuring the presence and expression of individual genes or entire genomes that overcome the previously described problems. In particular, the methods and devices of the instant invention make accurate and efficient use of arrayed oligonucleotides (called herein a universal detection array or "UDA") to avoid any requirements for cloning of complex mixtures of sequences into many individual samples of a single sequence, repetitive sequencing of sample components, electrophoretic separations, and so forth.

[0023] In a preferred embodiment, the search query defined by the sequence set is represented as a regular expression, which is used by regular expression search tools to search nucleic acid sequences represented as symbol strings. In an alternate embodiment, an index of subsequences present in the database of nucleotide sequences is first constructed. Second, using this index, sequences are searched for the regular expression representing a sequence set. This alternative embodiment is preferred in the case of repetitive searches of the same sequence database because it increases search efficiency.

[0024] The lengths of the subsequences in a sequence set are chosen in order to obtain adequate resolution and separation of the gene-calling methods. Resolution, defining how precisely a sequence set identifies a nucleic acid, is therefore related to how many sequences in the sequence database have a particular sequence set. Separation defines how accurately and uniquely the observation methods observe a sequence set. In the preferred embodiment, where a UDA of this invention observes the additional sequences in a subsample in parallel, separation improves with decreasing complexity of subsamples. Both these measures are improved by longer subsequences. However, longer subsequences result in increased numbers of subsamples (see below) necessary for adequate coverage. Generally, for nucleic acids derived from expressed human genes, preferred lengths for the subsequences are between 4 and 8.

[0026] In more detail, preferred methods for the first step produce a subsample by digesting the original sample with restriction endonuclease ("RE") enzymes that digest nucleic acids within their recognition sequence and produce single-stranded terminal overhangs. The primary subsequences recognized are therefore the recognition sequences of such REs. Complementary adapters are ligated to these terminal overhangs, either simultaneously with or sequentially to RE digestion. One such adapter preferably has a conjugated biotin (or other capture moiety) to aid in removing improperly digested or undigested fragments from the reaction products. The other adapter preferably has a subsequence which is the recognition site of a restriction endonuclease that digests nucleic acids outside of its recognition site (a Type IIS RE).

[0029] In preferred embodiments, techniques are employed to improve the specificity and strength of probe and fragment hybridization, especially in view of the length of the additional subsequences, which can be as short as 4 nucleotides. One technique employs stacking oligomers that hybridize to the probe adjacent to the hybridized fragments. Energetic base stacking interactions between the hybridized stacking oligomer and fragment improve overall hybridized duplex stability. Another technique employs a ligase enzyme to ligate nicks only in those hybridization structures that are fully and correctly hybridized, followed by a wash step to remove mis-hybridized, and, therefore, un-ligated fragments and stacking oligomers.

Problems solved by technology

Current observation and measurement methods suffer from one or more disadvantages that render them unnecessarily inaccurate, time consuming, labor intensive, or expensive.

Such disadvantages flow from requirements for, e.g., prior knowledge of gene sequences, cloning of complex mixtures of sequences into many individual samples each of a single sequence, repetitive sequencing of sample nucleic acids, electrophoretic separations of nucleic acid fragments, and so forth.

Since a single human cell is estimated to express 10,000-30,000 genes (Liang et al., Science, 257:967-971 (1992)), most of which remain unknown, single probe methods to identify all sequences in a complex sample are prohibitively cumbersome and time consuming.

USA 91:3072-3076 (1994)) also require that samples be arrayed into purified clones, making the methods inappropriate for complex mixtures.

This approach, at best providing only qualitative "fingerprints" of gene expression, suffers from well-known problems, including a high false positive rate, migration of multiple nucleic acid species within a single observed band, and non-quantitative results.

Such methods require sequencing and are statistically limited in their ability to discover rare transcripts.

Other methods for gene and gene-expression measurement, although unrelated to differential display, still have certain disadvantages, such as, e.g., requiring electrophoretic separation.

The length information of the signals of this method is, disadvantageously, observed electrophoretically.

These procedures can be unnecessarily labor intensive, slow, and uneconomical.

Subsequence recognition by hybridization performed in solution, however, often requires electrophoretic separation.

The chips are hybridized to samples of fluorescently tagged target DNAs, and are then imaged to determine to which oligonucleotides hybridization has occurred though some success has been reported with such chips, well-known problems remain, including those of obtaining unambiguous and reliable hybridization signals.

The need for such redundancy of immobilized probes poses serious throughput and cost limitations, especially in view of the 130,000 or so genes possibly expressed in human tissues.

Therefore, these described observational methods for gene-expression are not capable of rapidly, accurately, and economically observing and measuring the presence or expression of selected individual genes or of whole genomes.

Importantly, they have not been able to accurately and economically utilize the potential of arrayed oligonucleotides.

Images