The proper and harmonious expression of a large number of genes is a critical component of normal growth and development and the maintenance of proper health. Disruptions or changes in gene expression are responsible for many diseases. Using traditional methods to assay gene expression, researchers were able to survey a relatively small number of genes at a time. Microarrays allow scientists to analyze expression of many genes in a single experiment quickly and efficiently. A microarray works by exploiting the ability of a given mRNA molecule to bind specifically to, or hybridize to, the DNA template from which it originated.
DNA arrays are commonly used to study gene expression. In this type of study, mRNA is extracted from a sample (for example, blood cells or tumor tissue), converted to complementary DNA (cDNA) and tagged with a fluorescent label. In a typical microarray experiment, cDNA from one sample (sample A) is labeled with a first dye that fluoresces in the red and cDNA from another sample (sample B) is labeled with a different dye that fluoresces in the green. The fluorescent red and green cDNA samples are then applied to a microarray that contains DNA fragments (oligonucleotides) corresponding to thousands of genes. If a DNA sequence probe is present on the microarray and its complement is present in one or both samples, the sequences bind, and a fluorescent signal can be detected at the specific spot on the array, where the DNA sequence probe is located. The signals are generally picked up using a “scanner” which creates a digital image of the array. The red to green fluorescence ratio in each spot reflects the relative expression of a given gene in the two samples. The result of a gene expression experiment is referred to as a gene expression “profile” or “signature”.
This technology, though widely used, is not without its problems. Almost every procedure in the methodology is a potential source of fluctuation leading to a lot of noise in the system as a whole. The major sources of fluctuations to be expected are in mRNA preparation, reverse transcription leading to cDNA of varying lengths, systemic variation in pin geometry, random fluctuations in spot volume, target fixation, slide non-homogeneities due to unequal distribution of the probe, hybridization parameters and non-specific hybridization. Some of the errors mentioned above can be minimized by performing replicates of experiments or by using a flipped dye design.
Biological replicates are arrays that each use RNA samples from different individual organisms, pools of organisms or flasks of cells, but yet compare the same treatments or control/treatment combinations. Technical replicates are arrays that each use the same RNA samples and also the same treatment. Thus, in this setting, the only differences in measurements are due to technical differences in array processing. The rationale for the flipped dye design is that it allows for the estimation and removal of gene specific dye effects. These dye effects have been shown to be reproducible across independent arrays by the use of Control vs. Control arrays. Any deviation from a ratio of 1 in these arrays is due to either dye effect or residual error. However, none of these methods will account accurately for chip manufacturing error.
Therefore, there remains a need for the development of improved microarray technologies, and particularly technologies that allow researchers to control for errors and/or to normalize signals.
The present invention provides reagents and methods that are useful in normalizing and standardizing data from nucleic acid hybridization studies, and particularly from microarray-based hybridizations. The present invention teaches that it is useful to define nucleotide sequences that are “alien” to the sequence population under analysis. Such alien sequences may be included on microarrays and will not hybridize with the nucleic acid population under study. Alternatively or additionally, sequences complementary to the alien sequences may be mixed together with (i.e., “spiked” into) the hybridizing population in order to control for processing and hybridization events.
Use of the alien sequences (and/or their complements) according to the present invention provides a number of advantages. For instance, when an alien sequence is included in a microarray and its complement is not included in the hybridizing sample, the alien sequence may act as a negative control, revealing defects in hybridization conditions that could affect the experimental outcome.
Furthermore, when an alien oligonucleotide is present on an array, its complement may be added to the hybridizing sample, and processed and hybridized together with that sample, as a control for the processing/hybridization steps. If the alien oligonucleotide is present in spots at different locations on the chip, this strategy can also be used to control intra-chip hybridization variations.
Moreover, when the amount of anti-alien spiking nucleic acid (and/or alien oligonucleotide) is known in advance, the degree of anti-alien/alien hybridization may be relied upon to establish the amount of non-alien sequences present in the hybridizing sample based on the relative extent of their hybridization to complementary oligonucleotides. In fact, in some embodiments, multiple alien/anti-alien pairs at different amounts are utilized in order to provide multiple points for comparative quantitation of other nucleic acids. In certain preferred embodiments, the alien sequence probe and the probe detecting the target sequence to be quantified are mixed together in the same spot to allow in situ comparisons. This approach also provides a consistent standard (the fixed amount of alien probe) that can be relied upon to allow inter-slide comparisons and inter-experiment comparisons even when the experiments are carried out using rare samples (i.e., in a situations where the number of experimental replicates that can be performed for control purposes is limited), or over long time spans, etc.
Thus, alien sequence probes and their complements can be used to normalize the data obtained from array hybridizations. For instance, if every spot in an array contains a defined ratio of experimental probes to alien probes, the presence of the alien probes allows the researcher to control for variations between or among spots (e.g., by hybridizing the array with a sample containing anti-alien sequences that are differently labeled from the nucleic acid sequences under study).
Additionally, the presence of alien probes in microarray spots allows researchers to assess the quality and consistency of microarray fabrication and/or printing/spotting techniques. For example, when alien sequences are present in all or a representative collection of spots, the presence or absence of particular spots, overall spot morphology, and slide quality can often be assessed by hybridization (in parallel or simultaneously with experimental hybridization) with an anti-alien nucleic acid. Even random spotting of alien sequences can provide information about the overall integrity or uniformity of a slide. Often, however, it will at least be desirable to include alien sequences in one or more spots containing experimental samples so as to provide a direct assessment of an experimentally relevant spot.
Throughout the specification, several terms are employed, that are defined in the following paragraphs.
Alien gene—As used herein, the term “alien gene” refers to a nucleotide molecule comprised of at least two concatermerized alien sequences. The gene may contain multiple copies of a single alien sequence, or alternatively may contain a plurality of different alien sequences. An alien gene may be single or double stranded, and may contain or be associated with a promoter or other control sequence that will direct the production of a template of either strand of the gene. In particular, as will be clear from discussions herein, in some embodiments of the invention it will be desirable to produce an alien gene transcript that is an alien sequence, whereas in other embodiments it will be desirable to produce an alien gene transcript that is complementary to an alien sequence.
Alien sequence—A nucleotide sequence is considered “alien” to a particular source or collection of nucleic acids if it does not hybridize with nucleic acids in the source or collection. For example, if the source or collection is mRNA from normal kidney cells, an oligonucleotide will have a sequence that is “alien” to the mRNA if its complement is not present in the mRNA. Conversely, if the source or collection is cDNA from the same cells, then an oligonucleotide will have a sequence that is “alien” to the cDNA if its complement is not present in the cDNA. In certain preferred embodiments of the invention, the source or collection comprises expressed nucleic acids (e.g., mRNA or cDNA) of a target organism (e.g., mouse, dog, human, etc), tissue (e.g., breast, lung, colon, liver, brain, kidney, etc), or cell type (e.g., before or after exposure to a particular stimulus or treatment). Alternatively or additionally, the source or collection may preferably be a plurality of nucleic acids to be hybridized to an array.
Hybridizing sample—The terms “hybridizing sample” and “hybridizing mixture” are used herein interchangeably. They refer to the nucleic acid sample being or intended to be hybridized to a microarray. Those of ordinary skill in the art will appreciate that the hybridizing sample may contain DNA, RNA, or both, but most commonly contains cDNA. Those of ordinary skill in the art will further appreciate that the hybridizing sample typically contains nucleic acids whose hybridization with probes on an array is detectable. For example, in many embodiments, the hybridizing sample comprises or consists of detectably labeled nucleic acids.
Detectably labeled—The terms “labeled”, “detectably labeled” and “labeled with a detectable agent” are used herein interchangeably. They are used to specify that a nucleic acid molecule or individual nucleic acid segments from a sample can be detected and/or visualized following binding (i.e., hybridization) to probes immobilized on an array. Nucleic acid samples to be used in the methods of the invention may be detectably labeled before the hybridization reaction or a detectable label may be selected that binds to the hybridization product. Preferably, the detectable agent or moiety is selected such that it generates a signal which can be measured and whose intensity is related to the amount of hybridized nucleic acids. Preferably, the detectable agent or moiety is also selected such that it generates a localized signal, thereby allowing spatial resolution of the signal from each spot on the array. Methods for labeling nucleic acid molecules are well known in the art (see below for a more detailed description of such methods). Labeled nucleic acids can be prepared by incorporation of or conjugation to a label, that is directly or indirectly detectable by spectroscopic, photochemical, biochemical, immunochemical, radiochemical, electrical, optical, or chemical means. Suitable detectable agents include, but are not limited to: various ligands, radionuclides, fluorescent dyes, chemiluminescent agents, microparticles, enzymes, colorimetric labels, magnetic labels, and haptens. Detectable moieties can also be biological molecules such as molecular beacons and aptamer beacons.
Fluorescent Label—The terms “fluorophore”, “fluorescent moiety”, “fluorescent label”, “fluorescent dye” and “fluorescent labeling moiety” are used herein interchangeably. They refer to a molecule which, in solution and upon excitation with light of appropriate wavelength, emits light back. Numerous fluorescent dyes of a wide variety of structures and characteristics are suitable for use in the practice of this invention. Similarly, methods and materials are known for fluorescently labeling nucleic acids (see, for example, R. P. Haugland, “Molecular Probes: Handbook of Fluorescent Probes and Research Chemicals 1992-1994”, 5th Ed., 1994, Molecular Probes, Inc.). In choosing a fluorophore, it is generally preferred that the fluorescent molecule absorbs light and emits fluorescence with high efficiency (i.e., it has a high molar absorption coefficient and a high fluorescence quantum yield, respectively) and is photostable (i.e., it does not undergo significant degradation upon light excitation within the time necessary to perform the array-based hybridization). Suitable fluorescent labels for use in the practice of the methods of the invention include, for example, Cy-3™, Cy-5™, Texas red, FITC, Spectrum Red™, Spectrum Green™, Alexa-488, phycoerythrin, rhodamine, fluorescein, fluorescein isothiocyanine, carbocyanine, merocyanine, styryl dye, oxonol dye, BODIPY dye, and equivalents, analogues or derivatives of these molecules.
Microarray—The terms “microarray”, “chip” and “biochip” are used herein interchangeably. They refer to an arrangement, on a substrate surface, of multiple nucleic acid molecules of known or unknown sequences. These nucleic acid molecules are immobilized to discrete “spots” (i.e., defined locations or assigned positions) on the substrate surface. A discrete spot may contain a single nucleic acid molecule or a mixture of different nucleic acid molecules. Spots on an array may be arranged on the substrate surface at different densities. In general, microarrays with probe pitch smaller than 500 μm (i.e., density larger than 400 probes per cm2) are referred to as high density microarrays, otherwise, they are called low density microarrrays. Arrays come as two-dimensional probe matrices (or supports), which can be solid or porous, planar or non-planar, unitary or distributed. The term “micro-array” more specifically refers to an array that is miniaturized so as to require microscopic examination for visual evaluation. Arrays used in the methods of the invention are preferably microarrays. The present invention provides microarrays in which at least one spot contains an alien oligonucleotide. Other types of microarrays and sets of microarrays provided by the invention are described below.
Oligonucleotide—As used herein, the term “oligonucleotide”, refers to usually short strings of DNA or RNA to be used as hybridizing probes or nucleic acid molecule array elements. These short stretches of sequence are often synthesized chemically. As will be appreciated by those skilled in the art, the length of the oligonucleotide (i.e., the number of nucleotides) can vary widely, often depending on its intended function or use. Generally, oligonucleotides of at least 6 to 8 bases are used, with oligonucleotides ranging from about 10 to 500 bases being preferred, with from about 20 to 200 bases being particularly preferred, and 40 to 100 bases being especially preferred. Longer oligonucleotide probes are usually preferred in array-based hybridization reactions, since higher stringency hybridization and wash conditions can be used, which decreases or eliminates non-specific hybridization.
Probe—For the purposes of the present invention, a “probe” is a nucleic acid, often an oligonucleotide that is, or is intended to be, attached to a solid support in an array. Preferably, the probes that comprise a microarray or biochip are of a defined length and similarity. This allows for similar hybridization characteristics. As is well known to those skilled in the art, for the hybridization characteristics to be similar across a wide range of oligonucleotides, it is typically required that the probes on the array be of the substantially same length, have a similar percentage of Guanine to Cytosine content and lack any extensive runs of poly A, poly G, poly C, or poly T tracts. The goal of controlling these parameters is to produce probes that have similar melting and hybridization temperatures. Additionally, these probes should, preferably, lack length complementary regions and not form hairpin structures.
Target—The term “target” refers to nucleic acids intended to be hybridized (or bound) to probes immobilized on microarrays by sequence complementarity. As is well-known in the art, target nucleic acids may be obtained from a wide variety of organisms, tissues or cells. Methods and techniques for the extraction, manipulation and preparation of nucleic acids for hybridization reactions are well-known in the art (see, for example, J. Sambrook et al., “Molecular Cloning: A Laboratory Manual”, 1989, 2nd Ed., Cold Spring Harbour Laboratory Press: New York, N.Y.; “PCR Protocols: A Guide to Methods and Applications”, 1990, M. A. Innis (Ed.), Academic Press: New York, N.Y.; P. Tijssen “Hybridization with Nucleic Acid Probes—Laboratory Techniques in Biochemistry and Molecular Biology (Parts I and II)”, 1993, Elsevier Science; “PCR Strategies”, 1995, M. A. Innis (Ed.), Academic Press: New York, N.Y.; and “Short Protocols in Molecular Biology”, 2002, F. M. Ausubel (Ed.), 5th Ed., John Wiley & Sons).
Hybridization—The term “hybridization” has herein its art understood meaning and refers to the binding of two single stranded nucleic acids via complementary base pairing. A hybridization reaction is called specific when a nucleic acid molecule preferentially binds, duplexes, or hybridizes to a particular nucleic acid sequence under stringent conditions (e.g., in the presence of competitor nucleic acids with a lower degree of complementarity to the hybridizing strand).
High stringency conditions—For microarray-based hybridization, standard “high stringency conditions” are defined for solution phase hybridization as aqueous hybridization (i.e., free of formamide) in 6×SSC (where 20×SSC contains 3.0 M NaCl and 0.3 M sodium citrate), 1% SDS at 65° C. for at least 8 hours, followed by one or more washes in 0.2×SSC, 0.1% SDS at 65° C. “Moderate stringency conditions” are defined for solution phase hybridization as aqueous hybridization (i.e., free of formamide) in 6×SSC, 1% SDS at 65° C. for at least 8 hours, followed by one or more washes in 2×SSC, 0.1% SDS at room temperature.
The present invention provides reagents and methods that are useful in normalizing and standardizing data from nucleic acid hybridization studies, and particularly from microarray hybridizations. The present invention teaches that it is useful to define nucleotide sequences that are “alien” to the sequence population under analysis.
In particular, the use of such alien oligonucleotide sequences in micro-array based hybridization is herein described to be able to serve several distinct control purposes. For example, (1) when spotted on microarrays, alien sequences can serve as negative controls during the course of hybridization experimentation to assess the stringency (i.e., specificity) of target-to-probe hybridization. (2) Alien oligonucleotides spotted on micro-arrays, in combination with their complementary sequences used as in-spike controls can enable the experimenter to gauge the robustness of both the overall target labeling and hybridization efficiency. (3) When alien probe sequences are present within each sub-array on the biochip, they allow regional (intra-slide) effects of hybridization to be ascertained. (4) Alien oligonucleotides can also be used as in-spot controls and act as references so that inter-slide differences can be measured relative to a consistent control. (5) Detectably labeled alien sequences can be used to normalize the signal intensities of the samples under analysis on a per spot basis. Also, (6) in situ alien sequences may also be used to quality control the DNA microarray printing process.
In a first aspect, the present invention provides methods of identifying nucleotide sequences that are alien to a selected population.
Generating or Selecting Alien Sequences
As mentioned above, a nucleotide sequence is considered “alien” to a particular source or collection of nucleic acids if it does not hybridize with nucleic acids in the source or collection. For example, if the source or collection is mRNA or cDNA, then an oligonucleotide has a sequence that is “alien” to the mRNA or cDNA if its complement is not present in the mRNA or cDNA. Preferred alien oligonucleotides of the invention have complementary sequences that are maximally dissimilar from (i.e., non-identical to) those present in the source or collection.
When comparing polynucleotide sequences, two sequences are said to be “identical” if the sequence of nucleotides in the two sequences is the same when aligned for maximum correspondence. Comparisons between two sequences are typically performed by comparing the sequences over a comparison window to identify and compare local regions of sequence similarity. A “comparison window” refers to a segment of at least about 20 contiguous positions, usually 30 to about 75, or 40 to about 50, in which a sequence may be compared to a reference sequence of the same number of contiguous positions after the two sequences are optimally aligned.
Any of a wide variety of selection methods, systems or strategies that lead to the generation of oligonucleotides alien to a source or collection of nucleic acids can be used in the practice of the present invention. Such methods may, for example, be based on the use of an algorithm.
The present invention provides such an algorithm, in which the underlying logic is that of “partially reversing” the mathematical logic of the standard Hidden Markov Model. Such standard models are used to generate model sequences of DNA, RNA, proteins as well as other biological molecules, based on the statistics of known real (i.e., naturally occurring) sequences. Model sequences are generated based on sets of sequence symbol occurrences. For example, given the measured nearest neighbor frequencies (i.e., how often one nucleotide follows another) one then draws and outputs “randomly” from that set proportional to those frequencies. A very wide range of sequences statistics can be employed, from the simplest, the occurrence frequencies of the individual symbols, through all possible nearest neighbor frequencies to arbitrary spaced sequences frequencies.
A first approach used by the Applicants with the goal of generating “alien” or maximally dissimilar sequences from known real sequences was to perform a complete “reversal” of the statistics (i.e., to invert the sets of occurrence probability from most likely to least likely). However, when this strategy was tested over a very large set of sequences statistics, it did not work.
What did work in generating model sequences which are maximally dissimilar from those employed to obtain the sequence statistics, was to use a Markov process, in which, at an adjustable frequency, one draws from the measured real statistics but inversely proportional to those frequencies (or probability distributions). The sequence generated by this process contains, scattered throughout its length, intermittent highly improbable sequence patterns or subsequences. The frequency with which one switches between draws from the measured real sequence occurrence frequencies proportional to those frequencies and inversely proportional to those frequencies and inversely, ranges from one in five to one in ten. The selection of this ratio is partly a function of which sets of sequence statistics are used.
In the generation of maximally dissimilar DNA or mRNA complement sequences for microarray controls, preferably in the length range of 50 to 70 nucleotides, codon occurrence and codon boundary di-nucleotide frequencies were used for a range of inverse proportional inverse probability draws on these two statistics. This process was then followed by two filters, including: (1) a full genome sequence similarity search of all known or predicted protein coding regions, and (2) the calculation of TMs for all possible mRNA annealings for those with any sequence similarities above 60% identity and/or with matching runs longer than 18 nucleotides. All generated sequences with predicted annealing temperature above 37° C. or runs of twenty identities were eliminated. The TMs (i.e., midpoint disassociation temperatures) were calculated using multiple public domain software which included nucleotide stacking energies. This resulted in approximately one predicted “alien” or non-mRNA annealing oligo for every 5,000 genome coding regions in the higher animal and plant eukaryotic genomes currently known. Sets of these alien sequences were then synthesized and placed on “long oligo” microarray chips and physically tested for their annealing to real mRNA and/or cDNA samples. With rare exceptions (of one in ten), no detectable annealing was observed under standard experimental conditions for 70 mer oligo array chips for 21,000 mouse genes. These alien sequences then define a set of negative controls.
A set of microarray “alien positive controls” was then generated from the above set of alien oligo negative control sequences using the following algorithm. First all possible set of three to five sequentially concatenated alien oligos as defined above were generated in silico. These were investigated for the incidental creation of a sequence crossing the boundary between the concatenated alien oligos that have a significant match or predicted annealing TM above 37° C. to any of the non-alien oligos on the micro-array targeted. Only those that showed no such matches or higher TMs were selected. These oligos were then physically synthesized as “positive alien gene” controls and tested for their ability to only anneal to their complementary alien oligos.
In light of the inventive results described herein, those of ordinary skill in the art will appreciate that other algorithms may be employed or developed, for example, to include filter steps that, for example, verify the degree of “alien”ness of the selected sequence by comparing the generated oligonucleotide sequences to the organism's genome (if available) or cDNA by using any of a large number of sequence comparison programs.
A variety of methods for determining relationships between two or more sequences (e.g., identity, similarity and/or homology) are available, and well known in the art. The methods include manual alignment, computer assisted sequence alignment and combinations thereof. A number of algorithms (which are generally computer implemented) for performing sequence alignment are widely available, or can be produced by one of skill in the art. These methods include, e.g., the local homology algorithm of Smith and Waterman (Adv. Appl. Math., 1981, 2: 482); the homology alignment algorithm of Needleman and Wunsch (J. Mol. Biol., 1970, 48: 443); the search for similarity method of Pearson and Lipman (Proc. Natl. Acad. Sci. (USA), 1988, 85: 2444); and/or by computerized implementations of these algorithms (e.g., GAP, BESTFIT, FASTA, and TFASTA in the Wisconsin Genetics Software Package Release 7.0, Genetics Computer Group, 575 Science Dr., Madison, Wis.).
For example, a software for performing sequence identity (and sequence similarity) analysis using the BLAST algorithm is described in Altschul et al., J. Mol. Biol., 1990, 215: 403-410. This software is publicly available, e.g., through the National Center for Biotechnology Information on the World Wide Web at ncbi.nlm.nih.gov. This algorithm involves first identifying high scoring sequence pairs (HSPs) by identifying short words of length W in the query sequence, which either match or satisfy some positive-valued threshold score T when aligned with a word of the same length in a database sequence. T is referred to as the neighborhood word score threshold. These initial neighborhood word hits act as seeds for initiating searches to find longer HSPs containing them. The word hits are then extended in both directions along each sequence for as far as the cumulative alignment score can be increased. Cumulative scores are calculated using, for nucleotide sequences, the parameters M (reward score for a pair of matching residues; always>0) and N (penalty score for mismatching residues; always<0). For amino acid sequences, a scoring matrix is used to calculate the cumulative score. Extensions of the word hits in each direction are halted when: the cumulative alignment score falls off by the quantity X from its maximum achieved value; the cumulative score goes to zero or below, due to the accumulation of one or more negative-scoring residue alignments; or the end of either sequence is reached. The BLAST algorithm parameters W, T, and X determine the sensitivity and speed of the alignment. The BLASTN program (for nucleotide sequences) uses as defaults a wordlength (W) of 11, an expectation (E) of 10, a cutoff of 100, M=5, N=−4, and a comparison of both strands. For amino acid sequences, the BLASTP (BLAST Protein) program uses as defaults a wordlength (W) of 3, an expectation (E) of 10, and the BLOSUM62 scoring matrix (see, Henikoff & Henikoff, Proc. Natl. Acad. Sci. USA, 1989, 89:10915).
Additionally, the BLAST algorithm performs a statistical analysis of the similarity between two sequences (see, e.g., Karlin & Altschul, Proc. Natl. Acad. Sci. USA, 1993, 90: 5873-5787). One measure of similarity provided by the BLAST algorithm is the smallest sum probability (P(N)), which provides an indication of the probability by which a match between two nucleotide or amino acid sequences would occur by chance. For example, a nucleic acid is considered similar to a reference sequence if the smallest sum probability in a comparison of the test nucleic acid to the reference nucleic acid is less than about 0.1, or less than about 0.01, and or even less than about 0.001.
Another example of a useful sequence alignment algorithm is PILEUP. PILEUP creates a multiple sequence alignment from a group of related sequences using progressive, pairwise alignments. It can also plot a tree showing the clustering relationships used to create the alignment. PILEUP uses a simplification of the progressive alignment method of Feng & Doolittle (J. Mol. Evol. 1987, 35: 351-360). The method used is similar to the method described by Higgins & Sharp (CABIOS, 1989, 5: 151-153). The program can align, e.g., up to 300 sequences of a maximum length of 5,000 letters. The multiple alignment procedure begins with the pairwise alignment of the two most similar sequences, producing a cluster of two aligned sequences. This cluster can then be aligned to the next most related sequence or cluster of aligned sequences. Two clusters of sequences can be aligned by a simple extension of the pairwise alignment of two individual sequences. The final alignment is achieved by a series of progressive, pairwise alignments. The program can also be used to plot a dendogram or tree representation of clustering relationships. The program is run by designating specific sequences and their nucleotide coordinates for regions of sequence comparison.
An additional example of an algorithm that is suitable for multiple DNA sequence alignments is the CLUSTALW program (J. D. Thompson et al., Nucl. Acids. Res. 1994, 22: 4673-4680). CLUSTALW performs multiple pairwise comparisons between groups of sequences and assembles them into a multiple alignment based on homology. Gap open and Gap extension penalties can be, e.g., 10 and 0.05 respectively.
An algorithm for the selection of alien sequences may also include filter steps that check for TM, % GC content, low-complexity regions and self hybridization. A large number of software programs (including those described above) are available and can be used to carry out these steps.
Alien Oligonucleotide Preparation
In another aspect, the present invention provides isolated oligonucleotides or nucleic acids that are alien to a given source or collection of nucleic acids. As will be appreciated by one skilled in the art, alien oligonucleotides may be of different lengths, depending on their intended use (as negative control, normalization and/or quantification tool or as in-spike control). For example, alien oligonucleotides may contain a single alien sequence. Alternatively, an alien oligonucleotide may contain at least two alien sequences linked to one another. Inventive oligonucleotides provided herein also include those polynucleotides that contain anti-alien sequences. For example, as described herein, it will often be desirable to prepare anti-alien sequences for use in hybridization reactions. In some embodiments, such sequences are prepared by polymerization directed by an alien gene.
Alien and anti-alien oligonucleotides of the invention may be prepared by any of a variety of chemical techniques well-known in the art, including, for example, chemical synthesis and polymerization based on a template (see, for example, S. A. Narang et al., Meth. Enzymol. 1979, 68: 90-98; E. L. Brown et al., Meth. Enzymol. 1979, 68: 109-151; E. S. Belousov et al., Nucleic Acids Res. 1997, 25: 3440-3444; D. Guschin et al., Anal. Biochem. 1997, 250: 203-211; M. J. Blommers et al., Biochemistry, 1994, 33: 7886-7896; and K. Frenkel et al., Free Radic. Biol. Med. 1995, 19: 373-380; see also for example, U.S. Pat. No. 4,458,066).
For example, oligonucleotides may be prepared using an automated, solid-phase procedure based on the phosphoramidite approach. In such a method, each nucleotide is individually added to the 5′-end of the growing oligonucleotide chain, which is attached at the 3′-end to a solid support. The added nucleotides are in the form of trivalent 3′-phosphoramidites that are protected from polymerization by a dimethoxytrityl (or DMT) group at the 5′-position. After base base-induced phosphoramidite coupling, mild oxidation to give a pentavalent phosphotriester intermediate and DMT removal provides a new site for oligonucleotide elongation. The oligonucleotides are then cleaved off the solid support, and the phosphodiester and exocyclic amino groups are deprotected with ammonium hydroxide. These syntheses may be performed on commercial oligo synthesizers such as the Perkin Elmer/Applied Biosystems Division DNA synthesizer. Such a synthesis is described in Example 2.
Oligonucleotides can also be custom made and ordered from a variety of commercial sources well-known in the art, including, for example, the Midland Certified Reagent Company (mcrc@oligos.com), The Great American Gene Company (available on the World Wide Web at genco.com), ExpressGen Inc. (available on the World Wide Web at expressgen.com), Operon Technologies Inc. (Alameda, Calif.) and many others.
Purification of oligonucleotides of the invention, where necessary, may be carried out by any of a variety of methods well-known in the art. Purification of oligonucleotides is typically performed by either by native acrylamide gel electrophoresis or by anion-exchange HPLC as described, for example, by Pearson and Regnier (J. Chrom. 1983, 255: 137-149). The sequence of the synthetic oligonucleotides can be verified using the chemical degradation method of Maxam and Gilbert (in Grossman and Moldave (Eds.), Academic Press, New York, Methods in Enzymology, 1980, 65: 499-560).
The present invention provides nucleic acid arrays in which at least one spot contains an alien oligonucleotide. More specifically, inventive nucleic acids arrays comprise a solid support, and a plurality of nucleic acid probes attached to the solid support at discrete locations, wherein at least one the probes is an alien probe in that it has a sequence that is alien to a hybridizing mixture to be hybridized to the array.
Microarrays generally have sample spot sizes of less than 200 μm diameter, and generally contain thousands of spots per slide. For gene-expression analysis, each microarray preferably contains at least about 1,000, 5,000, 10,000, 50,000, 100,000, or 500,000 spots. The probes are printed (or attached) to the surface of the substrate, and the number of probes per unit area of the print surface is called the print density. The print surface corresponds to that area of the substrate on which the individual probes are printed, plus the surface area between the individual probes. If there are two or more groupings of a substantial number of probes on the substrate surface separated by surface area in which few or no probes are printed, the print surface includes the surface area between probes of a group but not the surface area of the substrate between groupings. For gene expression analysis, the print density is preferably high so that a large number of probes can fit on the substrate. Preferably, the print density is at least about 200, 500, 1,000, 5,000, 10,000, 20,000, or 40,000 probes per cm2.
There are two standard types of DNA microarray technology in terms of the nature of the arrayed DNA sequence. In the first format, probe cDNA sequences (typically 500 to 5,000 bases long) are immobilized to a solid surface and exposed to a plurality of targets either separately or in a mixture. In the second format, oligonucleotides (typically 20-80-mer oligos) or peptide nucleic acid (PNA) probes are synthesized either in situ (i.e., directly on-chip) or by conventional synthesis followed by on-chip attachment, and then exposed to labeled samples of nucleic acids. In the present invention, microarrays of the second type are preferably used.
In the practice of the methods of the invention, investigators may either buy commercially available arrays (for example, from Affymetrix Inc. (Santa Clara, Calif.), Illumina, Inc. (San Diego, Calif.), Spectral Genomics, Inc. (Houston, Tex.), and Vysis Corporation (Downers Grove, Ill.)), or generate their own starting microarrays (i.e., arrays to which at least one alien oligonucleotide is to be spotted). Methods of making and using arrays are well known in the art (see, for example, S. Kern and G. M. Hampton, Biotechniques, 1997, 23:120-124; M. Schummer et al., Biotechniques, 1997, 23:1087-1092; S. Solinas-Toldo et al., Genes, Chromosomes & Cancer, 1997, 20: 399-407; M. Johnston, Curr. Biol. 1998, 8: R171-R174; D. D. Bowtell, Nature Gen. 1999, Supp. 21:25-32; D. J. Lockhart and E. A. Winzeler, Nature, 2000, 405: 827-836; M. Cuzin, Transfus. Clin. Biol. 2001, 8:291-296; M. Gabig and G. Wegrzyn, Acta Biochim. Pol. 2001, 48: 615-622; and V. G. Cheung et al., Nature, 2001, 40: 953-958).
Arrays comprise a plurality of probes immobilized to discrete spots (i.e., defined locations or assigned positions) on a substrate surface. Substrate surfaces for use in the present invention can be made of any of a variety of rigid, semi-rigid or flexible materials that allow direct or indirect attachment (i.e., immobilization) of probes (including alien oligonucleotides) to the substrate surface. Suitable materials include, but are not limited to: cellulose (see, for example, U.S. Pat. No. 5,068,269), cellulose acetate (see, for example, U.S. Pat. No. 6,048,457), nitrocellulose, glass (see, for example, U.S. Pat. No. 5,843,767), quartz or other crystalline substrates such as gallium arsenide, silicones (see, for example, U.S. Pat. No. 6,096,817), various plastics and plastic copolymers (see, for example, U.S. Pat. Nos. 4,355,153; 4,652,613; and 6,024,872), various membranes and gels (see, for example, U.S. Pat. No. 5,795,557), and paramagnetic or supramagnetic microparticles (see, for example, U.S. Pat. No. 5,939,261). When fluorescence is to be detected, arrays comprising cyclo-olefin polymers may preferably be used (see, for example, U.S. Pat. No. 6,063,338).
The presence of reactive functional chemical groups (such as, for example, hydroxyl, carboxyl, amino groups and the like) on the material can be exploited to directly or indirectly attach probes including alien oligonucleotide sequences to the substrate surface. Methods of attachment (or immobilization) of oligonucleotides on substrate supports have been described and are well-known to those skilled in the art (see, for example, U. Maskos and E. M. Southern, Nucleic Acids Res. 1992, 20: 1679-1684; R. S. Matson et al., Anal. Biochem. 1995, 224; 110-116; R. J. Lipshutz et al., Nat. Genet. 1999, 21: 20-24; Y. H. Rogers et al., Anal. Biochem. 1999, 266: 23-30; M. A. Podyminogin et al., Nucleic Acids Res. 2001, 29: 5090-5098; Y. Belosludtsev et al., Anal. Biochem. 2001, 292: 250-256).
Methods of preparation of oligonucleotide-based arrays that can be used to attach probes to surface support of microarrays include: synthesis in situ using a combination of photolithography and oligonucleotide chemistry (see, for example, A. C. Pease et al., Proc. Natl. Acad. Sci. USA 1994, 91: 5022-5026; D. J. Lockhart et al., Nature Biotech. 1996, 14: 1675-1680; S. Singh-Gasson et al., Nat. Biotechn. 1999, 17: 974-978; M. C. Pirrung et al., Org. Lett. 2001, 3: 1105-1108; G. H. McGall et al., Methods Mol. Biol. 2001, 170; 71-101; A. D. Barone et al., Nucleosides Nucleotides Nucleic Acids, 2001, 20: 525-531; J. H. Butler et al., J. Am. Chem. Soc. 2001, 123: 8887-8894; E. F. Nuwaysir et al., Genome Res. 2002, 12: 1749-1755). The chemistry for light-directed oligonucleotide synthesis using photo labile protected 2′-deoxynucleoside phosphoramides has been developed by Affymetrix Inc. (Santa Clara, Calif.) and is well known in the art (see, for example, U.S. Pat. Nos. 5,424,186 and 6,582,908).
Alternatively or additionally, oligo probes may first be prepared or print-ready oligonucleotide (e.g., 60-70 mers) sets that are commercially available for human, mouse and other organism (see, for example, http://www.cgen.com, http://www.operon.com) may be obtained and then attached to the array surface. Similarly, alien oligonucleotides are first synthesized and then immobilized on the surface of a microarray.
In these cases, the preparation of microarrays is preferably carried out by high-speed printing robotics. The established robotic spotting technique (U.S. Pat. No. 5,807,522) uses a specially designed mechanical robot, which produces a probe spot on the microarray by dipping a pin head into a fluid containing an off-line synthesized nucleic acid molecule and then spotting it onto the slide at a pre-determined position. Washing and drying of the pins are required prior to the spotting of a different probe in the microarray. In current designs of such robotic systems, the spotting pin, and/or the stage carrying the microarray substrates move along the XYZ axes in coordination to deposit samples at controlled positions of the substrates.
In addition to the established quill-pin spotting technologies, there are a number of microarray fabrication techniques that are being developed. These include the inkjet technology and capillary spotting.
Example 2 describes the printing of alien oligonucleotides to the surface of oligo slides (CodeLink, Amersham Biosciences, Piscataway, N.J.), which also contain human and mouse positive control spots.
As mentioned above, microarrays provided by the present invention are arrays containing a plurality of oligo probes and in which at least one spot contains an alien oligonucleotide. In certain preferred embodiments, an alien oligonucleotide is printed at more than one spot on the array. For example, an inventive microarray may contain, in addition to a plurality of oligo probes, a representative collection of spots containing the same or different concentrations of the alien oligonucleotide. Alternatively, all the spots on an inventive microarray may contain the same or different concentrations of the alien oligonucleotide.
In other embodiments, an inventive microarray contains at least two different alien oligonucleotides. These alien oligonucleotides may be spotted randomly throughout the whole array or they may be present in specific areas of the substrate surface, for example, forming probe elements (i.e., sub-arrays) containing only one type of alien oligonucleotide.
In still other embodiments, an inventive microarray contains alien oligonucleotides of different sizes. For example, an inventive microarray may contain a first oligonucleotide comprising a single alien sequence and a second oligonucleotide comprising at least two different alien sequences. The presence of both types of alien oligonucleotides on the microarray may, for example, allow two different types of controls to be performed.
The present invention also provides sets of microarrays that all contain identical probe elements (i.e., defined sets of spots) except for one microarray (or part of one microarray), which contains no alien oligonucleotide and another microarray (or part of a microarray) that contains the same probe elements but with fixed amount(s) of alien oligonucleotide.
Labeling of Nucleic Acid Molecules
In certain embodiments, nucleic acid molecules of the hybridizing mixture are labeled with a detectable agent before hybridization. In other embodiments, complementary sequences of alien oligonucleotides (i.e., anti-alien oligonucleotides), which are added to the hybridization sample before hybridization, are also labeled. In both cases, the role of a detectable agent is to facilitate detection and to allow visualization of hybridized nucleic acids. Preferably, the detectable agent is selected such that it generates a signal which can be measured and whose intensity is related to the amount of labeled nucleic acids present in the sample being analyzed. The detectable agent is also preferably selected such that it generates a localized signal, thereby allowing spatial resolution of the signal from each spot on the array.
The association between the nucleic acid molecule and detectable agent can be covalent or non-covalent. Labeled nucleic acids can be prepared by incorporation of or conjugation to a detectable moiety. Labels can be attached directly to the nucleic acid or indirectly through a linker. Linkers or spacer arms of various lengths are known in the art and are commercially available, and can be selected to reduce steric hindrance, or to confer other useful or desired properties to the resulting labeled molecules (see, for example, E. S. Mansfield et al., Mol. Cell. Probes, 1995, 9: 145-156).
Many methods for labeling nucleic acid molecules are well-known in the art. For a review of labeling protocols, label detection techniques and recent developments in the field, see, for example, L. J. Kricka, Ann. Clin. Biochem. 2002, 39: 114-129; R. P. van Gijlswijk et al., Expert Rev. Mol. Diagn. 2001, 1: 81-91; and S. Joos et al., J. Biotechnol. 1994, 35: 135-153. Standard nucleic acid labeling methods include: incorporation of radioactive agents, direct attachment of fluorescent dyes or of enzymes; chemical modifications of nucleic acids making them detectable immunochemically or by other affinity reactions; and enzyme-mediated labeling methods, such as random priming, nick translation, PCR and tailing with terminal transferase. More recently developed nucleic acid labeling systems include, but are not limited to: ULS (Universal Linkage System; see, for example, R. J. Heetebrij et al., Cytogenet. Cell. Genet. 1999, 87: 47-52), photoreactive azido derivatives (see, for example, C. Neves et al., Bioconjugate Chem. 2000, 11: 51-55), and alkylating agents (see, for example, M. G. Sebestyen et al., Nat. Biotechnol. 1998, 16: 568-576).
Any of a wide variety of detectable agents can be used in the practice of the present invention. Suitable detectable agents include, but are not limited to: various ligands, radionuclides (such as, for example, 32P, 35S, 3H, 14C, 125I, 131I and the like); fluorescent dyes (for specific exemplary fluorescent dyes, see below); chemiluminescent agents (such as, for example, acridinium esters, stabilized dioxetanes and the like); microparticles (such as, for example, quantum dots, nanocrystals, phosphors and the like); enzymes (such as, for example, those used in an ELISA, e.g., horseradish peroxidase, beta-galactosidase, luciferase, alkaline phosphatase); colorimetric labels (such as, for example, dyes, colloidal gold and the like); magnetic labels (such as, for example, Dynabeads™); and biotin, dioxigenin or other haptens and proteins for which antisera or monoclonal antibodies are available.
In certain preferred embodiments, nucleic acid molecules (or anti-alien oligonucleotides) are fluorescently labeled. Numerous known fluorescent labeling moieties of a wide variety of chemical structures and physical characteristics are suitable for use in the practice of this invention. Suitable fluorescent dyes include, but are not limited to: Cy-3™, Cy-5™, Texas red, FITC, Alexa-488, phycoerythrin, rhodamine, fluorescein, fluorescein isothiocyanine, carbocyanine, merocyanine, styryl dye, oxonol dye, BODIPY dye (i.e., boron dipyrromethene difluoride fluorophore), and equivalents, analogues, derivatives or combinations of these molecules. Similarly, methods and materials are known for linking or incorporating fluorescent dyes to biomolecules such as nucleic acids (see, for example, R. P. Haugland, “Molecular Probes: Handbook of Fluorescent Probes and Research Chemicals 1992-1994”, 5th Ed., 1994, Molecular Probes, Inc.). Fluorescent labeling dyes as well as labeling kits are commercially available from, for example, Amersham Biosciences, Inc. (Piscataway, N.J.), Molecular Probes, Inc. (Eugene, Oreg.), and New England Biolabs, Inc. (Berverly, Mass.).
Favorable properties of fluorescent labeling agents to be used in the practice of the invention include high molar absorption coefficient, high fluorescence quantum yield, and photostability. Preferred labeling fluorophores exhibit absorption and emission wavelengths in the visible (i.e., between 400 and 750 nm) rather than in the ultraviolet range of the spectrum (i.e., lower than 400 nm).
Hybridization products may also be detected using one of the many variations of the biotin-avidin technique system, which that are well known in the art. Biotin labeling kits are commercially available, for example, from Roche Applied Science (Indianapolis, Ind.) and Perkin Elmer (Boston, Mass.).
Detectable moieties can also be biological molecules such as molecular beacons and aptamer beacons. Molecular beacons are nucleic acid molecules carrying a fluorophore and a non-fluorescent quencher on their 5′ and 3′ ends. In the absence of a complementary nucleic acid strand, the molecular beacon adopts a stem-loop (or hairpin) conformation, in which the fluorophore and quencher are in close proximity to each other, causing the fluorescence of the fluorophore to be efficiently quenched by FRET (i.e., fluorescence resonance energy transfer). Binding of a complementary sequence to the molecular beacon results in the opening of the stem-loop structure, which increases the physical distance between the fluorophore and quencher thus reducing the FRET efficiency and allowing emission of a fluorescence signal. The use of molecular beacons as detectable moieties is well-known in the art (see, for example, D. L. Sokol et al., Proc. Natl. Acad. Sci. USA, 1998, 95: 11538-11543; and U.S. Pat. Nos. 6,277,581 and 6,235,504). Aptamer beacons are similar to molecular beacons except that they can adopt two or more conformations (see, for example, O. K. Kaboev et al., Nucleic Acids Res. 2000, 28: E94; R. Yamamoto et al., Genes Cells, 2000, 5: 389-396; N. Hamaguchi et al., Anal. Biochem. 2001, 294: 126-131; S. K. Poddar and C. T. Le, Mol. Cell. Probes, 2001, 15: 161-167).
Multiple independent or interacting labels can also be incorporated into the nucleic acids. For example, both a fluorophore and a moiety that in proximity thereto acts to quench fluorescence can be included to report specific hybridization through release of fluorescence quenching (see, Tyagi et al., Nature Biotechnol. 1996, 14: 303-308; Tyagi et al., Nature Biotechnol. 1998, 16: 49-53; Kostrikis et al., Science, 1998, 279: 1228-1229; Marras et al., Genet. Anal. 1999, 14: 151-156; U.S. Pat. Nos. 5,846,726, and 5,925,517)
A “tail” of normal or modified nucleotides may also be added to nucleic acids for detectability purposes. A second hybridization with nucleic acid complementary to the tail and containing a detectable label (such as, for example, a fluorophore, an enzyme or bases that have been radioactively labeled) allows visualization of the nucleic acid molecules bound to the array (see, for example, system commercially available from Enzo Biochem Inc., New York, N.Y.).
The selection of a particular nucleic acid labeling technique will depend on the situation and will be governed by several factors, such as the ease and cost of the labeling method, the quality of sample labeling desired, the effects of the detectable moiety on the hybridization reaction (e.g., on the rate and/or efficiency of the hybridization process), the nature of the detection system to be used, the nature and intensity of the signal generated by the detectable label, and the like.
Hybridization
According to the methods provided, an inventive nucleic acid array (i.e., a microarray in which at least one spot contains an alien oligonucleotide) is contacted with a hybridizing mixture comprising a plurality of nucleic acids under conditions wherein the nucleic acids in the mixture hybridize to the probes on the array.
The hybridization reaction and washing step(s), if any, may be carried out under any of a variety of experimental conditions. Numerous hybridization and wash protocols have been described and are well-known in the art (see, for example, J. Sambrook et al., “Molecular Cloning: A Laboratory Manual”, 1989, 2nd Ed., Cold Spring Harbour Laboratory Press: New York; P. Tijssen “Hybridization with Nucleic Acid Probes—Laboratory Techniques in Biochemistry and Molecular Biology (Part II)”, Elsevier Science, 1993; and “Nucleic Acid Hybridization”, M. L. M. Anderson (Ed.), 1999, Springer Verlag: New York, N.Y.).
The hybridization and/or wash conditions may be adjusted by varying different factors such as the hybridization reaction time, the time of the washing step(s), the temperature of the hybridization reaction and/or of the washing process, the components of the hybridization and/or wash buffers, the concentrations of these components as well as the pH and ionic strength of the hybridization and/or wash buffers.
In certain cases, the specificity of hybridization may further be enhanced by inhibiting or removing repetitive sequences. By excluding repetitive sequences from the hybridization reaction or by suppressing their hybridization capacity, one prevents the signal from hybridized nucleic acids to be dominated by the signal originating from these repetitive-type sequences (which are statistically more likely to undergo hybridization).
Removing repetitive sequences from a mixture or disabling their hybridization capacity can be accomplished using any of a variety of methods well-known to those skilled in the art. Preferably, the hybridization capacity of highly repeated sequences is competitively inhibited by including, in the hybridization mixture, unlabeled blocking nucleic acids.
Microarray-based hybridization reactions in which alien oligonucleotides may serve as controls include a large variety of processes. For example, they may be useful in gene expression methods, such as those developed and used in pharmacogenomic research (see, for example, M. Srivastava et al., Mol. Med. 1999, 5: 753-767; and P. E. Blower et al., Pharmacogen. J. 2002, 2: 259-271); in drug discovery (see, for example, C. Debouk and P. N. Goodfellow, Nat. Genet. 1999, 21: 48-50; and A. Butte, Nat. Rev. Drug Discov. 2002, 1: 951-960), or in medicine and clinical research, for example, in cancer research (see, for example, J. DeRisi et al., Nat. Genet. 1996, 14: 457-460; C. S. Cooper, Breast Cancer Res. 2001, 3: 158-175; S. B. Hunter and C. S. Moreno, Front Biosci. 2002, 7: c74-c82; R. Todd and D. T. Wong, J. Dent. Res. 2002, 81: 89-97).
In another aspect, the inventive provides methods of using alien oligonucleotides and their complements in microarray-based hybridization experiments for different control purposes.
Alien Sequences as Negative Controls
In certain embodiments of the invention, alien oligonucleotide sequences are used to serve as a negative control during the course of the microarray experimentation. Negative controls are valuable when assessing the stringency of target-to-probe hybridization. For example, the selectivity of hybridization is known to be paramount to the accurate reflection of differential gene expression.
When present on a microarray, inventive alien oligonucleotides (i.e., molecules comprising sequences selected for their inability to hybridize nucleic acids of the source or collection under analysis) can act as negative controls. If a detectable signal can be measured from spots containing alien sequences, then hybridization conditions are not stringent and lead to significant cross-hybridization reactions, which, in turn, adversely affect the measured differential gene expression.
Use of Alien Sequences to Quantify Hybridization Sample Components
The present invention also provides methods that allow quantification of hybridizing sample components. Such methods are based on the use of microarrays containing alien oligonucleotides and on the addition of their complements (i.e., anti-alien sequences) to the hybridizing mixture before hybridization.
More specifically, inventive methods comprise providing a hybridizing mixture comprising a plurality of nucleic acids; and hybridizing the hybridizing mixture to a nucleic acid array of the invention, wherein the step of providing a hybridizing mixture comprises providing a mixture containing at least one anti-alien hybridizing nucleic acid whose sequence comprises a sequence complementary to the alien probe present on the inventive nucleic acid array.
In certain preferred embodiments, a known amount of an anti-alien oligonucleotide is added to a sample containing at least one experimental hybridizing nucleic acid of unknown quantity, and the mixture thus obtained is processed and prepared for hybridization to a microarray containing the alien oligonucleotide. The processing and preparation include labeling of both the anti-alien sequence and test nucleic acids with the same detectable agent. The degree of anti-alien/alien hybridization may be relied upon to establish the amount of test sequences present in the hybridizing sample based on the relative extent of their hybridization to complementary oligo probes present on the microarray.
In preferred embodiments, the degree of hybridization between the anti-alien and alien oligonucleotides and/or between the hybridizing nucleic acid and oligonucleotide probe present on the array is determined by measuring the signal intensities from the detectable label attached to the hybridized targets.
More specifically, if, for example, the target nucleic acids have been fluorescently labeled, the amount of a particular sequence in the hybridizing mixture is determined by comparing the intensity of the fluorescence signal measured for the hybridized sequence to the intensity of the fluorescence signal measured for the anti-alien sequence hybridized to the alien oligonucleotide present on the microarray.
In other preferred embodiments, an unknown amount of the anti-alien oligonucleotide is added to a nucleic acid sample to be analyzed and the resulting mixture is processed as above, before hybridization to a microarray containing a known amount of the alien oligonucleotide. The quantification of hybridization sample components may then be carried out as described above.
In other preferred embodiments of the invention, different amounts of multiple alien/anti-alien pairs are used for comparative quantification of nucleic acids of the test sample. Using amounts of multiple alien/anti-alien pairs, that vary from rare, to low, to abundant and highly abundant provides reference signal intensities for widely different ranges of target amounts (or concentrations), and therefore can help improve the accuracy of the quantification of test sequences. Such a method may be particularly useful when the signal intensity vs. detectable label amount (which is equivalent to hybridized target amount) exhibits a deviation from linearity in one or more concentration ranges.
Use of Alien Sequences for Normalization
Also provided by the present invention are methods wherein alien oligonucleotides are used as controls for in situ normalization.
At present, differential gene expression relies on changes in the relative abundance of any given mRNA between a test and reference total RNA sample. Usually ratios are derived that identify if a test sample mRNA is up- or down-regulated with respect to a reference sample, however in many instances no appropriate reference sample exists. Such a problem is typically encountered when samples are collected over extended periods of time (i.e., clinical studies) and need to be compared to a common reference or in diseased patients where no applicable reference is available.
In certain preferred embodiments, a microarray has spots containing a mixture of known amounts of the alien oligonucleotide and of a probe able to detect target (or hybridizing) sequences. Such an arrangement allows in situ comparisons. This approach also provides a consistent standard (the fixed amount of alien oligonucleotide) that can be relied upon to allow inter-slide comparisons and inter-experiment comparisons even when experiments are carried out with rare samples, or over a long time spans.
In these particular instances, an alien sequence can be used as an in-spot control and act as the reference so that inter-slide expression differences can be measured relative to a consistent control.
For instance, if every spot in an array has a defined mixture of experimental probes to alien probes, the presence of the alien oligonucleotides allows the researcher to control for variations between and among spots (e.g., by hybridizing the array with a sample containing anti-alien sequences that are differently labeled from the target sequences.
Those of ordinary skill in the art will appreciate that it is not essential that every spot on the array contain alien oligonucleotide, though it will typically be desirable that the alien oligo be present in a representative collection of spots, for example, so that the researcher can have reasonable confidence in the general uniformity of the spots. It will also be appreciated that, although convenient, it is not essential that every spot containing the alien sequence contain the same ratio of alien and experimental probes; so long as the ratio for each spot is defined and known.
In these methods, normalization is performed according to standard techniques.
As shown on the scheme presented in
Controlling Hybridization Sample Processing and Hybridization with Alien Sequences
Furthermore, when an alien oligonucleotide is present on an array, its complement may be added to the hybridizing sample, and processed (i.e., subjected to different treatments including labeling) together with the sample, and hybridized to an inventive microarray as a control for the processing/hybridization steps. If the alien oligonucleotide is present in spots at different locations on the chip, this strategy can also control intra-chip hybridization variation.
To give but one example, as described in the Examples, the present inventors have designed alien sequences that consist of four alien sequences that have been concatemerized behind a T7 promoter and to maintain polyadenylated tails. Upon transcription of the alien genes with T7 RNA polymerase, an alien transcript can be added to the total RNA input and act as an internal control during the course of cDNA generation, labeling, and hybridization. When alien probes, complementary to the alien gene, are included on the microarray, the experimenter can measure the extent of hybridization between the alien probe and the anti-alien nucleic acid in the labeled cDNA milieu to ascertain the overall labeling and hybridization efficiency. While this control does not definitively identify whether the labeling or hybridization may be at fault when there is a failure to detect fluorescent signal, it does allow the experimenter to identify if a problem has occurred and to compare the relative labeling efficiencies from experiment to experiment. One would anticipate that when the labeling and hybridization are successful, the relative signal intensity from the alien probe would be similar between slides. Similarly, regional effects of hybridization can be ascertained by including alien probe sequences within each sub-array on the chip. This comparative metric for inter-slide and intra-slide comparison is beneficial for quality control purposes.
Controlling for Array Manufacture using Alien Sequences
In another aspect, the invention provides methods that allow control of array manufacture. More specifically, when an alien oligonucleotide is present on an array, a standardized (i.e., a known amount, optionally labeled) complementary nucleic acid may be added to the hybridizing sample, and the extent of its hybridization to the alien sequence on the microarray can be used to assess the quantity of the array manufacture (e.g., the extent to which oligonucleotides were effectively coupled to the surface, etc).
Thus, according to the present invention, it is possible to analyze printed microarrays (e.g., prior to their experimental use, for example to ascertain if any spots are missing (and if so which ones), as well as to judge overall spot morphology and slide quality.
The following examples describe modes of making and practicing the present invention. However, it should be understood that these examples are for illustrative purposes only and are not meant to limit the scope of the invention. Furthermore, unless the description in an Example is presented in the past tense, the text, like the rest of the specification, is not intended to suggest that experiments were actually performed or data were actually obtained.
The present invention provides systems for identifying “alien” sequences that are not found in the relevant population of nucleic acids being hybridized to an array. For instance, the invention provides systems for identifying sequences that are not present in the cDNA of a selected organism.
In particular, a software program was developed that allows the user to generate “alien” cDNAs for a particular organism. The program, the algorithm of which was described above, takes in a list of all known cDNA sequences for that particular organism (e.g., mouse). From this list, the program calculates the codon frequency of the sequences as well as dinucleotide or transition sequences at the codon boundary. These files can be stored and are specific for the organism from which the frequencies are generated. The program then generates cDNA (with start and stop codons) using the above frequencies. A small percentage of the time (as may be specified by the user), the generated frequencies are flipped such that the least frequent codon is now generated in the middle of the sequence. Such a sequence should be different from any cDNA occurring in the genome. The degree of “alien”ness of the sequence can be verified by comparing the generated sequences to the organism's genome (if available) or cDNA by using BLAST or another sequence comparison program. Oligos are then generated from the sequences by using another software program which checks for Tm and % GC content. The generated oligos are also compared to the organism genome or cDNA to verify that they do not hybridize to any part of the genome.
For example,
Synthesis of alien oligonucleotides. Each of the 47 70 mer alien oligonucleotide probes depicted in
Preparation of oligo slide. Alien oligonucleotides were then printed and linked to the surface of oligos slides (CodeLink, Amersham Biosciences, Piscataway, N.J.), which also contained human and mouse positive control spots. All the plates were prepared following the same protocol.
Alien oligonucleotides were arrayed in Greiner 384-well flat-bottom plates (600 pmol of alien oligonucleotide per well). After resuspension in water to 20 μM, the oligonucleotides (5 μL) were re-arrayed into 384-well, Genetix polystyrene V-bottom plates, which were then allowed to dry in a chemical hood. Before printing, 5 μL of 1× Printing Buffer (150 mM sodium phosphate, 0.0005% Sarcosyl) were added to each well. The plates were incubated at 37° C. for 30 minutes to aid resuspension of DNA, vigorously shaken on a flat-bed shaker for 1 minute, and centrifuged at 2000 rpm for 3 minutes. These plates were then placed into an OmniGrid® 100 microarrayer (GeneMachines, San Carlos, Calif.) for the preparation of oligos slides.
After completion of each print run, the slides were removed from the microarrayer and placed overnight in a sealed humidification chamber containing a saturated brine solution and lined with moist paper towels. The slides were then transferred to a slide rack (25 slides per rack), which was placed into a container filled with Pre-warmed Blocking Solution (50 mM 2-aminoethanol; 0.1 M Tris pH 9, 0.1% N-Lauroyl sarcosine) to completely cover the slides, and then shaken for 15 minutes. The slides were rinsed twice with de-ionised water by transferring the slide rack to water filled containers. The slide rack was then transferred to another container filled with pre-warmed Washing Solution (4×SSC, 0.1% N-Lauroyl sarcosine) to completely cover the slides, and then shaken for 30 minutes. After the slides were rinsed twice with de-ionized water, they were dried by centrifugation at 800 rpm for 5 minutes, and stored in a dessicator.
Terminal Deoxynucleotidyl Transferase Quality Control. A first set of slides were treated with Terminal Deoxynucleotidyl Transferase in the presence of dCTP-Cy3, so that all oligonucleotides attached to the slide could be visualized and their attachment assessed. The labeling was performed by adding 10 μL of 5× reaction buffer (containing 500 mM sodium cacodylate, pH 7.2, 1 mM 2-mercaptoethanol, and 10 mM CoCl2), 0.5 μL of Cy3-dCTP (Amersham), 2 μL of Terminal Deoxynucleotidyl Transferase (Amersham, 12 units/mL) and water to a final volume of 124 μL. The reaction solution was briefly vortexed and spun. The slides were boiled for 10 minutes in ddH2O and dried with a gentle air stream. The Terminal Transferase hybridization procedure, which was performed using a GeneTac Hybridization station (BST Scientific, Singapore), included an incubation cycle carried out at 37° C. for 2 hours followed by three washing steps.
After the slides were rinsed with 0.06×SSC, and then dried by centrifugation, they were scanned within the next 24 hours using an Axon GenePix 4000B scanner (Axon Instruments, Union City, Calif.). The resulting images were analyzed using the GenePix 3.0 software package.
As shown in
A second set of slides was not treated with terminal deoxynucleotidyl transferase, and instead was hybridized with labeled mRNA from human (Stratagene's Universal RNA Human) and mouse (Stratagene's Universal RNA Mouse).
Labeling of Universal Mouse/Human RNA. Before hybridization, samples of both types of mRNA were labeled using the standard indirect labeling method developed by J. B. Randolph and A. S. Waggoner (Nucleic Acids Res. 1997, 25: 2923-2929). Human mRNA was labeled with Cy5™ and mouse mRNA was labeled with Cy3™. Briefly, aminoallyl dUTP was incorporated during the reverse transcription of the total RNAs. This modified cDNA in turn was labeled via a coupling between an N-hydroxysuccinimide activated ester of a fluorescent dye (Monoreactive Cy3 and Cy5 from Amersham) and the aminoallyl moiety of the dUTP, following a modified version of the Atlas Powerscript Fluorescent Labeling Kit (BD Biosciences Clontech, Palo Alto, Calif.) protocol.
Hybridization to alien oligonucleotide microarrays. Hybridizations were performed on a Genomic Solutions GeneTac Hybridization Station (BST Scientific). A competitive DNA mix (containing salmon sperm DNA, Poly-A DNA and optionally Cot-1 DNA when the nucleic acid population under analysis was human) was added to hybridizing mixtures before hybridization. After hybridization, the slides were rinsed with 0.06×SSC, dried by centrifugation and scanned within the next 24 hours as described above.
As shown in
The results presented in
Also, as shown in
As described herein, one advantage of using alien sequences in microarray experiments is that their complements may serve as an in-spike control, enabling the experimenter to gauge the robustness of the target labeling and hybridization. Specifically, if an alien oligonucleotide is present on a chip or slide, then a known amount of its complement may be added to the population of nucleic acids (e.g., mRNA or cDNA) to be hybridized to the slide. The population, now spiked with a known amount of anti-alien nucleic acid, is then labeled and hybridized to the chip or slide. Global problems in labeling or hybridization are revealed through the extent of alien/anti-alien hybridization on the chip or slide.
In order to create an in-spike control that would mimic an experimental cDNA sample to the greatest extent possible, three alien genes have been designed to consist of four different 70 mer alien sequences linked to one another in series and to a T7 promoter. The three alien genes also contained a polyadenylated tail to facilitate oligo(dT) priming. Alien gene A (321 bp), Alien gene B (322 bp) and Alien gene C (322 bp) are presented in
The alien gene shown in
More specifically, 10 ng of alien B was PCR amplified with a forward primer (5′-TTCTAATACGACTCACTATAGGGCATCTATCTATGTCAGTTACCGGC) and a reverse primer (5′-TTTTTTTTTTTTTTTTTTTTTTTTCTAATAACTGAGGTGATTTCCGAC) using the SuperMix High fidelity polymerase (Invitrogen, Carlsbad, Calif.) and the Manufacturer's suggested protocol (which included the following cycle program: 94° C. for 30 sec, 55° C. for 55 sec, and 72° C. for 1 min) was followed. The reaction was performed for 30 cycles followed by a 3 min. final elongation incubation. The PCR product was analyzed on a 1.5% agarose gel and quantified according to quantitative low range DNA markers (Invitrogen).
The PCR product was then used as a template for in vitro transcription. In a reaction volume of 50 μL, 500 nM of PCR product was combined with 200 mM HEPES, pH 7.5, 7 mM NTPs, 20 mM MgCl2, 40 mM dithiothreitol, 2 mM spermidine, 100 μg/mL bovine serum albumin (Roche, Nutley, N.J.), 8 units RNasin inhibitor (Promega, Madison Wis.), 0.5 units inorganic pyrophosphatase (Sigma, St. Louis, Mo.), and 500 units of T7 RNA polymerase (Epicentre, Madison, Wis.). The reaction was incubated for 16 h at 37° C. Following transcription, the reaction was phenol:chloroform extracted and LiCl precipitated. The pellet was rinsed with 70% aqueous ethanol, dissolved in 25 μL of buffer and quantified using UV spectroscopic methods.
The alien gene B run-off transcript was then reverse transcribed in the presence of amino-allyl dUTP (to allow for the incorporation of a label), using either a polyT primer or a collection of random hexamer primers. The resulting oligodT-primed cDNA was labeled with N-hydroxysuccinamide-Cy3; the resulting random-primed cDNA was labeled with N-hydroxysuccinamide-Cy5.
Microarrays were prepared by linking 8 different alien 70 mers, four of which were present in the alien gene and four of which were not, to a slide as described above in Example 2. As also described in Example 2, linkage of the 8 different oligonucleotides to the slide was assessed via enzymatic labeling with terminal transferase. As shown in
A comparable chip was then hybridized with a mixture of the labeled oligodT-primed cDNA and the labeled random-primed cDNA.
In order to demonstrate the use of alien sequences as internal controls for microarray spotting and hybridization, alien oligonucleotides were first shown to be able to effectively hybridize with their targets even when included in spots containing other oligonucleotides. Specifically, microarrays were constructed in which a single alien oligonucleotide, AO892 (5′GGTACGAATCTCCCATTGCATGGACAAATATAGTCCACGCATTGGACGCAC CCACCGATGGCTCTCCAAT), was spotted by itself in concentrations ranging from 2 to 20 μM, and was also spotted with a mixture of other 70 mer probes, whose concentrations also increased.
An 70 mer oligonucleotide whose sequence was complementary to that of AO892 was prepared, modified at the 5′-terminus with a C-6 amino linker, and labeled with N-hydroxysuccinimide Alexa-488. This labeled complement was hybridized to the array under standard hybridization conditions, and differences between its hybridization to the pure AO892 spots and the mixture spots were assessed. As can be seen in the insert of
Methods:
Microarray fabrication, hybridization and scanning: The process of microarray fabrication, alien oligo synthesis, hybridization and scanning was carried out by the Massachusetts General Hospital, DNA Core Group. The protocols used for each of the following steps are described in detail at their website (dnacore.mgh.harvard.edu/microarray/protocols.shtml).
Alien oligo synthesis: The alien oligonucleotide probes were synthesized using an Expedite DNA synthesizer following standard protocols of phosphoramidite chemistry. All oligonucleotides were modified at the 5′ terminus with a trifluoroacetamidohexyl-amino-C6-phosphoramidite which functionalizes the terminus and enables subsequent covalent attachment of the oligonucleotide to a CodeLink slide surface.
Preparation of the oligo slide: Briefly, alien oligos were arrayed in 384-well plates and mixed with printing buffer. These plates were then placed onto an Omnigrid 100 microarrayer for the preparation of oligo slides. After each print run, the slides were placed in a sealed humidification chamber. The slides were immersed in blocking solution, washed in 4×SSC and 0.1% N-lauroyl sacrosine and then stored in a dessicator. The alien oligos were printed along with a mouse oligonucleotide probe set which has 19,549 probes on the array providing complete coverage of the 2002 Mouse genome. Alien oligo 892 (“AO892”) was printed in known concentrations in all spots of the microarray that contained mouse probes. Oligos that make up the alien gene transcripts were printed in separate spots on the slide. To act as print quality control and to check the attachment of all nucleotides to the slide, a few slides were treated with terminal deoxynucleotidyl transferase in the presence of dCTP-Cy3.
Labeling of RNA: Before hybridization, mRNA samples were labeled using the standard indirect labeling method developed by Randolph and Waggoner, Stability, specificity and fluorescence brightness of multiply-labeled fluorescent DNA probes, Nucleic Acids Research, 15;25(14):2923-9, 1997. Briefly, aminoallyl dUTP was incorporated during the reverse transcription of total RNA. The modified cDNA was labeled via a coupling between an N-hydroxysuccinimide-activated ester of a fluorescent dye (Cy3 or Cy5) and the aminoallyl moiety of the dUTP. The anti-alien to oligo AO892 and the three alien gene transcripts were mixed in known concentrations with the extracted mouse RNA. The anti-alien to oligo AO892 was labeled with Alexa488 while the alien gene transcripts were labeled with both Cy5 and Cy3.
Hybridization reactions: Hybridizations were performed on a Genomic Solutions GeneTac Hybridization Station. Cy3 and Cy5-labeled RNA were mixed with a competitive DNA mix containing salmon-sperm DNA, Poly-A DNA and Cot-1 DNA before hybridization
Scanning: After hybridization and washing, the microarrays were scanned using the ProScanArray HT microarray scanner and the resulting images analyzed using ScanArray® Express v3.0 software.
Data Normalization and Filtering: Data from image analysis was stored for further processing in BASE (BioArray Software Environment). All spots flagged as unusable by the ScanArray software were excluded from further analysis. All array images were also analyzed manually to check for hybridization artifacts and to identify bad spots that had not been identified by the ScanArray program. The identified spots were also excluded from further analysis. Using BASE, all reference spots that had hybridization intensity readings less than 300 and all test sample spots that had hybridization intensity readings less than 50 were also removed from the dataset for analysis.
Typically, the first transformation applied to expression data, referred to as normalization, adjusted the individual hybridization intensities to balance them appropriately so that meaningful biological comparisons could be made. The filtered data was normalized in two ways depending on the presence or absence of information from alien/anti-alien hybridization.
Data normalization and replicate filtering in the absence of alien control data: Microarray data was normalized initially by scaling all individual intensities such that the total intensity was the same for both comparative samples (control and treatment) within a single array and across replicate arrays. This was based on the assumption that the starting amounts of RNA in each sample were equal. Using this approach, a normalization factor was calculated by summing the measured intensities in both channels:
where Gi and Ri are the measured green and red fluorescence intensities for the ith array element and Narray is the total number of elements represented in the microarray. One or both intensities were appropriately scaled to adjust for the normalization factor.
In addition to total intensity normalization, locally weighted linear regression (lowess) analysis was used to remove systematic, intensity-dependent effects in the data. The starting point for the lowess analysis, the ‘R-I’ (for ratio-intensity) plot, can reveal intensity-specific artifacts in the log2 (ratio) measurements. The RI or MA plot shows the measured log2 (Ri/Gi) for each element on the array as a function of the log10 (Ri*Gi) product intensities. Lowess detects systematic deviations in the R-I plot and corrects them by carrying out a local weighted linear regression as a function of the log10 (intensity) and subtracting the calculated best-fit average log2 (ratio) from the experimentally observed ratio for each data point. The data was normalized globally.
The replicates per treatment and per time point when available were then combined to reduce the complexity of the data set. Genes with only one data point across all replicates after initial selection were excluded. Genes with more than one replicate data point were then analyzed for outliers and discarded if necessary. The data was combined using the geometric mean of the replicate ratios.
Data normalization and replicate filtering in the presence of alien controls: Two different methodologies were used to normalize data using alien hybridization intensities. Due to experimental design, the amount of anti-alien to alien hybridization in every spot on the array should be equal. This implies that the recorded alien hybridization (Alexa488) intensities should also be equal. One normalization procedure is to calculate the average alien hybridization intensity across all spots on the array and then normalize the alien hybridization intensity at a spot to that average intensity. The normalization factor for each spot can then be used to scale the treatment and control intensities for that spot. This normalization algorithm can be applied globally or locally. Local normalization can be applied to each group of array elements deposited by a single spotting pen.
Another method of normalization is to scale all alien hybridization intensities to an arbitrary constant intensity value. In the analyses conducted here, the second method was used. All alien hybridization intensities were scaled to a uniform intensity value of 1000. The normalization factor used to scale each individual spot was then used to adjust the other channel intensities at that spot. Replicates were then combined and genes with only one data point across all replicates after initial selection were excluded. Genes with more than one data point were analyzed for outliers using intensities from all three channel. The data from replicates were combined by calculating the geometric means of the individual intensities.
Identifying differentially expressed genes: The log2 ratio of gene expression for each spot for each gene was calculated either using direct or indirect comparison. Assuming there are two samples A and B, while using direct comparisons, the ratio T of gene i in sample A to sample B is Ti=Ai/Bi, where Ai and Bi are the normalized intensity values.
Further assuming that U is the universal reference sample used in two separate microarray experiments 1 and 2 to compare sample A to sample B indirectly, if T1i is the ratio of intensities of gene i in sample A to gene i in the universal reference, and T2i is the ratio of intensities of gene i in sample B to gene i in the universal reference, then the ratio Ti of gene i in sample A to gene i in sample B is Ti=T1i/T2i.
When using the aliens as the reference channel for indirect comparison, the intensity of gene i in sample A and the intensity of gene i in sample B can be compared directly. This is possible as all spots in all arrays have been scaled such that the alien hybridization intensities are all equal to 1000. Therefore Ti=A1i/B2i, where A1i, and B2i are intensity normalized and scaled values.
The standard log2 ratios were then calculated for each gene in each of the above cases. The mean and standard deviation of the distribution of log2(ratio) values was then calculated. The Z-score value for each gene was then used to determine if the gene was differentially expressed. Genes with log2 ratios over 2 standard deviations from the mean were identified as differentially expressed and chosen for further analysis. This allowed us to identify genes that were expressed sufficiently above the noise without having to resort to an arbitrary minimum ratio value.
Results and Conclusions:
The alien oligos can be used as internal controls for microarray spotting and hybridization, by spotting them in a mixture with the probes used to hybridize to the sample. This arrangement allows for in situ comparisons of every spot on a microarray. The aliens thus spotted can also act as references for inter-slide expression measurement and for inter-experiment expression measurement even when the experiment has been carried out over a long time span. Spotting a known amount of oligo in every spot and hybridizing to it a known amount of anti-alien along with the experimental sample, allows one to normalize for variations between spots. This would also serve to control for errors in the hybridization and labeling steps and for controlling intra-chip hybridization variation.
To demonstrate this, alien genes were first shown to hybridize to their targets even when other probes were present in the same spot. A single alien oligo, AO892, was used for this experiment. A sequence complementary to the oligo was synthesized and labeled with Alexa488. This sample was then hybridized to a slide which had pure alien oligo spots as well as spots with mixtures of the alien and normal probes. There was no significant change in normalized signal intensity between the two types of spots (data not shown).
To determine whether alien AO892 could be used as an in-spot reference, it was tested against another sample that could be used for an indirect reference, Stratagene's Universal Mouse® Reference RNA mix. A twelve-slide experiment was designed and carried out using mouse liver and macrophage RNA samples. All slides had spots with mixtures of the alien oligo and probes for mouse RNA. Alien oligo AO892 was printed in known concentrations in spots of the microarray that contained mouse gene-specific probes. It was printed at 10% final concentration of the mouse gene-specific probes in that spot. In four slides, Universal Mouse Reference RNA was used as the reference sample and liver RNA was used as test. Another four slides used mouse macrophage RNA as test samples and Universal Mouse as reference. A transcript complementary to AO892 and labeled with Alexa488 was added to all pre-hybridization mixes of labeled cDNA. These set of slides permitted comparison of differential expression between mouse liver and mouse macrophage samples by using both the Universal Mouse Reference RNA as well as the aliens as references. The last four slides directly compared liver RNA samples to macrophage RNA samples. A dye-swap was incorporated in each set of experiments.
The RNA was labeled and then hybridized on a chip containing the probe mixtures. The intensity readings were collected and quantified. Genes with low intensities not significantly above background were excluded from analysis. This reduced the number of spots from 19,552 to 18,268 for aliens and to 8,667 for the Universal RNA. The log10 of the intensities was then calculated and their frequency plotted (see
When using Universal Mouse RNA as reference, the microarray data was normalized by scaling all individual intensities such that the total intensity of the all channels was the same across replicate arrays for that experiment. Data from the replicate arrays were then combined to identify outliers and reduce statistical variation. When using the alien channel as reference, a spot to spot comparison across the two experiments was done and all intensities adjusted such that the alien intensity was set to 1000. For the final analysis, spots were chosen such that data was available for both direct and indirect comparisons. 6,866 spots were selected for comparison through the alien channel and 5,322 through the Universal RNA channel. The data was compared using log2 ratios of test RNA intensity to reference intensity and plotted (see
In the most widely used experimental design for microarrays, all the direct comparisons are made to a single reference sample. By following this method, the path connecting any two samples is always two steps. Thus, all comparisons are made with equal efficiency. In experiments that analyze RNA samples from two different conditions or two different treatments and when these samples derive from a series of time points, the most commonly used reference is the wild-type or untreated sample. This is inefficient because fully half of the measurements are made on the reference sample, which is presumably of little or no interest. Alien sequences could be used as a common reference in this experiment design as well. In this example, we have designed an experiment to compare fracture healing in wild-type and TNF-α receptor-deficient mice that would also allow us to test the use of the alien sequences as a common reference.
A total of 56 DNA arrays were divided into five sections (Table 1). RNA extracted from the tibia of wild type mice before fracture was used as the universal reference. All microarrays had the alien oligo AO892 mixed in with the gene-specific probes. The complementary sequence to AO892 was labeled with Alexa 488 and mixed into all hybridizing samples. The alien oligos that can bind the three alien transcripts to be used as in-spike controls were deposited in separate spots on the array. The alien transcripts were mixed with the sample and reference RNA before labeling in all experiments other than those in Experiment E of Table 1.
In Experiment A, the different time points of fracture healing for the TNF-α receptor knockout mice were compared to the reference. There were four replicates per time point, including a dye swap. In Experiment B, the time points of fracture healing in wild-type mice were compared to the reference again with four replicates per time point. The microarray datasets in Experiments A and B enabled a differential expression comparison of fracture healing in the transgenic mice as compared to the wild type mice at each time point, using either the common reference channel or the alien channel as control. The two references, wild type mice at time zero and the alien channel can also be used to compare across time points to generate a time series profile of gene expression during fracture healing.
Experiments C and D compared healthy tissues in wild type mice and knockout strains with themselves. This method identified genes that could cause problems during analysis. Ideally, ratios of the test channel intensity to the reference channel intensity in the case of these experiments should be 1. However, this was not true for some genes, due to factors beyond the control of the experiment. These genes were removed from the dataset before analysis. Also, differential mRNA expression between healthy knockout strains and knockout strains undergoing fracture healing can be measured using datasets from Experiments A and C.
Experiment E checked whether the alien oligos cross-hybridize to RNA from the two test samples. The test samples used here were not mixed with the in-spike alien transcripts. Analysis of channel 1 intensities from spots that contain the only alien oligos as well as analysis of channel 2 intensities of spots that don't have in-spike controls showed any non-specific hybridization to the alien sequences. The RNA was labeled and then hybridized on a chip containing the probe mixtures. The intensity readings were collected and quantified. Genes with low intensities in each of the reference channels were filtered out from the dataset. This reduced the dataset by approximately 10% when using the aliens as reference as opposed to more than 50% when using the sample from unfractured tibia.
The data was normalized as discussed in the Methods section of Example 5 above, using both the sample from unfractured tibia as well as the alien reference. Data from the microarrays in Experiment A was concatenated with data from microarrays in Experiment B and a common list of genes for which information was available from both set of experiments for each time point was identified. This helped in performing an indirect comparison of the genes in each of the knock-out time points to those in the wild-type. Differentially expressed genes were then identified using both indirect comparisons to the sample of unfractured tibia as well as the alien channel. Table 2 compares the data available from indirect comparisons for each of the time-points of fracture healing.
As can be seen from the data in Table 2, more genes were available for analysis when using the alien sequences as reference. Most of the missing data in either method is due to the initial filtering step when spots with low intensities are removed. Since there were more genes available for analysis while using the aliens as reference, that method also identified more genes as differentially expressed as compared to using the sample from unfractured tibia as reference. There were some genes that were identified as differentially expressed by one method but not by the other. RT-PCR experiments would need to be performed to verify which of the methods provided better results. Some cytokine-related genes were identified as differentially expressed only when the alien sequences were used as reference. These genes showed little or no expression in the sample from unfractured tibia.
Data from the in-spike control spots on the microarray was also analyzed. The three alien transcripts were spiked into the test samples at different concentrations. The mean normalized log10[intensity] values for the spike-in control probes was used to define a standard curve relating signal intensity to copy number (see
Number | Date | Country | Kind |
---|---|---|---|
PCT/US04/01911 | Jan 2004 | WO | international |
This application claims priority to Provisional Patent Application No. 60/441,832, filed Jan. 22, 2003, U.S. patent application Ser. No. 10/763,039 filed Jan. 22, 2004, and PCT International Application No. PCT/US04/01911, each of which is incorporated herein by reference in its entirety.
Number | Date | Country | |
---|---|---|---|
60441832 | Jan 2003 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 10763039 | Jan 2004 | US |
Child | 11224573 | Sep 2005 | US |