Ancient DNA: Methods and Protocols (34 page)

Castor fi ber
) (see Chapter

22 ). Accordingly, deeper

sequencing of these libraries may be necessary to reach suffi cient coverage (e.g., 20× coverage might be the desired coverage of a resequencing experiment) for a targeted region.

21 Target Enrichment via DNA Hybridization Capture

179

Table 1

The sequencing throughput required and hybridization capture protocol recommended will depend on the desired target region size. Additionally, the expected percentage of reads that map to target and the desired coverage will infl uence the amount of data (in Megabases, Mb, or kilobases, kb) that need to be sequenced.

The amount of endogenous DNA is diffi cult to estimate for aDNA extracts and infl uences the percentage of reads that will map to target. Therefore, the values presented here should be seen as minimums

2.5 Mb (e.g., any nuclear 16 kb (e.g., mitochondrial 500 bp (e.g., mitochondrial Region size

region)

genome)

control region)

Percentage of

30

30

20

reads that

map to target

Desired

20×

20×

20×

coverage

Bp to sequence 167 Mb

1 Mb

50 kb

Recommended NimbleGen arrays,

Custom self-made bait

Custom self-made bait

hybridization

Agilent Sure Select,

prepared using long—

prepared using PCR

capture

MYselect custom

range PCR and

and hybridization in

protocol

probes

hybridization in

solution; Primer

solution

extension capture (PEC)

Blocking oligonucleotides may also be used as part of the

hybridization protocol. Because all of the DNA fragments in the library will have the same universal adapters ligated to their ends, they may hybridize to each other during the enrichment protocol, forming long nonsensical chains. Blocking oligonucleotides cover the ends of the sequences that contain the adapters, preventing accidental hybridization between adapters and thereby also preventing nontarget DNA molecules from being pulled down along with target molecules. The particular blocking oligonucleotides required will depend on the sequencing platform to be used.

An outline of different enrichment protocols is provided in T
able 2 . The hybridization mixtur
e (containing the DNA library and the bait) can either be incubated in solution or immobilized in a solid phase on arrays or beads. Hybridization in solution and immobilization on beads both require conventional tubes and hybridization ovens, but incubation may also take place in a thermal cycler. If arrays are used, these need to be placed in special racks for rotation. Hybridization in solution requires a subsequent bead capture step, which is not necessarily required for either immobilization approaches. It has been suggested that hybridization in solution may be more effi cient for libraries with fragment lengths shorter than 500 bp, as is expected for most aDNA libraries
( 5 )
.

180

S. Horn

)

14
(

get.

etches

lated oligos

to magnetic beads

and hybridizes

with the tar

of bp

On beads

6

Self-made biotiny—

DNA bait is bound

DNA str

Up to thousands

) (
6

-

ray

oar

e ar

get.

sequence

captur

to the micr

ray and hybridizes with the

tar

ound 60 bp

5

Roche Nimblegen

DNA bait is bound

DNA oligos

Ar

e

eSelect

ray and

get.

rays

oar

ray

DNA Captur

Ar

bound to the

micr

hybridizes with

the tar

Immobilized

hybridization

On Ar

4

Agilent Sur

DNA bait is

DNA oligos

25–60 bp

)

(
10

e (PEC)

ed via

e elongated by a

captur

ar

polymerase and

captur

attached biotin.

ound 30 bp

3

Primer extension

Primers hybridize,

DNA oligos

Ar

ed

)

(
15

etches

get and captur

oligos

hybridize with the

tar

via attached biotin.

2

Self-made biotinylated

DNA bait is used to

DNA str

Up to thousands of bp

ed

) (
5

e on beads after hybridization

eSelect

get and captur

Sur

hybridize with the

tar

via attached biotin.

Hybridization

in solution

Captur

1

MYselect or Agilent

RNA bait is used to

RNA oligos

120–200 bp

obes

2

of

oach no.

enrichment

Table

Approaches of DNA hybridization capture for target enrichment prior to sequencing echnology

Mode

hybridization

Immobilization

Appr

T

Principle of

Bait, or pr

Length of bait

21 Target Enrichment via DNA Hybridization Capture

181

Immobilization of the bait and tight physical clustering of bait molecules, as is common on arrays, may result in steric interference between target and nontarget molecules. The resulting pulldown of nontarget DNA could cause fewer sequencing reads to map to the desired target region. Finally, generating bait rather than purchasing bait may reduce the cost of the enrichment considerably.

The protocol presented in this chapter describes how to generate self-made bait for longer and shorter contiguous targets. For later immobilization on streptavidin-coated beads, biotin needs to be introduced into the bait. This is achieved by the ligation of biotinylated adapters to sheared long-range PCR products. Alternatively, biotin can be introduced into shorter amplicons during a biotinylating PCR step.

Serial enrichments, where enrichment is performed more than once for a single library, can be applied to aDNA libraries that contain very low amounts of endogenous DNA compared to contaminating or backgr
ound DNA

( 11 )
. In such instances, a single enrichment step may be insuffi cient to provide adequate coverage of the target loci. The case study reported in Chapter 22 used two enrichment steps to target an approximately 500-bp stretch of mitochondrial DNA of beavers.

PCR amplifi cation of libraries prior to, during, or after hybridization is all potentially problematic, but may be nonetheless unavoidable in many aDNA applications. Potential drawbacks of amplifi cation include a skewed representation of the original library due to preferential amplifi cation of certain molecules, jumping PCR artifacts such as chimeras, and additional errors introduced by polymerases. Commercial hybridization kits both in solution and on arrays require 3–4 m g of DNA in a library
( 12 )
, which cannot be achieved from most ancient samples without amplifi cation. In addition, amplifi cation is generally necessary to produce suffi cient DNA for sequencing on either the Illumina or SOLiD platforms (however, as little as 10 6 molecules per 1/16th lane may be suffi -

cient for 454 sequencing). Thus, when using self-made bait or PEC primers
( 10 )
for hybridization (approaches no. 2, 3, and 6 in T
able 2
) with subsequent 454 sequencing, amplifi cation steps may be avoidable. Because the potential problems associated with amplifi cation are most likely to occur when the library quality is poor, care should be taken to select samples with the best quality and quantity of endogenous DNA, as may be identifi ed using quantitative PCR.

Manufacturers provide a variety of arrays with capture probes made from DNA and RNA as well as in-solution capture kits
( 13
) (see Note 1 and Table 2). Instead of purchasing a kit, however, DNA hybridization capture can be performed with standard tools in any molecular biology lab. I present a protocol to generate self-made bait to target genomic regions of interest. For this, it is possible to use the products of long-range PCR, thereby covering 182

S. Horn

larger target regions. Bait can also be produced during a regular PCR for shorter targets. I provide a protocol for hybridizing a library to this self-made bait, and for the subsequent bead capture step that immobilizes the reaction. After elution, the enriched library can be directly sequenced on a high-throughput sequencing platform.

2. Materials

All reagents and plasticware should be sterile, DNA and DNAse free.

1. aDNA library.

2. If bait is to be produced from long-range PCR products: (a)

Sonicator (e.g., Diagenode or Covaris).

(b)

Two complementary oligonucleotides, one of them

5 ¢ -biotinylated.

(c)

Oligo hybridization buffer (10×): 500 mM NaCl, 10 mM

Tris-Cl (pH 8.0), 1 mM EDTA (pH 8.0).

(d)

Tango buffer (10×,
e.g.
Fermentas).

(e)

T4 DNA ligase (5 U/ m L, supplied with 10× T4 DNA

ligase buffer and 50% PEG-4000 solution).

(f)

T4 DNA polymerase (5 U/ m L).

(g)

T4 polynucleotide kinase (10 U/ m L).

(h)

ATP (100 mM).

(i)

Bst
DNA polymerase, large fragment (supplied with 10×

buffer).

or

2. If bait is to be produced by biotinylating PCR:

(a)

Biotin-dUTP (100 m M).

(b)

dNTP mix containing 25 mM of dATP, dGTP, and dCTP,

but 24.5 mM of dTTP (mix 10 m L each of 100 mM dATP,

dGTP, and dCTP with 9.8 m L 100 mM dTTP and 0.2 m L

H O).

2

(c)

A polymerase incorporating biotinylated nucleotides (e.g., taq polymerase).

3. Streptavidin covered magnetic beads (e.g., Dynabeads M270, Invitrogen).

4. Tween-20.

5. EBT and TET: elution buffer from any kit and 1× TE buffer, including 0.05% Tween-20.

21 Target Enrichment via DNA Hybridization Capture

183

6. 1× bind and wash (BWT) buffer : 1 M NaCl, 10 mM Tris-Cl, 1 mM EDTA, 0.05% Tween-20, pH 8.0.

7. Hot wash (HW) buffer : 200 m L 10× PCR buffer, 200 m L

MgCl (25 mM), 1.6 mL H O.

2

2

8. Phusion High Fidelity PCR master mix (New England Biolabs).

9. SPRI beads (Agencourt AMPure XP) or MinElute kit

(Qiagen).

10. Hybridization buffer and blocking agent (e.g., from the Agilent aCGH kit, Cat. no. 5188-5220).

11. Barrier/fi lter tips and PCR reaction tubes/plates.

12. DNA Spectrophotometer.

13. Magnetic rack for SPRI bead cleanups and capture with magnetic beads.

14. Hybridization oven.

15. A thermal cycler with heated lid.

16. Laboratory fi lm (e.g., Parafi lm).

3. Methods

Design and order blocking oligonucleotides. The sequences of blocking oligonucleotides correspond to the sequences of the respective adapters and can include ambiguity codes for barcodes which may vary between samples. Be sure to include the oligonucleotides to cover adapters in 5’–3’ as well as in 3’–5’ direction; examples are given in
( 11 )
(see Notes 2 and 3).

Produce biotinylated bait DNA for the enrichment using PCR

products
( 14, 15
) . It is recommended to exclude repetitive regions from the PCR by placing the primers appropriately (see Note 4).

Generate DNA bait from long-range PCR products for target

regions of kilobases in size, following manufacturers’ instructions for long-range PCR

1. Prepare a biotinylated double-stranded adapter from two complementary oligonucleotides, one carrying a biotin at the 5 ¢ -end (see Note 5) in the following mix:

(a)

40 m L of oligo 1 (500 m M).

(b)

40 m L of oligo 2 (500 m M).

(c)

10 m L 10× oligo hybridization buffer.

(d)

10 m L H O.

2

Heat the mixture to 95°C for 5 s, then ramp to 15°C at the rate of −0.1°C/sec. The resulting adapter has a concentration of 200 m M.

184

S. Horn

2. Fragment long-range PCR products with ultrasound in a sonicator twice for 7 minutes at “high” to obtain a fragment size of around 200–500 bp. Check the size of the obtained DNA on an agarose gel (1–2%) and, if necessary, repeat the fragmentation.

3. Purify the fragmented long-range PCR product using a

MinElute column (see Note 6) and measure the concentration on a DNA spectrophotometer. Use up to 500 ng per reaction

in the next step. Several reactions may be necessary to produce large amounts of bait exceeding 1 m g.

4. Set up a blunt-end repair, include per reaction:

(a)

7.12 m L H O.

2

(b)

7 m L Buffer Tango (10×).

(c)

0.28 m L dNTPs (25 mM each).

(d)

0.7 m L ATP (100 mM).

(e)

3.5 m L T4 polynucleotide kinase (10 U/ m L).

(f)

1.4 m L T4 DNA polymerase (5 U/ m L).

Add 20 m L of master mix to 50 m L of purifi ed long-range PCR

product. Mix gently and incubate in a thermal cycler for 15 min at 25°C followed by 5 min at 12°C.

5. Place the reaction on ice or immediately perform a cleanup using a MinElute column and elute in 20 m L EBT.