Comprehensive Summaries of Uppsala Dissertations
from the Faculty of Medicine 1231
Genetic
Analyses using
Rolling Circle or PCR
Amplified
Padlock Probes
BY
JOHAN BANÉR
ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2003
CONTENTS
INTRODUCTION...1
INVESTIGATIONS OF DNA AND RNA ...2
Polymerase chain reaction ...2
Real-time PCR...3
Hybridization of nucleic acids on arrays ...4
Identification of sequence variants ...5
Ligase-based assays...5
Padlock probes ...6
Rolling circle amplification of padlock probes ...7
Multiplexed genotyping using allele-specific padlock probes ...8
PRESENT INVESTIGATIONS...10
Paper I. Signal amplification of padlock probes by rolling circle replication ...10
Comments on Paper I ...11
Paper II. Highly multiplexed genotyping with sequence-tagged molecular inversion probes and DNA microarrays ...12
Comments on Paper II ...13
Paper III. Serial circle amplification – A novel quantitative DNA amplification method for genotyping and expression analyses...14
Comments on Paper III...15
Paper IV. Parallel analysis of TCR Vβ gene expression using padlock probes and microarrays ...16
Comments on Paper IV ...17
DISCUSSION...18
Applications of RCA ...18
Multiplex genetic studies...19
ACKNOWLEDGEMENTS...21
ABBREVIATIONS
bp base pair
cDNA complementary DNA
CV coefficient of variation
DNA deoxyribonucleic acid
FRET fluorescence resonance energy transfer HRCA hyperbranched rolling circle amplification
LN lymph node
nt nucleotide OLA oligonucleotide ligation assay PCR polymerase chain reaction RCA rolling circle amplification RCR rolling circle replication
RNA ribonucleic acid
SCA serial circle amplification SEB Staphylococcal enterotoxin B
SNP single-nucleotide polymorphism TCR T cell receptor
TDLN tumor draining lymph node TIL tumor infiltrating lymphocytes
INTRODUCTION
In time for celebrating the revelation of the double helical structure of DNA fifty years ago1, a complete sequence of the human genome will be
published with the detailed order of its about 3x109 base pairs. The
information has been made accessible through the Internet along with links to identified genes and other annotated features2. This sequence represents a
consensus of those found among different individuals. The most common form of sequence variation is the single nucleotide polymorphism (SNP) which occurs on average every 1,000 bp. Those found among humans now exceed several millions and are also stored in databases, e.g. dbSNP3. SNPs
arise through mutations, some of them affecting phenotypical traits such as response to the environment or to medication, but most have no effect on the phenotype. Much interest has been focused on these variations and they are now used in many research areas. One example is diagnostic pharmacogenetic studies of SNPs that affect protein-drug interactions4.
Another is evolutionary studies, where a particular set of SNPs can be linked to common ancestral alleles found in geographically separated populations or in different species5. A field where SNPs have received increasing
attention is association and haplotype studies6. The underlying thought is to
identify associations of particular genes involved in diseases using SNPs as markers7. It has been demonstrated that SNPs are co-inherited as separate
blocks8-10 which has reduced the estimations of the number of SNPs needed
for whole-genome association studies, and the current view is approximately 300,000 (REF. 8).
Methods to genotype large numbers of sequence variations are under intense development and a variety of these now exist. The following section will list some methods that are relevant for this thesis.
INVESTIGATIONS OF DNA AND RNA
Terms that are often used when evaluating methods are specificity, selectivity and sensitivity. Throughout this thesis they will be used in the following context. Specificity represents a method’s capacity to identify a target of interest among a multitude of others. For example, investigations of a particular sequence directly in the human genome require high specificity because of the complexity constituted by its 3x109 base pairs. The specificity
of oligonucleotide probes depend on length and sequence composition, but is also influenced by factors such as salt, temperature and even electric current11. Unspecific interactions can cause signal to background ratios to
decrease or render false positive results. A related term is selectivity, which represents the ability to choose between some given variants. In methods like the oligonucleotide ligation assay (OLA), the selectivity of a ligase determines the outcome of an attempt to join two nearby oligonucleotide ends that are either matched or mismatched to their complement. A highly selective reaction step may improve over-all specificity of an assay. Sensitivity, finally, defines a method’s capability to detect targets present in very low amounts. Sensitivity is an important prerequisite to identify e.g. single copy genes in total genomic DNA or low abundant proteins in a serum sample.
Other important qualities are reproducibility, throughput, scalability and cost. Combinations of these determine the usefulness of a method for a particular purpose. For example, studies of abundant sequences in situ may require only moderate sensitivity while specificity and selectivity are crucial to distinguish related sequences.
Polymerase chain reaction
The polymerase chain reaction (PCR) is widely used in genetic research because of its high specificity and sensitivity. Most genotyping methods depend on PCR prior to, or concomitant with a reaction aimed at distinguishing a sequence variation12. PCR employs a thermostable DNA
polymerase that copies targets by extending hybridized primers. Each cycle of denaturation, annealing of primer and extension uses previously
synthesized products as templates resulting in an exponential amplification13
given sufficient reagents. During the course of amplification, re-association of products are competing with annealing of new primers, which in an early phase is negligible but in later cycles will cease the amplification efficiency, and the reaction will finally reach a plateau phase. Various techniques have been developed to monitor the amplification in real time (see discussion below).
Much effort has been invested in optimizing the reaction for multiplexed amplifications, and some degree of multiplexing has also been achieved14,15. However, the major limitation is the potential cross-reactivity
that arises when combining pairs of PCR primers in one reaction16 (Figure
1). PCR involving more than one amplicon that differ in length or sequence composition can also result in different amplification efficiencies17.
Figure 1. Possible primer combinations in a multiplex PCR using three primer pairs. 3’-ends are indicated by arrows, and desired combinations by thicker grey lines. Total number of interactions can be expressed as 2n2+n, where n is number of
primer pairs, and thus 21 combinations are possible in this example.
Real-time PCR
The polymerase chain reaction has proven very useful for quantitative measurements of nucleic acids because of its extensive dynamic range which spans at least 5 logs. Quantitative data can be obtained by monitoring the reaction in real-time and different detection techniques have been developed. The TaqMan approach applies a thermostable DNA polymerase exhibiting a 5’-3’ exonucleolytic activity, which during primer extension degrades a prehybridized probe. A fluorophore is thereby separated from a quencher, and by recording emitted light a correlation to the initial target amount can be made18-20.
Molecular beacons have also been used for real-time monitoring of PCR amplification. These are probes that in their native states form an intramolecular hairpin loop that brings a quencher in proximity to a fluorophore. Hybridization of the loop sequence to a target displaces the stem duplex, separating the reporter molecules, and emitted light can be recorded21. Molecular beacons have the advantage over TaqMan probes that
they can be used to monitor both PCR and rolling circle amplification (RCA) in real-time. Several other FRET-based reporting systems have been developed, e.g. scorpion probes22 and sunrise primers23 that have been used
to monitor PCR or hyperbranched rolling circle amplification (HRCA)24.
Hybridization of nucleic acids on arrays
The introduction of DNA microarrays had a major impact on throughput in genetic research, allowing thousands of different sequences to be analyzed at the same time25-27. Arrays are manufactured either by deposition of
oligonucleotides by pins or ink-jet printers, or through synthesis directly on the surface28-30. An example of the latter strategy involves the use of
photolithographic masks that direct deprotection of synthesized features prior to nucleoside coupling, and this technology has been used to make arrays with around 500,000 features31. Generic microarrays have been
constructed by using well characterized oligonucleotides as addressers, so called tag sequences that lack relation to biological sequences and exhibit low cross reactivity among each other. Public availability of these tag sequences have allowed researches to standardize arrays for different assays32,33. However, hybridization on solid phases exhibit different kinetics
than hybridization in solution34 and this affect the sensitivity of arrays,
requiring extensive hybridization times to detect targets at low concentrations. Also, to maintain sufficient specificity during complex hybridizations, the differences in Tm between analytes must not be too
extensive.
Array based expression analyses are often performed using surface attached cDNAs or oligonucleotides, or as recently described using arrayed beads coupled with oligonucleotides35,36. The most common application is
expression profiling where expression patterns of mRNA from different tissues or samples are compared. Total mRNA is then converted to labeled cDNA and hybridized together with a distinctly labeled control sample, for example to identify or classify tumor tissues37. The dynamic range of these
hybridizations is typically 103 (REF. 31), making comparisons between vast
Identification of sequence variants
A number of methods exist for discriminating between allelic variants based on a few reaction principles as reviewed by Syvänen12. Practically all
techniques rely on target preamplification by PCR, but a few are applicable directly on genomic DNA with a subsequent amplification step38. Some
methods may be better suited for high throughput investigations of a few genotypes in many individuals39,40, while highly multiplexed genotyping
methods have continued to challenge researchers.
The principle of allele-specific oligonucleotide (ASO)41 hybridization
has been successful in combination with real-time monitoring during target amplification39,40,42. Although a small number of different targets can be
amplified in parallel, the availability of spectrally distinct fluorophores limits the number of real-time analyses that can be performed in the same reaction43,44. Up to ten different genotypes have been distinguished using
wavelength-shifting molecular beacons in a single-tube assay45.
The minisequencing reaction is a widely adopted method for genotyping SNPs. A primer is annealed on preamplified targets directly downstream of the nucleotide to be analyzed, and a polymerase extends the primer with a labeled allele-selective nucleotide that also terminates the extension46. Different formats have been developed, adopted to microtiter
wells47, gel-based size distinction48, solution-based assays using FRET49,50 or
fluorescence polarization, and array-based methods51-53. In the related
pyrosequencing method, incorporation of nucleotides results in the appearance of a flash of light via a series of enzymatic reactions54, and the
method can be applied for small-scale multiplex SNP and haplotype studies55.
Ligase-based assays
Ligase-based detection of sequence variants relies on the fidelity of ligases to differentiate between matched and mismatched substrates. Tth ligase has a 1500-fold higher fidelity for joining of 20 nt long matched substrates compared to mismatched56, while the corresponding fidelity for T4 DNA
ligase is a 1000-fold57. Different ligases have different properties, e.g.
thermophilic ligases (such as Tth) permit higher temperatures to be used as well as ability to repeat denaturation, annealing and ligation for higher yield, while T4 DNA ligase can be used for RNA templated joining of oligonucleotide ends36,58. DNA ligases are more selective to mismatches at
The oligonucleotide ligation assay (OLA)59 and the ligase chain
reaction (LCR)60 are established methods that distinguish alleles through
ligation. OLA is being successfully used in clinical diagnostics61, but it
requires either amplification of targets prior to ligation, or amplification of ligation products62 to generate detectable signals. Parallel genotyping of tens
of loci has been demonstrated on preamplified targets in combination with gel separation of the size coded products63. Yeakely et al. applied OLA
directly on cellular mRNA, followed by PCR amplification with common primers and array readout, and were able to detect transcripts and splice variants in a highly multiplexed format36. LCR applies a thermostable ligase
for repeated rounds of annealing and ligation, and can also result in an exponential amplification if pairs of oligonucleotides complementary to both strands of the target sequence are used. LCR can be performed directly in genomic DNA60 or on preamplified targets64.
Padlock probes
A variant of OLA was developed by Nilsson et al. where the two probe ends that do not participate in ligation were connected, forming so called padlock probes65. These probes are oligonucleotides typically 70-100 nt in size,
designed to recognize a target sequence so that its two ends meet and form a nicked ligation substrate (Figure 2). Sealing of the nick converts the probe to a circular molecule, thereby changing its physical nature to a molecule different from those present before the reaction. Selection for reacted probes can then be performed by e.g. stringent washes65, exonucleolysis66 (Paper II),
or polymerization reactions66,67. The segment connecting the two target
complementary arms (termed back piece) can be used for different purposes68. Direct labeling (e.g. with dinitrophenyl-11-UTP or
diogoxigenin-11-UTP) allows circularized probes to be visualized in situ using antibodies directed against the haptens69,70. The back piece can also contain sequences
for amplification using standard primers24,66, sequences for detection by
labeled hybridization probes71,72 or molecular beacons73 (Paper III), or tag
sequences for analysis on microarrays (Paper II, III and IV).
Padlock probes are typically synthesized by phosphoramidite chemistry, but the yield is often reduced because of their length. Many impurities such as shorter, truncated or depurinated oligonucleotides are also produced and have to be removed74. Other means to generate padlock probes
is enzymatic synthesis70, or ligation of smaller segments followed by
preparative gel electrophoresis. The latter is suitable when many different probes sharing some segments have to be produced, while the former is better for longer probes (> 120 nt) and for incorporation of modified nucleotides.
Figure 2. A padlock probe with its target (light grey). Point of ligation is indicated by an arrow. A typical padlock probe is 70-100 nucleotides long of which the target-complementary arms constitute ∼20 nucleotides each.
Rolling circle amplification of padlock probes
The rolling circle mechanism is used by several viruses and plasmids for replication of their genomes, and the mechanism has later been applied on small circular oligonucleotides75,76. Polymerization is initiated from a
hybridized 3’-end and proceeds linearly for an extensive time period and is, unlike PCR, not product inhibited. Double-stranded DNA has a maximal flexure angle of 2.4 degrees per base pair, which makes circles smaller than 150 nucleotides unable to be double-stranded along the total circumference77. Therefore, a typical padlock probe would be partially
double-stranded during RCA, with nucleotide incorporation continuously accompanied by disruption of base pairs at the other end. This has been confirmed using small circles and polymerases that despite inherent 5’-3’ exonuclease activity generated long polymerization products75,76.
Polymerases with low processivity, e.g. Klenow78, are less suitable for RCA
of padlock probes, while the highly processive Φ29 DNA polymerase has proved very efficient.
The RCA product is composed of complementary repeats of the template circle, and it is single-stranded which makes the reaction easily monitored in real-time using modified molecular beacons73. Very high
precision has been obtained when amount of circular molecules were quantified using RCA, with a CV of around 2%73. Size digested products can
also be applied directly to arrays and visualized with labeled probes in a sandwich-type assay. RCA has been demonstrated on solid surfaces using immobilized primers to which circular oligonucleotides were hybridized71,79
(Jarvius, J. and Ericrsson, O. unpublished), from primers attached to immobilized antibodies72,80, to give information of cellular localization of
proteins or protein complexes (Gullberg, M et al., in preparation), and in situ81.
Lizardi et al. included in the reaction a second primer complementary to the RCA product and obtained thereby a hyperbranched rolling circle amplification (HRCA) that proceeded faster than a linear amplification71,82.
Each product monomer can in this format template extension of the second primer, which subsequently displaces upstream primers rendering these single-stranded and accessible to the first primer. This isothermal amplification is sensitive enough to detect single-copy genes in the human genome71 or low copy-number infectious agents83. HRCA produces
diffusible and partially double-stranded products detectable on agarose gels or by fluorescence24,84, but the method is less well suited for in situ or array
applications.
Multiplexed genotyping using allele-specific padlock probes
Padlock probes have proven quite useful in the area of genotyping allelic variants. The dual recognition by the probe ends provides sufficient specificity to distinguish single nucleotide variants directly in the total human genome70,71,85, and allows multiple probes to be combined in the same
ligation reaction with minimal risk of cross-reactions. Reacted padlock probes can be amplified by PCR to generate detectable signals24,66,70,84, and
by using a standard primer pair, multiple amplification products can simultaneously be differentiated through hybridization to immobilized and unique sequence tags, specific for each padlock probe. In an early design for multiplexed genotyping (later modified in Paper II), one padlock probes was made for each allele at 13 loci in the human ATP7B gene. All probes shared a common primer sequence and either of two sequences indicative of which allele was being recognized. Each pair of probes also shared a tag sequence common for the two allele-specific variants, but unique for each locus (Figure 3, box). After thermocycled ligation and exonucleolytic removal of dimerized or remaining linear probes, a PCR primer set with one common and two differently labeled allele-specific ones was used to amplify all circularized padlock probes. The PCR products were then sorted on microarrays, and genotypes determined by the fluorescence ratio at each tag-complementary array position (Figure 3). The strategy was used to genotype 29 individuals with results in complete agreement with previous data (Banér et al., in preparation). This strategy is further compared with the one presented in Paper II on page 13.
Figure 3. Multiplexed genotyping using allele-specific padlock probes. The box illustrates a pair of allele-specific padlock probes of which one is circularized on a matched target, while the other is degraded by exonucleolysis. An allele-specific fluorescent PCR primer labels the amplification products from the circularized probe, and the products are hybridized to a microarray. Thirteen genotypes from six individuals (differentiated by symbols) are plotted in the graph below. Fluorescence from each allele-specific product (labeled with FITC or TAMRA fluorophores) is expressed as a ratio of the average signal from individuals homozygous for the corresponding allele, and plotted on x- and y-axis respectively. Lines separate heterozygous genotypes from homozygous and are arbitrarily drawn. Only two fmol of each padlock probe and around 1 ng of genomic DNA is used per genotype in this experiment. Padlock probes, ligation and exonuclease protocols are identical to that described in Paper III. PCR amplification was performed by transferring six microliter of the exonuclease reactions to 24 µl PCR mixes with final concentrations of 30 mM Tris-HCl pH 8.8, 18 mM KCl, 3 mM MgCl2, 0.08% Triton X-100, 200 µM
dNTP, 400 nM of each of the three PCR primers and 0.1 U/µl HotStart Pfu polymerase (Stratagene). Reactions were placed in a thermal cycler at 95°C for 2
min, and then cycled 26 times between 98°C for 30 sec and 55°C for 2 sec. Preparation of arrays and hybridization protocol is described in Paper III.
PRESENT INVESTIGATIONS
The papers herein focus on two amplification methods of circularized padlock probes, RCA and PCR. Paper I discuss RCA of padlock probes, what effect the target strand impose on the reaction, and also suggests a mechanism how this is resolved. A method for increased sensitivity of RCA is presented in Paper III. This serial circle amplification (SCA) method results in a polynomial amplification allowing highly precise quantifications of nucleic acid targets as well as ability to perform multiplexed genotyping. Paper II presents a strategy for highly multiplexed genotyping where padlock probes are molecularly inverted prior to amplification using common PCR primers, and the products simultaneously analyzed in array formats. An application of the multiplexed genotyping strategy is presented in Paper IV where TCR Vβ repertoires are simultaneously analyzed in samples collected from tumor tissues, and compared with normal control tissues to identify potential immune responses.
Paper I. Signal amplification of padlock probes by rolling circle replication
Johan Banér1, Mats Nilsson1, Maritha Mendel-Hartvig and Ulf Landegren. (1998).
Nucleic Acids Research 26(22): 5073-5078
1 The first two authors should be regarded as joint First Authors.
This paper describes how a signal from circularized padlock probes can be amplified by a method termed rolling circle replication (RCR). It also demonstrates what influence the target strand has on the reaction, and it suggests a mechanism by which padlock probes are released from their targets during extension of the primer, given that the target’s 3’-end is located nearby the padlock ligation site.
Oligonucleotide targets of different lengths or a closed circular single-stranded M13 plasmid were used to template ligation, and either a separate primer or the target’s 3’-end was used to initiate polymerization by the Φ29 DNA polymerase. The RCR products were shown to be ligation dependent tandem repeats of monomers, complementary to the padlock probe by the appearance of unit sized restriction fragments when analyzed in
polyacrylamide gels. The Φ29 polymerization rate was calculated by comparing the incorporation of radioactive nucleotides using linear and circular padlock probes and estimated to 1,000 nt/min, while polymerization on circular probes continued for at least 12 hours with an estimated half-life of 11 hours.
We found that RCR was greatly inhibited when the target ends were located distant (∼3.6 kb) from the padlock ligation site, or totally inhibited if no free ends were present (M13 plasmid). A target with ends extending 10 nt 5’ and 3’ from the padlock hybridization site delayed polymerization slightly compared to a completely hybridized target. An explanation for this may be the 3’-5’ exonuclease activity associated with Φ29 polymerase, which could degrade protruding ends and transform them into primers as soon as they became hybridized to the padlock. However, it was not the sole explanation since five phosphorothioate bonds in the target 3’-end also permitted RCR, suggesting a target displacing activity of the Φ29 polymerase. We concluded that if padlock probes are to be amplified by RCR, the target will impose a significant block unless its 3’-end can be passed through the probe circle, a process made easier the closer it is to the padlock ligation site.
Comments on Paper I
Several groups have reported RCA of padlock probes ligated on circular or long strands of genomic DNA in the trail of this paper24,84,86. Kuhn et al.
observed some 5-fold decrease of RCA efficiency compared to unlinked circles, using Sequenase DNA polymerase and padlock probes catenated to pseudorotaxane targets. They were not able to obtain RCA using Φ29 DNA polymerase from these constructs, and therefore proposed that the use of different polymerases may explain their results86. Zhang et al. also remarked
that the different choice of polymerase is a possible explanation, although they used Bst DNA polymerase to amplify padlock probes circularized on the linear EBV genome84. Likewise, Thomas et al. used the same polymerase
to generate HRCA of padlock probes circularized on linearized plasmids24.
We were, however, unable to repeat their results either with Φ29 or Sequenase DNA polymerase and the M13 plasmid as ligation template. We cannot exclude that the explanations given by Kuhn et al. and Zhang et al. regarding polymerases are correct, especially concerning Bst DNA polymerase. If the probes remained linked to their targets in these experiments, it is unclear how a polymerase could pursue polymerization under such complex topological conditions.
Paper II. Highly multiplexed genotyping with sequence-tagged molecular inversion probes and DNA microarrays
Paul Hardenbol, Johan Banér, Maneesh Jain, Mats Nilsson, Eugeni A. Namsaraev, George A. Karlin-Neumann, Hossein Fakhrai-Rad, Mostafa Ronaghi, Thomas D. Willis, Ulf Landegren and Ronald W. Davis. (Accepted in Nature Biotechnology).
This paper describes a novel strategy for highly multiplexed genotyping using a variant of padlock probes, the so-called molecular inversion probes. The approach is shown to enable simultaneous typing of more than 1,000 probes in a single reaction tube by amplifying reacted probes with common PCR primers and sorting the products on universal tag sequence DNA chips. Exonuclease conditions for enzymatic selection of circularized padlock probes are also presented.
Initially we designed two padlock probes for each locus with the 3’-ends positioned at the polymorphic nucleotide as presented on page 8. A small-scale study successfully demonstrated that the method could be applied for multiplexed genotyping. To enable a large scale-up, the need for probes was halved by redesigning the probes to locus-specific ones (instead of as before, allele-specific). The 3’ variant nucleotide was now left out, creating a gap between the target complementary arms. The gap was sealed in four separate allele-specific polymerization and ligation reactions in presence of one of four nucleotides. Circularized probes were then released from the genomic DNA by cleaving the probes at uracil residues situated between the common primer sequences, in effect inverting the probes before amplification. After separate PCR amplifications, the products were applied to separate DNA chips and each genotype obtained by comparing the signals from the respective reactions.
Assay performance was determined by selecting 1121 SNPs located in a linkage peak for IgA nephropathy87. Sixteen percent of the synthesized
probes were nonfunctional for any of several reasons, including errors in the database, probe design, probe synthesis failure, or failure of the assay itself. Call rate using the active 938 probes was 95% determined from 25 different individuals assayed with a median signal to noise ratio of 17. Accuracy was measured by concordance with independent Sanger sequencing to 99.4%, and a similar accuracy was achieved by comparing with Pyrosequencing data. Additional assay performances were determined including repeatability and scalability, as well as performance in a two-color detection format on two chips.
The intramolecular nature of these dual recognition probes, assisted by a combination of nucleotide-selective incorporation and ligation steps, allows higher multiplexing than any other current approach. Circularization requires that the molecule is extended by a single nucleotide in an
allele-specific reaction, and misincorporations may still be discriminated as poor ligation substrates. The exonuclease treatment greatly suppresses detection of failures in the previous reaction steps. The molecular inversion strategy offers inherent quality control since probe damage or loss of function affects both alleles equally, achieved by monitoring all four nucleotide calls. Assay costs are amortized by the high degree of multiplexing involved, leading to an inexpensive assay.
Comments on Paper II
Two similar strategies of multiplex genotyping using padlock probes have evolved: one employing allele-specific padlock probes (described on page 8) and the other employing locus-specific probes (Paper II). The two differ in a few aspects but are based on the same principle; sets of padlock probes are simultaneously allowed to react directly with genomic DNA followed by amplification of circular probes using a common primer pair, and the products are analyzed on microarrays. The multiplex PCR problem (discussed on page 3) is thereby circumvented. However, the differences between the two strategies need a few comments.
Because of the basic probe design, half as many oligonucleotides can be synthesized using locus-specific probes, with reduced reagent costs as a benefit. One probe per polymorphism eliminates need for optimizing individual probe imbalances, and failure of probe construction will affect both allele calls equally, resulting in a failed call rather than an incorrect genotype.
Allele discrimination is accomplished either by ligation or in a combined gap-fill and ligation reaction, respectively. Two selective enzyme reactions are likely to be more discriminative than one. However, in the latter strategy the reactions must be performed in separate nucleotide-specific tubes for each individual, thereby consuming two or fourfold the amount of probes, reagents and samples. Genotyping using allele-specific padlock probes is performed in a single vessel per individual. Moreover, the inversion strategy does not allow thermocycled ligation for higher yield, although the genomic DNA needs no digestion prior to ligation.
Amplification of reacted probes is also performed slightly differently. If the genotyping reaction has been performed with locus-specific probes and four separate nucleotide incorporations, there is only need for one fluorophore-labeled PCR primer, but this format then requires four separate arrays. While both the corresponding two-color based format and the strategy using allele-specific probes require differently labeled primers, the former needs two chips per individual and a post-staining procedure. The allele-specific probe strategy presented on page 8 was performed using
in-house made arrays allowing 12 individuals to be analyzed on the same slide. About 500 genotypes per hybridization chamber can be analyzed using this format, while a 56-well format allows 100 genotypes to be analyzed per chamber53.
Paper III. Serial circle amplification – A novel quantitative DNA amplification method for genotyping and expression analyses
Fredrik Dahl, Johan Banér, Mats Gullberg, Maritha Mendel-Hartvig, Jonas Jarvius, Ulf Landegren and Mats Nilsson. (Manuscript).
Paper III presents a novel method to amplify padlock probes based on rolling circle amplification, termed serial circle amplification (SCA). The method is described in detail, and its application in multiplexed genotyping and quantitative expression analysis is demonstrated.
In SCA, individual monomers of the rolling circle polymerization product are via a ‘restriction cassette’ converted to new circular molecules, and these are subsequently used as replication templates in further polymerizations. The process may then be repeated if desired. After an initial RCA, a complementary oligonucleotide is added along with a restriction enzyme, digesting the polymerization product to monomers that in turn can be recircularized using the same oligonucleotide as a template. Through new rounds of polymerization, digestion and recircularization, a polynomial amplification results that is directly proportional to the initial number of circles and duration of each polymerization event. The final product is single-stranded and readily available for detection with molecular beacons or by array hybridization. Recircularization is unlikely to give rise to cross-reactive products because of the intramolecular nature of padlock probes, which makes the method useful for multiplexed SNP or expression studies.
In radioactive experiments, both the 1st and 2nd generation SCA
generated a conversion rate of RCA products to circles of at least 99%. This was confirmed in real-time by the similar increase of first and third generation RCA products from circularized probes not diluted, and a million fold diluted, respectively. Molecular beacons with complementary polarities to those previously used demonstrated no evidence of hyperbranched rolling circle amplification (HRCA). Thermal profiles of the amplifications were also compared with a true HRCA in the presence of a double-stranded intercalating dye. A sharp increase in fluorescence was observed from the HRCA reaction when temperature dropped below 85°C, indicative of high
molecular double-stranded DNA, while no such profile was observed in any SCA reaction.
To investigate if SCA could be used for multiplexed genotyping, a set of 24 different padlock probes were circularized using oligonucleotide targets and amplified either by RCA or a 7-cycle PCR, or diluted a million fold and amplified by SCA or a 27-cycle PCR. Polymerization products from SCAs were monomerized and the four reactions hybridized to microarrays. The relative representation of each product diverged only modestly between the 1st and 3rd generation of amplification, indicating that
different circles were amplified with equivalent efficiency during SCA. Similar divergence was also seen from the reactions amplified by PCR. The same probe set was therefore used to genotype ten genomic DNA samples, and the results were in concordance with previous data demonstrating that SCA is applicable to multiplexed genotyping studies.
The sensitivity of SCA was examined using a dilution series of pre-formed test circles ranging from 7,500 to 120,000 molecules, with a reference circle present in the 3rd generation of amplification. Two
differently labeled molecular beacons were used to monitor the amplifications. Proportional increase in signal from the test circles was detected with an average CV of 25%.
The method is an attractive alternative to other amplification techniques because of its low cross-reactivity and easily tuned level of amplification.
Comments on Paper III
Contrary to PCR, the progression of RCA is unaffected by previously synthesized products. SCA can give rise to a quite viscous solution, indicating that reagents suffice for production of µg/µl of DNA. However, during recircularization in SCA, high product concentrations may lead to dimerization and thereby reduce the over-all amplification rate. This is because the conversion efficiency of monomers to new circles depends on the total concentration of monomer ends present in the reaction. Once a monomer end has hybridized to a target oligonucleotide, the local concentration of the other end can be estimated at 2.8 µM, assuming 100-mer padlock probes with 20 nt target-complementary arms88. A total
monomer concentration in that order will result in approximately equal fraction of circularized and dimerized ligation products. For maximal product yield, the last recircularization should therefore be conducted at concentrations below the theoretical circularization vs. dimerization threshold, although also other factors are expected to limit synthesis at this stage, e.g. the concentration of dNTP or inhibition by pyrophosphates.
Paper IV. Parallel analysis of TCR Vβ gene expression using padlock
probes and microarrays
Johan Banér, Per Marits, Mats Nilsson, Fredrik Dahl, Ola Winqvist and Ulf Landegren. (Manuscript).
Background on TCR
T-cells play an important role in the immune system through their ability to recognize antigenic peptides via the T-cell receptor (TCR). The majority of cells produce TCRs constituted of αβ heterodimeric chains, but a minor proportion of cells also express γδ chains. During T-cell development the TCR β locus is rearranged to include combinations of variable (V), diversity (D) and joining (J) segments. Transcripts of V genes are then spliced to a constant (C) region to generate a functional mRNA89. The CDR3 region is
formed by the conjunction of V-J (for TCR α and γ) or V-D-J (for TCR β and δ), and is thought to mediate recognition of MHC-bound peptides89.
T-cells circulating the lymphatic system constitute a repertoire derived from the 34 expressed subfamilies of Vβ genes. This distribution is skewed upon encounter with e.g. tumor specific antigens, which cause clonal expansion of T-cells expressing a particular TCR Vβ gene. Comparison of Vβ distributions from different tissues traditionally involves subfamily-specific PCRs, but analysis of the entire repertoire render these assays labor intensive.
Results of Paper IV
Paper IV describes a method for parallel expression analyses of Vβ gene segments in TCRs using 25 Vβ subfamily-specific padlock probes. Each probe contained a common primer sequence and a family-specific sequence used for array hybridization. Probe function and array performance were investigated using fragmented B-cell DNA, resulting in standard deviations of individual Vβ-family representations ranging from 0.3 to 2.8%. A threshold for a changed distribution was set to three standard deviations compared to respective Vβ-family in a control sample. To investigate if the method could detect differential changes in Vβ repertoires, B-cell genomic DNA was spiked with increasing amounts of oligonucleotide targets for some selected padlock probes. All additions resulted in distribution changes greater than the thresholds for concerned Vβ, respectively. In general, four-fold increase in added target amounts within the series could be differentiated for all concerned Vβ, and in some cases also two-fold increases.
We then cultured two lymph node samples in vitro with or without Staphylococcal enterotoxin B (SEB), a superantigen known to specifically induce expression of Vβ3, Vβ12, Vβ14, Vβ15, Vβ17 and Vβ20 (REF. 90),
and applied the method on prepared cDNA. Significant increase in representation of Vβ3, Vβ14, Vβ15, Vβ17, and Vβ20 was seen in lymph node 1, and Vβ14 and Vβ20 in nymph node 2. We also isolated T-cells from different tissues of a bladder cancer patient and prepared cDNA. The Vβ repertoire from tumor infiltrating lymphocytes (TILs) was compared with that from a tumor draining lymph node (TDLN), and also with that from an irrelevant lymph node (Ir LN). Predominant use of Vβ2 and Vβ23 was seen in the samples from TILs and TDLN, suggesting a tumor specific response of these families. This profile was not seen in the irrelevant lymph node sample.
The method is thus a promising tool for identifying potential tumor reactive T-cell clones by comparing Vβ repertoires from different tissues. We believe that this method will enable higher throughput in analyses of Vβ repertoires compared with other assays.
Comments on Paper IV
V-gene distributions are also being evaluated in flow-cytometric analyses, using fluorescence labeled monoclonal antibodies against Vβ chains. Up to four-color cytometric assays can be performed by combining differently labeled monoclonal antibodies to gain information about the Vβ repertoire91.
Drawbacks of this method include the limited number of fluorophores that can be used simultaneously, instability of antibodies during storage and the limited availability of monoclonal antibodies for other chains, especially the diverse Vα chain. The commonly employed PCR approach is used for Vβ, Vγ and Vδ expression, while the polymorphic nature of the α locus has restricted diagnostic analyses of these TCRs91.
The strategy employed in Paper IV allows all Vβ families to be analyzed in the same reaction tube by amplifying circularized padlock probes with a common PCR primer pair. Although a substantial mix of different templates, e.g. linear oligonucleotides36 or padlock probes (Paper II
and IV), can be amplified simultaneously, the use of common primers will eventually end the entire amplification through formation of hybrids between different products. Although this effect preserves the relative concentration of products, low abundant ones might go undetected. Also, solid-phase hybridization of abundant amplification products will saturate their corresponding microarray spots, and therefore only report binding capacity of that spot. Transcripts differing in concentration over a broad range are therefore difficult to detect quantitatively using end-point measurements on arrays. An alternative strategy is to use solid phase RCA of circularized padlock probes, further outlined under Discussion.
DISCUSSION
Applications of RCA
Rolling circle amplification of circular padlock probes is readily performed in solution, but also in situ from cytological targets or from immobilized primers. Nallur et al. reported a 65% efficiency of hybridization and priming of circular oligonucleotides from a solid phase compared to conventional oligonucleotide hybridizations, and the amplification reaction was observed to proceed for at least two hours79. They also reported a 1,000-fold increase
in sensitivity on porous surfaces, such as the CodeLink slides used in Paper II, III and IV.
Solid phase RCA would therefore be advantageous to use for expression studies, and could offer increased dynamic range compared to traditional array-based techniques. For example, the padlock probes used to investigate Vβ distributions in Paper IV can, after circularization and exonuclease treatment, be hybridized to immobilized oligonucleotides on an array. A subsequent amplification gives information not only about which Vβ transcripts were present, but also their relative amounts over an extensive range by reading the array for total fluorescence. Preliminary data indicate a two-hundred fold higher sensitivity using RCA of pre-circularized padlock probes, compared to hybridization alone (Eriksson, O. et al. unpublished). Nevertheless, scarce amounts of transcripts may still result in low concentration of circularized probes and undetectable signals. A preceding amplification step is therefore desirable, and the SCA method is ideal for this purpose for several reasons, but mainly the ability to regenerate circular molecules.
In situ studies require not only sensitive detection methods but also localized signals, practically excluding PCR, although some successful studies have been presented92. The major limitations of in situ PCR have
been diffusible products and high background93. Padlock probes are useful
reagents for in situ studies because of its high specificity and resistance to superstringent washes65,69,70, but also because of the possibility to generate
hybridization probes. However, for reasons described in Paper I, a primer-initiated RCA requires a nearby target end to enable release of the padlock probe, thereby risking loss of the positional information. On the other hand, by covalently attaching the product to the target itself, truly localized signal amplification can be obtained that is highly resistant to washes. To achieve this, a physical break must be created downstream of the target sequence onto which a padlock probe has been circularized. Using the 3’-5’ exonucleolytic activity of Φ29 DNA polymerase, the protruding 3’-end can be degraded and eventually used as primer with the circularized padlock probe as template94. This procedure is expected to have specificity
advantages compared to related techniques81,95.
Multiplex genetic studies
As mentioned in the introduction, interest has been drawn to whole-genome association studies using SNPs as markers to identify genetic variants associated with an inherited predisposition for disease. Development of methods for highly multiplex genetic studies has unfortunately been slow, mainly due to the dependency of preceding PCR amplifications. To date, no method exists with acceptable capacity or cost per SNP to undertake such large-scale studies.
On the other hand, several commercial vendors have demonstrated promising genotyping capacity. In a highly multiplex assay, where allele-specific primers were extended and ligated to addresser oligonucleotides followed by common-primer PCR amplification, more than 1,000 SNPs were analyzed simultaneously in an array format62. Notably, in this assay, as
is the case when using padlock probes, the selective reactions are performed directly in genomic DNA. Thereby, the limitations imposed by multiplex PCR may be avoided. As discussed above, padlock probes are expected to have advantages over assays requiring two independent probe hybridizations. In fact, recent advances now permits 7,500 genotypes to be analyzed in parallel using the molecular inversion strategy (Tom Willis, ParAllele Bioscience, personal communication).
Nevertheless, further scale-up is required to realize whole-genome analyses in single runs. An attempt to achieve this using the multiplex padlock probe strategy could be to design the probes to separate pools. A common primer pair amplifies all probes within a pool while each pool uses different primer pairs, thereby creating a multiplex of multiplex genotyping assay. A critical step in performing multiplex PCR concerns primer design, but the multiplex of multiplex strategy would be alleviated by the liberty to select primer pairs for minimal cross-reactivity. Moreover, this strategy also
requires that ten-fold more PCR product is generated compared to the simplex reaction.
Whether the SCA method is applicable at high multiplexing levels will be influenced by a number of factors. The efficiency of circularizing padlock probes in a multiplexed ligation reaction will depend on, among other things, concentrations of initial probes and the multiplexing level. The resulting amount of circular molecules will determine the total concentration of products, so that lower number of initial circles per locus allows a proportionally higher multiplexing level. Amount of products required for array detection will determine the level of amplification needed, thereby affecting preceding reaction steps. Therefore, the SCA method may prove capable of higher multiplexing levels than any method to date because of its ability to produce large amounts of DNA.
DNA microarray hybridization is currently the superior technique to analyze large sets of distinct nucleic acid molecules27. One of the limiting
factors that arise in large-scale genotyping analyses is the availability of suitable tag sequences. In the selection procedure described by Shoemaker et al.32, only 50,000 tags were approved out of the total number of 20-mers
(∼1012); those with similar hybridization characteristics but minimal
cross-reactivity. Thus, tag length may have to be increased to suffice for whole-genome analysis. Furthermore, padlock probe design and synthesis are demanding in an extensive scale-up, which has to be approached by automated and computer-assisted processes for sequence optimization and tag selection (Stenberg, J et al. in preparation).
ACKNOWLEDGEMENTS
This work has been carried out at the department of Genetics and Pathology, first at BMC and later at the Rudbeck Laboratory. I am grateful to many people who have contributed in different ways.
I would like to thank Ulf Landegren for giving me the opportunity to work in his group. Ulf has been an excellent supervisor and an endless source of ideas, creating an exciting working environment.
Thanks to all members of the department, especially Ulf Pettersson, Ulf Gyllensten, Britt-Marie Carlberg, Lena Åslund, William Schannong and Viktor Persson, Tommy Östlund and Eddie Lundh, Elisabeth Sandberg and Maria Hedefalk. Thanks also to the former members Jeanette Backman for all administrative help, and Elsy Johnsen for early help in the lab and golf discussions.
Marek Kwiatkowski for letting me start off as summer worker and giving me insight in the world of chemistry, and Anders Isaksson for good collaboration and discussions.
The members of Ulf’s group: Mats Nilsson for being an excellent co-supervisor and stimulating colleague, Simon Gummisvingen Fredriksson for great fun during these years even though you finished earlier than me, and Mats Gullberg for sharing office during the last half-year. Thanks to Jonas Jarvius for all possible assistance and for saving the old Ford from the graveyard. Fredrik Dahl, formally known as Kungen av Escobar, for continuous production of cakes and for great golf, although you need more practice. Sigrun Gustafsdottir for enlightening the parties, Johan Stenberg, Olle Eriksson, Ola Söderberg, Jonas Melin, Adam Strömstedt, Henrik Johansson, for making this group very pleasant. Thanks also to the former members, Maritha Mendel-Hartvig, Dan-Oscar Antson, Lotta Olsson, and finally Anette Hagberg for being a nice golf partner and an excellent person to trick.
Thanks to all former and present colleagues at the department; Eva Lindholm, Kristina Lagerstedt, Anh-Nhi Tran, Maja Nyström, Álvaro Rada, Andrej Semenyuk, Benjamin Bakall, Joakim Klar, Hans Matsson, Mattias Jansson, Miriam Entesarian, Larry Mansouri, Mårten Fryknäs, Malin Larsson, Ulrika Wickenberg, Hanna Andréasson, Anna-Maria Divne, Malin Engelmark, Martin Moberg, Martina Nilsson, Patrik Magnusson, Max Ingman, Veronica Magnusson, Cecilia Johansson, Bo Johannesson, Niklas Nordquist, Jenny Jonsson, Fredrik Hjelm, Dieter Fuchs, Marti Tammi, Charlotte Stenh, Åsa Johansson and Jenny von Salomé.
The people at Akademiska Sjukhuset; Ann-Christine Syvänen, Erik Waldenström, Ulrika Liljedahl, Sneivar Sigurdsson, and Katarina Lindroos. Thanks also to Per Marits for all Vβ.
I would like to express gratitude to all people I have met during my years in Uppsala, and also to Stockholm Nation where I have spent a few Thursday evenings. Fredrik de Geer for happily sharing a flat at Studentvägen, keeping count of toast consumption and confirming the rumor that my cooking is better, and for maintaining presence of kapten, who also Ludwig Holmgren deserves credit for. Thanks to Fredrik Österberg for being an outstanding former labpartner, gofpartner and for your endless stream of omtentor. I would like to thank Öfvre Fjellstedtska Studenthemmet for providing excellent housing during these years and all fun people who stayed there. Especially Per Schrotti, Klas Winberg, Gerhard de Geer, Claes Holmqvist, Richard Torgersson, Madeleine Koskull and Peter Tårtan von Satzger.
My work during these years has been made possible by generous grants from The Beijer Foundation, the Swedish Research Council for Natural and Engineering Sciences, the Swedish Cancer Fund, and by Polysaccharide Research AB (Uppsala).
REFERENCES
1. Watson, J.D. & Crick, F.H.C. Molecular structure of nucleic acids. A structure for deoxyribose nucleic acid. Nature 171, 737-738 (1953). 2. Baxevanis, A.D. The Molecular Biology Database Collection: 2003
update. Nucleic Acids Res 31, 1-12 (2003). 3. dbSNP, http://www.ncbi.nlm.nih.gov/SNP/.
4. Evans, W.E. & Relling, M.V. Pharmacogenomics: translating functional genomics into rational therapeutics. Science 286, 487-491 (1999). 5. Hacia, J.G. et al. Determination of ancestral alleles for human
single-nucleotide polymorphisms using high-density oligosingle-nucleotide arrays. Nat Genet 22, 164-167 (1999).
6. Risch, N. & Merikangas, K. The future of genetic studies of complex human diseases. Science 273, 1516-1517 (1996).
7. Collins, F.S., Guyer, M.S. & Charkravarti, A. Variations on a theme: cataloging human DNA sequence variation. Science 278, 1580-1581 (1997).
8. Patil, N. et al. Blocks of limited haplotype diversity revealed by high-resolution scanning of human chromosome 21. Science 294, 1719-1723. (2001).
9. Daly, M.J., Rioux, J.D., Schaffner, S.F., Hudson, T.J. & Lander, E.S. High-resolution haplotype structure in the human genome. Nat Genet
29, 229-232 (2001).
10. Johnson, G.C. et al. Haplotype tagging for the identification of common disease genes. Nat Genet 29, 233-237 (2001).
11. Edman, C.F. et al. Electric field directed nucleic acid hybridization on microchips. Nucleic Acids Res 25, 4907-4914 (1997).
12. Syvänen, A.C. Accessing genetic variation: genotyping single nucleotide polymorphisms. Nat Rev Genet 2, 930-942. (2001). 13. Saiki, R.K. et al. Enzymatic amplification of beta-globin genomic
sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 230, 1350-1354. (1985).
14. Pastinen, T. et al. A system for specific, high-throughput genotyping by allele-specific primer extension on microarrays. Genome Res 10, 1031-1042. (2000).
15. Broude, N.E., Zhang, L., Woodward, K., Englert, D. & Cantor, C.R. Multiplex allele-specific target amplification based on PCR
16. Landegren, U. & Nilsson, M. Locked on target: Strategies for future gene diagnostics. Ann. Med. 29, 585-590 (1997).
17. McDowell, D.G., Burns, N.A. & Parkes, H.C. Localised sequence regions possessing high melting temperatures prevent the amplification of a DNA mimic in competitive PCR. Nucleic Acids Res 26, 3340-3347 (1998).
18. Lee, L.G., Connell, C.R. & Bloch, W. Allelic discrimination by nick-translation PCR with fluorogenic probes. Nucleic Acids Res 21, 3761-3766 (1993).
19. Livak, K.J., Flood, S.J., Marmaro, J., Giusti, W. & Deetz, K. Oligonucleotides with fluorescent dyes at opposite ends provide a quenched probe system useful for detecting PCR product and nucleic acid hybridization. PCR Methods Appl 4, 357-362 (1995).
20. Fredriksson, S. et al. Protein detection using proximity-dependent DNA ligation assays. Nat Biotechnol 20, 473-477 (2002).
21. Tyagi, S. & Kramer, F.R. Molecular beacons: probes that fluoresce upon hybridization. Nat Biotechnol 14, 303-308 (1996).
22. Whitcombe, D., Theaker, J., Guy, S.P., Brown, T. & Little, S. Detection of PCR products using self-probing amplicons and fluorescence. Nat Biotechnol 17, 804-807. (1999).
23. Nazarenko, I.A., Bhatnagar, S.K. & Hohman, R.J. A closed tube format for amplification and detection of DNA based on energy transfer. Nucleic Acids Res 25, 2516-2521 (1997).
24. Thomas, D.C., Nardone, G.A. & Randall, S.K. Amplification of padlock probes for DNA diagnostics by cascade rolling circle
amplification or the polymerase chain reaction. Arch Pathol Lab Med
123, 1170-1176 (1999).
25. Schena, M., Shalon, D., Davis, R.W. & Brown, P.O. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270, 467-470 (1995).
26. Lockhart, D.J. et al. Expression monitoring by hybridization to high-density oligonucleotide arrays. Nat Biotechnol 14, 1675-1680. (1996). 27. Giaever, G. et al. Functional profiling of the Saccharomyces cerevisiae
genome. Nature 418, 387-391 (2002).
28. Pease, A.C. et al. Light-generated oligonucleotide arrays for rapid DNA sequence analysis. Proc Natl Acad Sci U S A 91, 5022-5026 (1994). 29. Kwiatkowski, M., Fredriksson, S., Isaksson, A., Nilsson, M. &
Landegren, U. Inversion of in situ synthesized oligonucleotides: improved reagents for hybridization and primer extension in DNA microarrays. Nucleic Acids Res 27, 4710-4714. (1999).
30. Pirrung, M.C. How to make a DNA chip. Angew Chem Int Edit 41, 1277-+ (2002).
31. Lipshutz, R.J., Fodor, S.P., Gingeras, T.R. & Lockhart, D.J. High density synthetic oligonucleotide arrays. Nat Genet 21, 20-24 (1999). 32. Shoemaker, D.D., Lashkari, D.A., Morris, D., Mittmann, M. & Davis,
highly parallel molecular bar-coding strategy. Nat Genet 14, 450-456. (1996).
33. Gerry, N.P. et al. Universal DNA microarray method for multiplex detection of low abundance point mutations. J Mol Biol 292, 251-262 (1999).
34. Wetmur, J.G. DNA probes: applications of the principles of nucleic acid hybridization. Crit Rev Biochem Mol Biol 26, 227-259 (1991).
35. Brenner, S. et al. Gene expression analysis by massively parallel signature sequencing (MPSS) on microbead arrays. Nat Biotechnol 18, 630-634 (2000).
36. Yeakley, J.M. et al. Profiling alternative splicing on fiber-optic arrays. Nat Biotechnol 20, 353-358. (2002).
37. Golub, T.R. et al. Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science 286, 531-537 (1999).
38. Lyamichev, V. et al. Polymorphism identification and quantitative detection of genomic DNA by invasive cleavage of oligonucleotide probes. Nat Biotechnol 17, 292-296 (1999).
39. Livak, K.J. Allelic discrimination using fluorogenic probes and the 5' nuclease assay. Genet Anal 14, 143-149 (1999).
40. Beaudet, L., Bedard, J., Breton, B., Mercuri, R.J. & Budarf, M.L. Homogeneous assays for single-nucleotide polymorphism typing using AlphaScreen. Genome Res 11, 600-608 (2001).
41. Wallace, R.B. et al. Hybridization of synthetic
oligodeoxyribonucleotides to phiX174 DNA: the effect of single base pair mismatch. Nucleic Acids Res. 6, 3543-3557 (1979).
42. Tyagi, S., Bratu, D.P. & Kramer, F.R. Multicolor molecular beacons for allele discrimination. Nat Biotechnol 16, 49-53 (1998).
43. Marras, S.A., Kramer, F.R. & Tyagi, S. Multiplex detection of single-nucleotide variations using molecular beacons. Genet Anal 14, 151-156 (1999).
44. Lee, L.G. et al. Seven-color, homogeneous detection of six PCR products. Biotechniques 27, 342-349 (1999).
45. Tyagi, S., Marras, S.A. & Kramer, F.R. Wavelength-shifting molecular beacons. Nat Biotechnol 18, 1191-1196 (2000).
46. Syvänen, A.C., Aalto-Setala, K., Harju, L., Kontula, K. & Soderlund, H. A primer-guided nucleotide incorporation assay in the genotyping of apolipoprotein E. Genomics 8, 684-692. (1990).
47. Syvänen, A.C., Sajantila, A. & Lukka, M. Identification of individuals by analysis of biallelic DNA markers, using PCR and solid-phase minisequencing. Am J Hum Genet 52, 46-59 (1993).
48. Pastinen, T., Partanen, J. & Syvänen, A.C. Multiplex, fluorescent, solid-phase minisequencing for efficient screening of DNA sequence
variation. Clin Chem 42, 1391-1397 (1996).
49. Chen, X. & Kwok, P.Y. Template-directed dye-terminator
based on fluorescence resonance energy transfer. Nucleic Acids Res 25, 347-353 (1997).
50. Chen, X., Zehnbauer, B., Gnirke, A. & Kwok, P.Y. Fluorescence energy transfer detection as a homogeneous DNA diagnostic method. Proc Natl Acad Sci U S A 94, 10756-10761 (1997).
51. Pastinen, T., Kurg, A., Metspalu, A., Peltonen, L. & Syvänen, A.C. Minisequencing: a specific tool for DNA analysis and diagnostics on oligonucleotide arrays. Genome Res 7, 606-614 (1997).
52. Shumaker, J.M., Metspalu, A. & Caskey, C.T. Mutation detection by solid phase primer extension. Hum Mutat 7, 346-354 (1996).
53. Lindroos, K., Sigurdsson, S., Johansson, K., Ronnblom, L. & Syvänen, A.C. Multiplex SNP genotyping in pooled DNA samples by a four-colour microarray system. Nucleic Acids Res 30, e70 (2002). 54. Nyrén, P., Pettersson, B. & Uhlen, M. Solid phase DNA
minisequencing by an enzymatic luminometric inorganic pyrophosphate detection assay. Anal Biochem 208, 171-175 (1993).
55. Pourmand, N., Elahi, E., Davis, R.W. & Ronaghi, M. Multiplex Pyrosequencing. Nucleic Acids Res 30, e31 (2002).
56. Luo, J., Bergstrom, D.E. & Barany, F. Improving the fidelity of Thermus thermophilus DNA ligase. Nucleic Acids Res. 24, 3071-3078 (1996).
57. Nilsson, M. et al. Making ends meet in genetic analysis using padlock probes. Hum Mutat 19, 410-415. (2002).
58. Nilsson, M., Barbany, G., Antson, D.O., Gertow, K. & Landegren, U. Enhanced detection and distinction of RNA by enzymatic probe ligation. Nat Biotechnol 18, 791-793 (2000).
59. Landegren, U., Kaiser, R., Sanders, J. & Hood, L. A ligase-mediated gene detection technique. Science 241, 1077-1080 (1988).
60. Barany, F. Genetic disease detection and DNA amplification using cloned thermostabile ligase. Proc. Natl. Acad. Sci. USA 88, 189-193 (1991).
61. Dequeker, E. & Cassiman, J.J. Genetic testing and quality control in diagnostic laboratories. Nat Genet 25, 259-260 (2000).
62. Oliphant, A., Barker, D.L., Stuelpnagel, J.R. & Chee, M.S. BeadArray technology: enabling an accurate, cost-effective approach to high-throughput genotyping. Biotechniques Suppl, 56-58, 60-51 (2002). 63. Grossman, P.D. et al. High-density multiplex detection of nucleic acid
sequences: oligonucleotide ligation assay and sequence-coded separation. Nucleic Acids Res 22, 4527-4534 (1994).
64. Day, D.J., Speiser, P.W., White, P.C. & Barany, F. Detection of steroid 21-hydroxylase alleles using gene-specific PCR and a multiplexed ligation detection reaction. Genomics 29, 152-162 (1995).
65. Nilsson, M. et al. Padlock probes: circularizing oligonucleotides for localized DNA detection. Science 265, 2085-2088 (1994).
66. Zhang, D.Y., Brandwein, M., Hsuih, T.C. & Li, H. Amplification of target-specific, ligation-dependent circular probe. Gene 211, 277-285 (1998).
67. Banér, J., Nilsson, M., Mendel-Hartvig, M. & Landegren, U. Signal amplification of padlock probes by rolling circle replication. Nucleic Acids Res 26, 5073-5078 (1998).
68. Banér, J. et al. More keys to padlock probes: mechanisms for high-throughput nucleic acid analysis. Curr Opin Biotechnol 12, 11-15 (2001).
69. Nilsson, M. et al. Padlock probes reveal single-nucleotide differences, parent of origin and in situ distribution of centromeric sequences in human chromosomes 13 and 21 [see comments]. Nat Genet 16, 252-255 (1997).
70. Antson, D.O., Isaksson, A., Landegren, U. & Nilsson, M.
PCR-generated padlock probes detect single nucleotide variation in genomic DNA. Nucleic Acids Res 28, E58 (2000).
71. Lizardi, P.M. et al. Mutation detection and single-molecule counting using isothermal rolling-circle amplification. Nat Genet 19, 225-232 (1998).
72. Schweitzer, B. et al. Immunoassays with rolling circle DNA
amplification: A versatile platform for ultrasensitive antigen detection. Proc Natl Acad Sci U S A 97, 10113-10119 (2000).
73. Nilsson, M., Gullberg, M., Dahl, F., Szuhai, K. & Raap, K. Real-time monitoring of rolling-circle amplification using a modified molecular beacon design. Nucleic Acids Res 30, 1-6 (2002).
74. Kwiatkowski, M., Nilsson, M. & Landegren, U. Synthesis of full-length oligonucleotides: cleavage of apurinic molecules on a novel support. Nucleic Acids Res 24, 4632-4638 (1996).
75. Fire, A. & Xu, S.Q. Rolling replication of short DNA circles. Proc Natl Acad Sci U S A 92, 4641-4645 (1995).
76. Liu, D., Daubendiek, S.L., Zillman, M.A., Ryan, K. & Kool, E.T. Rolling circle DNA synthesis: small oligonucleotides as efficient templates for DNA polymerases. J. Am. Chem. Soc. 118, 1587-1594 (1996).
77. Crothers, D.M., Drak, J., Kahn, J.D. & Levene, S.D. DNA binding, flexibility, and helical repeat by cyclization kinetics. Methods Enzymol.
212, 3-29 (1992).
78. Polesky, A.H., Steitz, T.A., Grindley, N.D.F. & Joyce, C.M. Identification of residues critical for the polymerase activity of the Klenow fragment of DNA polymerase I from Escherichia coli. J. Biol. Chem. 265, 14579-14591 (1990).
79. Nallur, G. et al. Signal amplification by rolling circle amplification on DNA microarrays. Nucleic Acids Res 29, E118 (2001).
80. Kingsmore, S.F. & Patel, D.D. Multiplexed protein profiling on antibody-based microarrays by rolling circle amplification. Curr Opin Biotechnol 14, 74-81 (2003).
81. Christian, A.T. et al. Detection of DNA point mutations and mRNA expression levels by rolling circle amplification in individual cells. Proc Natl Acad Sci U S A 98, 14238-14243 (2001).
82. Isaksson, A. & Landegren, U. Accessing genomic information: alternatives to PCR. Curr Opin Biotechnol 10, 11-15. (1999). 83. Zhang, W. et al. Detection of Chlamydia trachomatis by isothermal
ramification amplification method: a feasibility study. J Clin Microbiol
40, 128-132 (2002).
84. Zhang, D.Y., Zhang, W., Li, X. & Konomi, Y. Detection of rare DNA targets by isothermal ramification amplification. Gene 274, 209-216 (2001).
85. Faruqi, A.F. et al. High-throughput genotyping of single nucleotide polymorphisms with rolling circle amplification. BMC Genomics 2, 4 (2001).
86. Kuhn, H., Demidov, V.V. & Frank-Kamenetskii, M.D. Rolling-circle amplification under topological constraints. Nucleic Acids Res 30, 574-580 (2002).
87. Gharavi, A.G. et al. IgA nephropathy, the most common cause of glomerulonephritis, is linked to 6q22-23. Nat Genet 26, 354-357 (2000). 88. Nilsson, M. in Genetics and Pathology 50 (Uppsala University,
Uppsala; 1998).
89. Davis, M.M. & Bjorkman, P.J. T-cell antigen receptor genes and T-cell recognition. Nature 334, 395-402 (1988).
90. Choi, Y.W. et al. Interaction of Staphylococcus aureus toxin "superantigens" with human T cells. Proc Natl Acad Sci U S A 86, 8941-8945 (1989).
91. Hodges, E., Krishna, M.T., Pickard, C. & Smith, J.L. Diagnostic role of tests for T cell receptor (TCR) genes. J Clin Pathol 56, 1-11 (2003). 92. Harrer, T., Schwinger, E. & Mennicke, K. A new technique for cyclic in
situ amplification and a case report about amplification of a single copy gene sequence in human metaphase chromosomes through
PCR-PRINS. Hum Mutat 17, 131-140 (2001).
93. Long, A.A. & Komminoth, P. In situ PCR. An overview. Methods Mol Biol 71, 141-161 (1997).
94. Mendel-Hartvig, M. in Genetics and Pathology 41 (Uppsala University, Uppsala; 2002).
95. Zhong, X.B., Lizardi, P.M., Huang, X.H., Bray-Ward, P.L. & Ward, D.C. Visualization of oligonucleotide probes and point mutations in interphase nuclei and DNA fibers using rolling circle DNA
