Polony DNA Sequencing UNIT 7.

This unit provides protocols for performing polony DNA sequencing, a nonelec-
trophoretic sequencing method that eliminates in vivo cloning artifacts and affords a
lower cost per base than conventional Sanger sequencing (Shendure et al., 2005). The
first step in polony DNA sequencing is to construct an in vitro paired-tag library from
genomic DNA (Basic Protocol 1; Fig. 7.8.1). Next, library molecules are clonally am-
plified on microbeads by emulsion PCR (Basic Protocol 2; Dressman et al., 2003). This
clonal amplification yields polymerase colonies, or polonies, that can be sequenced. A
Support Protocol describes a method used to optimize the amount of library DNA used
as the ePCR template. Amplicon-bearing beads are then enriched (Basic Protocol 3) and
two-dimensionally arrayed within an acrylamide gel matrix on a microscope coverglass
(Basic Protocol 4). Finally, millions of short reads are generated in parallel from the
microbeads via a cyclic DNA sequencing strategy that utilizes T4 DNA ligase to selec-
tively tag each microbead with fluorescent labels that correlate with the unique nucleotide
sequence present on any given bead (Basic Protocol 5).

The combination of protocols can be thought of as a modular platform, with each major
step (library construction, emulsion PCR, sequence readout) independent enough to be
replaced by an alternate. For example, if paired tags are not necessary to place reads
(i.e., to match sequences to a reference genome sequence), single-tag libraries can be
constructed (e.g., for barcode sequencing or RNA quantitation). The emulsion PCR
protocol presented in this unit is a variation of that published by Dressman et al. (2003),
but should not be thought of as the only applicable amplification method. Different
sequencing biochemistries can also be used; in fact, during development of the polony
sequencing platform, a number of alternate polymerase- and ligase-driven approaches
were evaluated.

To implement the protocols presented in this unit, various instruments (e.g., fluo-
rescent microscope, autosampler, and flow cell) and computer hardware and soft-
ware are required. A discussion of the critical features of the setup is pro-
vided (see Critical Parameters and Troubleshooting), but the reader is directed to
http://arep.med.harvard.edu/Polonator for more detailed information, including full
schematics and drawings. Instructions are available there for using the hardware and
software, as are binary and source codes for the software. All software is free and
open-source, and all hardware is commercially available.

Figure 7.8.1 General procedure for polony sequencing. A paired-tag genomic library is used
as template for emulsion PCR on microbeads to generate polymerase colonies (polonies). The
beads are cast on a coverslip in a thin layer of polyacrylamide, and put through iterative cycles of
single-base sequencing.

DNA Sequencing
Contributed by Gregory J. Porreca, Jay Shendure, and George M. Church 7.8.1
Current Protocols in Molecular Biology (2006) 7.8.1-7.8.22
Copyright

C 2006 by John Wiley & Sons, Inc.
Supplement 76
 BASIC CONSTRUCTION OF A SHOTGUN PAIRED-TAG GENOMIC LIBRARY
PROTOCOL 1
Polony genomic resequencing is generally performed on a shotgun paired-tag library
(Shendure et al., 2005). Each library molecule is 135 bp in length, and has two 17- to
18-bp paired genomic tags separated and flanked by common sequences (Fig. 7.8.2).
The key step in this protocol is the circularization of randomly sheared and size-selected
genomic DNA around a synthetic insert-linker. This links sequences that are separated
on the genome by a defined distance distribution.

Materials
Genomic DNA
Buffer EB (Qiagen)
End-It DNA end repair kit (Epicentre), including:
10× buffer
10× ATP
10× dNTP mix
Enzyme mix
10× PCR buffer without MgCl2 (Invitrogen)
50 mM MgCl2
100 mM dATP
5 U/µl Taq DNA polymerase
20 mg/ml glycogen
Oligonucleotides:
100 µM T30-T: 5 -phosphorylated-GTCGGAGGCCAAGGCGGCCGTACGT
CCAACT-3 (purified by HPLC)
100 µM T30-B: 5 -phosphorylated-GTTGGACGTACGGCCGCCTTGGCC
TCCGACT -3 (purified by HPLC)
1 mM N6 oligonucleotides: 5 -NNNN*N*N-3 (*signifies phosphorothioate
linkage; IDT)
100 µM FDV-B: 5 -ATCACCGACTGCCCATAGAGAGGAAAGCGGAGGC
GTAGTGGTT-3 (purified by HPLC)
100 µM FDV-T: 5 -AACCACTACGCCTCCGCTTTCCTCTCTATGGGCAGT
CGGTGAT-3 (purified by HPLC)
100 µM RDV-B: 5 -AGAGAATGAGGAACCCGGGGCAGTT-3 (purified by
HPLC)
100 µM RDV-T: 5 -AACTGCCCCGGGTTCCTCATTCTCT-3 (purified by
HPLC)
Quick ligation kit (NEB), including:
2× Quick ligation buffer
Quick T4 DNA ligase
20 U/µl Exonuclease I (NEB)
100 U/µl Exonuclease III (NEB)
TE buffer, pH 8.0 (APPENDIX 2)
25 mM dNTP mix (25 mM each nucleotide)
10× RepliPhi phi29 reaction buffer (Epicentre Technologies)
100 U/µl RepliPhi phi29 polymerase (Epicentre)
10× NEBuffer 4 (NEB)
1.6 mM S-adenosylmethionine (SAM; diluted from 32 mM stock in 1× NEBuffer
4; NEB)
2 U/µl MmeI (NEB)
2000 U/µl T4 DNA ligase and 10× buffer (NEB)
40% (w/v) polyethylene glycol 8000 (PEG) in H2 O
10× NEBuffer 2 (NEB)
10 U/µl E. coli DNA polymerase I (NEB)
Polony DNA
Sequencing 5 U/µl Platinum Taq DNA polymerase (Invitrogen)

7.8.2
Supplement 76 Current Protocols in Molecular Biology
 Qiaquick columns (Qiaquick PCR cleanup kit; Qiagen)
NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies)
Microcon-30 column (Millipore)
Thermal cycler
Additional reagents and equipment for DNA quantitation (APPENDIX 3D),
phenol/chloroform extraction and ethanol precipitation (UNIT 2.1A), and agarose
and polyacryamide gel electrophoresis (UNITS 2.5A & 2.7)
Prepare genomic DNA for circularization
1. Shear 15 µg genomic DNA to desired size distribution.
The Hydroshear (http://www.genomicsolutions.com) is the recommended instrument for
this purpose since, when the manufacturer’s instructions are followed, it produces a
population of fragments having a relatively tight size distribution and ends that can be
efficiently blunted. The authors typically shear DNA at a concentration of ∼60 ng/µl.

2. Purify DNA on Qiaquick columns (≤10 µg per column) as per manufacturer’s

instructions.
Use as few columns as possible.

3. To repair the DNA ends (i.e., blunt them following shearing), combine the DNA with
the components of the End-It DNA end repair kit as follows:

10 µg column-purified, sheared genomic DNA in buffer EB

25 µl 10× buffer
25 µl 10× ATP
25 µl 10× dNTP mix
5 µl enzyme mix
H2 O to a final volume of 250 µl.

Incubate 1 hr at room temperature.

4. Purify DNA on a Qiaquick column per manufacturer’s instructions, except elute with
90 µl buffer EB.
5. Quantitate DNA using a NanoDrop ND-1000 spectrophotometer (see APPENDIX 3D).
Concentration should be ∼100 ng/µl.

6. Incubate 15 min at 70◦ C to eliminate any residual enzyme activity.

7. Add an adenosine tail by combining the following components:

2.2 µg end-repaired DNA

10 µl 10× PCR buffer without MgCl2
6 µl 50 mM MgCl2
0.5 µl 100 mM dATP
0.5 µl 5 U/µl Taq DNA polymerase
H2 O to 100 µl.

Incubate 30 min at 70◦ C then cool to 4◦ C. Transfer to ice.

In the A-tailing reaction, Taq polymerase adds a single A to the 3 ends of the blunted
DNA in a template-independent fashion.

DNA Sequencing

7.8.3
Current Protocols in Molecular Biology Supplement 76
 8. Extract DNA from the reaction mixture with phenol/chloroform and precipitate in
ethanol using 20 µg glycogen as carrier (UNIT 2.1A). Resuspend pellet in 40 µl buffer
EB.
Glycogen should be used as a carrier in all ethanol precipitation steps in this protocol.

9. Separate the sheared, A-tailed DNA on a 6% polyacrylamide gel with TBE buffer
(UNIT 2.7). Stain with ethidium bromide.
For Invitrogen precast gels, mix 20 µl DNA with 5 µl of 5× loading buffer (Novex) and
load across two 0.5-cm lanes.

10. Cut out the desired bands and extract DNA using the crush-and-soak method (UNIT 2.7).
Precipitate in ethanol using glycogen as a carrier, and resuspend in 10 µl buffer EB.
For a paired-tag library with an average intertag distance of 1 kb, excise and pool gel
bands centered at 1 kb with a total width of no more than 500 bp. Exposure of the DNA
to UV light during visualization should be minimized.
11. Quantitate the size-selected DNA by separating on an agarose or polyacrylamide gel
(UNIT 2.5A or 2.7) and comparing the brightness of the smear to that of a known amount
of a molecular weight standard.
The concentration should be ∼20 to 40 ng/µl.

Circularize genomic material around synthetic oligonucleotides

12. Anneal oligonucleotides T30-T and T30-B in a thermal cycler by combining equal
amounts and heating to 95◦ C for 10 min and then slowly cooling to room temperature
over the course of 1 hr.
The final concentration of the T30 insert-linker should be 100 µM. T30 has T 3 overhangs
designed to be complementary to the A 3 overhangs of the genomic DNA.

13. Ligate genomic DNA fragments in the presence of annealed T30 using the Quick
ligation kit as follows:

170 ng size-selected DNA (0.25 pmol at ∼1 kb)

0.8 µl 1 µM annealed T30 (0.8 pmol)
40 µl 2× Quick ligation buffer
4.0 µl Quick T4 DNA ligase
H2 O to a final volume of 80 µl.

Incubate reaction 10 min at room temperature then heat-inactivate 10 min at 65◦ C.

Use the heat-inactivated reaction mixture directly in the next step without further
purification.
The ligation should be performed under conditions favoring formation of monomeric
recombinant circles (i.e., genomic DNA–T30). To this end, T30 should be present at a
three-fold molar excess to genomic fragments.

14. Eliminate all noncircularized material by exonucleolysis as follows:

1.0 µl 20 U/µl Exonuclease I

0.1 µl 100 U/µl Exonuclease III
80 µl heat-inactivated reaction mix
TE buffer, pH 8.0, to a final volume of 90 µl.

Incubate 45 min at 37◦ C, and then heat-inactivate 20 min at 80◦ C. Use this reaction
mixture directly in the next step without further purification.
Polony DNA
Sequencing

7.8.4
Supplement 76 Current Protocols in Molecular Biology
Amplify circular DNA
15. Prepare the master mix for hyper-branched rolling-circle amplification of the circular
DNA:
12 µl 25 mM dNTP mix
30 µl 10× RepliPhi phi29 reaction buffer
15 µl 1 mM N6 oligonucleotides
213 µl H2 O
30 µl circularized DNA.
Split into 6 tubes of 50 µl each.
16. Denature material by heating to 95◦ C for 5 min, then anneal by rapidly cooling to
4◦ C. Add 2.5 µl phi29 polymerase to each tube, keeping on ice. Incubate overnight
at 30◦ C.
17. Pool amplified circular DNA into a single tube. Purify DNA using a Microcon-30
column according to the manufacturer’s instructions, washing with 1 ml TE buffer,
pH 8.0. Wash the membrane several times to maximize recovery.
18. Resuspend the DNA using 750 µl buffer EB, preheated to 50◦ C. Quantitate the DNA
using a NanoDrop ND-1000 spectrophotometer.
Concentration should be ∼230 ng/µl.
Release paired tags
19. Digest the amplified DNA with MmeI to release the 70-bp T30-linked paired tags:
40 µg amplified circular DNA
100 µl 10× NEBuffer 4
1.6 µl of 1.6 mM SAM
60 µl MmeI (2 U/µl)
H2 O to 1000 µl.
Split into eight tubes of 125 µl each on ice, and then incubate at 37◦ C for 30 min.
MmeI cuts 18 to 19 bp from its recognition site, which is 1 bp from the start of the genomic
fragment. The released fragments contain the common T30 sequence flanked by two 17-
to 18-bp tags of genomic DNA.
20. Immediately extract the DNA with phenol/chloroform and precipitate with ethanol
using glycogen as a carrier. Resuspend pellet in a total volume of 80 µl TE buffer,
pH 8.0.
21. Purify the 70-bp paired-insert library as in steps 9 and 10, resuspending precipitated
DNA in 20 µl TE buffer, pH 8.0.
The sample should be split across four 0.5-cm lanes of a precast 6% polyacrylamide gel
in TBE buffer.
22. Quantitate the gel-purified DNA by electrophoresis (step 11).
DNA concentration should be ∼12 ng/µl.
Add emulsion PCR adapters
23. To perform the blunting reaction (to remove 2-nt 3 overhangs from the MmeI diges-
tion), combine the DNA with components of the End-It DNA end repair kit as follows:
8.5 µl paired-insert library (∼100 ng DNA)
1.25 µl 10× buffer
1.25 µl 10× ATP
1.25 µl 10× dNTP mix
0.25 µl enzyme mix. DNA Sequencing
Incubate at room temperature for 45 min.
7.8.5
Current Protocols in Molecular Biology Supplement 76
 24. Put on ice and add TE buffer, pH 8.0 to bring volume to 50 µl. Extract the DNA
with phenol/chloroform and precipitate in ethanol using 20 µg glycogen as carrier.
Resuspend DNA in 8 µl TE, pH 8.0.
25. Prepare emulsion PCR amplification primer sequences by annealing oligonucleotides
FDV-T and FDV-B and oligonucleotides RDV-T and RDV-B as in step 12.
The final concentration of each oligonucleotide should be 50 µM. FDV and RDV are not
5 -phosphorylated, so they will not ligate to one another.

26. Prepare the ligation reaction mixture by mixing the following components on ice:

8.0 µl blunt-ended library DNA (∼100 ng, 2 pmol)

1.0 µl 50 µM FDV (50 pmol)
1.0 µl 50 µM RDV (50 pmol)
2.5 µl 10× T4 DNA ligase buffer (∼0.5×)
21.1 µl 40% PEG
12.3 µl H2 O.

27. Mix on ice, then add 2.0 µl of 2000 U/µl T4 DNA ligase and incubate overnight at
16◦ C to ligate primer sequences to the library molecules.
After ligation, several species will be present: FDV-insert-FDV, FDV-insert-RDV, and
RDV-insert-RDV.

28. Increase volume to 100 µl with TE buffer, pH 8.0. Extract the DNA with phe-
nol/chloroform, precipitate in ethanol using glycogen as a carrier, and resuspend in
10 µl buffer EB.
29. Purify the 135-bp FDV-RDV paired-insert library as in steps 9 and 10, resuspending
the precipitated DNA in 10 µl TE buffer, pH 8.0.
30. Assemble on ice the reaction mixture used to remove nicks in double-stranded library
molecules by nick translation:

10.0 µl library DNA

0.5 µl 25 mM dNTP mix
2.5 µl 10× NEBuffer 2
1.0 µl 10 U/µl E. coli DNA polymerase I
11.0 µl H2 O.

Incubate 30 min at 16◦ C.

31. Increase volume to 100 µl with TE buffer, pH 8.0. Extract the DNA with phe-
nol/chloroform, precipitate in ethanol using glycogen as a carrier, and resuspend in
10 µl TE buffer, pH 8.0.
Amplify FDV/RDV-adapted library
32. Prepare the master mix for PCR amplification (for eight tubes of 50 µl each):
50 µl 10× PCR buffer without MgCl2
4.0 µl 25 mM dNTP mix
15 µl 50 mM MgCl2
426 µl H2 O
1.0 µl 100 µM FDV-T
1.0 µl 100 µM RDV-T
2.0 µl 5 U/µl Platinum Taq DNA polymerase.
Polony DNA
Sequencing

7.8.6
Supplement 76 Current Protocols in Molecular Biology
33. Split the master mix into eight tubes of 50 µl each, and add 0.5 µl FDV/RDV-adapted
library material to each tube.
34. Carry out PCR using the following amplification cycles:

Initial step: 2 min 94◦ C (denaturation)

12 cycles: 30 sec 94◦ C (denaturation)
30 sec 55◦ C (annealing)
90 sec 72◦ C (extension).
The number of cycles is kept low to minimize PCR amplification bias. Due to suppression
PCR effects (Siebert et al., 1995), FDV/FDV- and RDV/RDV-adapted molecules are not
amplified.
35. Purify DNA using Qiaquick columns according to the manufacturer’s instructions
for PCR cleanup.
36. Purify the amplified 135-bp FDV-RDV paired-insert library as in steps 9 and 10,
resuspending the precipitated DNA in 10 µl TE buffer, pH 8.0.
When excising the 135-bp library band, be careful to avoid contamination by marker
DNA, if used.
37. Quantitate the final DNA library by electrophoresis (step 11).
The concentration should be ∼2 ng/µl.

EMULSION PCR OF PAIRED-TAG LIBRARY ON MICROBEADS BASIC

PROTOCOL 2
Emulsion PCR (ePCR) is used to generate clonally amplified DNA fragments suitable
for polony sequencing. First, a biotinylated PCR primer is immobilized on streptavidin-
coated beads. Next, billions of microreactors are formed by the emulsification of a PCR
mix in mineral oil (Dressman et al., 2003). When one bead (bearing an immobilized
PCR primer) and one template molecule are trapped in a compartment, amplification
will proceed and result in tens of thousands of copies of the template molecule attached
to the bead. The result is a complex mixture of clonally amplified templates (polonies)
anchored to beads. Each bead bears numerous copies of the same template, but different
beads bear different templates.

Materials
MyOne C1 1-µm paramagnetic, streptavidin-coated beads (Dynal)
Bind and wash (B&W) buffer (see recipe)
Oligonucleotides:
1 mM PR1-F-2BIO: 5 -dual-biotin-CCACTACGCCTCCGCTTTCCTCTCTAT
GGGCAGTCGGTGAT-3
2 mM PR1-R: 5 -CTGCCCCGGGTTCCTCATTCTCT-3
10 µM PR1-3LF: 5 -CCTCTCTATGGGCAGTCGGTGAT-3
TE buffer, pH 8.0 (APPENDIX 2)
Light mineral oil (Sigma)
10% Span 80 (see recipe)
Tween 80
Triton X-100
10× PCR buffer without MgCl2 (Invitrogen)
50 mM MgCl2
25 mM dNTP mix (25 mM each nucleotide)
5 U/µl Platinum Taq DNA polymerase (Invitrogen)
Template DNA (library DNA at appropriate concentration; see Basic Protocol 1
and Support Protocol) DNA Sequencing

7.8.7
Current Protocols in Molecular Biology Supplement 76
 NXS buffer (see recipe)
0.1 M NaOH
1.5-ml microcentrifuge tubes
Magnetic particle concentrator (MPC; Dynal)
2-ml cryogenic vials (Corning no. 430661)
Stir bar, flea-size (VWR no. 58948-353)
Magnetic stirrer (closed-loop; VWR)
200-µl eight-tube PCR strips
Thermal cycler
Bind forward PCR primer to microbeads
1. To 30 µl MyOne beads in a 1.5-ml microcentrifuge tube, add 30 µl B&W buffer.
2. Mix and then remove all liquid using a magnetic particle concentrator (MPC).
Vortex to mix in all steps except those where enzyme activity is required. In those cases
mix gently by flicking or rotating the tube.
3. Wash two times in 60 µl B&W buffer using an MPC, and resuspend in 60 µl B&W
buffer.
4. Add 1.2 µl PR1-F-2BIO, mix, and incubate 20 min at room temperature.
Mix periodically by pipetting to prevent settling of beads.
PR1-F-2BIO is synthesized with a dual biotin at the 5 end to tightly bind the streptavidin-
coated microbeads. Dual biotin is necessary to ensure the template remains on the bead
for the duration of the sequencing run.
5. Remove all liquid using an MPC, and then wash twice in 100 µl B&W buffer and
once in 100 µl TE buffer, pH 8.0.
6. Remove all liquid using an MPC and resuspend in 60 µl TE buffer.
Concentration should now be ∼5 × 109 beads/ml.
Perform emulsion PCR reaction
7. Prepare the oil phase by assembling six 1.5-ml microcentrifuge tubes containing the
following:
545 µl light mineral oil
450 µl 10% Span 80
4.0 µl Tween 80
0.5 µl Triton X-100.
Vortex each tube individually for at least 20 sec to thoroughly mix all components.
Watch the pipet tips carefully to ensure that viscous reagents are delivered into the tube.
Use reverse pipetting (see http://www.rainin.com/pdf/rainin classic manual.pdf) for all
oil-phase components.
8. Prepare the aqueous phase by combining the following:

96 µl 10× PCR buffer without MgCl2

359 µl 50 mM MgCl2
135 µl 25 mM dNTP mix
12.0 µl 2 mM PR1-R
4.8 µl 10 µM PR1-3LF
60 µl microbeads from step 6
246 µl H2 O
Polony DNA
Sequencing 54 µl 5 U/µl Platinum Taq DNA polymerase
1.0 µl template DNA.
7.8.8
Supplement 76 Current Protocols in Molecular Biology
 9. Place twelve 2-ml cryogenic vials in the centers of three magnetic stirrers set to
1400 rpm.
10. Place a flea-size stir bar in each vial.
11. Add 400 µl oil phase to each vial.
12. Add 75 µl aqueous phase, dropwise over 1 min, to each vial.
Be sure each drop of liquid hits the bottom of the tube, not the walls.
13. Stir for 30 min using a magnetic stirrer.
14. Pipet the contents of each cryogenic vial into a separate eight-tube PCR strip (50 µl
per tube) for a total of twelve strips, or 96 tubes.
Aspirate and dispense slowly to minimize the amount of emulsion that sticks to the walls
of the pipet tip.
15. Carry out PCR using the following amplification cycles:

Initial step: 2 min 94◦ C (denaturation)

120 cycles: 15 sec 94◦ C (denaturation)
30 sec 57◦ C (annealing)
75 sec 70◦ C (extension)
Finish: 2 min 72◦ C
Hold: indefinitely 4◦ C.

16. Pool contents of each strip of eight 200-µl tubes into a separate 1.5-ml microcen-
trifuge tube (twelve tubes). Immediately proceed with recovery.
Recover beads from emulsion
17. Add 800 µl NXS buffer to each tube.
18. Vortex each tube individually at maximum speed for at least 20 sec.
19. Centrifuge all tubes 90 sec at 16,110 × g, room temperature.
20. Aspirate all but 100 µl supernatant (being sure to remove all of the top oil phase).
21. Repeat steps 17 to 20 two more times.
22. Pellet beads in tubes using an MPC, and aspirate all liquid.
23. Resuspend each pellet in 25 µl TE buffer, pH 8.0, and transfer all to a single new
tube. Mix by pipetting up and down.
All twelve tubes should be pooled into this one new tube. Try to minimize the amount of
oil and detergent carried over into the new tube from the walls of the old tube.
24. Pellet the beads using an MPC, aspirate all liquid, and add 250 µl TE buffer, pH 8.0.
Vortex briefly, then pulse centrifuge to pull the liquid into the bottom of the tube.
Repeat two more times.
25. Pellet beads using an MPC, and aspirate all liquid.
26. Remove the nonbiotinylated strand from the beads by incubating in 100 µl of 0.1 M
NaOH for 20 min at room temperature. Periodically mix by pipetting to prevent
beads from settling.
27. Pellet beads using an MPC, and remove all liquid.
28. Wash once with 100 µl 0.1 M NaOH and three times with 150 µl TE buffer. Mix
each wash by vortexing to ensure removal of residual NaOH.
DNA Sequencing
29. Pellet beads using an MPC, aspirate all liquid, and resuspend in 60 µl TE buffer.
7.8.9
Current Protocols in Molecular Biology Supplement 76
 BASIC ENRICHMENT FOR AMPLICON-BEARING BEADS
PROTOCOL 3
The result of ePCR is a population of beads, most (∼80%) of which do not bear an
amplified template. In order to maximize the number of amplicon-bearing beads on the
sequencing array, an enrichment step is required. Fluorescence-activated cell sorting
can be employed to select for these beads by hybridizing a fluorescently labeled probe
specific to amplicons. Alternately, the centrifugation technique presented below can be
used for more rapid and cost-effective enrichment.

In this protocol, the dense paramagnetic ePCR beads are incubated with low-density
polystyrene beads bearing an oligonucleotide complementary to a common amplicon
sequence. Amplicon-bearing beads hybridize to these capture beads, while non-amplicon-
bearing beads do not. The density differential between captured and uncaptured ePCR
beads is exploited for enrichment by centrifugation through a viscous 60% glycerol
solution; the former remain in the supernatant while the latter form a pellet.

Materials
Spherotech particles (3-µm streptavidin-coated polystyrene particles; Spherotech
no. SVP-30-5)
Bind and wash (B&W) buffer (see recipe)
1 mM PR1-BIOXL:
5 -biotinTEG-CGTACCCCGCTTGGTCTTTCTCCCGTACCCCGCTTGG
TCTTTCTCCCTGCCCCGGGTTCCTCATTCTCT-3
TE buffer, pH 8.0 (APPENDIX 2)
ePCR beads (Basic Protocol 2)
Glycerol
0.1 M NaOH
Magnetic particle concentrator (MPC; Dynal)
1.5-ml microcentrifuge tubes
Bind capture oligonucleotide to capture beads
1. Centrifuge 400 µl Spherotech particles for 30 sec at 16,110 × g, room temperature.
2. Remove all liquid, and wash pellet with 400 µl B&W buffer.
3. Centrifuge for 30 sec at 16,110 × g.
4. Remove all liquid, and resuspend in 400 µl B&W buffer.
5. Add 8 µl of 1 mM PR1-BIOXL and incubate at room temperature for 20 min.
6. Wash three times in 800 µl TE buffer, pH 8.0, pelleting each time by centrifugation.
Resuspend final pellet in 80 µl B&W buffer.
Anneal capture beads to emulsion PCR beads
7. Add 60 µl ePCR beads to 80 µl capture beads and incubate at 56◦ C for 10 min.
8. During the incubation, prepare 1 ml of fresh 60% (v/v) glycerol in H2 O. Vortex to
mix thoroughly.
Must be made fresh each day.

9. Add 150 µl of 60% glycerol to each of four 1.5-ml microcentrifuge tubes.

10. Layer 35 µl bead mixture from step 7 onto the top of the glycerol solution in each
of the four tubes.
Be careful to pipet gently so the bead solution does not mix with glycerol solution.
Polony DNA
Sequencing

7.8.10
Supplement 76 Current Protocols in Molecular Biology
Separate bead populations by centrifugation
11. Centrifuge tubes for 1 min at 16,110 × g.
The amplicon-bearing ePCR beads are hybridized to the complementary sequences car-
ried on the low-density polystyrene capture beads and will remain in the supernatant; the
ePCR beads without amplified sequences will form a pellet in the bottom of the tube.

12. Transfer all the supernatant to four new microcentrifuge tubes, being careful not to
disturb the pellets.
If necessary, wash the walls of the tubes with the supernatant to collect any beads bound
there.

13. To each recovered supernatant tube, add 1 ml H2 O and vortex until thoroughly mixed.
14. Centrifuge for 3 min at 16,110 × g, room temperature, to pellet all beads.
15. Aspirate, being careful to not disturb the pellet, and then place tubes on an MPC to
hold the pellet firmly in place. Remove any remaining liquid.
16. Separate ePCR beads from capture beads by adding 120 µl of 0.1 M NaOH and
incubating 10 min at room temperature to denature the hybridized DNA strands.
17. Using the MPC, collect the amplicon-bearing ePCR beads into a pellet. Remove and
discard all liquid.
The liquid contains the released capture beads.

18. Wash the pellet twice by adding 120 µl of 0.1 M NaOH, vortexing briefly, centrifuging
briefly to pull the liquid to the bottom of the tube, and collecting the pellet using the
MPC.
19. Resuspend in 6.5 µl TE buffer, pH 8.0.
CASTING A POLONY BEAD ARRAY BASIC
PROTOCOL 4
The amplified, enriched beads are cast in a two-dimensional array to fix their positions
during the subsequent sequencing reactions by mixing with acrylamide and pouring into
a shallow mold formed by a Teflon-masked microscope slide. The coverslip placed on
top of the gel is treated with Bind Silane, which covalently attaches the gel to the glass
coverslip. The acrylamide gel mixture is formulated such that it takes 30 min or more
to polymerize, during which time the beads settle into a single monolayer before their
positions become fixed. This allows efficient imaging of all beads in the gel, since they
all reside in the same focal plane of the optics used for imaging.

Once polymerized, the gel is approximately 30 microns thick, while the beads are only
1 micron in diameter. To ensure the DNA-bearing beads are accessible to reagents for
enzymatic sequencing reactions, the gel is cast such that the beads will be fixed at the
exposed surface (i.e., opposite the side attached to the Bind Silane–treated glass).

Materials
1% Triton X-100 in H2 O
Glacial acetic acid
Bind Silane (Promega)
100% ethanol
Enriched ePCR beads in TE buffer, pH 8.0 (Basic Protocol 3)
40% acrylamide/bisacrylamide (19:1) solution
5% TEMED
0.5% ammonium persulfate (APS)
Wash 1 (see recipe) DNA Sequencing

7.8.11
Current Protocols in Molecular Biology Supplement 76
 Round coverslips, 40 mm diameter, no. 1.5 (Bioptechs no. 40-1313-0319)
1000-ml plastic beaker
Shaker
Vacuum desiccator
Teflon-masked microscope slides (Erie Scientific, no. ER-203W)
Treat coverslips with Bind Silane
1. Wash round coverslips in 1% Triton X-100 for 20 min with gentle shaking.
2. Prepare silane solution in a 1000-ml plastic beaker:
500 ml H2 O
110 µl glacial acetic acid
2 ml Bind Silane.

Stir for 15 min until Bind Silane droplets have completely disappeared.
3. Immerse coverslip in silane solution and incubate 1 hr at room temperature with
gentle shaking.
4. Wash thoroughly with H2 O, then rinse with 100% ethanol.
5. Allow to dry completely in vacuum desiccator. Store coverslips up to 1 month under
vacuum.
Cast enriched ePCR beads
6. Prepare bead/acrylamide mixture:

6.5 µl enriched ePCR beads in TE buffer

1.25 µl 40% acrylamide/bisacrylamide solution (19:1)
0.5 µl 5% TEMED
0.5 µl H2 O.

Mix by pipetting, then add 0.75 µl of 0.5% APS.

7. Immediately pipet 9.5 µl onto a Teflon-masked microscope slide and cover with a
silane-coated coverslip. Allow to polymerize for 45 min.
The coverslip should spread the liquid across the entire surface of the circular well on
the Teflon-masked slide. During polymerization, the beads will settle to the surface of the
gel closest to the microscope slide.
8. Invert the slide/coverslip stack and remove microscope slide from gel by gently
lifting with a razor blade near the edge of the coverslip.
The gel should remain attached to the coverslip.

9. Immerse the coverslip, with polony bead array attached, in wash 1 until ready for
use, up to ∼1 month at 4◦ C.

BASIC DNA SEQUENCING BY LIGATION WITH DEGENERATE FLUORESCENT

PROTOCOL 5 NONAMERS
An anchor primer is first hybridized to a known sequence, or priming site, within the
single-stranded template. This primer is synthesized with deoxyuridine in place of thymi-
dine. A ligation reaction is then performed with a pool of fully degenerate, fluorescently
labeled nonanucleotides. During the ligation reaction, each bead becomes tagged with
a fluorophore that indicates the identity of the base present at the position being inter-
Polony DNA rogated. After imaging the array to determine which fluorophore labels each bead, the
Sequencing
anchor-primer/fluorescent-nonamer complex is removed from the template by reacting
7.8.12
Supplement 76 Current Protocols in Molecular Biology
Figure 7.8.2 Schematic of the 135-bp bead-bound template and sequences of oligonucleotides used in sequencing
reactions (Basic Protocol 5). For each reaction, the combination of anchor primer(s) and fluorescent nonamers is specified
by the position in the tag being sequenced. All anchor primers should be synthesized with deoxyuridine to allow enzymatic
removal of the primers. Fluorescent nonamers should be synthesized by Integrated DNA Technologies (IDT) with fully
degenerate positions hand mixed to ensure equal ratios of each nucleotide at each position. The T30 sequence is shown
as T30L and T30R to emphasize that the sequencing primers do not bind the full length of T30. Abbreviations: DMA and
DMB, distal minus A and B strands; DP, distal plus; PM, proximal minus; PPA and PPB, proximal plus A and B strands;
PR1-F and PR1-R, forward and reverse PR1 primers.

with the uracil-specific excision reagent (USER) enzyme. The USER enzyme excises
all deoxyuridines from the anchor primer, resulting in short fragments that dissociate at
moderate temperatures (<60◦ C) due to their low melting points (Tm ), and leaving the
bead-bound template ready for another position to be interrogated.

The combination of degenerate nonamer pool and anchor primer specifies which position
in the template the reaction will interrogate (see Fig. 7.8.2). Whether the anchor primer
or the nonanucleotides should be synthesized with a 5 phosphate is dependent on which
position in the template is being queried with a given combination of anchor primer and
nonamer. When sequencing the 5 side of a tag, the anchor is phosphorylated, and when
sequencing the 3 side, the nonamers are phosphorylated.

At the start of a sequencing run, all 3 ends of unextended forward ePCR primers and tem-
plates are capped by the addition of dideoxynucleotides using terminal deoxytransferase.
This capping reaction is necessary to prevent nonspecific ligation of labeled nonamers to
bead-bound ePCR primers and templates.

This protocol describes the ligation and stripping steps required to query the base at a
specific position in an iterative fashion. When using the autosampler and flow cell system,
multiple solutions can be prepared for one or several cycles and used with a computer
program preset with the appropriate times, temperatures, and volumes indicated in the
protocol steps. Imaging and analysis procedures are discussed in Critical Parameters and
Troubleshooting. DNA Sequencing

7.8.13
Current Protocols in Molecular Biology Supplement 76
 Materials
Bead array with template DNA (Basic Protocol 4)
5× tailing buffer (Invitrogen)
1.25 mM ddNTP mix (1.25 mM each dideoxynucleotide)
15 U/µl terminal deoxytransferase (recombinant; Invitrogen)
Wash 1 (see recipe)
6× SSPE (APPENDIX 2)/0.01% Triton X-100
Anchor primer (see Fig. 7.8.2)
2000 U/µl T4 DNA ligase and 10× buffer (NEB)
100 µM degenerate fluorescent nonamer mix (Integrated DNA Technologies)
TE buffer, pH 8.0 (APPENDIX 2)
1 U/µl USER enzyme mix (NEB)
Automated fluorescent microscope with flow cell (see Table 7.8.1 for parts list)
Autosampler (see Table 7.8.1)
1-ml disposable syringe
0.45-µm pore size, 4-mm cellulose acetate syringe filter (VWR)
Perform dideoxynucleotide capping reaction
1. Prepare the capping reaction mix:

114 µl H2 O
30 µl 5× tailing buffer
4 µl 1.25 mM ddNTP mix
2 µl 15 U/µl terminal deoxytransferase.

2. Place the bead array with template DNA on a coverslip into the flow cell mounted
on the microscope stage. Prime fluid lines with wash 1.
3. Add the reaction mix to the appropriate well of the autosampler. Incubate 60 min at
35◦ C.
Hybridize anchor primer to array
4. Prepare the hybridization mix by combining:

150 µl 6× SSPE with 0.01% Triton X-100

1.5 µl 1 mM anchor primer.

5. Add hybridization mix to the appropriate well of the autosampler. Incubate 5 min at
56◦ C, then 2 min at 42◦ C.
Ligate fluorescent nonamers to visualize complexes
6. Prepare ligation reaction on ice:

15 µl 10× T4 DNA ligase buffer

12.0 µl 100 µM degenerate fluorescent nonamer mix
3.0 µl 2000 U/µl T4 DNA ligase
120 µl H2 O.

When preparing multiple ligation reaction mixtures (for sequential cycles) at once, be
sure they are consumed within 6 hr, since T4 DNA ligase is unstable once mixed with
reaction buffer. In addition, only one mix containing 5 -phosphorylated nonamers should
be prepared, and it should be used in the first sequencing cycle performed.

Polony DNA
Sequencing

7.8.14
Supplement 76 Current Protocols in Molecular Biology
 7. Aspirate the solution into a 1-ml syringe, attach a 0.45-µm filter to the Luer syringe
adapter, and pass the contents through the filter into the appropriate well of the
autosampler. Incubate 30 min at 32◦ C.
The nonamer solutions contain large aggregates of fluorescent dye. Filtering is essential
to ensure these aggregates are not introduced into the flow cell, since they will gradually
accumulate by sticking to the gel, interfere with imaging and data extraction, and poison
subsequent reactions.
Incubation temperature and duration can be optimized for variations in signal. If signal is
low, a slightly longer time or higher temperature can be used; if signal is very high, cycle
time can be decreased by decreasing incubation time appropriately. Although increased
concentrations of nonamer may also drive the reaction to completion faster, it should be
noted that cost will increase and fidelity may be affected.
8. Image using fluorescence microscopy to determine the nonamer ligated to each bead
and a single position of one tag.
The automated microscope used for polony sequencing includes a number of add-on com-
ponents both for imaging (e.g., motorized stage, piezo objective positioner, illumination
wheels) and for fluidics (e.g., flow cell assembly, autosampler, digital I/O modules). Com-
plete details are beyond the scope of this unit. All major components are listed in Table
7.8.1, and a discussion is provided in Critical Parameters and Troubleshooting. More
detailed information on assembly (including fittings, tubing, and more) can be found at
http://arep.med.harvard.edu/Polonator.

Strip fluorescent complexes from the beads

9. Prepare stripping reaction mix by combining:

300 µl TE buffer, pH 8.0

6 µl USER enzyme.

Stripping is performed by reacting the array twice with the stripping solution, hence
the volume is double that of a standard reaction. This can be adjusted according to the
specific performance observed. If stripping is observed to be complete after one round of
reaction and heating, the second stripping can be eliminated.

10. Add wash 1 and the stripping mix to the appropriate wells of the autosampler.
Perform stripping protocol as follows:

incubate at 37◦ C for 5 min

incubate at 56◦ C for 1 min
wash
incubate at 37◦ C for 5 min
wash.

The autosampler software for stripping implements this protocol by default.

Query next position

11. Repeat steps 4 to 8 using a different combination of either anchor primer or fluores-
cent nonamer or both for up to 26 cycles, stripping the fluorescent complexes as in
steps 9 and 10 between sequencing cycles.
The read length limit of T4 DNA ligase is 6 or 7 bp from the ligation junction, for a total
of 13 bp per tag.

DNA Sequencing

7.8.15
Current Protocols in Molecular Biology Supplement 76
 SUPPORT TITRATION OF TEMPLATE FOR CLONAL AMPLIFICATION BY ePCR
PROTOCOL
Before performing emulsion PCR on a new sample for the first time, the appropriate
amount of template must be determined empirically. Since emulsion PCR generates
clonally amplified beads by a limiting dilution of template, one can be sure most beads
are clonal by finding the template concentration at which approximately 15% to 20% of
beads bear an amplicon. Generally, the correct amount of template is somewhere near
1 fmol per reaction. Sample four different concentrations and, if necessary, interpolate
to determine the final amount. For example, given that the concentration of the library
material is known, amounts of 5 fmol, 1 fmol, 0.2 fmol, and 0.05 fmol per reaction should
be tried. Somewhere in this range, a linear increase in amplified beads will be observed
with a linear increase in template.

After amplification of the four template samples, bead arrays are cast on microscope
slides instead of coverslips (see Basic Protocol 4) for ease of handling. Amplified beads
are identified by hybridizing a fluorescently labeled probe complementary to a sequence
present in the central (T30) sequence. The percent of amplified beads can then be
determined by fluorescence microscopy.

Additional Materials (also see Basic Protocols 2 and 4)

Fluorescently labeled probes:
100 µM T30-P2-Cy5-A: 5 -Cy5-AGUUGGACGUACGGCC-3
100 µM T30-P2-Cy5-B: 5 -Cy5-AGUCGGAGGUCAAGGC-3
Heat block or slide thermal cycler
Microscope for bright-field and fluorescence microscopy, with CCD camera and
filter set for Cy5
Amplify template dilutions
1. Prepare four emulsion PCR reactions as in Basic Protocol 2, using four different
amounts of template DNA in the aqueous phase (step 8). Scale down each reaction
twelve-fold, so that each template amount is prepared in a single cryogenic vial (step
9) and amplified in a single eight-tube strip (step 14).
This will result in four strips of beads, one per template amount tested. Be sure to scale
down the amount of DNA used because each reaction is one-twelfth the size of a standard
emulsion PCR reaction.

2. When recovering the beads from the emulsion, be sure to not combine the four strips
as in step 23.
The result will be 4 tubes of beads.

3. Resuspend each tube in 6.5 µl TE buffer, pH 8.0, and proceed to labeling of amplicons
without performing enrichment.
Cast bead arrays
4. Cast each sample as a separate bead array on a Teflon-coated microscope slide. Use
the same procedure as in Basic Protocol 4, but treat the microscope slide with Bind
Silane instead of the coverslip, and invert the stack during polymerization so that the
beads are immobilized next to the coverslip instead of the slide.
Hybridize fluorescent probe
5. Prepare the hybridization mixture:

100 µl of 6× SSPE with 0.01% Triton X-100

1.0 µl of 100 µM T30-P2-Cy5-A (5 -Cy5-AGUUGGACGUACGGCC-3 )
Polony DNA 1.0 µl of 100 µM T30-P2-Cy5-B (5 -Cy5-AGUCGGAGGUCAAGGC-3 )
Sequencing

7.8.16
Supplement 76 Current Protocols in Molecular Biology
 6. Dry the Teflon surface of the slide with a Kimwipe, being careful to not touch the
gel.
7. Pipet the hybridization mixture onto the gel.
The mixture will stay contained on top of the gel if the slide was dried properly.

8. Place the slide face up on a heat block or slide thermal cycler and incubate at 56◦ C
for 5 min.
9. Wash for 5 min in fresh wash 1 to remove free fluorescent probe.
Identify amplified beads
10. Place an array under a fluorescence microscope and acquire an image under trans-
mitted bright-field illumination.
Magnification of 20× is sufficient for the pixel size of most CCD cameras. Long working
distance optics with relatively low NA are sufficient when imaging bead arrays at low
density (∼20,000 beads per field of view). When higher densities are desired (∼150,000
to 200,000 per field of view) high-NA dry optics should be used.
Depending on the particular optics used, the beads will appear as black circles or as
white points surrounded by dark halos.

11. Acquire a fluorescence image at the same position using appropriate excitation and
emission filters for Cy5.
12. Superimpose the two images in false color and determine the percentage of beads
that are labeled.
13. Calculate the amount of library material to use in each ePCR based on the percentage
of beads that bear labeled amplicons.

REAGENTS AND SOLUTIONS

Use deionized, distilled water in all recipes and protocol steps. For common stock solutions, see
APPENDIX 2; for suppliers, see APPENDIX 4.

Bind and wash buffer

5 mM Tris·Cl, pH 7.5 (APPENDIX 2)
0.5 mM EDTA, pH 8.0 (APPENDIX 2)
1.0 M NaCl
Store up to 6 months at room temperature
NXS buffer
10 mM Tris·Cl, pH 7.5 (APPENDIX 2)
1 mM EDTA, pH 8.0 (APPENDIX 2)
100 mM NaCl
1% Triton X-100
1% (w/v) SDS
Mix ingredients thoroughly
Store up to 6 months at room temperature
Span 80, 10% (v/v)
10% (v/v) Span 80 in light mineral oil
Measure ingredients by positive-displacement syringe or pipet. Vortex to mix thor-
oughly and incubate under vacuum to remove air bubbles. Store up to 1 week at
room temperature.
DNA Sequencing

7.8.17
Current Protocols in Molecular Biology Supplement 76
 Wash 1
10 mM Tris·Cl, pH 7.5 (APPENDIX 2)
50 mM KCl
2 mM EDTA, pH 8.0 (APPENDIX 2)
0.01% (v/v) Triton X-100
Mix ingredients thoroughly
Store up to 6 months at room temperature

COMMENTARY
Background Information unit, polonies are generated from a genomic
Since the dideoxy Sanger method (Sanger paired-tag library on magnetic beads and are
et al., 1977) was introduced almost 30 years then enriched by hybridization to a common
ago, DNA sequencing has become a standard sequence. The enriched beads are cast in an
(albeit expensive) tool in the biologist’s reper- array in a polyacrylamide matrix, where they
toire. With the list of reference genome se- are sequenced.
quences growing, attention is shifting toward The focus of this unit is genomic resequenc-
the study of how a reference genome dif- ing. By using a reference genome as a scaffold
fers from derivative experimental genomes of onto which reads are placed and interrogated
interest (e.g., adaptive evolution of bacterial for differences, the genome in question can
strains, drug resistance/sensitivity). In many be effectively resequenced using much shorter
cases, to fully enable these studies, the cost reads than if a reference did not exist. Ad-
per quality base of DNA sequence must be ditionally, paired tags allow for placement of
lower than that obtained with state-of-the-art reads too short to otherwise span repetitive re-
Sanger sequencing. gions. All that is required is that a read-pair
Polony sequencing aims to address this be unique enough, given its intertag distance
need by leveraging highly multiplex processes distribution, that it places to only one loca-
during both sample preparation and sequenc- tion on the reference genome. Shendure et al.
ing itself (Shendure et al., 2005). This allows (2005) demonstrated that 90% coverage of the
an effective decrease in reagent volumes per E. coli genome is feasible with paired 13-bp
reaction from the microliter scale to the fem- reads having an intertag distance distribution
toliter scale, while also dramatically reduc- of 1 kb ± 250 bp. In addition to genomic re-
ing data acquisition time per base. These re- sequencing, polony technology in general has
ductions in time and materials translate to a been used for haplotyping (Mitra et al., 2003;
lower cost per base and, as an added bonus, Turner et al., 2006; Zhang et al., 2006) and
higher throughput. As a cyclic array method splicing analysis (Zhu et al., 2003).
(for a more thorough discussion, see Shen-
dure et al., 2004), sequence is generated base-
by-base through successive cycles of ligation Critical Parameters and
and imaging. Each cycle of ligation and imag- Troubleshooting
ing sequences a single base in the template. Paired-tag library
Each 17- to 18-bp genomic tag in the sequenc- The key to successful execution of the li-
ing template yields 13 bp of sequence when brary protocol is to achieve yields at each step
queried by T4 DNA ligase (with the central 4 at least as high as indicated in the protocols. In
to 5 bp of the tag remaining unsequenced). The particular, gel purifications, ethanol precipita-
result of a sequencing run is one or more mil- tions, and column purifications must be repro-
lion discontiguous reads of 26 bp each. While ducibly efficient. DNA quantification should
each is much shorter than a Sanger read, for be performed as indicated in the protocol, be-
many applications this length is sufficient. cause NanoDrop readings are not reliable in
Polony sequencing derives its name from the authors’ hands following ethanol precipi-
the sequencing of DNA molecules that are tation.
clonally amplified in or on a solid matrix. The most critical step in the paired-tag li-
Each amplified clone represents a polymerase brary protocol is the circularization of the ge-
colony or “polony.” Polonies can be ampli- nomic fragment around the synthetic linker-
fied from template molecules immobilized in insert. Shendure et al. (2005) found this to
Polony DNA a polyacrylamide matrix (Mitra and Church, be the rate-limiting step in terms of the num-
Sequencing 1999) or on beads (Diehl et al., 2006). In this ber of unique sequencing templates that were
7.8.18
Supplement 76 Current Protocols in Molecular Biology
generated in the library construction. This li- A jig can be used to hold the coverslip
brary protocol is inefficient from a DNA mass in the correct position relative to the Teflon
standpoint because it represents the product well of the microscope slide. A 0.5-in.-thick
of a number of inefficient individual steps. block of polyethylene can be machined with
Additionally, the release of a pair of 18- a shallow pocket that will constrain move-
bp tags from each fragment following circu- ment of both the slide and the coverslip. This
larization results in a nominal mass loss of ensures that the array always forms in the
∼96% for 1000-bp fragments (linearly worse center of the coverslip, and thus is always
for longer fragments). This loss necessitates properly placed in the center of the flow cell
the hyperbranched rolling-circle amplification chamber.
performed after circularization.
Software, hardware, and sequence analysis
Emulsion PCR
To perform polony sequencing using the
When performing emulsion PCR, it is criti-
protocols presented here, repetitive cycles of
cal not to deviate in any way from the protocol
sequencing biochemistry and imaging are per-
provided until it works to the point of pro-
formed until the desired read length is reached
ducing clonal, strongly amplified beads from
(e.g., 26 cycles for 26 bp). Then, fluorescence
stable emulsions. Once reliable, the method
data must be extracted from the ∼50,000
can be used as a control in further optimiza-
2-MB images generated by the instrument,
tions. Inconsistent stability from one batch to
normalized, and converted to sequence. In
the next is usually due to either inconsistent
the case of genomic resequencing, these short
measuring of reagents (especially detergents)
reads must be placed onto (or matched to)
or using oil phase components (e.g., 10% Span
a reference genome to determine which po-
80) that were not freshly prepared.
sitions differ. The process is summarized
Reverse amplification primer (nonbiotiny-
below.
lated) is present at a very high concentration
The polony sequencing instrument is an in-
in the reaction. This primer must be HPLC-
verted epifluorescence microscope which has
purified to prevent formation of primer-dimer
motorized, encoded x, y, and z axes, as well as
product on beads (which may also bear library
motorized fluorescence excitation and emis-
molecule amplicons). Although primer-dimer
sion wavelength selection. The motorized fea-
formation is also a function of sequence, the
tures are controlled by a series of C++ binary
PR1 sequences yield especially strong ePCR
files (available with source codes at http://
amplicons relative to several other primer
arep.med.harvard.edu/Polonator). Attached
pairs.
to the microscope is the fluidics/temperature
The interested reader is directed to excel-
control system, which performs the biochem-
lent papers (Dressman et al., 2003; Diehl et al.,
istry in the flow cell attached to the automated
2005, 2006; Li et al., 2006) which cover the
stage. It is controlled by a series of Matlab
technique of emulsion PCR on beads in more
scripts, also available at http://arep.med.
detail.
harvard.edu/Polonator.
Casting a polony bead array A list of the major components necessary
When preparing Bind Silane–treated slides to build a duplicate of the instrument used by
or coverslips, be sure to use Bind Silane which the authors is shown in Table 7.8.1; a full
has not reached its expiration date. Use Erie parts list (including fittings, tubing, and de-
Teflon-masked slides as the gel molds. Their tails for assembly) is available at http://arep.
Teflon coating is unusually hydrophobic and med.harvard.edu/Polonator. In building an in-
uniform in surface flatness, both of which are strument, the recommendation is to make
necessary to keep the gel from leaking out the no substitutions for parts, even when the
top of the circular well before polymerization. particular components seem to have simi-
Be certain that no bubbles are present in the lar performance characteristics. Subtle fea-
gel after it has been spread between the slide tures present in each component listed are
and the coverslip. A small bubble noticed in the in many cases necessary for the system to
gel after lowering the coverslip onto the slide function as expected. For example, Nikon is
can be removed by gently lifting the edge of the only microscope manufacturer that man-
the coverslip slightly with the edge of a razor ufactures a 20× plan apochromatic objec-
blade to move the bubble to the edge of the gel tive with both a phase contrast ring (neces-
and slowly lowering the coverslip back onto sary for imaging densely packed bead arrays)
the slide. If this process is performed carefully, and a 1.0-mm working distance (necessary
DNA Sequencing
the gel will not spread outside the Teflon well. for adequate clearance around the flow cell).
7.8.19
Current Protocols in Molecular Biology Supplement 76
 Table 7.8.1 Parts List for Automated Microscope Used in Polony Sequencing

Quantity Part no. Manufacturer Description

1 86022-SPR Chroma Cy3/Cy5 polychroic filter w/matched

individual exciters/emitters
1 41004* Chroma Custom TxRed filter set (S572/23x,
Q595LP, S630/60m)
2 CFL Chroma Empty filter cube for TE2000
1 91002-02 Hamamatsu Camera, 1000x1000, electron
multiplied, 30 fps readout
1 500-H117E1T4 Prior Stage, motorized encoded x,y for
TE2000; 1-mm screw pitch
1 500-H30XY-E Prior Controller for x,y stage w/encoder
connectors
1 500-CS152V2 Prior Joystick for manual x,y control
1 LB-LS/OF30 Sutter Lambda light source, 300 W xenon
1 LLG Sutter Liquid light guide for LS, 2 meters
1 LG-N27 Sutter Adapter to attach light guide to
TE2000 (no epi arm)
1 LB10-3 Sutter Control unit for 3 wheels and
2 shutters
1 LB10-NWIQ Sutter 10-position wheel for 25-mm filters
with smart shutter
1 LB10-NEW Sutter 10-position wheel for 25-mm filters -
setup for emission
1 10-N27-EM Sutter Adapter to attach wheel to side port
of TE2000
1 IQ25-SA Sutter Smart shutter for transmitted light
1 10-N27-EC Sutter Adapter to attach transmitted light
shutter to Nikon HMX-II
1 O661301 Sutter Extra xenon bulb, 300 W
1 Pifoc 400-µm travel piezo positioner
w/controller
1 MEA51001 Nikon TE2000-E2 inverted microscope
1 MEF55010 Nikon T-HUBC hub controller
1 MEF55000 Nikon T-RP remote control pad
1 MPF52061 Nikon Universal power supply, 110-240V
1 MAE15001 Nikon Lamphouse HMX2
1 MEE59900 Nikon T-DH dia-illuminator, 100W
1 MAE15002 Nikon Collector lens for QH100
1 MBF13240 Nikon Lamp socket F/HMX 100W QI
1 MXA20425 Nikon FXA-HMX2 adpt f/halogen
1 84125 Nikon Microphot-FXA L.L. 12V-100W BU
1 91141-IN Nikon T-BP R100 optical path prism
1 MEF42252 Nikon TE2-PS 100W power supply
Polony DNA continued
Sequencing

7.8.20
Supplement 76 Current Protocols in Molecular Biology
Table 7.8.1 Parts List for Automated Microscope Used in Polony Sequencing, continued

Quantity Part no. Manufacturer Description

2 79035 Nikon Power cord

1 91155 Nikon Null modem cable
1 MEP51300 Nikon T-N6 sextuple nosepiece
1 MBN11710 Nikon Filter 45mm NCB11
1 MEL3000 Nikon Diaphot condenser turret
1 MEL36200 Nikon Diaphot condenser Lwd lens
1 MEH41200 Nikon Te-C Lwd Ph2 module
1 MEV51100 Nikon T-FLMC motorized cassette holder
1 MRD30200 Nikon Cfi plan apo dm 20× objective
1 060319-2-0303-20 Bioptechs Temperature controller, no alarm
1 060319-2-03-30 Bioptechs FCS2 chamber only, 30-mm aperture
1 060319-2-1242HG Bioptechs Low dead volume top, 3-in.
(0.020-in. i.d.) stainless-steel tubing,
drilled
1 060319-2-0049 Bioptechs FCS2 connector assembly, modified
for slide only
5 130119-5-292 Bioptechs Specialty microaqueduct slide - bus
bars, 2 holes, no grooves
20 1907-250 Bioptechs Shiny gasket, harvard die 452458
1 40-1313-0319 Bioptechs 40-mm round coverslips, 250/pk
1 A HMS machine Fluidics bulkheada
shop
2 B HMS machine Flow cell support armsa
shop
1 C HMS machine A/D I/O modulea
shop
1 D HMS machine Round coverslip caddya
shop
1 AL719-P Alcott Chro- Harvard autosampler
matography
1 3000XL Cavro Syringe pump: RS232 1/4-28 3-port,
500-µl
a Drawings and schematics available at http://arep.med.harvard.edu/Polonator/2006.

Neither of these specifications is adver- At the conclusion of the sequencing run,

tised as important by most vendors, but for images from each cycle must be aligned to
this particular application, both are indeed each other such that bead pixels from one im-
important. age correspond to the same bead-pixels in all
At the start of the instrument run, a C++ bi- other images from that same raster location.
nary file is executed on the computer attached Bright-field images are then thresholded and
to the microscope to determine the proper focal used to define the image pixels which coincide
position for each raster location. In this way, with beads. The mean intensity of each bead
during each subsequent cycle of imaging, the in each image is computed and saved to disk.
microscope can quickly acquire images from These fluorescence values are normalized and
all 500 to 700 raster locations without the need converted to four-dimensional unit vectors.
to refocus. When plotted in four-dimensional space, the
DNA Sequencing

7.8.21
Current Protocols in Molecular Biology Supplement 76
 data for each cycle will naturally form four tions in the plasma of patients with colorectal tu-
clusters corresponding to the four base iden- mors. Proc. Natl. Acad. Sci. U.S.A. 102:16368-
16373.
tities (A, C, G, or T). Bases are called by as-
signing each bead to the base identity of the Diehl, F., Li, M., He, Y., Kinsler, K.W., Vogelstein,
B., and Dressman, D. 2006. BEAMing: Single-
cluster to which it is nearest by Euclidean dis-
molecule PCR on microparticles in water-in-oil
tance. The quality score of each call is the emulsions. Nat. Methods 3:551-559.
Euclidean distance from the bead to the center
Dressman, D., Yan, H., Traverso, G., Kinzler, K.W.,
of its assigned cluster. and Vogelstein, B. 2003. Transforming single
To determine which sequenced bases differ DNA molecules into fluorescent magnetic par-
from bases in the reference genome, the polony ticles for detection and enumeration of ge-
reads must be placed on the reference sequence netic variations. Proc. Natl. Acad. Sci. U.S.A.
100:8817-8822.
while allowing for mismatches. The output of
this algorithm is the list of input polony reads Li, M., Diehl, F., Dressman, D., Vogelstein, B., and
Kinzler, K.W. 2006. BEAMing up for detection
annotated with the genomic position to which
and quantification of rare sequence variants. Nat.
each places, if any. The critical constraint is Methods 3:95-97.
that a given read only place to one location on
Mitra, R.D. and Church, G.M. 1999. In situ lo-
the reference genome, such that bases in the calized amplification and contact replication of
experimental genome can be unambiguously many individual DNA molecules. Nucleic Acids
assigned. Once a list of placements has been Res. 37:e34.
computed, the experimental genome sequence Mitra, R.D., Butty, V.L., Shendure, J., Benjamin,
can be easily derived by sorting the reads by R., Williams, B.R., Housman, D.E., and Church,
location and reporting the call at each location. G.M. 2003. Digital genotyping and haplotyping
with polymerase colonies. Proc. Natl. Acad. Sci.
When more than one base is called at a given
U.S.A. 100:5926-5931.
location (e.g., from two different reads), qual-
Sanger, F., Nicklen, S., and Coulson, A.R. 1977.
ity scores of each raw base call can be used to
DNA sequencing with chain-terminating in-
resolve the ambiguity. hibitors. Proc. Natl. Acad. Sci. U.S.A. 74:5463-
5467.
Anticipated Results Shendure, J., Mitra, R.D., Varma, C., and Church,
The raw data produced by each sequenc- G.M. 2004. Advanced sequencing technologies:
ing run is ∼50,000 2-MB images. These are Methods and goals. Nat. Rev. Genet. 5:335-
processed automatically by software avail- 344.
able at http://arep.med.harvard.edu/Polonator Shendure, J., Porreca, G.J., Reppas, N.B., Lin, X.,
to yield several million 26-bp reads, each an- McCutcheon, J.P., Rosenbaum, A.M., Wang,
M.D., Zhang, K., Mitra, R.D., and Church, G.M.
notated with quality scores for each raw base. 2005. Accurate multiplex polony sequencing of
However, the number of reads produced de- an evolved bacterial genome. Science 309:1728-
pends on the number of microbeads arrayed on 1732.
the coverglass and the fraction of those beads Siebert, P.D., Chenchik, A., Kellogg, D.E.,
that bear useable amplicons. Lukyanov, K.A., and Lukyanov, S.A. 1995. An
improved PCR method for walking in uncloned
genomic DNA. Nucleic Acids Res. 23:1087-
Time Considerations 1088.
Library construction takes ∼1 week if long
Turner, D.J., Shendure, J., Porreca, G., Church,
incubations are performed overnight. The pro- G., Green, P., Tyler-Smith, C., and Hurles,
tocol can be stopped, if necessary, after any M.E. 2006. Assaying chromosomal inversions
purification step. Titration of template con- by single-molecule haplotyping. Nat. Methods
centration for emulsion PCR takes 2 days. 3:439-445.
Emulsion PCR takes part of two days. In ex- Zhang, K., Zhu, J., Shendure, J., Porreca, G.J.,
perienced hands, two plates of emulsion PCR Aach, J.D., Mitra, R.D., and Church, G.M.
can be prepared, recovered, and enriched in 2 2006. Long-range polony haplotyping of in-
dividual human chromosome molecules. Nat
days by one person. The sequencing run takes Genet. 38:382-387.
3 hours to set up. Each sequencing cycle takes
Zhu, J., Shendure, J., Mitra, R.D., and Church, G.M.
1.75 hr, for a total of 49 hours if performed 2003. Single molecule profiling of alternative
continuously. pre-mRNA splicing. Science 301:836-838 .

Literature Cited
Diehl, F., Li, M., Dressman, D., He, Y., Shen, D., Contributed by Gregory J. Porreca, Jay
Szabo, S., Diaz, L.A. Jr., Goodman, S.N., David, Shendure, and George M. Church
Polony DNA K.A., Juhl, H., Kinzler, K.W., and Vogelstein, Harvard Medical School
Sequencing B. 2005. Detection and quantification of muta- Boston, Massachusetts
7.8.22
Supplement 76 Current Protocols in Molecular Biology