6500 Harbour Heights Pkwy., Suite #303
Mukilteo, WA 98275 • USA
800-985-CBMX • 425-493-2000
http://www.combimatrix.com

What is the principle of Hyb & Seq?
Can Hyb & Seq be applied for microarray chips other than CombiMatrix CustomArray?
What are the major applications of the Hyb & Seq method?
How does Hyb & Seq improve SNP detection in comparison with Tm-based approaches?
What is the typical error rate associated with usage of Hyb & Seq for re-sequencing? How do regions of secondary structure affect Hyb & Seq performance?
Can I use Hyb & Seq to perform re-sequencing in both forward and reverse directions?
Can I perform a multiplex assay of several targets on the same chip?

How should I design CustomArray probes for Hyb & Seq?
Where should I put mismatches: in the middle or on the terminal end of probes?
How long should CustomArray probes be for Hyb & Seq?
What kind of control probes do I need to include in the CustomArray design for Hyb & Seq?
What are the controls factory-built on CustomArray chips?
How many nucleotide positions of the target DNA sequence can be analyzed using one CustomArray of 12k probes?

What are the size limits for the amplified target DNA sequence to be used in Hyb & Seq?
What is the ideal length of the amplified target DNA sequence to be used in Hyb & Seq?
How do I obtain specific primers to amplify the target DNA sequence of interest?
What kind of detection primer can I use?
How much would the presence of an additional SNP affect detection of the SNP of interest? How many additional SNPs could be tolerated to still enable detection of the SNP of interest by Hyb & Seq?
Why should I use single-strand DNA targets?
Can I use denatured double-strand DNA targets?
What is the recommended protocol for generation of single stranded DNA targets?
Can I use alternative protocols to generate single stranded DNA targets?
How many micrograms of DNA target can I put on one CustomArray?
Will long hybridization times (over 1 hour) improve performance?

What type of software can I use to interpret the results of Hyb & Seq?
Do I need to perform background subtraction and normalization for Hyb & Seq data?

WHAT IS THE PRINCIPLE OF HYB & SEQ?
The principle of Hyb & Seq is mismatch detection by enzymatic reaction performed on the microarray chip. A target DNA sequence is amplified with incorporation of a common tag antisense to the detection primer. A CustomArray 12K chip is synthesized with oligonucleotide probes that are complimentary to the target sequence and carry terminal mismatches. The amplified target sequence is hybridized to the CustomArray chip in the presence of the common labeled detection primer (Cy3-tag). Oligonucleotide probes of the chip and the detection primer anneal to the target. Subsequent incubation with enzymes results in extension of the detection primer and ligation to the probes, but only in the cases of exact match between the probe terminal nucleotide and the target sequence. After the enzymatic reactions are complete, DNA hybrids are denatured by stringent washing. The labeled detection primers are washed away from the chip except for those ligated to the probes. Therefore, signal is only detected for the probes that have exact matches to the target sequence.

CAN HYB & SEQ BE APPLIED FOR MICROARRAY CHIPS OTHER THAN COMBIMATRIX CUSTOMARRAY?
Currently, we are not aware of any other chip manufacturer that has an enzymatic protocol validated on their system where you receive an array and conduct the experiment yourself.

WHAT ARE MAJOR APPLICATIONS OF HYB & SEQ METHOD?
The applications for Hyb & Seq include: SNP detection, genotyping, DNA re-sequencing, and detection of DNA/DNA hybridization upon difficult and/or variable conditions.

HOW DOES HYB & SEQ IMPROVE SNP DETECTION IN COMPARISON WITH Tm-BASED APPROACHES?
Hyb & Seq™ overcomes the following serious problems that are inherent to SNP analysis conducted with traditional thermal-based array applications:

1. How to define a single optimal condition that is suitable for all varying probe sequences
2. How to distinguish certain ‘difficult’ mismatches such as G-G or G-T
3. How to detect mismatches in a GC-rich environment, even when probes are standardized to the same Tm, A-T-rich probes will be longer than G-C-rich probes and internal mismatches will be more difficult to detect.

In addition, due to stringent washing, the Hyb & Seq protocol can reduce typical backgrounds caused by cross-hybridization and non-specific DNA binding that are seen in most hybridizations.

WHAT IS THE TYPICAL ERROR RATE ASSOCIATED WITH USAGE OF HYB & SEQ RESEQUENCING? HOW DO REGIONS OF SECONDARY STRUCTURE AFFECT HYB & SEQ PERFORMANCE?
Hyb & Seq demonstrates 90-100% accuracy with re-sequencing. Most errors seem to be associated with secondary structure such as hairpins and palindromes. When hairpins and palindromes are present in the single-strand target DNA, they tend to self-hybridize within the target sequence, rather than hybridize to the corresponding probes on the array. The result will be a decreased signal for that portion of the sequence. (See "Target Hybridization" Below)

CAN I USE HYB & SEQ TO PERFORM RESEQUENCING IN BOTH FORWARD AND REVERSE DIRECTIONS?
You cannot re-sequence both strands of the same target DNA sequence on the same array simultaneously because then your single-strand DNA targets would hybridize to each other. However, you can use both strands individually with two different arrays (carrying the probes in appropriate orientation).

CAN I PERFORM A MULTIPLEX ASSAY OF SEVERAL TARGETS ON THE SAME CHIP?
A multiplex assay of a mixture of different single-strand DNA targets could be done on the same array providing that the targets do not cross-hybridize to each other. We have tested a mix of up to four different targets on the same array for SNP analysis and two targets for influenza genotyping.

HOW SHOULD I DESIGN CUSTOMARRAY PROBES FOR HYB & SEQ?
Probes are bound to the chip at the 3' end; thus, the 5' end of the probe is used for match/mismatch detection. Probes can be synthesized with a specific melting temperature and with a 3' linker. For SNP detection and re-sequencing, four identical probes are synthesized, each having a different 5' end (A, G, C, or T).

WHERE SHOULD I PUT MISMATCHES: IN THE MIDDLE OR ON THE TERMINAL END OF PROBES?
For Hyb & Seq, the mismatches are placed at the 5' end of the probe. This allows for ligation of the extended labeled detection primer.

HOW LONG SHOULD CUSTOMARRAY PROBES BE FOR HYB & SEQ?
We have successfully used probes with melting temperatures of 50 to 70ºC. Depending upon G-C content, the probes can range from 14 to 35-mer (optimal: 17 to 28-mer). If shorter probes are used, the addition of a 5 to 10-mer oligo(dT) linker will prevent the effects of proximity to the chip surface.

WHAT KIND OF CONTROL PROBES DO I NEED TO INCLUDE IN THE CUSTOMARRAY DESIGN FOR HYB & SEQ?
For each assayed nucleotide position, the chips will have four identical probes with the 5' termini variable (A,G, T or C) . One variant will be the wild-type oligonucleotide, one the mutation (e.g. known SNP) and the remaining two will be mismatches. These mismatches should not hybridize with the probe and can therefore be used as controls for background. Sequencing chips will also have four probes ending in the four nucleotides A, C, T, and G.

WHAT ARE THE CONTROLS FACTORY-BUILT ON CUSTOMARRAY CHIPS?
All CustomArray™ chips carry a set of factory-built controls including two types of synthesized probes: negative controls (NC), and quality controls (QC). Some positions are also left as 'no-synthesis' controls. All these probes are necessary for array quality control; thus, their layout cannot be changed. The exact sequences of these probes are shown in the table 1 (Cross-reference to this table in the CustomArray FAQ). You could BLAST these control sequences against your target sequence to estimate cross-hybridization.

HOW MANY NUCLEOTIDE POSITIONS OF THE TARGET DNA SEQUENCE CAN BE ANALYSED USING ONE CUSTOMARRAY OF 12K PROBES?
One nucleotide position of the target sequence requires 4 probes, thus, 12000 probes are sufficient to analyze 3000 nucleotide positions (3 kb).

WHAT ARE THE SIZE LIMITS FOR THE AMPLIFIED TARGET DNA SEQUENCE TO BE USED IN HYB & SEQ?
We have successfully used amplified DNA fragments from 100 to 1000 bp. However, usage of large DNA fragments seems to lower the signals, to increase variability among signals obtained with different probe sequences, and affect the reproducibility among samples. Short target fragments may work better, because they contain more probe binding sites per nanogram than long targets (see Fig. A).

Selection of short fragments for amplification is also very important for pathogen genotyping. For example, RNA virus targets are amplified with reverse transcription followed by PCR. The resultant products are biased toward the first strand synthesis primer; thus, the subsequent PCR amplification would be more robust for small rather than long targets (see Fig. B).

Figure A

Figure B

WHAT IS THE IDEAL LENGTH OF THE AMPLIFIED TARGET DNA SEQUENCE TO BE USED IN HYB & SEQ?
We recommend using amplified DNA fragments of 100-200 bp, though larger targets could also be used successfully.

HOW DO I OBTAIN SPECIFIC PRIMERS TO AMPLIFY THE TARGET DNA SEQUENCE OF INTEREST?
Primers can be ordered from any oligonucleotide service. The forward primer should contain a sequence tag added to the 5’ end. We use the specific tag sequence TAATACGACTCACTATAGGG.

WHAT KIND OF DETECTION PRIMER CAN I USE?
We use the complement of the specific tag sequence with an attached fluorochrome label (5'/LABEL/GCATCCTAATACGACTCACTATAGG).

HOW MUCH WOULD THE PRESENCE OF AN ADDITIONAL SNP AFFECT DETECTION OF THE SNP OF INTEREST? HOW MANY ADDITIONAL SNPS COULD BE TOLERATED TO STILL ENABLE DETECTION OF THE SNP OF INTEREST BY HYB & SEQ?
If additional SNPs correspond to the middle of probe sequence, an increase in probe length could reduce the SNP effect on annealing. The probe length that we have successfully used ranges from 17 to 30 nt (Tm= 50 to 55oC) with a 10-T linker. The probe length and Tm could be increased to account for the internal SNP. Once annealing is achieved, the additional mismatches would not affect the discrimination of the terminal nucleotide, because it is based on ligation reaction rather than the strength of target hybridization.

If multiple SNPs are adjacent to the terminal nucleotide, it may be necessary to design probes covering all possible sequence variants. We have encountered this type of problem with an assay of CpG island methylation. Multiple Cs can be methylated in a small area; thus, we designed probes for all potential sequence variants to increase the robustness of the assay.

WHY SHOULD I USE SINGLE-STRAND DNA TARGETS?
Single-strand target hybridization is very efficient: up to 100% of target can bind to array probes. Double-strand target will not hybridize to probes without a denaturizing step and self-annealing will be favored over single-strand binding to array probes. Thus, even with a denaturation step, the double strand target binding cannot be as robust as the single-strand target binding (See "Single vs. Double Strand Target" Below).

Figure C - Single vs. Double Stranded Target

CAN I USE DENATURED DOUBLE-STRAND DNA TARGETS?
We have compared denatured double strand versus single strand targets and found that single strand target annealing to the array probes is much greater, resulting in higher signals and better mismatch detection.

WHAT IS THE RECOMMENDED PROTOCOL FOR GENERATION OF SINGLE-STRAND DNA TARGETS?
The recommended method is based on two amplification steps. First is performed to amplify the double-strand DNA fragment and to incorporate specific tag at one end. The resultant product is cleaned using a PCR product purification kit (e.g. from Qiagen) to remove primers. It is then subjected to a second amplification with only one, reverse specific primer (not the tagged one) for 40 to 50 cycles to produce the single-strand DNA fragment.

CAN I USE ALTERNATIVE PROTOCOLS TO GENERATE SINGLE-STRAND DNA TARGETS?
An alternative approach is to perform PCR with two primers, one of which is tagged with the specific tag sequence and labeled with biotin. The resulting two strands can be separated by removing the biotin-tagged strand by binding to streptavidin beads and then eluting with 0.1 N NaOH.

HOW MANY MICROGRAMS OF DNA TARGET CAN I PUT ON ONE CUSTOMARRAY?
We use approximately 10 to 20 nanograms / microliter of target in a volume of 50 to 100 microliters, but amounts can vary.

WILL LONG HYBRIDIZATION TIMES (OVER 1 h) IMPROVE PREFORMANCE?
We have found that 1 h hybridizations are sufficient for accurate and reproducible detection.

WHAT TYPE OF SOFTWARE CAN I USE TO INTERPRET THE RESULTS OF HYB & SEQ?
CombiMatrix provides free software for the extraction of data from scanned images. Subsequently, we perform a simple data analysis in MS Excel using a set of macro commands. For each probe, it compares signal intensities from four oligonucleotides that correspond to all possible terminal nucleotide variants, and finds the highest one to make a base call. We can share this set of macro commands if you contact our customer support service.

DO I NEED TO PERFORM BACKGROUND SUBTRACTION AND NORMALIZATION FOR HYB & SEQ DATA?
Of the four oligonucleotides that represent each probe, two will not hybridize and can serve as controls. Thus, we do not recommend any additional normalization method.

If you have a questions that have not been answered here, please email:

Lifterslip™ is a trademark of Erie Scientific Corporation.

Date Last Modified: 05/02/2007