Download Repair of Site-Specific DNA Double-Strand Breaks in

The Plant Cell, Vol. 26: 2156–2167, May 2014, www.plantcell.org ã 2014 American Society of Plant Biologists. All rights reserved.
Repair of Site-Specific DNA Double-Strand Breaks in Barley
Occurs via Diverse Pathways Primarily Involving the
Sister Chromatid
W
Giang T.H. Vu,a Hieu X. Cao,a Koichi Watanabe,a Goetz Hensel,a Frank R. Blattner,a Jochen Kumlehn,a
and Ingo Schuberta,b,1
a Leibniz
b Faculty
Institute of Plant Genetics and Crop Plant Research, D-06466 Gatersleben, Germany
of Science and Central European Institute of Technology, Masaryk University, CZ-61137 Brno, Czech Republic
ORCID IDs: 0000-0001-7080-7983 (J.K.); 0000-0002-6300-2068 (I.S.)
DNA double-strand break (DSB) repair mechanisms differ in their requirements for a homologous repair template and in the
accuracy of the result. We aimed to quantify the outcome of repair of a single targeted DSB in somatic cells of young barley
(Hordeum vulgare) plants. Amplicon sequencing of three reporter constructs revealed 47 to 58% of reads as repaired via
nonhomologous end-joining (NHEJ) with deletions and/or small (1 to 3 bp) insertions. Alternative NHEJ revealed 2 to 5 bp
microhomology (15.7% of cases) or new replication-mediated short duplications at sealed breaks. Although deletions outweigh
insertions in barley, this bias was less pronounced and deleted sequences were shorter than in Arabidopsis thaliana. Between
17 and 33% of reads likely represent restoration of the original sequence. Depending on the construct, 20 to 33% of reads arose
via gene conversion (homologous recombination). Remarkably, <1 to >8% of reads apparently display synthesis-dependent
strand annealing linked with NHEJ, inserting 4 to 61 bp, mostly originating from the surrounding of breakpoints. Positional
coincidence of >81% of sister chromatid exchanges with target loci is unprecedented for higher eukaryotes and indicates
that most repair events for staggered DSBs, at least in barley, involve the sister chromatid and occur during S or G2 phase
of the cell cycle.
INTRODUCTION
DNA double-strand breaks (DSBs) can be caused directly by endonucleases, by ionizing irradiation and other S phase–independent
clastogens, or indirectly, when the DNA replication fork collapses
due to interference with single-strand DNA nicks or gaps, for instance, at positions of excision repair (Rothkamm et al., 2003;
Schubert et al., 2004). DSBs, if not repaired, are lethal for dividing
cells due to loss of acentric fragments and instability of centric
break ends. Living organisms have evolved a range of DSB repair
mechanisms. DSBs can be repaired correctly to restore the
prebreak status or result in genome variability by conversion,
inversion, insertion, deletion, or translocation (Figure 1; Schubert
et al., 2004). There are two main groups of DSB repair pathways,
which either use extended homologous sequences as template
(homologous recombination [HR]) or have no or only little (#25 bp)
sequence homology requirement (nonhomologous DNA endjoining [NHEJ] or microhomology-mediated end-joining [MMEJ]).
HR and NHEJ/MMEJ pathways may operate in both competitive
and collaborative manners, depending on the cell cycle phase,
cell type, and genome organization, providing situations that differ
in the availability and use of homologous repair templates (McVey
and Lee, 2008; Shrivastav et al., 2008). DSB repair by HR is usually
1 Address
correspondence to [email protected].
The author responsible for distribution of materials integral to the findings
presented in this article in accordance with the policy described in the
Instructions for Authors (www.plantcell.org) is: Giang T.H. Vu ([email protected]).
W
Online version contains Web-only data.
www.plantcell.org/cgi/doi/10.1105/tpc.114.126607
error-free if it utilizes the identical intact sister chromatid as template, but it may also use the homologous chromosome or ectopic
(nonallelic) homologous sequences as template (reviewed in
Puchta, 2005; Heyer et al., 2010; Jasin and Rothstein, 2013). In the
latter case, HR is error-prone. HR is the preferred pathway in late S
and G2 phase cells, when sister chromatids are available and tied
together by cohesins or actively find each other via the SMC5/6
complex (Watanabe et al., 2009). HR is initiated by the 59 to 39
resection of DNA ends at the DSB to produce single-stranded
overhanging 39 ends that invade a DNA duplex containing a homologous sequence (Symington and Gautier, 2011). Using homologous templates, the 39 tail within a D-loop is extended by
DNA polymerase. The newly synthesized 39 end of the invading
strand may dissociate from the template strand and anneal to the
39 overhang of the opposite break end through complementary
base pairing, followed by ligation of both strands. This type of
event is called synthesis-dependent strand annealing (SDSA) and
results in noncrossover products that represent gene conversion.
If the second 39 overhang of the DSB also invades the D-loop,
a double Holliday junction is formed that can be resolved as
noncrossover or as crossover (representing an exchange of
flanking sequences). Depending on the template involved in Holliday junctions, the crossover-type resolution can result in a sister
chromatid exchange (SCE), a crossing over between homologs, or
a reciprocal translocation. Alternatively, end resection may provide
an intermediate for nonconservative single-strand annealing (SSA)
between complementary single strands of the same or of different
DSBs, for NHEJ or for MMEJ (Jasin and Rothstein, 2013). Depending on the interacting double helices, various rearrangements
such as deletions and/or translocations can result from SSA, NHEJ,
Quantification of DSB Repair Products
2157
Figure 1. Scheme of DSB Repair Pathways and Their Potential Results.
Invasion (*) into an undamaged double-helix (sister chromatid, homolog or heterolog with ectopical homologous regions). cNHEJ, classical nonhomologous end-joining; Alt-NHEJ, alternative nonhomologous end-joining; CA, chromosome aberration; CO, crossing over.
and MMEJ, respectively (Figure 1). NHEJ could also precisely restore the prebreak situation, e.g., by direct ligation of compatible
ends of an endonuclease-mediated break, as postulated by Lin
et al. (2013). SSA, NHEJ, and MMEJ may join DNA break ends
throughout the cell cycle, but predominantly occur during G1,
when sister chromatids are not available (reviewed in Lieber, 2010).
Moreover, both ends of a DSB may interact separately and follow
different repair pathways, which result in various types (and combinations) of rearrangements. A nonreciprocal recombination repair
(DSB-induced conservative replication from the proximal break
end up to the terminus of the template chromatid [BIR]; Haber,
1999), as postulated for yeast (Saccharomyces cerevisiae), has not
yet been proven for higher eukaryotes (Schubert et al., 2011).
Genetic transformation represents an outcome of erroneous
DSB repair. Gene targeting (GT) on the basis of sequence homology between the transgene and the target locus is an important goal for scientific as well as for breeding purposes. So far,
GT works well in the moss Physcomitrella patens (Schaefer, 2001;
Kamisugi et al., 2006) but is rather inefficient in angiosperms (Britt
and May, 2003; Reiss, 2003). The GT efficiency can be increased
by DSBs in the target sequence (e.g., Fauser et al., 2012; Puchta
and Fauser, 2013; Qi et al., 2013). To increase GT in angiosperms
in a controlled manner, it is necessary to know the circumstances
under which DSBs are repaired along different pathways, with the
aim of supporting favorable and avoiding unfavorable conditions
as much as possible.
The different pathways and results of DSB repair have been
studied in various eukaryotic systems. Because no system was
investigated comprehensively for all potential mechanisms, we
have only insufficient knowledge about the quantitative contribution of the individual options for DSB repair. Therefore, we
aimed to estimate the proportions of different DSB repair
pathways at the molecular and chromosomal level by tracing the
fate of DSBs targeted to three b-glucuronidase reporter transgene variants (GU.US, IU.GUS, and DU.GUS; Orel et al., 2003) in
the monocotyledoneous plant barley (Hordeum vulgare). The
reporter genes are interrupted by 38 bp in GU.US and 30 bp in
DU.GUS and IU.GUS. All spacer sequences include the 18-bp
recognition site for the yeast restriction endonuclease I-SceI,
which is simultaneously expressed (Figures 2 and 3). The GUS
constructs were originally generated to test for restoration of
enzyme activity via SSA or SDSA. However, it was not clear (1)
what happened in the majority of cells, i.e., where the fate of the
locus could not be traced by GUS staining indicating a functional
gene, (2) how representative the obtained results are for other
plants, and (3) whether the repair processes are intramolecular
events or whether the sister chromatid plays a substantial role.
Therefore, we analyzed sequences of various barcoded amplicon
libraries obtained from the three tester lines to estimate the contribution of SSA, SDSA, NHEJ, and MMEJ to DSB repair in these
constructs. Additionally, we measured the relative sister chromatid exchange frequency at the target position of one of these
2158
The Plant Cell
Figure 2. Transgene Constructs and Their Locations in Barley.
(A) Schemes of recombination substrates, expected recombination products (restoration of a functional GUS gene from GU.US via SSA linked with
a deletion or from IU/DU.GUS via SDSA = gene conversion) and the I-SceI gene.
(B) Chromosomal localization of the recombination substrate loci by FISH (red arrows).
lines to get a first insight into which proportion of conversions, insertions, deletions, and reciprocal exchanges between sister
chromatids occur during DSB repair in distinct sequence contexts.
We found similarities and differences compared with previous reports for other eukaryotes and provide interpretations based on
microscopic evaluation of SCEs at target loci after DSB induction.
RESULTS
DSB Repair Reporter Assay
For quantitative analysis of distinct DSB repair pathways (NHEJ,
MMEJ, SSA, SDSA, and SCE; Figure 1) by means of a reporter
gene, lines harboring the corresponding recombination substrate
(BG190E04 with the DU.GUS substrate locus on 4HS end;
BG192E42 with IU.GUS in the middle of the long arm of 1H, both
in particular for SDSA detection; BG189E13 with a GU.US locus
terminally on 5HL, for SSA detection) and a line ectopically expressing the restriction endonuclease I-SceI (BG213E03) were
established (Figure 2). The number and chromosomal position of
the transgenic reporter gene loci were determined by fluorescence
in situ hybridization (FISH) (Figure 2).
Single-locus plants were propagated by selfing for three generations (up to T3). Based on segregation patterns in T2, T1 lines
harboring transgenic recombination substrates homozygously
were selected for crossing with a homozygous line expressing the
Quantification of DSB Repair Products
2159
monitor the deletions of ;200 to 300 bp and insertions on either
site of the DSB of ;300 bp. Larger deletions (;900 bp for the
DU.GUS and IU.GUS constructs and ;1400 bp for the GU.US
construct) or deletions combined with insertions could be detected in the libraries from longer amplicons (12A [1626 bp], 12B
[1002 bp], 11C [1118 bp], 12C [1002 bp], and 11D [1118 bp]). PCR
products from homozygous plants invariably revealed the expected sequences of original constructs. Thus, the large number
of mutated sequences within double-hemizygous plants indicates
efficient in vivo cleavage activity of I-SceI.
NHEJ Is the Dominant DSBs Repair Pathway in Barley
Figure 3. Amplicon Libraries Derived from the Three Recombination
Substrates.
The 454 amplicon libraries harboring constructs containing break sites
(I-SceI) at GU.US, DU.GUS, and IU.GUS were constructed and sequenced
in order to validate how the DSBs were repaired.
restriction endonuclease I-SceI from a barley codon adapted
transgene. The expression of I-SceI in the three resulting
double-hemizygous hybrids (DU.GUS/WT–I-SceI/WT, IU.GUS/
WT–I-SceI/WT, and GU.US/WT–I-SceI/WT) has been confirmed
via quantitative RT-PCR (Supplemental Figure 1).
Leaves of adult plants were exposed to GUS staining and
revealed in GU.US/WT–I-SceI/WT hybrid plants two regions with
the expected blue color (Supplemental Figure 2A). Within the
roots of hybrids of the cross DU.GUS/WT 3 I-SceI/WT, 1.15%,
and of GU.US/WT 3 I-SceI/WT hybrids, 1.48%, of the total area
revealed blue spots (Supplemental Figure 2B and Supplemental
Table 1). These values probably underestimate the frequency of
restoration of a functional GUS gene because blue spots of the
size of single of very few cells would hardly be detectable. Thus,
GUS staining alone would not suffice for quantitative estimation
of SDSA versus SSA events in barley.
For further analysis of recombination events in the three doublehemizygous hybrids, we designed and massively sequenced (with
the Roche 454 GS-FLX Titanium platform) two amplicon libraries
of the GU.US and three amplicon libraries each of the DU.GUS
and IU.GUS constructs, all containing the I-SceI recognition site
(Figure 3), in order to validate how the DSBs were repaired. The
short-range libraries (11E [398 bp], 12E [493 bp], and 12F [493 bp]),
amplified with primers close to the I-SceI site, allowed us to
DSB repair by NHEJ is usually accompanied by loss or gain (or
loss and gain) of nucleotides. Therefore, we evaluated the efficiency of DSB repair via NHEJ by testing for short deletions (<30
bp; often linked with classical NHEJ) and longer deletions (indicating alternative end joining; Deriano and Roth, 2013) both
accompanied by small (#3 bp) insertions at the break sites and
ligated with (MMEJ) or without using microhomology of #25 bp.
From a total of 7668 informative reads obtained from the two
libraries of the GU.US construct, only 212 reads (2.76%) contained
small modifications of 1- to 3-bp (deletions, insertions, substitutions, and deletions combined with insertions) at the I-SceI
site (Figure 4A; Supplemental Data Set 1). Due to possible
sequence errors at homopolymer sequences, the inclusion of
insertions and/or deletions of 1 to 3 bp might lead to a slight
overestimation of the true indel (insertion and deletion) frequency among this category.
Of the informative reads derived from the GU.US libraries,
55.37% (4246 reads) contain deletions within the range of 4 to
10 bp (Figure 4A; Supplemental Data Set 1), with deletions of
10 bp (3758 reads = 49.01%) representing the largest proportion.
This could be due to either an early repair event and amplification
of the resulting sequence during subsequent cell divisions or due
to independent, recurrent events that are favored for structural
reasons. Only 12 sequences (0.16% of reads) showed mutations
comprising 11 to 30 bp (Figure 4A). One large deletion of 1047 bp
appeared in 16 sequence reads (Figure 4A). Sequences indicating
restoration of a functional GUS sequence by a deletion of 652 bp
around the breakpoint via SSA were not detected. Overall,
58.50% of the GU.US sequence reads (4486 reads) appeared to
be the result of NHEJ repair accompanied by either deletions,
small insertions (1 to 3 bp), or combination of both adjacent to the
I-SceI site (Figure 4B; Supplemental Data Set 1). A spot check of
presumed NHEJ repair products of the GU.US construct by
sequencing PCR fragments also revealed small deletions at the
I-SceI site in Arabidopsis thaliana (Serra et al., 2013).
Even in the case of the DU.GUS and IU.GUS constructs for
which ectopic homologous sequences in direct or inverse orientation are expected to be used to repair the break via gene
conversion, most DSBs were repaired by NHEJ. Of the total informative sequence reads analyzed for DU.GUS (9522 reads) and
IU.GUS (7171 reads), 47.02 and 53.70%, respectively, displayed
either deletions or small insertions (1 to 3 bp) or both (Figure 4;
Supplemental Data Set 2). These data suggest that NHEJ and/or
MMEJ represent the most abundant DSB repair pathways in
barley.
2160
The Plant Cell
Figure 4. Quantitative Estimation of DSB Repair Products in Barley after Amplicon Sequencing.
(A) Proportion of mutation types among reads of the three constructs (columns) and number of repair classes C100 to M1000+ (colored lines). Both
presentations comprise classic NHEJ (C100 to M30) and, based on longer deletions (Deriano and Roth, 2013), alternative NHEJ (M100 to M1000+) with
or without the use of microhomology.
(B) Proportion of repair types among the reads detected in the three constructs. Precise SCE or uncut positions cannot be distinguished by sequencing,
but uncut sequences should be infrequent (see text). Red, immediate ligation (cNHEJ); blue, cNHEJ and aNHEJ; green (32.86 and 20.56), gene
conversion restoring the functional GUS gene via SDSA (=HR); violet, insertion via an SDSA-like mechanism in combination with NHEJ.
Repair of DSBs by Immediate Ligation
The Role of Microhomology in NHEJ
Direct ligation of break ends after cleavage by I-SceI would
restore the I-SceI recognition site. We detected 33.42% of
reads (2563 reads) from GU.US, 17.56% (1672 reads) from
DU.GUS, and 24.86% (1783 reads) from IU.GUS that displayed
the original I-SceI site (Figure 4B; Supplemental Tables 1 and 2).
Previously, transgenic barley plants showed constitutive expression of green fluorescent protein under the control of the
maize (Zea mays) UBIQUITIN-1 promoter (Furtado and Henry,
2005). Therefore, we presume I-SceI nuclease activity in all
somatic cells and cleavage of nearly all recognition sites. The
breaks in reads with the original I-SceI site would have been
repaired, restoring the site, by either immediate ligation or by
precise SCE. Distinguishing between these two types of repair
is not possible by sequence analysis but requires microscopy
analysis (see below). DSB repair by immediate ligation is
a variant of classical NHEJ (c-NHEJ) that in humans depends
on the DNA repair proteins Ku70, XRCC4, and DNA LIGASE IV
(LIG4) (Lin et al., 2013). It should occur with the same relative
frequency for all three constructs. The actual proportion of
reads that reveal the original I-SceI site ranges from 17.56%
(DU.GUS) up to 33.42% in the GU.US-derived amplicon libraries (Figure 4B), suggesting that immediate ligation (or
precise SCE) is substantially involved in repair of DSBs caused
by endonucleases in barley. Theoretically, the efficiency of that
process might vary between chromosomal positions, possibly
depending on the structural environment; however, we expect
no substantial deviations because the three different constructs
at different chromosomal positions display (in principle) similar
results (Figure 4).
Immediate and correct ligation of staggered breaks after cleavage
by restriction endonucleases is prevented if break ends become
shortened by exonucleases and complementary single-stranded
break ends for correct annealing are no longer available, a situation resembling blunt-end DSBs. Then, DSBs can still be sealed
directly without sequence homology (c-NHEJ), or single-stranded
break ends find (or generate by replication) short complementary
sequences to anneal before ligation (MMEJ; Yu and McVey,
2010). To address the role of microhomology at break ends for
end-joining repair of DSBs, we quantified microhomologous
sequences at sites of ligation among the reads of all three
constructs.
We found 25 different end-joining classes, representing single
or recurrent events, that utilized 2 to 5 bp microhomology for
end-joining of DSBs among 1125 DU.GUS construct-derived
reads from all three libraries (Supplemental Data Set 3). Of these,
58 reads were ligated using 2 bp homology observed in nine
classes. The majority of reads (875 sequences) of seven NHEJ
classes used 3 bp microhomology for end-joining, while junctions with microhomology of 4 bp were detected in 184 reads
(eight classes). In addition to 2- to 4-bp microhomology ends,
MMEJ has also been observed in one class in which eight reads
have been ligated through annealing of 5 bp microhomology.
Similar observations were made for the other constructs. Two–
to four–base pair microhomology was used for end-joining in
441 reads (15 classes) of IU.GUS and for 516 sequences (five
classes) of GU.US construct libraries (Supplemental Data Set 3).
In all these cases, single-strand resection generated short
single-stranded DNA stretches complementary to the end of the
Quantification of DSB Repair Products
opposite strand. Short-range end resections that caused deletions
of <20 bp were most frequently found for all three constructs.
However, 16 reads of the GU.US construct library (12A) revealed
a long-range end resection of 1047 bp before ligation involving 2-bp
microhomology. Interestingly, seven out of eight DU.GUS-derived
events that displayed long-range end resection (>500 bp deletion)
were ligated using 2- to 5-bp microhomology (Supplemental Data
Sets 2 and 3). Also, among IU.GUS construct reads, a single longrange end resection (639 bp deletion) was repaired by sealing break
ends at 3-bp homology.
It is important for cells to have an emergency mechanism to
quickly repair DSBs in cycling cells without extended deletions.
Our studies suggest that in many cases microhomology plays
an important role in repairing DSBs, thus avoiding the risk of cell
lethality that might be caused by long-range end resections. In
some cases, microhomology was generated by small replication-
2161
mediated insertions, leading to short duplications (Figure 5A), as
suggested by Yu and McVey (2010).
Homologous Recombination by Gene Conversion
We next determined the frequency of HR using ectopic homologous sequences DU (direct U) and IU (inverted U) as template for
repair via conversion of the mutated U harboring the DSB. This
type of repair (SDSA) should result in loss of 30 bp comprising
18 bp of I-SceI recognition site and 12-bp borders in the constructs
DU.GUS and IU.GUS using the homologous sequence on the
same chromatid (intrachromatid recombination) or on the sister
chromatid, in the latter case, with or without SCE.
Because products of gene conversion within the libraries 12E
and 12F of the DU.GUS and IU.GUS constructs (Figure 2, Table 1)
are identical to the homologous templates, we quantified the
Figure 5. Examples of Different Types of MMEJ/NHEJ in Barley.
(A) Duplicated microhomology (9 bp) by replication-mediated insertion of 3 bp (orange), exemplified by read D10 I3 of the DU.GUS construct
(Supplemental Data Set 3) (dotted regions represent 27 bp).
(B) Combination of an SDSA-like mechanism linked with NHEJ as exemplified by read D14 I12 of the GU.US construct (Supplemental Data Set 2).
2162
The Plant Cell
463-bp (template U) and 493-bp (mutated U) products of libraries 12E and 12F (Figure 2) within amplicon samples of
homozygous lines DU.GUS and IU.GUS in comparison to
those of double-hemizygous plants DU.GUS 3 I-SceI and
IU.GUS 3 I-SceI using the Agilent High Sensitivity DNA chips
and the Agilent 2100 Bioanalyzer. An increase of the 463-bp
fragment among amplicon products above that obtained with
the homozygous parental lines was considered to be a result of
gene conversion events. Thus, we estimated 2681 reads of the
DU.GUS 3 I-SceI library 12E and 413 reads of the IU.GUS 3
I-SceI library 12F to represent products of DSB repair by gene
conversion (Supplemental Data Set 2).
In total, from all corresponding libraries 3129 reads (32.86%)
of DU.GUS and 1474 reads (20.56%) of IU.GUS after DSB repair
represent HR by gene conversion (SDSA). However, we could
not distinguish whether these products resulted from gene
conversion in cis (intrachromatid recombination) or in trans (noncrossing-over between sister chromatids). Probably, the inverse
orientation of the conversion template of the IU.GUS construct,
requiring a loop formation for SDSA, is less efficiently used than
the directly oriented “U” sequence of the DU.GUS construct.
Sequence Conversion of Short Sequences around the DSB
Sites Is Linked with NHEJ
We also examined the repair products from all three constructs
containing sequences inserted into the breakpoint that failed to
restore the functional gene by homologous recombination via
SDSA. The aim was to find out the possible origin and mode of
insertion of 4 bp or longer.
We found 244 sequence reads (2.56%; Figure 4B) of the
DU.GUS construct displaying insertions of 4 to 61 bp. These
reads belonged to at least 25 SDSA repair classes, of which six
classes (65 reads) revealed captured sequences from the left
side (upstream) of I-SceI–induced DSBs, nine classes (146
reads) from the right side (downstream) of the break sites, and
10 classes (33 reads) that had captured a sequence of unknown
origin (Supplemental Data Set 2).
Of the IU.GUS construct, 44 sequence reads belonging to at
least eight repair classes captured sequences of unknown origin
(4 to 22 bp). Two classes were detected by a total of 15 reads
with 4 and 6 bp captured from downstream of the break sites.
Two classes (two reads each) displayed a deletion of 427 and
48 bp starting from 65 and 381 bp upstream of the break site,
respectively (Supplemental Data Set 2). These two classes most
likely reflect transient deletions caused by left-sided endresection, which became newly resynthesized from the right
side (according to the sister chromatid) up to positions 65 and
381, respectively, from where the actual deletion, linked with
NHEJ, started.
Among GU.US construct-derived libraries, 619 sequence
reads (8.07%) representing five repair classes showed insertions
of 4 to 15 bp, copied from 5 to 23 bp downstream of the break
site (Figure 4B; Supplemental Data Set 1).
We were particularly interested in finding out how frequently
known repetitive sequences, such as retrotransposons, were
captured into the DSBs within the large barley genome. However, so far we did not detect any sequence that had an insertion
larger than 20 bp homologous to a known retrotransposon
sequence.
Our results demonstrate that in addition to gene conversion
restoring a functional GUS sequence in IU/DU.GUS, a majority
of the SDSA-like repair classes for all three constructs revealed
short conversion tracts, with insertions preferentially originating
from the right side of the breakpoint. Likely these conversion
events were formed via a double-strand gap at the I-SceI break
site, yielding deletions on one or both break ends (Figure 5B;
Supplemental Tables 1 and 2). The shortened 39 end(s) of the
cleaved site then apparently formed a D-loop with the donor
double helix in cis or with that of the sister chromatid (mostly
downstream of the break) to replicate the insert sequence before
ligation via N.HEJ. This scenario results in at least a partial duplication of the inserted sequence (Figure 5B) and can explain
most of the insertions $4 bp.
SCE but Not Chromosome Interchanges Are Frequently
Linked with Repair of Single DSBs
All pathways leading to repair a single targeted DSBs as described
above could represent either intramolecular reactions involving
only the DNA double helix of the damaged chromatid or, after
replication, intermolecular reactions involving the identical double
helix of the sister chromatid. To get an idea about the order of
magnitude regarding the frequency of involvement of the sister
chromatid in DSB repair processes, we comparatively investigated
the SCE frequency in young primary root meristems of double
Table 1. The Amplicon Libraries That Have Been Sequenced
Construct
GU.US
DU.GUS
IU.GUS
a
Library ID
(Bar Coding
Sequence)
Amplicon
Size (bp)
Primer Sequences (F)
Primer Sequences (R)
De-Bar-Coded
Reads
Informative
Reads
11E (GTGATC)
12A (CTACTC)
12E (CATCTC)
12B (TAGCTC)
11C (GCGATC)
12F (TCTCTC)
12C (ACGCTC)
11D (TCGATC)
398
1,626
493
1,002
1,118
493
1,002
1,118
GU4F: GGAATGGTGATTACCGACGAAAAC
DU1F: TCCCAATTCGATCTACATCCGTCCTG
GU2F: ACTATGCCGGAATCCATCGCAG
GU1F: AATTCGATCTACATCCGTCCTG
GU2F: ACTATGCCGGAATCCATCGCAG
GU2F: ACTATGCCGGAATCCATCGCAG
GU1F: AATTCGATCTACATCCGTCCTG
GU2F: ACTATGCCGGAATCCATCGCAG
GU4R: GAATATCTGCATCGGCGAACTGATC
DU1R: CCAGTCCATTAATGCGTGGTCGTG
DU1R: CCAGTCCATTAATGCGTGGTCGTG
DU1R: CCAGTCCATTAATGCGTGGTCGTG
GU3R: GCCATGCACACTGATACTCTTCAC
DU1R: CCAGTCCATTAATGCGTGGTCGTG
DU1R: CCAGTCCATTAATGCGTGGTCGTG
GU3R: GCCATGCACACTGATACTCTTCAC
29,768
18,867a
62,239
5,862
1,832
33,782
15,988
3,205
7,652
16a
8,023
1,380
119
3,136
3,736
299
Almost all reads of the GU.US library 12A are noninformative because they do not overlap the break position (all 16 informative reads of 579 bp
revealed the same 1047-bp deletion, much larger than the 625 bp that would have restored a functional GUS gene), suggesting that large deletions may
occur at random and are rare among the repair products of the GU.US construct. Apparently, the efficiency of libraries strongly decreases with the
length of the amplicon.
Quantification of DSB Repair Products
hemizygous GU.US 3 I-SceI plants versus the homozygous
GU.US line after incorporation of the base analog ethylyldeoxyuridine
(EdU) during the last but one replication before arresting metaphases. In particular, the frequency of SCEs at the target position
for DSBs at the very end of the long arm of chromosome 5H was
monitored. In the control line (GU.US), the average SCE frequency
per chromosome 5H (1.8) is the same as for other chromosomes,
while in the I-SceI–expressing plants, it shows a few more SCEs
than other chromosomes (2.6 versus 2.1). Twenty-two out of 54
chromosomes 5H (of which one half [=27] should harbor the
target transgene in double-hemizygous plants) displayed an SCE
at the very end of 5HL after DSB induction, while only 4 out of 32
5H chromosomes of the homozygous GU.US control did so. This
indicates that 81.5% of 5HL ends, harboring the GU.US construct
in double hemizygous plants, showed an SCE at the target locus
after DSB induction (Figure 6). The background frequency at the
end of 5HL without DSB induction was 12.5% (P < 0.001, Fisher’s
test). Previous studies have shown that SCEs in barley (Schubert
et al., 1980), as in most other tested species (Schubert and Rieger,
1981), show a length proportional distribution along chromosomes
and no clustering toward the chromosome ends. Because gene
conversion using the sister helix as the template is not necessarily
linked with a reciprocal exchange, 81.5% of SCEs most likely even
underestimates the use of the sister helix as template for DSB repair. Thus, our results indicate that within the GU.US construct,
most DSBs are repaired by an interchromatid process using the
sister helix as a template and are linked with reciprocal exchange.
In contrast to the presence of two or more DSBs per nucleus,
a single DSB per nucleus is not assumed to significantly increase the frequency of reciprocal chromosome translocations
(Puchta, 1999; Richardson and Jasin, 2000; Pacher et al., 2007).
We observed chromatin bridges in 1.6% of 126 anaphases in
GU.US control meristems and 3.8% of 105 anaphases of the
I-SceI expressing plants (not significant at a 99% confidence
interval). This result indicates that end resection, deleting the
right and/or the left part of the reporter construct, and subsequent
long-range conversion (as found in field bean; see Schubert et al.,
2011) using the homologous chromosome as template is,
when it occurs, rarely linked with reciprocal interchromosomal
exchange.
DISCUSSION
To date, the outcome of DSB repair in angiosperms has been
studied mainly in Arabidopsis (and tobacco [Nicotiana tabacum])
by means of negative and positive selection systems and targeted endonuclease-mediated DSB induction. The constructs
used in these experiments were designed such that the formation of a functional gene, replacing the variant interrupted by an
I-SceI recognition site by deletion of the interrupting sequence
via SSA, or by conversion from an uninterrupted ectopic gene
fragment, is detectable by restoration of GUS activity. Alternatively, repair-mediated loss of a transgenic marker gene (negative selection) has been used to interpret repair processes that
did not restore the original sequence context. Orel et al. (2003)
showed that in Arabidopsis, SSA restored a functional gene
about 5 times more efficiently than SDSA and that the template
used for gene conversion by SDSA was independent of its
2163
Figure 6. SCEs after Targeted DSB Induction.
Complete metaphase of barley displaying 44 SCEs with one chromosome, 5H, showing an SCE in distal position of its long arm (arrow) where
the GU.US construct is located (Fig. 1B). The inset shows other 5H
chromosomes, each with a terminal SCE (arrows).
orientation toward the target site (DU.GUS versus IU.GUS). Blue
spots indicating GUS activity appeared 1 to 2 orders of magnitude above the spontaneous frequency after transient DSB
induction. The interpretation of repair pathways was supported
by DNA gel blot hybridization, PCR, and, in individual cases, by
sequencing of PCR products.
Recombinant calli of tobacco revealed a frequent use of
microhomology for joining of I-SceI–mediated DSBs; however,
only 1 out of ;10,000 DSBs was repaired via SDSA by means of
allelic or ectopic homology (Gisler et al., 2002), while in one-third
of cases, marker gene loss was due to intrachromosomal HR via
SSA (Siebert and Puchta, 2002). Furthermore, a comparison of
recombinant calli of Arabidopsis and of tobacco using a negative selection marker showed that 40% of repair-mediated deletions in tobacco were associated with insertions, while in the
20-fold smaller Arabidopsis genome, insertions were barely
detectable, and most of the repair-mediated deletions were
much larger than in tobacco (within the scale of detectability of
up to 2300 bp). Several NHEJ pathways were observed by
tracing the fate of linearized plasmids in tobacco protoplasts
(Gorbunova and Levy, 1997).
Here, we describe similarities as well as differences in repair of
targeted DSBs compared with previous reports for other eukaryotes. Our sequencing approach allowed a higher resolution
compared with what could be obtained by GUS staining. We
provide interpretations of DSB repair based on evaluation of
SCEs at target loci after DSB induction and regarding links of
SDSA-like mechanisms with NHEJ. So far, we could only report
on target sequences at a single time point in young barley plants
2164
The Plant Cell
(10 to 15 cm shoots) constitutively expressing the endonuclease
I-SceI. Therefore, we cannot unambiguously determine to what
degree identical reads represent a single early or a recurrent
event. We call groups of sequence-identical reads “classes.”
The high frequency of reads that indicate gene conversion
using the homologous template sequence of the DU.GUS and
the IU.GUS construct more likely goes back to several independent
rather than to a single very early event.
Sequence reads identical to the original sequence could represent either uncut positions or restoration of the original sequence
by immediate intramolecular ligation or by precise SCE. We consider the majority of these fractions (comprising ;17 to 33% of
reads) to represent multiple individual ligation events rather than
uncut target positions. We argue that the GU.US construct displaying 33.42% of reads with the original sequence yielded an
SCE at 81.5% of the target loci. Keeping additionally in mind that
involvement of the sister chromatid in DSB repair in the case of
gene conversion is not necessarily accompanied by an exchange;
this observation excludes the possibility that one-third of reads
represent uncut positions. Rather, the largest proportion of these
reads might represent individual immediate ligation events within
one chromatid or between sister chromatids (SCE).
Similar to what is found in Chinese hamster cells (Johnson and
Jasin, 2000), for all three constructs, the largest proportion of
reads (47 to 58%) revealed NHEJ with deletions and/or small (1 to
3 bp) insertions. In 45 out of 287 NHEJ repair classes, microhomology of 2 to 5 bp was used for end-joining. Also, duplicative
microhomology due to re-replication of short sequences (Yu and
McVey, 2010) was found to be associated with joining of break
ends (Figure 5A). Although all three constructs showed a deletion
bias (on average 6.5 classes with a net deletion per one repair
class with a net insertion), within the frame of resolution of 900 to
1400 bp, respectively, by far the most deletions were smaller than
100 bp, and thus clearly below the size of deletions reported for
Arabidopsis calluses (Kirik et al., 2000). The lack of reads with
a 652-bp deletion within the GU.US library 12A that would have
resulted in a functional GUS gene by SSA (which is 5 times more
efficient than functional GUS formation via SDSA in Arabidopsis;
Orel et al. 2003) also supports the infrequent occurrence of large
deletions during DSB repair in barley. Thus, our results are concordant with the hypothesis that very small genomes are the result of a stronger bias toward long deletions during DSB repair
(Kirik et al., 2000; Puchta, 2005). Possibly, the pronounced use of
microhomology for end-joining prevents a more frequent appearance of large deletions in barley.
Immediate ligation has previously been quantified for human
cells (Lin et al., 2013). These authors claim that nearly all I-SceI–
mediated DSBs are directly ligated. This is much more than we
estimated for the target constructs in barley (;17 to 33% of
reads); however, their data are based on a small deletion (30 bp)
between two closely adjacent recognition sites.
In this study, we observed insertions of 4 to 61 bp linked with
NHEJ, which is usually not detectable by positive selection for
restored functional reporter genes. Of 42 such repair models,
eight revealed sequences from upstream of the break, 16 from
downstream of the break, and 18 classes included insertions
from unknown genomic positions. These (mis)repair products
suggest short-range conversion by a mechanism similar to SDSA
(Figure 5B) that creates partial duplication before NHEJ. Alternatively, extrachromosomal DNA may have been inserted.
However, such a process would depend on the less probable
availability of the corresponding oligonucleotides.
Most remarkable is the high coincidence of SCEs with the
GU.US target locus positioned at the end of the long arm of chromosome 5H. This observation suggests that, in contrast to the
presumed predominant intrachromatid repair in Arabidopsis (Orel
et al., 2003) and in tobacco (Siebert and Puchta, 2002), barley
uses the sister chromatid in the majority of cases for repair of
staggered DSBs, apparently independent of whether immediate
ligation, SDSA, NHEJ, or MMHJ is followed to seal the break. A
few hints that SCE could be a major pathway of DSB repair came
from the observation of increased SCE frequency after treatment
of mammalian cells with restriction endonucleases (Natarajan
et al., 1985) and from Rad51 (recombinase)-dependent equal SCE
in broken centric monoplasmids of yeast (González-Barrera et al.,
2003). In contrast, in cultivated Chinese hamster cells, “the sister
chromatid acts as a repair template in a substantial proportion of
DSB repair events,” but “the outcome.is primarily gene conversion unassociated with reciprocal exchange” (Johnson and Jasin,
2000). Furthermore, the frequent involvement of SCE in DSB repair
indicates that S and G2 are the preferred cell cycle phases to
perform DSB repair in barley. Perhaps the I-SceI site is preferentially cut during replication in S or in (early) G2, when chromatin is
easily accessible. A minor fraction of DSBs induced and repaired
during G1 should represent either intrastrand repair or long-range
conversion with the homologous chromosome as a template, deleting the reporter construct, at least from the right or the left of the
break position. Such a long range SDSA using the homolog is very
rare in tobacco (Gisler et al., 2002) and would not be detectable
among our amplicons. If it occurs, it should usually not be linked
with a reciprocal exchange because we observed no significant
increase in anaphase bridges. Postponing repair of G1 DSBs until
S/G2, as an alternative explanation, does not seem plausible.
The absence of EdU labeling at both sister chromatids distal
to the I-SceI recognition site would be indicative of classical BIR.
Because we did not find such a labeling pattern, BIR is not
frequently involved in DSB repair in barley.
Based on these results, it will be of interest to test in future
experiments (1) whether SCE is involved independently of the
genomic target locus position, (2) whether other developmental
stages reveal similar proportions of various DSB repair pathways
in barley, (3) whether blunt end and complementary end repair show
the same efficiency and accuracy (as they do with plasmids in human cells, where DSB repair is only impaired at non-complementary
break ends; Smith et al., 2001), (4) what results can be obtained
using the sequencing approach in Arabidopsis and (5) which impact
various repair gene mutants display on the contribution of distinct
DSB repair pathways, and (6) what is the outcome of more than one
I-SceI–mediated DSB per nucleus.
METHODS
Plant Materials and Growth Conditions
To estimate the proportions of the various pathways and end-points of
DSB repair, three constructs (GU.US, DU.GUS, and IU.GUS) with an
Quantification of DSB Repair Products
I-SceI restriction site in the “U” sequence and a suitable repair template
for SDSA or SSA were generated as target genes (Figure 2). In detail, the
vector pUbi-ABM (DNA Cloning Service) was mutagenized to remove the
original Acc65I site. A new SfiI-Acc65I site was introduced in front of, and
the 59 end of the GUS gene behind the maize (Zea mays) UBIQUITIN-1
promoter. A 1.9-kb BamHI/HindIII GUS fragment of p35S_DU-GUS
(GenBank: JX475905.1) and a 2.5-kb BamHI/HindIII GU.US fragment of
p35S_GU.US (GenBank: JX475904.1) were inserted downstream of the
UBIQUITIN-1 promoter of the mutagenized vector pUbi-ABMDAcc65I,
generating two plasmids, pUbi_GUS and pUbi_GU.US, respectively.
Subsequently, upstream of the UBIQUITIN-1 promoter of the plasmid
pUbi_GUS, a 1.05-kb Acc65I/Acc65I U fragment was inserted from the
p35S_DU-GUS to obtain pUbi_DU.GUS and pUbi_IU.GUS, respectively,
depending on the orientation of the U fragment. The three cassettes
pUbi_GU.US, pUbi_DU.GUS, and pUbi_IU.GUS were transferred as SfiI/
HindIII fragments into the respective restriction sites of the binary vector
p6U (DNA Cloning Service), resulting in the three reporter constructs
GU.US, DU.GUS, and IU.GUS. In addition, an I-SceI expression cassette
was constructed by introducing a barley codon–optimized coding sequence of I-SceI as a BamHI/SalI fragment into the vector pUbi-ABM
between the UBIQUITIN-1 promoter and the terminator of the Agrobacterium tumefaciens NOPALINE SYNTHASE gene. The resulting cassette pUbi_I-SceI was then transferred into the binary vector p7U (DNA
Cloning Service) using SfiI and HindIII restriction sites. Then, all three reporter constructs and the I-SceI construct were stably introduced into
barley (Hordeum vulgare cv ‘Golden Promise’) via Agrobacterium-mediated
gene transfer to immature embryos as described previously (Hensel et al.,
2009). Plants were regenerated on selective media, and the presence of the
transgene was confirmed by PCR. The number of inserted T-DNA copies
was determined by DNA gel blot hybridization and the chromosomal
position of the transgene by FISH as described (Ma et al., 2010).
The three homozygous lines (BG190E04, BG192E42, BG189E13)
were crossed with a homozygous line ectopically expressing a codonoptimized gene for restriction endonuclease I-SceI (BG213E03). Barley
plants with the genetic background of cv ’Golden Promise’ used in this
study were grown in soil at 22°C under long day conditions (16 h light/8 h
dark).
Detection of I-SceI and GUS Gene Activity
To confirm transcription of the I-SceI gene, total RNA from homozygous
and double-hemizygous plants was extracted using The RNeasy plant
mini kit (Qiagen). One microgram of total RNA was used for cDNA synthesis using a first-strand cDNA synthesis kit (Thermo Science). Triplicate
quantitative RT-PCR assays using SYBR green were done with an IQ5
cycler (Bio-Rad) with 5 mL 1:10 diluted cDNA to amplify ADP-ribosylation
factor 1 (primers: 59-GCTCCACAGGATGCTGAATG-39 and 59-ATGCCCTCGTACAACCCTTC-39) and I-SceI (primers: 59-AAGAACGCGTCAATCACCTG-39 and 59-ATTTGCCTCCGTCATCCATG-39) sequences. The
following program was used: initial denaturation, 10 min at 95°C; then
40 cycles with 10 s denaturation at 95°C, 15 s annealing at 60°C, and
20-s elongation at 72°C.
Pieces from leaves of adult plants of the hybrids to be tested were covered
with substrate (100 mM sodium phosphate buffer, pH 7.0, containing 0.1%
Triton X-100, 10 mM EDTA, 1 mM X-gluc, and 1.4 mM potassium ferricyanide), vacuum infiltrated for 1 min (ILMVAC, Laboratory Vacuum System,
LVS 301 Zp), and incubated overnight at 37°C. Then, chlorophyll was extracted by several washes in 70% ethanol at 37°C. Plants were examined on
a Leica MZFLIII stereomicroscope (Leica Microsystems). For the seedling
assay, three grains of each hybrid were germinated on wet filter paper at 24°C
in the dark for 5 d and treated and examined as above. As a negative control,
cv Golden Promise, and as positive control, transgenic barley plants constitutively expressing the GUS gene under the control of the maize UBIQUITIN-1 promoter, were used.
2165
454 Amplicon Library Preparation
In total, eight 454 amplicon libraries from three double-hemizygous hybrids
expressing the I-SceI restriction endonuclease and harboring the break site
(I-SceI) at GU.US, DU.GUS, and IU.GUS sequences, respectively, were
constructed (Figure 3). Library preparation was done as described (Meyer
et al., 2008).
DNA samples were isolated using a DNeasy Plant Mini Kit (Qiagen)
from leaves of plants 10 d after sowing. PCR using a BioMix kit (Bioline)
was performed from 50 ng genomic DNA with primers shown in
Supplemental Data Set 1 (20 mL/reaction). The following program was
used for amplification: 3 min denaturation at 94°C, 23 cycles of 45 s at
94°C, 30 s at 59°C, and 1 min at 72°C, and final extension of 5 min at 72°C.
Then, each amplicon sample was blunt-end repaired. After that, specific
bar-coding adapters (Table 1) were ligated to both ends of the molecules
to distinguish the individual libraries. Nicks resulting from the adaptor
ligation were filled by Bst polymerase (NEB) before the eight barcoded
samples were pooled in equimolar ratios. A single-stranded 454 sequencing library was prepared for the pooled sample.
Conversion frequency in DU.GUS 12E and IU.GUS 12F libraries was
estimated via quantification of 463- and 493-bp PCR products with
capillary gel electrophoresis of the Agilent 2100 Bioanalyzer system using
the Agilent High Sensitivity DNA kit. The ratios of molar concentrations
(pM) between 463- and 493-bp fragments were compared between the
homozygous lines and the double-hemizygous plants expressing the
I-SceI enzyme and harboring the break site (I-SceI) at DU.GUS and
IU.GUS. In DU.GUS lines, these ratios were 1.1 in control sample and 2.15
in 12E sample. Therefore, ;33.4% (2681 reads) of 8023 informative reads
in 12E were considered to be products of gene conversion (Supplemental
Data Set 2). These ratios were 2.74 in control sample and 3.30 in IU.GUS
12F sample. Thus, we calculated that ;13.2% (413 reads) resulted from
gene conversion in 12F (Supplemental Data Set 2).
Sequence Analysis
Each library was sequenced using Roche’s 454 technology (GS FLX
Titanium XL+) to obtain between 1832 and 98,911 single-end reads for
each amplicon library (Supplemental Data Set 1). After removing the bar
code and trimming low-quality bases using default quality settings of CLC
Genomics Workbench (version 5), the remaining sequence reads of
>200 bp were used for further analysis. The high-quality sequences from
the same sequencing directions of each amplicon library were clustered
using USEARCH version 6.0.307_win32 with UPARSE pipeline (Edgar,
2013). The three main clustering and filtering steps included (1) dereplicating (removal) of shorter reads of otherwise identical sequences to
reduce the size of the data set and the running time of clustering analysis,
(2) denoising sequences by forming clusters of 99% identity and selecting
the most abundant sequence as representative of the cluster, and (3)
filtering out chimeric reads and reads with presumable sequencing errors
that appear only once. For each amplicon, all denoised and high-confident
sequences (informative reads) from both sequencing directions were
aligned against the parental sequence constructs as a reference using
ClustalW in order to curate manually homopolymer sequencing errors
and to detect sequence polymorphisms indicative of distinct repair
mechanisms.
SCE and Mitotic Chromatin Bridges
For measuring the relative SCE frequency at the target locus compared
with the overall SCE frequency, 1- to 2-cm-long roots of young seedlings
of double-hemizygous tester plants harboring the GU.US construct close
to the long arm end of satellite chromosome 5H and expressing the I-SceI
endonuclease were exposed for 17 h (one replication cycle) to 20 mM of
the base analog EdU, followed by 19 h (second replication cycle) in
2166
The Plant Cell
Hoagland solution (Merck) and 2.5 h in 0.05% colchicine for metaphase
enrichment before fixation in ethanol:acetic acid (3:1) and squashing of root
tips. After this treatment, only one DNA strand of the two sister helices of
a metaphase chromosome should contain EdU, thus enabling visualization of
SCEs by detection of incorporated EdU. For detection of differential incorporation of EdU, the click iT Imaging Kit (Invitrogen) was applied according
to the manufacturer’s instruction. Microscopy evaluation of SCEs was performed with an epifluorescence microscope (Zeiss Axiophot) using a 3100/
1.45 Zeiss alpha-plan-fluar objective and a Sony (DXC-950P) camera. For the
control, SCEs were counted from homozygous GU.US plants (without I-SceI
expression).
Untreated fixed and squashed root tip meristems were inspected after
49,6-diamidino-2-phenylindole staining (2 mg/mL in antifade Vectashield;
Vector Laboratories) for the frequency of bridges among more than 100
ana-/telophases in double-hemizygous GU.US/WT–I-SceI/WT plants
versus homozygous GU.US plants.
Accession Numbers
Sequence data of original three reporter constructs and the I-SceI expression
plasmid were deposited in the GenBank library under the following accession
numbers: KJ817199 to KJ817202. The barley ADP-ribosylation factor 1
sequence (AJ508228) was used as the reference in quantitative RT-PCR.
Supplemental Data
The following materials are available in the online version of this article.
Supplemental Figure 1. Real-Time RT-PCR Analysis of I-SceI Gene
Expression in Leaves of the Double-Hemizygous GU.US/WT–I-SceI/
WT, DU.GUS/WT–I-SceI/WT, DU.GUS/WT–I-SceI/WT Plants.
Supplemental Figure 2. GUS Assay of Leaves of Adult Plants and
Seedlings of Hybrids and Controls.
Supplemental Table 1. GUS Assay of Hybrid and Control Plant
Seedlings.
Supplemental Data Set 1. Repair Classes Detected in GU.US
Construct-Derived Libraries.
Supplemental Data Set 2. Repair Classes Detected in DU.GUS and
IU.GUS Construct-Derived Libraries.
Supplemental Data Set 3. Repair Classes That Utilized Microhomology for End-Joining of DSBs among Reads of DU.GUS, IU.GUS, and
GU.US Construct-Derived Libraries.
ACKNOWLEDGMENTS
We thank Holger Puchta (Karlsruhe University), Bernd Reiss (MPIPZ
Cologne), and Jörg Fuchs and Florian M. Mette, (IPK Gatersleben) for
reading the article and for critical discussions. We also thank Armin
Meister (IPK) for help with statistics, Veit Schubert (IPK) for providing SIM
images of chromosomes with SCEs, and Andrea Kunze, Sibylle Freist,
Cornelia Marthe, and Sandra Wolf for excellent technical assistance. This
study is supported by the German Research Foundation (SCH 951/17-1).
I.S. was partially supported by the European Social Fund (CZ.1.07/
2.3.00/20.0189). The generation of transgenic lines was partly supported
by GABI-PRECISE (BMBF) and by the IPK. The authors declare no
conflict of interest.
AUTHOR CONTRIBUTIONS
G.T.H.V. and I.S. designed research. G.T.H.V., H.X.C., K.W., G.H., F.R.B.,
J.K., and I.S. performed research. G.T.H.V., H.X.C., K.W., and I.S.
analyzed data. G.T.H.V. and I.S. wrote the article. All authors read and
approved the article.
Received April 14, 2014; revised April 24, 2014; accepted May 7, 2014;
published May 29, 2014.
REFERENCES
Britt, A.B., and May, G.D. (2003). Re-engineering plant gene targeting.
Trends Plant Sci. 8: 90–95.
Deriano, L., and Roth, D.B. (2013). Modernizing the nonhomologous
end-joining repertoire: alternative and classical NHEJ share the stage.
Annu. Rev. Genet. 47: 433–455.
Edgar, R.C. (2013). UPARSE: highly accurate OTU sequences from
microbial amplicon reads. Nat. Methods 10: 996–998.
Fauser, F., Roth, N., Pacher, M., Ilg, G., Sánchez-Fernández, R.,
Biesgen, C., and Puchta, H. (2012). In planta gene targeting. Proc.
Natl. Acad. Sci. USA 109: 7535–7540.
Furtado, A., and Henry, R.J. (2005). The wheat Em promoter drives
reporter gene expression in embryo and aleurone tissue of transgenic
barley and rice. Plant Biotechnol. J. 3: 421–434.
Gisler, B., Salomon, S., and Puchta, H. (2002). The role of doublestrand break-induced allelic homologous recombination in somatic
plant cells. Plant J. 32: 277–284.
González-Barrera, S., Cortés-Ledesma, F., Wellinger, R.E., and Aguilera,
A. (2003). Equal sister chromatid exchange is a major mechanism of
double-strand break repair in yeast. Mol. Cell 11: 1661–1671.
Gorbunova, V., and Levy, A.A. (1997). Non-homologous DNA end
joining in plant cells is associated with deletions and filler DNA
insertions. Nucleic Acids Res. 25: 4650–4657.
Haber, J.E. (1999). DNA recombination: the replication connection.
Trends Biochem. Sci. 24: 271–275.
Hensel, G., Kastner, C., Oleszczuk, S., Riechen, J., and Kumlehn,
J. (2009). Agrobacterium-mediated gene transfer to cereal crop
plants: current protocols for barley, wheat, triticale, and maize. Int.
J. Plant Genomics 2009: 835608.
Heyer, W.D., Ehmsen, K.T., and Liu, J. (2010). Regulation of homologous
recombination in eukaryotes. Annu. Rev. Genet. 44: 113–139.
Jasin, M., and Rothstein, R. (2013). Repair of strand breaks by homologous
recombination. Cold Spring Harb. Perspect. Biol. 5: a012740.
Johnson, R.D., and Jasin, M. (2000). Sister chromatid gene
conversion is a prominent double-strand break repair pathway in
mammalian cells. EMBO J. 19: 3398–3407.
Kamisugi, Y., Schlink, K., Rensing, S.A., Schween, G., von
Stackelberg, M., Cuming, A.C., Reski, R., and Cove, D.J. (2006).
The mechanism of gene targeting in Physcomitrella patens: homologous
recombination, concatenation and multiple integration. Nucleic Acids
Res. 34: 6205–6214.
Kirik, A., Salomon, S., and Puchta, H. (2000). Species-specific
double-strand break repair and genome evolution in plants. EMBO
J. 19: 5562–5566.
Lieber, M.R. (2010). The mechanism of double-strand DNA break
repair by the nonhomologous DNA end-joining pathway. Annu. Rev.
Biochem. 79: 181–211.
Lin, W.Y., Wilson, J.H., and Lin, Y. (2013). Repair of chromosomal
double-strand breaks by precise ligation in human cells. DNA
Repair (Amst.) 12: 480–487.
Ma, L., Vu, G.T., Schubert, V., Watanabe, K., Stein, N., Houben, A., and
Schubert, I. (2010). Synteny between Brachypodium distachyon and
Hordeum vulgare as revealed by FISH. Chromosome Res. 18: 841–850.
Quantification of DSB Repair Products
McVey, M., and Lee, S.E. (2008). MMEJ repair of double-strand
breaks (director’s cut): deleted sequences and alternative endings.
Trends Genet. 24: 529–538.
Meyer, M., Stenzel, U., and Hofreiter, M. (2008). Parallel tagged
sequencing on the 454 platform. Nat. Protoc. 3: 267–278.
Natarajan, A.T., Mullenders, L.H., Meijers, M., and Mukherjee, U.
(1985). Induction of sister-chromatid exchanges by restriction
endonucleases. Mutat. Res. 144: 33–39.
Orel, N., Kyryk, A., and Puchta, H. (2003). Different pathways of
homologous recombination are used for the repair of double-strand
breaks within tandemly arranged sequences in the plant genome.
Plant J. 35: 604–612.
Pacher, M., Schmidt-Puchta, W., and Puchta, H. (2007). Two
unlinked double-strand breaks can induce reciprocal exchanges in
plant genomes via homologous recombination and nonhomologous
end joining. Genetics 175: 21–29.
Puchta, H. (1999). Double-strand break-induced recombination between
ectopic homologous sequences in somatic plant cells. Genetics 152:
1173–1181.
Puchta, H. (2005). The repair of double-strand breaks in plants: mechanisms
and consequences for genome evolution. J. Exp. Bot. 56: 1–14.
Puchta, H., and Fauser, F. (2013). Gene targeting in plants: 25 years
later. Int. J. Dev. Biol. 57: 629–637.
Qi, Y., Zhang, Y., Zhang, F., Baller, J.A., Cleland, S.C., Ryu, Y.,
Starker, C.G., and Voytas, D.F. (2013). Increasing frequencies of
site-specific mutagenesis and gene targeting in Arabidopsis by
manipulating DNA repair pathways. Genome Res. 23: 547–554.
Reiss, B. (2003). Homologous recombination and gene targeting in
plant cells. Int. Rev. Cytol. 228: 85–139.
Richardson, C., and Jasin, M. (2000). Frequent chromosomal
translocations induced by DNA double-strand breaks. Nature 405:
697–700.
Rothkamm, K., Krüger, I., Thompson, L.H., and Löbrich, M. (2003).
Pathways of DNA double-strand break repair during the mammalian
cell cycle. Mol. Cell. Biol. 23: 5706–5715.
2167
Schaefer, D.G. (2001). Gene targeting in Physcomitrella patens. Curr.
Opin. Plant Biol. 4: 143–150.
Schubert, I., and Rieger, R. (1981). Sister chromatid exchanges and
heterochromatin. Hum. Genet. 57: 119–130.
Schubert, I., Künzel, G., Bretschneider, H., Rieger, R., and Nicoloff, H.
(1980). Sister chromatid exchanges in barley. Theor. Appl. Genet. 56: 1–4.
Schubert, I., Pecinka, A., Meister, A., Schubert, V., Klatte, M., and
Jovtchev, G. (2004). DNA damage processing and aberration
formation in plants. Cytogenet. Genome Res. 104: 104–108.
Schubert, I., Schubert, V., and Fuchs, J. (2011). No evidence for
“break-induced replication” in a higher plant - but break-induced
conversion may occur. Front. Plant Sci. 2: 8.
Serra, H., Da Ines, O., Degroote, F., Gallego, M.E., and White, C.I.
(2013). Roles of XRCC2, RAD51B and RAD51D in RAD51-independent
SSA recombination. PLoS Genet. 9: e1003971.
Shrivastav, M., De Haro, L.P., and Nickoloff, J.A. (2008). Regulation
of DNA double-strand break repair pathway choice. Cell Res. 18:
134–147.
Siebert, R., and Puchta, H. (2002). Efficient repair of genomic doublestrand breaks by homologous recombination between directly
repeated sequences in the plant genome. Plant Cell 14: 1121–1131.
Smith, J., Baldeyron, C., De Oliveira, I., Sala-Trepat, M., and
Papadopoulo, D. (2001). The influence of DNA double-strand break
structure on end-joining in human cells. Nucleic Acids Res. 29:
4783–4792.
Symington, L.S., and Gautier, J. (2011). Double-strand break end
resection and repair pathway choice. Annu. Rev. Genet. 45: 247–271.
Watanabe, K., Pacher, M., Dukowic, S., Schubert, V., Puchta, H.,
and Schubert, I. (2009). The STRUCTURAL MAINTENANCE OF
CHROMOSOMES 5/6 complex promotes sister chromatid alignment
and homologous recombination after DNA damage in Arabidopsis
thaliana. Plant Cell 21: 2688–2699.
Yu, A.M., and McVey, M. (2010). Synthesis-dependent microhomologymediated end joining accounts for multiple types of repair junctions.
Nucleic Acids Res. 38: 5706–5717.
Repair of Site-Specific DNA Double-Strand Breaks in Barley Occurs via Diverse Pathways
Primarily Involving the Sister Chromatid
Giang T.H. Vu, Hieu X. Cao, Koichi Watanabe, Goetz Hensel, Frank R. Blattner, Jochen Kumlehn and
Ingo Schubert
Plant Cell 2014;26;2156-2167; originally published online May 29, 2014;
DOI 10.1105/tpc.114.126607
This information is current as of August 3, 2017
Supplemental Data
/content/suppl/2014/05/08/tpc.114.126607.DC1.html
References
This article cites 42 articles, 10 of which can be accessed free at:
/content/26/5/2156.full.html#ref-list-1
Permissions
https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X
eTOCs
Sign up for eTOCs at:
http://www.plantcell.org/cgi/alerts/ctmain
CiteTrack Alerts
Sign up for CiteTrack Alerts at:
http://www.plantcell.org/cgi/alerts/ctmain
Subscription Information
Subscription Information for The Plant Cell and Plant Physiology is available at:
http://www.aspb.org/publications/subscriptions.cfm
© American Society of Plant Biologists
ADVANCING THE SCIENCE OF PLANT BIOLOGY