Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
METHOD FOR ENLARGING THE RANGE OF APPLICABILITY OF MULTIPLE DISPLACEMENT AMPLIFICATION OF LINEAR DNA
Document Type and Number:
WIPO Patent Application WO/2006/111784
Kind Code:
A1
Abstract:
Underrepresentation of templates of DNA sequences mapping within about 5 kb from the ends, reportedly following to Multiple Displacement Amplification (MDA) is overcome by end-to-end ligation of the templates prior to MDA. This leads to their random oligomerization, permutation and eventually circularization, thus favoring the copying of the whole length of the templates. The amplification through end joining is accomplished in a massive way and helps enlarging the range of applicability of MDA, by facilitating the study of valuable but insufficient or otherwise inadequate material for which the termini underrepresentation may mean the loss of relevant parts of the original information.

Inventors:
SGARAMELLA VITTORIO (IT)
PANELLI SIMONA (IT)
DAMIANI GIUSEPPE (IT)
Application Number:
PCT/IB2005/001048
Publication Date:
October 26, 2006
Filing Date:
April 20, 2005
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
PARCO TECNOLOGICO PADANO SRL (IT)
SGARAMELLA VITTORIO (IT)
PANELLI SIMONA (IT)
DAMIANI GIUSEPPE (IT)
International Classes:
C12Q1/68
Domestic Patent References:
WO2005012550A22005-02-10
WO2003033724A22003-04-24
Foreign References:
US6280949B12001-08-28
US20040248105A12004-12-09
Other References:
TZVETKOV MLADEN V ET AL: "Genome-wide single-nucleotide polymorphism arrays demonstrate high fidelity of multiple displacement-based whole-genome amplification.", ELECTROPHORESIS. FEB 2005, vol. 26, no. 3, February 2005 (2005-02-01), pages 710 - 715, XP002371239, ISSN: 0173-0835
LUTHRA RAJYALAKSHMI ET AL: "Isothermal multiple displacement amplification: a highly reliable approach for generating unlimited high molecular weight genomic DNA from clinical specimens.", THE JOURNAL OF MOLECULAR DIAGNOSTICS : JMD. AUG 2004, vol. 6, no. 3, August 2004 (2004-08-01), pages 236 - 242, XP002371240, ISSN: 1525-1578
Attorney, Agent or Firm:
Sandri, Sandro (Via Locatelli 20, Verona, IT)
Download PDF:
Claims:
CLAIMS
1. Method for enlarging the range of applicability of Multiple Displacement Amplification (MDA) of linear DNA, comprising the following subsequent steps: a) providing short, linear DNA templates; b) ligation of said templates; c) multiple displacement amplification of the ligated templates.
2. Method according to claim 1, wherein said linear DNA is constituted by any natural linear DNA and its restriction fragments and any DNA copy of RNA.
3. Method according to any one of the preceding claims, including random oligomerization of the templates.
4. Method according to claim 3, including permutation of the templates.
5. Method according to claim 4, including circularization of the templates.
6. Method according to any one of the preceding claims, wherein said templates are ligated endtoend.
7. Method according to any one of the preceding claims, wherein the DNA to be amplified is restricted into large fragments through restriction endonucleases.
Description:
Method for Enlarging the Range of Applicability of Multiple Displacement Amplification of Linear DNA

* * * * *

Technical Field of the Invention

The present invention relates to a method for enlarging the range of applicability of Multiple Displacement Amplification (MDA) of linear DNA. More particularly, the present invention relates to a method which is applied to linear DNA templates, to their restriction fragments, to their shearing products properly modified at their ends, whereby said templates undergo a ligation protocol before MDA. The invention is mainly applied in the field of Whole Genome Amplification (WGA) .

Background art

Depending on the employed techniques, standard analytical scans require amounts of DNA in the range of nanograms (10 ~9 , or one billionth of a gram) or even picograms (one thousandth of a nanogram) per signal, be it a band, a spot, a drop, a well, etc.

This means that given the average size of 1-10 kb of a sequence to be analysed, one needs around one million cells with a genome of the order of Gb (or billion bases, such as the human genome ) .

Specific tissues or organs are expected to be of the range of ten-to-hundred billions of cells. Whereas their proteins and RNA are certainly different among each other, their genomes are probably similar but possibly not identical among each other.

Divergences are intrinsic to development. Additional causes of divergence could be due to physiopathological

events: they may further reduce the numbers of identical cells in a given population (tissue, organ, etc).

The importance of pushing the analysis down to a very limited number of cells, possibly to just one, rests on the fact that the peculiarities of few cells, likely to be involved in, if not the direct determinants of, novel acquired properties or features, may be averaged out, diluted and even be blurred away if the analysis is performed on several cells. The problem becomes more and more serious the larger is the number of cells subjected to the analysis. This may severely frustrate the chances of analytical procedures aimed at the detection of minor or early differences in as few as possible progenitor cells, as highly desirable in situations where an early detection is of extreme importance to mount effective utilizations (stem cells) or interventions (diagnosis).

Therefore, in order to exploit the most popular analytical techniques, ideally one would need to amplify the genome of a few if not of a single cell by a factor close to 100.000-fold in a faithful and representative way.

Among the available techniques, PCR is largely the preferred but has been shown to suffer of strong bias as to level of amplification of different sequences: such bias covers a four orders of magnitude range (Hosono S. et al. Unbiased whole genome amplification directly from clinical samples. Genome Res. 13, 954, 2003 )

A more recent procedure aiming at a Whole Genome Amplification (WGA) exploits the unusually high processivity and fidelity of two prokaryotic DNA polymerases, extracted form B. subtilis phage φ29 (Blanco L. et al. Highly efficient DNA synthesis by phage φ29 DNA polymerase. Symmetrical mode of DNA replication. J. Biol. Chem. 264: 8935-8940, 1989) and from B. stearothermophilus (Hafner G.

J. et al. Isothermal amplification and multimerization of DNA by Bst DNA polymerase. BioTechniques 30: 852-867, 2001), primed mainly by a set of random hexanucleotides thioderivatized at the 3' end for an improved resistance to exonucleolytic degradation (Barker D. L. et al. Two methods of whole genome amplification enable accurate genotyping across a 2320-SNP linkage panel. Genome Res. 14: 901-907, 2004).

It has been reported that this amplification involves a multiple displacement of the newly replicated strands enacted by the growing primers: hence the expression Multiple Displacement Amplification, or MDA (Lizardi P. M. et al. Mutation detection and single-molecule counting using isothermal rolling circle amplification. Nat. Genet. 19: 225-232, 1998).

An extensive survey of the extant MDA protocols indicates that the procedure affords a remarkable fidelity of amplification across genomes as large as the human one, but that the fidelity becomes wanting within few thousand bases (kb) from the ends of linear templates (Dean F. B. et al. Comprehensive human genome amplification using multiple displacement amplification. PNAS 99, 5261, 2002).

In particular, experiments carried out by the applicants show that terminal sequences of linear templates appear underrepresented after MDA, what could jeopardise the use of MDA on several important templates such as cDNA, or forensic and archaeological samples.

This reaction has been thoroughly investigated using in particular a model consisting of the λ phage genome and its HindIII restriction fragments as templates, whereby amplified products have been analyzed by restriction with the original enzyme as well as frequent cutters, and gel electrophoresis .

The result of the amplification of these linear DNA is

an underrepresentation of sequences mapping within about 5 kb from the ends .

Description of the Invention The present invention takes the limitations of the background art into due account , and aims to improve substantially the MDA procedure by exploiting the circularisation of genomic fragments or indeed of any DNA duplexes following their random oligomerization. This is achieved by carrying out a method for enlarging the range of applicability of Multiple Displacement Amplification (MDA) of linear DNA having the features disclosed in claim 1.

The depending claims outline advantageous forms of embodiment of said method.

According to the invention, the genome to be amplified is restricted into fragments by means of one of the several restriction endonucleases .

The wide range of A-T content and size of their target sequences confers a desirable flexibility as to the sites of cleavage and the length of the fragments .

The fragments are brought to appropriate concentrations and then converted first into randomly combined oligomers and eventually into circular structures with the catalysis of the T4 or any other DNA ligase .

After oligomerization-circularization , the mix is amplified : since the amplification uses as substrates the oligomerized-circularized structures ( which are both randomly recombined and devoid of ends ) , this procedure abrogates the end effect of MDA and leads to a faithful and massive amplif ication of the whole genome , whereby the recovered fragments map either internally or near the template ends .

Given the pivotal employment of the ligase to improve

the well established MDA, the combined approach according to the present invention is conventionally called "LIMDA".

Illustration of drawings • Figure 1 shows restriction patterns resolved on agarose

(A) and acrylamide (B) gels. In all cases, lanes are the following: l=native λ DNA;

2=uncut λ DNA+MDA; 3= λ/Hindlll+MDA;

4= λ/Hindlll+ligase+MDA; m=dYIW marker.

Bands not detectable in lanes 2 and 3 are signalled respectively with circles and arrows. • Figure 2 shows the position, on the λ restriction map, of the bands lost in the case of the Pstl+Hindlll (A) or Rsal+Hindlll (B) digestion. In this figure: numbers in italics refer to the map position of the

HindIII sites; numbers in bold above the maps refer to the MW of the lost or underrepresented fragments; bands not detectable when subjecting to MDA the uncut λ

DNA are indicated with circles; and bands lost when amplifying λ/Hindlll fragments are signalled with arrows.

Figure 3 illustrates the principle of the ligation- based MDA procedure according to the invention in the case of λ/Hindlll fragments. The white and black squares represent, respectively, the cosL and cosR sequences .

Description of a preferred embodiment of the invention

As previously stated, even if MDA-DNA has been successfully used for many applications, problems have been

reported, and in particular sequences near the ends of linear chromosomes appear under-represented after MDA.

Both incomplete priming events and abortive chain terminations could contribute to this problem, which limits the applications of MDA, and jeopardises its use on short, linear DNA (e.g., degraded genomes, as forensic and archaeological samples, and cDNA) .

In these cases, the end underrepresentation causes the loss of substantial information. Figures 1 and 2 show how the applicants studied this end problem using λ phage DNA (48.5 kb). MDA was performed using the Genomiphi ® kit (Amersham Biosciences) first on the uncut λ DNA (1 ng) , and then on its seven HindIII fragments (1 ng). The amplified samples have subsequently been compared to the native DNA to assess the representation of the various genome portions after MDA.

To this aim, restriction analysis with frequent cutters (Rsal: 113 sites on λ; Pstl: 28 sites) in combination with HindIII was used.

Restrictions with HindIII alone were also performed.

Digestions were visualized on either agarose or acrylamide gels, for discriminating both high and low MW.

HindIII and Hindlll+Pstl digestions were visualized on 1% agarose gels, Hindlll+Rsal patterns on 1.4% gels.

The electrophoresis was run at 80 V for 2 hours in TAE, and visualized using ethidium bromide.

4% polyacrylamide gels ( acrylamide :bisacrylamide 37.5:1) were used in TBE at 120 V for 30 minutes, and 200 V for 3 hours .

Patterns were visualized using Vistra-green ® (Amersham Biosciences) .

In figure 2 it may be observed how in the first case (circles), Pst fragments confined within 5 kb from the 5'

end present a very diminished intensity after MDA (A) .

On the other hand, in the second case (arrows), the λ/Hind+MDA sample subjected to Pstl+Hindlll digestion loses the 704 and 547 bp fragments (A) . Both derive from Pst fragments containing internal Hind sites. The same sample, after Rsal+Hindlll digestion, loses several fragments (B), all close to Hindlll sites.

Other fragments, among which the terminal ones, could not be unambiguously identified because of co-migration and gel-resolution problems.

They were thus excluded from the analysis.

Turning back to the figures, Figure 1 shows the patterns of the original λ DNA (lanes 1) and of the uncut DNA after MDA (lanes 2). Restriction patterns are easily recognized in amplified samples, suggesting that MDA produces double-stranded DNA whose length covers nearly the whole λ genome.

The disappearance of a terminal band (the 3' one of the Hindlll pattern, 4.4 kb) may be observed in the MDA sample (Figure IA, white circle).

Other bands present diminished intensity in MDA-DNA.

This is more easily visible in the Pstl+Hindlll pattern (Figure IB, bands marked with circles).

In accordance with the end underrepresentation phenomenon, all these fragments map in the first 5 kb at the 5' end of the genome (Figure 2A).

Recognizable patterns are still present after restriction of MDA-treated λ/Hindlll fragments (Figure 1, lanes 3), except when digesting with Hindlll, as expected. Concerning the other digestions, some fragments are no longer detectable (Figure IB, signalled with arrows).

These are Rsal or Pstl fragments containing, or close to, Hindlll sites (Figure 2), again in concordance with the termini underrepresentation.

Thus, the analysis carried out by the applicants shows that after MDA the sequences in proximity of the template ends are lost or underrepresented.

It has been defined that on λ genome this phenomenon affects regions within about 5 kb from the ends.

The behaviour of the Hindlll+Pstl 5' terminal fragments (Figures IB and 2A, circles) makes it likely that premature detachments of the growing chain give a contribution to the end underrepresentation. The relevant bands are much less intense when amplifying the uncut DNA, while their intensities become comparable to those of the native DNA after MDA of the shorter Hind fragments .

In order to overcome the end problem, the present invention provides for a simple protocol based on ligation of linear templates before MDA, to produce oligomers with permuted sequences as to the order of the constituting fragments, to be eventually circularized.

In accordance with the invention, 3 Weiss units of T4 DNA ligase (Invitrogen) were used to ligate the cohesive ends of λ/Hindlll fragments (7.5 ng/μl).

Also other restriction enzymes have been tried with comparable results (not shown).

Reactions were incubated overnight at 3O 0 C and analysed on 1% agarose gels.

750 pg of the ligated sample were MD-amplified.

For all MDA reactions, 10-20 μl product have been obtained.

The product was then compared to the original DNA as previously described.

Lanes 4 of Figure 1 refer to the ligated sample.

It may be noted that its restriction patterns are the most similar to those of the unamplified DNA.

Nearly all the bands not detectable in lanes 2-3 are

recovered .

A special consideration is due for the ends of λ DNA, cosR and cosL , which after HindIII digestion are borne respectively by the 23 kb and 4.4 fragments . Once ligated, they are resolved only by specialised enzymes . The λ/Hindlll 3 ' band ( 4 . 4 kb ) is not rescued after ligation (Figure IA) .

In this sample , the 4 . 4 kb fragment , after cosL/R joining, forms a 27 .4 kb band, indistinguishable from the 23 kb one , as they both lie in the non-resolving region of the gel .

The overall principle which has been applied in carrying out the above described tests is shown in Fig. 3.

According to the present invention , the end-to-end ligation of templates he lps overcoming the termini underrepresentation, and enlarges the MDA applications .

It is useful to obtain faithful amplifications of small , linear templates that otherwise could not be successfully amplified. Both cohesive and blunt ends can be satisfactorily ligated before MDA.

For a wide range of applications the fill-in of the ends , followed by a blunt-end ligation, possibly mediated by appropriate synthetic adaptors or linkers , can be a suitable strategy to obtain a faithful MDA.

This can be the case of degraded genomic samples , for which the protocol according to the present invention provides a simpler alternative to the OmniPlex ® technology ( 1 ) for performing WGA. Recently a method has been described for isothermal in vitro amplification of specific regions . It exploits DNA helicase to generate single-stranded template for specific primer hybridization and subsequent extension.

However, its application could be problematic if the

template is limited or degraded. On the other hand, the procedure according to the present invention, through the fill-in and blunt end ligation of linear DNA before MDA, can substitute DNA extraction as the first step of a totally isothermal process for amplifying specific DNA sequences regardless of the template scarcity and/or degradation.