Limits...
Mate pair sequencing of whole-genome-amplified DNA following laser capture microdissection of prostate cancer.

Murphy SJ, Cheville JC, Zarei S, Johnson SH, Sikkink RA, Kosari F, Feldman AL, Eckloff BW, Karnes RJ, Vasmatzis G - DNA Res. (2012)

Bottom Line: Sequencing data predicted genome coverage and depths similar to unamplified genomic DNA, with limited repetition and bias predicted in WGA protocols.Mapping algorithms developed in our laboratory predicted high-confidence rearrangements and selected events each demonstrated the predicted fusion junctions upon validation.Rearrangements were additionally confirmed in unamplified tissue and evaluated in adjacent benign-appearing tissues.

View Article: PubMed Central - PubMed

Affiliation: Department of Molecular Medicine, Mayo Clinic, Medical Sciences Building 2, 200 First St., SW, Rochester, MN 55905, USA. murphy.stephen@mayo.edu

ABSTRACT
High-throughput next-generation sequencing provides a revolutionary platform to unravel the precise DNA aberrations concealed within subgroups of tumour cells. However, in many instances, the limited number of cells makes the application of this technology in tumour heterogeneity studies a challenge. In order to address these limitations, we present a novel methodology to partner laser capture microdissection (LCM) with sequencing platforms, through a whole-genome amplification (WGA) protocol performed in situ directly on LCM engrafted cells. We further adapted current Illumina mate pair (MP) sequencing protocols to the input of WGA DNA and used this technology to investigate large genomic rearrangements in adjacent Gleason Pattern 3 and 4 prostate tumours separately collected by LCM. Sequencing data predicted genome coverage and depths similar to unamplified genomic DNA, with limited repetition and bias predicted in WGA protocols. Mapping algorithms developed in our laboratory predicted high-confidence rearrangements and selected events each demonstrated the predicted fusion junctions upon validation. Rearrangements were additionally confirmed in unamplified tissue and evaluated in adjacent benign-appearing tissues. A detailed understanding of gene fusions that characterize cancer will be critical in the development of biomarkers to predict the clinical outcome. The described methodology provides a mechanism of efficiently defining these events in limited pure populations of tumour tissue, aiding in the derivation of genomic aberrations that initiate cancer and drive cancer progression.

Show MeSH

Related in: MedlinePlus

MP NGS of WGA DNA products. (a) A schematic of the MP library assembly, detailing the initial fragmentation of the gDNA to 3–5 kb and biotinylation of the termini of these fragments to enable isolation and sequencing of the ligated terminal regions after circularization and a second fragmentation to ∼500 bp. The bridged coverage is demonstrated as the span of the initial 3–5 kb fragments and the resulting terminal MP fusion products and conventional PE fragments depicted. (b) The bridged length of paired reads mapping to the reference genome from (i) an unamplified gDNA sample of T-cell origin and (ii) the WGA GP4 sample. (c) Bridged (open squares) and base pair coverage (open triangles) for individual chromosomes from the WGA GP3 (green) and GP4 (blue) samples and the unamplified T-cell control (red). (d) Percentage replication in the GP3 (green square), GP4 (blue triangle) and unamplified (red star) samples MP data. (e) Percentage replication (blue), together with bridged (red) and sequence (green) fold coverage for 22 additional WGA prostate cancer DNA samples following LCM isolation on the HiSeq platform. (f) Frequency coverage plots for chromosomes 4 (i and iii) and 10 (ii and iv) of unamplified DNA (i and ii) and WGA G4 DNA (iii and iv) sequencing data.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3473372&req=5

DSS021F2: MP NGS of WGA DNA products. (a) A schematic of the MP library assembly, detailing the initial fragmentation of the gDNA to 3–5 kb and biotinylation of the termini of these fragments to enable isolation and sequencing of the ligated terminal regions after circularization and a second fragmentation to ∼500 bp. The bridged coverage is demonstrated as the span of the initial 3–5 kb fragments and the resulting terminal MP fusion products and conventional PE fragments depicted. (b) The bridged length of paired reads mapping to the reference genome from (i) an unamplified gDNA sample of T-cell origin and (ii) the WGA GP4 sample. (c) Bridged (open squares) and base pair coverage (open triangles) for individual chromosomes from the WGA GP3 (green) and GP4 (blue) samples and the unamplified T-cell control (red). (d) Percentage replication in the GP3 (green square), GP4 (blue triangle) and unamplified (red star) samples MP data. (e) Percentage replication (blue), together with bridged (red) and sequence (green) fold coverage for 22 additional WGA prostate cancer DNA samples following LCM isolation on the HiSeq platform. (f) Frequency coverage plots for chromosomes 4 (i and iii) and 10 (ii and iv) of unamplified DNA (i and ii) and WGA G4 DNA (iii and iv) sequencing data.

Mentions: MP libraries were assembled for the prostate GP3 and GP4 WGA DNA samples using the modified MP library assembly protocol described. The conventional Illumina protocols were adapted to account for the different physical properties of the WGA DNA. The MP libraries were sequenced on single lanes of the Illumina GAIIx platform. MP sequencing of the GP3 and GP4 cancers generated 33.0 and 29.4 million mappable read pairs, respectively, consistent with results from our MP sequencing of conventional unamplified gDNA on the GAIIx sequencer, which yielded averages of 27.1 million reads. A set of algorithms was developed to detect large chromosomal aberrations with low false-positive rates. The algorithms were specifically designed to handle NGS from mate-pair protocols.7 The algorithms describe coverage in two ways: conventional base-pair coverage from the 75–100-bp reads and the theoretical bridged-coverage considering the original 3–5-kb span separating two MP reads (Fig. 2a). MP reads mapping with the expected region span allows us to infer that the correct sequence span lies between the reads. Fragment mapping profiles are presented in Fig. 2b. The predominant fragment population reflects the MP reads with 3–5 kb bridged spans. The second narrower peaks at around −300 bp consist of co-purifying PE library fragments derived from internal fragments of the circles not spanning the ligation junctions (Fig. 2a). The PE peaks map algorithmically with negative values due to the different polarity of the read pairs stemming from the circularization in the MP methodology.5 The fragment mapping profile for unamplified (Fig. 2b, i) and WGA (Fig. 2b, ii) DNA are effectively equivalent, with the WGA DNA mapping with just slightly reduced span of 2.5–5 kb indicative of the starting input of the smaller DNA fragments (Fig. 1c). Although the proportion of PE mapping fragments is still higher with the WGA DNA, additional refinements of the MP protocol could reduce this further.Figure 2.


Mate pair sequencing of whole-genome-amplified DNA following laser capture microdissection of prostate cancer.

Murphy SJ, Cheville JC, Zarei S, Johnson SH, Sikkink RA, Kosari F, Feldman AL, Eckloff BW, Karnes RJ, Vasmatzis G - DNA Res. (2012)

MP NGS of WGA DNA products. (a) A schematic of the MP library assembly, detailing the initial fragmentation of the gDNA to 3–5 kb and biotinylation of the termini of these fragments to enable isolation and sequencing of the ligated terminal regions after circularization and a second fragmentation to ∼500 bp. The bridged coverage is demonstrated as the span of the initial 3–5 kb fragments and the resulting terminal MP fusion products and conventional PE fragments depicted. (b) The bridged length of paired reads mapping to the reference genome from (i) an unamplified gDNA sample of T-cell origin and (ii) the WGA GP4 sample. (c) Bridged (open squares) and base pair coverage (open triangles) for individual chromosomes from the WGA GP3 (green) and GP4 (blue) samples and the unamplified T-cell control (red). (d) Percentage replication in the GP3 (green square), GP4 (blue triangle) and unamplified (red star) samples MP data. (e) Percentage replication (blue), together with bridged (red) and sequence (green) fold coverage for 22 additional WGA prostate cancer DNA samples following LCM isolation on the HiSeq platform. (f) Frequency coverage plots for chromosomes 4 (i and iii) and 10 (ii and iv) of unamplified DNA (i and ii) and WGA G4 DNA (iii and iv) sequencing data.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3473372&req=5

DSS021F2: MP NGS of WGA DNA products. (a) A schematic of the MP library assembly, detailing the initial fragmentation of the gDNA to 3–5 kb and biotinylation of the termini of these fragments to enable isolation and sequencing of the ligated terminal regions after circularization and a second fragmentation to ∼500 bp. The bridged coverage is demonstrated as the span of the initial 3–5 kb fragments and the resulting terminal MP fusion products and conventional PE fragments depicted. (b) The bridged length of paired reads mapping to the reference genome from (i) an unamplified gDNA sample of T-cell origin and (ii) the WGA GP4 sample. (c) Bridged (open squares) and base pair coverage (open triangles) for individual chromosomes from the WGA GP3 (green) and GP4 (blue) samples and the unamplified T-cell control (red). (d) Percentage replication in the GP3 (green square), GP4 (blue triangle) and unamplified (red star) samples MP data. (e) Percentage replication (blue), together with bridged (red) and sequence (green) fold coverage for 22 additional WGA prostate cancer DNA samples following LCM isolation on the HiSeq platform. (f) Frequency coverage plots for chromosomes 4 (i and iii) and 10 (ii and iv) of unamplified DNA (i and ii) and WGA G4 DNA (iii and iv) sequencing data.
Mentions: MP libraries were assembled for the prostate GP3 and GP4 WGA DNA samples using the modified MP library assembly protocol described. The conventional Illumina protocols were adapted to account for the different physical properties of the WGA DNA. The MP libraries were sequenced on single lanes of the Illumina GAIIx platform. MP sequencing of the GP3 and GP4 cancers generated 33.0 and 29.4 million mappable read pairs, respectively, consistent with results from our MP sequencing of conventional unamplified gDNA on the GAIIx sequencer, which yielded averages of 27.1 million reads. A set of algorithms was developed to detect large chromosomal aberrations with low false-positive rates. The algorithms were specifically designed to handle NGS from mate-pair protocols.7 The algorithms describe coverage in two ways: conventional base-pair coverage from the 75–100-bp reads and the theoretical bridged-coverage considering the original 3–5-kb span separating two MP reads (Fig. 2a). MP reads mapping with the expected region span allows us to infer that the correct sequence span lies between the reads. Fragment mapping profiles are presented in Fig. 2b. The predominant fragment population reflects the MP reads with 3–5 kb bridged spans. The second narrower peaks at around −300 bp consist of co-purifying PE library fragments derived from internal fragments of the circles not spanning the ligation junctions (Fig. 2a). The PE peaks map algorithmically with negative values due to the different polarity of the read pairs stemming from the circularization in the MP methodology.5 The fragment mapping profile for unamplified (Fig. 2b, i) and WGA (Fig. 2b, ii) DNA are effectively equivalent, with the WGA DNA mapping with just slightly reduced span of 2.5–5 kb indicative of the starting input of the smaller DNA fragments (Fig. 1c). Although the proportion of PE mapping fragments is still higher with the WGA DNA, additional refinements of the MP protocol could reduce this further.Figure 2.

Bottom Line: Sequencing data predicted genome coverage and depths similar to unamplified genomic DNA, with limited repetition and bias predicted in WGA protocols.Mapping algorithms developed in our laboratory predicted high-confidence rearrangements and selected events each demonstrated the predicted fusion junctions upon validation.Rearrangements were additionally confirmed in unamplified tissue and evaluated in adjacent benign-appearing tissues.

View Article: PubMed Central - PubMed

Affiliation: Department of Molecular Medicine, Mayo Clinic, Medical Sciences Building 2, 200 First St., SW, Rochester, MN 55905, USA. murphy.stephen@mayo.edu

ABSTRACT
High-throughput next-generation sequencing provides a revolutionary platform to unravel the precise DNA aberrations concealed within subgroups of tumour cells. However, in many instances, the limited number of cells makes the application of this technology in tumour heterogeneity studies a challenge. In order to address these limitations, we present a novel methodology to partner laser capture microdissection (LCM) with sequencing platforms, through a whole-genome amplification (WGA) protocol performed in situ directly on LCM engrafted cells. We further adapted current Illumina mate pair (MP) sequencing protocols to the input of WGA DNA and used this technology to investigate large genomic rearrangements in adjacent Gleason Pattern 3 and 4 prostate tumours separately collected by LCM. Sequencing data predicted genome coverage and depths similar to unamplified genomic DNA, with limited repetition and bias predicted in WGA protocols. Mapping algorithms developed in our laboratory predicted high-confidence rearrangements and selected events each demonstrated the predicted fusion junctions upon validation. Rearrangements were additionally confirmed in unamplified tissue and evaluated in adjacent benign-appearing tissues. A detailed understanding of gene fusions that characterize cancer will be critical in the development of biomarkers to predict the clinical outcome. The described methodology provides a mechanism of efficiently defining these events in limited pure populations of tumour tissue, aiding in the derivation of genomic aberrations that initiate cancer and drive cancer progression.

Show MeSH
Related in: MedlinePlus