Journal of Experimental Botany, Vol. 65, No. 10, pp. 2747–2756, 2014
doi:10.1093/jxb/ert470 Advance Access publication 21 January, 2014
Review paper
A maize root tip system to study DNA replication
programmes in somatic and endocycling nuclei during plant
development
Hank W. Bass1, Emily E. Wear2, Tae-Jin Lee2, Gregg G. Hoffman1, Hardeep K. Gumber1, George C. Allen3,
William F. Thompson2 and Linda Hanley-Bowdoin4,*
1 Department of Biological Science, Florida State University, Tallahassee, FL 32306-4295, USA
Department of Plant and Microbial Biology, North Carolina State University, Raleigh, NC 27695, USA
3 Department of Horticultural Science, North Carolina State University, Raleigh, NC 27695, USA
4 Department of Molecular and Structural Biochemistry, North Carolina State University, Raleigh, NC 27695, USA
2 * To whom correspondence should be addressed. E-mail: [email protected]
Received 1 October 2013; Revised 26 November 2013; Accepted 5 December 2013
Abstract
The progress of nuclear DNA replication is complex in both time and space, and may reflect several levels of chromatin structure and 3-dimensional organization within the nucleus. To understand the relationship between DNA replication and developmental programmes, it is important to examine replication and nuclear substructure in different
developmental contexts including natural cell-cycle progressions in situ. Plant meristems offer an ideal opportunity
to analyse such processes in the context of normal growth of an organism. Our current understanding of large-scale
chromosomal DNA replication has been limited by the lack of appropriate tools to visualize DNA replication with high
resolution at defined points within S phase. In this perspective, we discuss a promising new system that can be used
to visualize DNA replication in isolated maize (Zea mays L.) root tip nuclei after in planta pulse labelling with the thymidine analogue, 5-ethynyl-2′-deoxyuridine (EdU). Mixed populations of EdU-labelled nuclei are then separated by flow
cytometry into sequential stages of S phase and examined directly using 3-dimensional deconvolution microscopy to
characterize spatial patterns of plant DNA replication. Combining spatiotemporal analyses with studies of replication
and epigenetic inheritance at the molecular level enables an integrated experimental approach to problems of mitotic
inheritance and cellular differentiation.
Key words: DNA replication, EdU, endocycle, flow cytometry, high-resolution microscopy, maize, mitotic cell cycle, root
development, S phase.
Introduction
Eukaryotic DNA replication is distributed in both time and
space, with different regions of the genome replicating at different times in S phase and at different locations within the
nucleus (reviewed by Bryant and Aves, 2011; Gilbert et al.,
2010). In multicellular organisms, gene-dense euchromatic
regions replicate early, and repetitive heterochromatic regions
replicate later within the S phase of the cell cycle (Rhind
and Gilbert, 2013). It has been reported in mammals and
Xenopus that early replication occurs in the interior of the
nucleus while late replication mostly occurs at the periphery
(Pope and Gilbert, 2013). There is also evidence in mammals
that clusters of replicons, known as replication domains, are
regulated as units that can be visualized in cells pulse-labelled
with nucleotide analogues (Gilbert et al., 2010).
Chromatin structure and replication programmes change
during development and cell differentiation. Several examples
of developmental regulation of replication timing have been
described in mammals (Gilbert et al., 2010; Pope and Gilbert,
© The Author 2014. Published by Oxford University Press on behalf of the Society for Experimental Biology. All rights reserved.
For permissions, please email: [email protected]
2748 | Bass et al.
2013) and Drosophila (Nordman and Orr-Weaver, 2012). Less
is known in plants, but there is evidence that replication time
and origin use are subject to hormonal control (Ishida et al.,
2010; Bryant and Aves, 2011). A significant amount of information has been gathered about the proteins and epigenetic
modifications associated with plant DNA replication (Shultz
et al., 2007; Costas et al., 2011a,b), and a replication timing
profile has been produced for Arabidopsis chromosome 4 for
cultured cells (Lee et al., 2010). However, our understanding
of the larger-scale dynamics of chromosomal DNA replication is still very limited for plants.
Although some aspects of the DNA replication process are
broadly conserved across eukaryotes (Bryant and Aves, 2011;
DePamphilis and Bell, 2011), studying replication in plants
presents unique opportunities. Unlike animals, plant meristems are typically indeterminate and continue to give rise to
new organs after the initial embryonic stage, allowing studies
of genome duplication and maintenance in organized tissues
rather than in cells artificially maintained in culture (Costas
et al., 2011b). In addition, many plant tissues undergo developmentally programmed endocycles, in which S phase is
uncoupled from mitosis resulting in multiple rounds of chromosomal DNA replication and increased ploidy. We have
only begun to understand potential differences in chromatin
structure, replication programmes, and epigenetic inheritance
between mitotic and endocycle S phases in plants (Hayashi
et al., 2013).
Maize (Zea mays) has several features, including well-characterized genetics and a fully sequenced genome, that make it
an excellent plant system for studying DNA replication. The
relatively large and distinct chromosomes of maize are well
suited for studies of cytogenetics and subnuclear structure
(Birchler and Bass 2009). In addition, given its prominence
as a crop species worldwide, understanding how maize replicates its genome may have important agricultural implications. This is especially true for important seed tissues such
as the endosperm of cereal grain species, which undergo
programmed endoreduplication and produce large amounts
of valuable products such as seed protein, starch, and oils
(Sabelli and Larkins, 2009)
The maize B73 genome is similar in size (2.3 Gb) to
the human (3.2 Gb) and mouse (2.5 Gb) genomes and is
comprised of about 85% transposable elements (Lander
et al., 2001; Waterston et al., 2002; Schnable et al., 2009).
Interestingly, the maize genome is c.17-times larger than the
genome of Arabidopsis thaliana (Poethig, 2001), which has
been used as a model for plant DNA replication studies (Van’t
Hof et al., 1978; Jacob et al., 2010; Lee et al., 2010; Costas
et al., 2011a,b). In spite of the large difference in genome size,
it is estimated that the maize and Arabidopsis genomes contain 32 500 and 25 000 genes, respectively—a difference of
only about 30%.
Comparison of DNA replication in maize and Arabidopsis
will lead to a better understanding of chromatin dynamics in
higher plants. As a first step towards such an understanding,
we discuss in this perspective paper some of the advantages,
challenges and opportunities associated with investigating
the spatiotemporal aspects of DNA replication in maize. In
particular, we highlight the utility of our recently developed
maize root tip system using a combination of fluorescent
labelling, flow cytometry, and high-resolution 3D microscopy
to characterize spatiotemporal replication patterns in nuclei
from naturally cycling cells.
Isolating and staging S-phase nuclei for
analysis
EdU as a tool for plant research
In recent years, 5-ethynyl-2′-deoxyuridine (EdU), a nucleoside
analogue of thymidine, has been used increasingly to detect
cells actively replicating their DNA. This technology takes
advantage of a copper(I)-catalysed azide-alkyne cycloaddition (often referred to as ‘click chemistry’) first described
by Rostovtsev et al. (2002) and Tornøe et al. (2002). Click
chemistry can be used to detect EdU incorporated into newly
replicated DNA by reaction of the terminal alkyne on EdU
with an azide attached to a fluorescent dye such as Alexafluor
488 (Salic and Mitchison, 2008). Darzynkiewicz et al. (2011)
reviewed the benefits of using EdU for detecting DNA synthesis compared to earlier techniques. In particular, the elimination of the acid or heat denaturation treatment needed for
antibody detection of 5-bromo-2′-deoxyuridine, a commonly
used thymidine analogue, greatly enhances maintenance of
DNA and protein structural integrity (Darzynkiewicz et al.,
2011).
Kotogány et al. (2010) were the first to look in depth at the
feasibility of using EdU in plant suspension cultures and roots
for flow cytometry and microscopy. Since then, a number of
research groups have used EdU for plant cell-cycle studies,
establishing its utility in a variety of plant tissues (Table 1).
Most of these studies have addressed three broad areas: (1) the
proliferation patterns and fraction of cells entering S phase
in various plant tissues (Ayaydin et al., 2011; Kelliher and
Walbot, 2011; Ilina et al., 2012; Kuznetsova and Sheval, 2013);
(2) changes in DNA replication in response to various mutations, transgenes or chemical treatments (Vanstraelen et al.,
2009; Zhou et al., 2010; Schubert et al., 2011; Tresch et al.,
2011; Zhu et al., 2011; Xu et al., 2012, 2013; Bazin et al., 2013;
Xiong et al., 2013); and (3) the characterization of replication
patterns at a chromosome or a chromatin level (Heckmann
et al., 2011; Greer et al., 2012; Xu et al., 2012; Hayashi et al.,
2013; She et al., 2013). More recently, EdU labelling over discrete time intervals has been used in combination with confocal microscopy (Hayashi et al., 2013) or flow cytometry (E.
Wear, L. Hanley-Bowdoin, and W. Thompson, unpublished
data) to determine the average length of S phase in plant
tissues.
EdU-labelling techniques have been combined with immunocytology and fluorescent in situ hybridization (FISH) to
explore patterns of DNA replication, replication-associated
epigenetic marks, and chromatin packaging in plant nuclei.
Multiple groups have studied the progression of premeiotic replication in germ line cells using EdU in conjuction
with DNA probes or antibodies that target centromeric or
New approaches to studying plant DNA replication | 2749
Table 1. Reports of EdU labelling of nascent DNA in plant systems
Plant species
Name
Tissue
Applications
References
Arabidopsis thaliana
Arabidopsis thaliana
Arabidopsis thaliana
Thale cress
Thale cress
Thale cress
Cell culture
Inflorescence
Root
Flow cytometry, microscopy
Microscopy
Microscopy
Arabidopsis thaliana
Cucumis sativus
Cucurbita pepo
Hordeum vulgare
Luzula elegans
Medicago sativa
Thale cress
Cucumber
Squash
Barley
Wood rush
Alfalfa
Root
Root
Hypocotyl and cotyledon
Inflorescence
Apical meristem
Cell culture
Flow cytometry
Microscopy
Microscopy
Microscopy
Microscopy
Flow cytometry, microscopy
Medicago truncatula
Nicotiana tabacum
Barrel medic
Tobacco
Root
Cell culture
Flow cytometry
Microscopy
Oryza sativa
Oryza sativa
Rice
Rice
Cell culture
Root
Flow cytometry, Microscopy
Microscopy
Secale cereale
Secale cereale × Triticum
aestivum
Solanum lycopersicum
Vicia faba
Vitis berlandieri × V. rupestris
Zea mays
Zea mays
Rye
Wheat–rye hybrid
Root
Tiller
Microscopy
Microscopy
Kotogány et al. (2010)
She et al. (2013)
Vanstraelen et al. (2009), Zhou
et al. (2010), Zhu et al. (2011),
Hayashi et al. (2013), Xiong
et al. (2013), Xu et al. (2013)
Xu et al. (2012)
Hayashi et al. (2013)
Ilina et al. (2012)
Higgins (2013)
Heckmann et al. (2012)
Kotogány et al. (2010), Ayaydin
et al. (2011)
Bazin et al. (2013)
Kotogány et al. (2010), Tresch
et al. (2011), Hayashi et al. (2013)
Kotogány et al. (2010)
Kotogány et al. (2010), Hayashi
et al. (2013)
Klemme et al. (2013)
Greer et al. (2012)
Tomato
Fava bean
Grape hybrid
Maize
Maize
Root
Root
Cell culture
Cell culture
Immature tassel
Microscopy
Microscopy
Microscopy
Microscopy
Microscopy
Kuznetsova and Sheval (2013)
Schubert et al. (2011)
Kotogány et al. (2010)
Kotogány et al. (2010)
Kelliher and Walbot (2010)
telomeric DNA, histone proteins, or developmentally related
proteins (Greer et al., 2012; Higgins, 2013; She et al., 2013).
In particular, Higgins presented detailed methods for EdU
labelling and isolating barley inflorescences and subsequent
immunolocalization and FISH preparations. Others have
investigated the colocalization of EdU with DNA probes
or histone marks and the overall chromosomal organization
following mitotic replication on holocentric chromosomes
(Heckmann et al., 2011) or supernumerary B chromosomes
(Klemme et al., 2013). There has also been interest in identifying ‘early’ and ‘late’ replication patterns and associating
them with euchromatic and heterochromatic regions of DNA
(Heckmann et al., 2011; Greer et al., 2012; Xu et al., 2012;
Hayashi et al., 2013; She et al., 2013).
Most of the studies in Table 1 used microscopy to visually
stage plant nuclei in S phase, which only distinguishes large
differences in EdU labelling and does not provide measures
of DNA content. As a consequence, visual staging of EdUlabelled nuclei cannot resolve them into different parts of S
phase. The few studies in Table 1 that combined EdU labelling with flow cytometry were analytical in nature and did not
use flow sorting to enrich for discrete S-phase populations.
Hence, more precise methods for staging plant nuclei within
S phase are needed for fine-scale observations and measurements of DNA replication patterns both spatially and in relation to chromatin features. To address this need, we developed
a method whereby EdU-labelled nuclei are isolated using
a flow cytometer to separate S-phase nuclei into sequential
stages for 3D microscopic analysis. We combined this strategy with analyses of nuclei from cells at different positions in
the developmental gradient of the growing maize root tip to
establish a new experimental system for studying plant DNA
replication. Approaches to visualize and analyse DNA replication patterns in time and space are emphasized here. We are
currently using these approaches to compare DNA replication programmes in mitotic and endocycling plant cells.
Flow sorting for temporal resolution within S phase
Flow cytometers were originally developed for use with
mammalian blood cells but evolved to include a wide range
of applications in the plant sciences (reviewed by Galbraith,
2010). The introduction of the chopping procedure to release
intact nuclei from plant tissue homogenates (Galbraith et al.,
1983) made it possible to run a single particle suspension
through a flow cytometer and sort nuclei with various fluorescent tags. This isolation method has been adapted to fixed
tissues through the use of gentle blending (Lee et al., 2010),
after which nuclei can be subjected to the click reaction to
couple a fluorescent Alexa-488 azide to EdU incorporated
into nascent DNA.
Many DNA replication studies have used a flow cytometer
to isolate different fractions of S phase based on DNA content as measured by 4′,6-diamidino-2-phenylindole (DAPI)
2750 | Bass et al.
or propidium iodide staining (Chen et al., 1995; Woodfine
et al., 2004; Lee et al., 2010). While this is a useful technique,
it is not suitable for some applications because it does not separate replicating from non-replicating cells or nuclei (Fig. 1B
and D). By adding a second parameter to the flow sorting (e.g.
EdU incorporation measured after coupling to an Alexa-488
azide), we can readily distinguish nuclei in S phase from other
cell-cycle phases. EdU-labelled nuclei, even those in very early
or very late S phase, are resolved by flow cytometry as an arc
distinct from unlabelled G1 and G2 nuclei in a bivariate plot
of DAPI versus Alexa-488 (Fig. 1A and C). Gating of the
arc can be used to further separate S phase into early, middle, and late stages (Fig. 1A, rectangles labelled E, M, and
L). This approach for staging nuclei allows us to start with
an asynchronous population of nuclei and avoid artefacts
sometimes associated with synchronization procedures such
as sucrose starvation or incubation with nucleotide synthesis
inhibitors (Cooper, 2003).
Some technical considerations must be addressed when
combining flow sorting and high-resolution microscopy. First,
the fixation strategy and buffers must be carefully selected
to balance the need to ensure a single particle suspension
for the flow cytometer with the need to preserve chromatin
structure and nuclear architecture for downstream analysis.
Second, depending on the experimental goals, the length of
the EdU pulse can be important. For example, a short pulse
will label DNA proximal to replication forks, while a long
pulse will label more of the genome and include sequences
Fig. 1. Labelling actively replicating DNA in maize root tips. (A–D) Root tips from 3-d-old seedlings were pulse-labelled with EdU for 20 min, followed by
fixation and nuclei isolation and the nuclei were analysed by flow cytometry using a BD Influx instrument equipped with 355 nm (UV) and 488 nm (blue)
lasers. Bivariate plots of DNA content (based on DAPI fluorescence in the 460 ± 50 nm detection range) and EdU incorporation (based on Alexa fluor (AF)
488 fluorescence in the 530 ± 40 nm detection range) are shown in A and C. The corresponding univariate histograms of DNA content are shown in B and D,
respectively, from 0–1 mm root segments (A, B) and 1–3 mm segments (C, D); the gates corresponding to early (E), middle (M), and late (L) subpopulations of
S phase are indicated as rectangles for the mitotic cell cycle (A) and the endocycle (C). (E) Bivariate flow cytometry profiles of mitotic and endocycling nuclei
from consecutive 1-mm root segments up to 5 mm labelled with EdU for 1 h: the identical black boxes in each profile indicate the gates used for analysis of
the percentage of nuclei with 2C, 4C, and 8C DNA content shown in the root schematic below. (F) Merged confocal images of root tip cross-sections taken
at the points indicated by the black arrowheads, showing DAPI-stained DNA (red), and EdU label in newly replicated DNA (green). Bar, 100 μm.
New approaches to studying plant DNA replication | 2751
that are distal as well as proximal to active forks. In many
cases, knowing the duration of S phase in the material under
study will allow a more informed decision on pulse length.
Programmed endocycling in plants and the
developmental biology of roots
Endocycling: a widespread phenomenon in plants
A shift from a mitotic cell cycle to endoreduplication or an
endocycle, in which cells replicate their DNA without subsequent cell divisions, has been documented during development for roots, shoots, leaves, and endosperm of diverse plant
species (Joubes and Chevalier, 2000; Lee et al., 2009; Hayashi
et al., 2013). The formation of many specialized plant tissues and organs is known to require endocycling. Examples
include trichomes, leaf epidermal cells, maize endosperm,
and tomato fruit among many others (Joubes and Chevalier,
2000). Recently, Breuer et al. (2010) reviewed some of the
developmental controls of the endocycle in Arabidopsis leaves
and roots, emphasizing the importance of the endocycle for
normal organ growth. They and others have also noted that
the transition from the mitotic cell cycle to the endocycle
is often coupled with a shift from cell proliferation to cell
expansion and differentiation (Breuer et al., 2010; Hayashi
et al., 2013).
Opportunity afforded by root tip organization
Developmentally distinct regions along the maize root have
been identified morphologically (Baluska et al., 1990; Baluska
and Mancuso, 2013). The tip contains dividing cells, followed
by a region termed the ‘transition zone’ or ‘post-mitotic isodiametric zone’, located approximately 1–2.5 mm from the
root apex. In this region, cell division mostly halts and cells
begin to increase in length and width before reaching a zone
of rapid elongation. Endopolyploidy has been observed for
maize roots (Biradar et al., 1993; Ogawa et al., 2010). However,
there are only a few reports that integrate ploidy information
with plant root developmental programmes (Hayashi et al.,
2013). Using flow cytometry, we found that the first 10 mm
of the maize root tip contain 2C, 4C, and 8C cells, but very
few cells of higher ploidy (data not shown). To increase our
resolution, we dissected a series of 1-mm segments from the
apex up to 5 mm of EdU-labelled maize roots and examined
the distribution of isolated nuclei in the mitotic cell cycle and
the endocycle by 2-colour flow cytometry, as summarized in
Fig. 1. The flow cytometry data were first filtered to remove
cellular debris based on plots of side scatter (90° angle light
scatter) and DAPI fluorescence. Identical gates were then set
to analyse the percentage of 2C, 4C, and 8C nuclei in each
root segment (Fig. 1E). This analysis revealed that the first
1 mm of the root consists primarily of cells in the mitotic
cycle with 37% 2C, 52% 4C, and 4% 8C nuclei (n = 39 351).
The endocycle is increasingly prominent beyond 1 mm, with
the 2–3 mm segment containing 9% 2C, 44% 4C, and 38% 8C
nuclei (n = 27 238). Both types of replication have essentially
ceased by 4–5 mm from the tip. Strikingly, most maize root
cells exit the cell cycle with either 4C (43%) or 8C (36%)
DNA content, with only a small fraction (10%) having 2C
content. We also observed the distribution of EdU-labelled
cells across different cell types by imaging cross-sections at
intervals along the root (Fig. 1F).
Our experiments established the feasibility of combining
careful dissection and flow sorting to target either the mitotic
cycle or the endocycle for a molecular analysis of DNA replication in an organized plant tissue. There are several advantages of studying DNA replication patterns using root tips
instead of cell cultures. By focusing on zones with active division, we can obtain preparations highly enriched in nuclei
undergoing replication. Fixed root tips also provide relatively
clean nuclei preparations, ideal for flow sorting. In addition, the spatial separation of mitotic and endocycling cells,
in combination with flow sorting, affords an opportunity to
compare DNA replication programmes before and after a
normal developmental transition between the two cell-cycle
types.
Collection and analysis of 3D fluorescent
images of nuclei
Flow sorting considerations for 3D imaging using
multiple wavelengths
Formaldehyde is an excellent fixative for chromatin and has
been widely used for decades to preserve nuclear architecture and chromatin structure for microscopy (Belmont et al.,
1989; Swedlow et al., 1993). We have used deconvolution
microscopy to obtain high-resolution fluorescent 3D images
of fixed nuclei using optical sectioning. In the past, we often
employed a 3D acrylamide FISH technique to preserve nuclei
within cellular or tissue contexts (Bass et al., 1997, 2000a,b,
2003). However, we found that nuclei fixed in situ and then
purified from the maize root tip system retained their generally spherical shape without acrylamide embedding, which
considerably simplifies the mounting and imaging procedures. We also found that nuclei fixation with formaldehyde
can be carried out in the presence of various buffers, including phosphate-buffered saline or Buffer A. Buffer A was specifically developed to preserve the native chromatin state for
cytological analysis of nuclear architecture and chromatin
structure (Sedat and Manuelidis, 1978; Belmont et al., 1989).
The formaldehyde plus Buffer A chromatin fixation has been
adapted for use in plants and is also compatible with additional staining techniques such FISH or immunocytochemistry (Howe et al. 2013). Phosphate-buffered saline has the
advantage of being more compatible with flow cytometry
and, as such, is our buffer of choice for experiments that combine sorting and microscopy. In addition, nuclei separated
by 2-colour flow cytometry are already fixed, labelled with
Alexa-488, stained with DAPI and, thus, ready for immediate
2-colour imaging. Fixed and sorted nuclei are stable in various buffers, and can be stored for later use at 4°C for a few
months in air-tight, light-free tubes (data not shown).
2752 | Bass et al.
Other useful features of wide-field epifluorescence
deconvolution microscopy with a sensitive CCD camera
include the ability to collect two additional wavelengths
using conventional red (Cy3, Rhodamine, Texas Red,
ROX, Alexa-546) or far-red (Cy5, Alexa-694) fluorophores and obtain intensity data over a wide linear range
(1–4096 counts/pixel before deconvolution) for every
wavelength. These capacities enhance comparison of
EdU-labelling patterns with those of other chromatin or
nuclear feature staining methods such as in situ hybridization or immunocytochemistry.
Display of 3D image data on 2D media
Collection of multicolour data adds extra dimensions of
information. For each wavelength, a series of optical sections
are photographed using a CCD camera at defined Z-axis step
sizes, providing 3D digital data. The different wavelengths
represent a fourth dimension. For 3D imaging of fixed samples, separate images are collected for each wavelength at
each focal plane, before the stage is moved to the next optical
section. The interleaved image data are sorted by wavelength
and subjected to computational deconvolution to correct for
chromatic aberrations (Chen et al., 1995). In this way, photon
counts are conserved and redistributed within the 3D space
of a cube of images, accomplishing a level of deblurring and
brightening that distributes pixel signal intensities over a wide
dynamic range. Linear scaling up to 32 000 counts can be
achieved without the use of photomultiplier tubes or artificial gain. As a result, the deconvolved images contain considerable substructure (for example, within the heterochromatic
knobs, as shown in Fig. 2). This substructure is visible even
for regions of the 3D dataset that may have narrow portions
(10%) of the intensity range in any given experiment.
In Fig. 2, a representative 2-colour (DAPI and Alexa-488)
3D image of a single nucleus from a somatic cell in late S phase
illustrates several ways to visualize the information contained
in a 3D dataset using 2D media, such as computer screens
or the printed page. The 3D dataset is comprised of optical
sections of pixels with X–Y dimensions of 0.07 × 0.07 μm,
just below the resolution limit of this type of microscopy.
The spacing between adjacent images is separated by a userchosen Z-spacing at point of initial imaging, typically 0.2,
0.25, or 0.3 μm Z-step sizes. These dimensions provide oversampling in X, Y, and Z, and are optimized for the tradeoff
between image file size and resolution. The X–Y magnification and the Z step size together define the dimensions of the
individual volumetric pixels (voxels) of the dataset. These
voxels have a single assigned spatial position but wavelengthspecific intensity values. As a consequence, the data for different wavelengths can be viewed separately (Fig. 2B) or in
pseudocolour combination. Images are typically adjusted for
brightness and contrast using interactive linear scaling prior
to producing a particular image for display.
A valuable method for 3D image display is to produce a
through-focus projection of the entire data stack, as shown
in Fig. 2B. Two methods of doing this are shown for comparison. The ‘average’ projection preserves the true average
value of all the voxels at a given X–Y position, whereas
the ‘maximum intensity’ method displays only the single
maximum voxel value regardless of where it comes from in
the Z-stack. The maximum intensity view appears sharper
because it is a composite of the brightest features, but the
average projection is mathematically more correct and, thus,
more suitable for quantitative measurements. Another view
that provides a useful overview of a multicolour 3D image
involves the production and display of sequential projections
(Fig. 2C). Here, the 3D cube of data is divided into equalsized thick segments along the Z-axis and the resulting thick
optical sections are displayed as a series of images representing adjacent through-focus projections. This allows for ready
inspection of the entire 3D data in several parts, without
having to view all 60 optical sections separately. Objects that
appear adjacent in whole-nucleus projections (Fig. 2B, double arrows) can often be seen to be spatially separate when
viewed in sequential projections (Fig. 2C, double arrows).
We found that both the DAPI and Alexa-488 (EdU) signals
are very stable through additional staining techniques such
as FISH (Fig. 2D), allowing for further colocalization studies within the nucleus. The images in Fig. 2 reveal variation
in DNA replication patterns across the nucleus, including the
previously reported tendency for heterochromatic knobs to replicate late in S phase (Pryor et al. 1980). The late replication
of knobs is seen as overlapping signals for the Alexa-488 and
the knob FISH probe (‘k’ in Fig. 2D indicates the locations of
two late-replicating knobs). We also note that 3D imaging may
reveal important but subtle differences in replication timing patterns, such as those associated with early versus middle S phase
(H. Bass, G. Hoffman, T-J. Lee, E. Wear, S. Joseph, G. Allen,
L. Hanley-Bowdoin, and W. Thompson, unpublished data).
Other useful ways to visualize 3D data include production of
stereo pair projections (not shown) or movie files that move your
view through or around the 3D dataset (Supplementary Table
S1, available at JXB online). Exporting of 3D images to common movie file formats allows one to scroll through the optical sections (step-through movies) or ‘spin’ the 3D data using
a series of projections viewed from different angles about the
Y-axis (spinning projection movies). The 3D dataset shown for
the FISH-stained nucleus (Fig. 2D) was used to produce stepthrough and spinning projections movies of single or combined
wavelengths (Supplementary Table S1 available at JXB online).
Interacting with 3D data in this way can increase comprehension of 3D multiwavelength data, as illustrated by the patchy
labelling pattern in this late S-phase nucleus. This example also
demonstrates that Click-It Alexa-488 labelling of EdU provides
lasting signals marking the pulse-labelled sites of DNA replication. Combining 3D microscopy with flow-sorted temporal
separation based on DNA content gives unprecedented access
to spatiotemporal information on DNA replication timing.
Spatial and fluorescence signal measurements from
image datasets
Another useful feature of 3D digital data collection is that
the datasets can be used as raw material for experiments
New approaches to studying plant DNA replication | 2753
Fig. 2. Collection and display of 3D image data from a representative somatic nucleus pulse-labelled during late S phase. Root tips were pulse-labelled with
EdU and fixed with formaldehyde. Nuclei isolated from the labelled root tips were stained for EdU incorporation using a Click-It EdU Alexa-488 imaging kit,
separated by flow cytometry, and either imaged directly (A–C, E) or subjected to 3D acrylamide FISH (D) to visualize the location of knob and centromere
sequences; images in A–C are all from a single 3D dataset comprised of two-colour optical sections that imaged a single nucleus from late S phase. (A)
Digital data are recorded in voxels (3D pixels with real-space X, Y, and Z dimensions). (B) Whole-nucleus through-focus projections generated by two
common methods are compared: projections using the average method are shown in the left column, while those in the right column were made with
the maximum intensity method. (C) Sequential projections illustrate the entire data set as a series of partial projections with each projection representing
one-fifth of the Z-sections after 3D cropping to the edges of the nucleus: this approach optimizes the information–space relationship for showing 3D data
in 2D formats; double arrows indicate loci (also depicted in B and C) that may appear close together in projection, but spatially separate along the Z-axis.
(D) Suitability of DAPI-stained and EdU-Alexa-488-labelled nuclei for 3D FISH using the method described by Howe et al. (2013). The knob and centromere
probes are fluorescent oligonucleotide probes (knob probe ‘NUBI-R’ is 5′-Rhodamine-AACATATGTGGGGTGAGGTGTATG-3′ and centromere probe ‘MCCY’
is 5′-Cy5-GAAAAACGAAGAAATGGTTCTGGTG-3′, from Koumbaris and Bass, 2003). k indicates the locations of two late-replicating knobs. Bars, 3 μm.
allowing for measurements and quantitative hypothesis
testing that take advantage of the spatial resolution and
dynamic range of signal intensity data. In the case of DNA
replication time during S phase, it is often of interest to
compare the replicated DNA captured with EdU pulse
labelling to the total genomic DNA. Here, two types of
2754 | Bass et al.
measurements prove useful. One type measures the correlation of signals (DAPI versus Alexa-488) across the population of volumetric pixels, while the other extracts real-space
measurements—such as volumes and centres of space and
‘mass’–as summarized in Fig. 3. Here we illustrate the use
of such techniques to quantify the differences between early
and middle S phase. For instance, the early S-phase nucleus
(Fig. 3A left) appears to have more separation between the
DAPI and Alexa-488 signals than does the nucleus from
middle S phase (Fig. 3A right). This difference was quantified using Pearson’s correlation coefficient (Fig. 3 legend)
and the visual impression was verified numerically with a
statistical R-value.
We also measured nuclear volumes and the non-­uniform
distribution of signals using image segmentation (Fig. 2B
legend). The nuclear volume was segmented using computer-assisted or manual edge tracings and the resulting polygons—one per optical section—were connected
to produce a solid surface object with spatial coordinates
linked to that dataset. The solid object serves to delineate
the inside and outside of the nucleus, providing information on nuclear volume, and the degree to which the DAPI
and Alexa-488 signals are codistributed across the entire
3D dataset.
Together, the two approaches illustrated here—image
projections for visual data inspection and collection of signal intensity and real-space measurements for quantitative
analysis—are useful starting points for 3D cytology and
particularly valuable for DNA replication studies. As large
digital datasets become more common, sharing and reanalysing archived cytological data is expected to become more
commonplace (Allan et al., 2012), and approaches similar to
those illustrated here are expected to come into widespread
use for 3D datasets.
Fig. 3. Signal correlation, spatial, and volumetric measurements on 3D
image data. Nuclei from maize root tip cells were pulse-labelled for DNA
replication and flow sorted as described for Fig. 2. Optical sections were
collected to generate a 3D dataset. Following computational deblurring
by iterative deconvolution (Chen et al., 1995), information content in
the dataset can be obtained and analysed computationally using image
analysis software packages (DeltaVision, API/GE Health Care in this
example) to measure colocalization of signals (A) or volumetric data
following 3D image segmentation (B). (A) An example of analysis with
Pearson’s correlation coefficient is shown for images from two different
nuclei from different stages (early or middle S phase, as indicated under
each image). Scatter plots can be used to show the relationship of DAPI
(displayed as red in image and plot Y-axis) and Alexa-488 (displayed
as green in image and plot X-axis) signals for the population of signals
in a single optical section (shown here) or for an entire 3D dataset (not
shown). The resulting bivariate plot allows for a correlation coefficient
(R-value) to be calculated. In this case, there is a lower correlation
between DAPI and Alexa-488 at early S phase (R = 0.696) than at middle
S phase (R = 0.842). This type of measurement allows for comparative
analysis across different images and is particularly well suited for diffuse,
scattered, or spatially distributed signals. (B) For spatial measurements,
nuclei are segmented by manual or automated edgefinding, drawing a
polygon around the perimeter of the nucleus that is displayed for each
section (polygon displayed as red in grey-scale images, and white in
pseudocoloured image). A single polygon is applied to all wavelengths
New insights and future prospects
Using EdU and click chemistry to fluorescently label nascent DNA is an important technical advance in the general field of DNA replication. Although the EdU-labelling
technology has been increasingly applied to plant systems
(Table 1), its full potential is only now being realized with
its adaptation to other technologies like flow cytometry.
In this report, we discuss the experimental advantages
for combined application of EdU and flow cytometry to
label and fractionate replicating nuclei from maize root
tips for DNA replication studies. Unlike cell-culture-based
of each optical section. Once completed, the series of all polygons (blue
lines) are connected to form a closed surface that defines a solid object
(green) that segments the voxel data into regions inside or outside of the
object. The solid object is then used to calculate parameters such as
real-space volumes (μm3), the centre of space (given as a single point with
X-, Y-, and Z-values common to all wavelengths), or the centre of intensity
(comparable to ‘centre of mass’, an intensity-weighted centre of mass
with unique X-, Y-, and Z-values for each wavelength). The volumetric data
provide quantitative values for observations that are otherwise estimates or
subjective impressions.
New approaches to studying plant DNA replication | 2755
systems typically used to study eukaryotic DNA replication, maize root tips provide a source of actively dividing
nuclei undergoing a natural developmental programme in
the context of an organized tissue. In particular, the maize
root tip system can be used to examine DNA replication
during the developmental transition from a mitotic cell
cycle to an endocycle—a common transition during plant
cell differentiation.
As described here, the maize root tip system can be combined readily with high-resolution microscopy to compare
the spatial patterns of DNA replication during S phase in
mitotic and endocycling cells. The maize root tip system can
also be adapted to genomic protocols like DNA-IP, ChIP,
and Hi-C to study replication programmes, chromatin
landscapes, and long-distance interactions in EdU-labelled
nuclei. These types of studies will help define the regions of
the maize genome that replicate during early, middle, or late
S phase and any associated chromatin features. Comparison
of the temporal, spatial, and epigenetic properties of replicating DNA in maize root tips will provide insight into general characteristics of a plant nucleus in S phase and those
that change when a cell transitions from a proliferating state
to an endocycling state during differentiation.
Knowledge of the mechanisms that mediate and control
DNA replication is essential for understanding plant growth
and development. We now have the tools to explore these
important processes in plants and to use this information to
address important agronomic questions such as reducing phenotypic variation associated with passage through S phase or
increasing grain yield by modulating DNA replication during
endosperm development.
Supplementary material
Supplementary data are available at JXB online.
Supplementary Table S1. Movies from a 4-wavelength 3D
deconvolution image dataset of a nucleus in late S phase
showing EdU-labelled DNA synthesis along with knob and
centromere repeat FISH signals.
Acknowledgements
This work was supported by the National Science Foundation (grants IOS1025830 to L.H.-B., W.F.T., and G.C.A., and IOS-1025954 to H.W.B.).
References
Allan C, Burel JM, Moore J, et al. 2012. OMERO: flexible, model-driven
data management for experimental biology. Nature Methods 9, 245–253.
Ayaydin F, Kotogány E, Abraham E, Horvath GV. 2011.
Synchronization of Medicago sativa cell suspension culture. Methods in
Molecular Biology 761, 227–238.
Baluska F, Kubica S, Hauskrecht M. 1990. Postmitotic isodiametric
cell-growth in the maize root apex. Planta 181, 269–274.
Baluska F, Mancuso S. 2013. Root apex transition zone as oscillatory
zone. Frontiers in Plant Science 4, 354.
Bass HW, Bordoli SJ, Foss EM. 2003. The desynaptic (dy) and
desynaptic1 (dsy1) mutations in maize (Zea mays L) cause distinct
telomere-misplacement phenotypes during meiotic prophase. Journal of
Experimental Botany 54, 39–46.
Bass HW, Marshall WF, Sedat JW, Agard DA, Cande WZ. 1997.
Telomeres cluster de novo before the initiation of synapsis: a threedimensional spatial analysis of telomere positions before and during
meiotic prophase. Journal of Cell Biology 137, 5–18.
Bass HW, Nagar S, Hanley-Bowdoin L, Robertson D. 2000a.
Chromosome condensation induced by geminivirus infection of mature
plant cells. Journal of Cell Science 113, 1149–1160.
Bass HW, Riera-Lizarazu O, Ananiev EV, Bordoli SJ, Rines HW,
Phillips RL, Sedat JW, Agard DA, Cande WZ. 2000b. Evidence for the
coincident initiation of homolog pairing and synapsis during the telomereclustering (bouquet) stage of meiotic prophase. Journal of Cell Science
113, 1033–1042.
Bazin J, Khan GA, Combier JP, et al. 2013. miR396 affects
mycorrhization and root meristem activity in the legume Medicago
truncatula. The Plant Journal 74, 920–934.
Belmont AS, Braunfeld MB, Sedat JW, Agard DA. 1989. Large-scale
chromatin structural domains within mitotic and interphase chromosomes
in vivo and in vitro. Chromosoma 98, 129–143.
Biradar DP, Rayburn AL, Bullock DG. 1993. Endopolyploidy in diploid
and tetraploid maize (Zea mays L). Annals of Botany 71, 417–421.
Birchler JA, Bass HA. 2009. Cytogenetics and chromosomal
structural diversity. In: Bennetzen J, Hake S, editors, The maize
handbook, Volume II: Domestication, Genetics and Genomics. USA:
Springer, pp 163–167.
Breuer C, Ishida T, Sugimoto K. 2010. Developmental control of
endocycles and cell growth in plants. Current Opinion in Plant Biology 13,
654–660.
Bryant JA, Aves SJ. 2011. Initiation of DNA replication: functional and
evolutionary aspects. Annals of Botany 107, 1119–1126.
Chen H, Swedlow JR, Grote MA, Sedat JW, Agard DA. 1995. The
collection, processing, and display of digital three-dimensional images
of biological specimens. In: Pawley JB, editor, Handbook of biological
confocal microscopy. New York, NY: Plenum Press, pp 197–210.
Cooper S. 2003. Rethinking synchronization of mammalian cells for cell
cycle analysis. Cellular and Molecular Life Sciences 60, 1099–1106.
Costas C, Desvoyes B, Gutierrez C. 2011a. A chromatin perspective
of plant cell cycle progression. Biochimica et Biophysica Acta 1809,
379–387.
Costas C, Sanchez MD, Sequeira-Mendes J, Gutierrez C. 2011b.
Progress in understanding DNA replication control. Plant Science 181,
203–209.
Darzynkiewicz Z, Traganos F, Zhao H, Halicka HD, Li JW. 2011.
Cytometry of DNA replication and RNA synthesis: historical perspective
and recent advances based on ‘click chemistry’. Cytometry Part A 79A,
328–337.
DePamphilis ML, Bell SD. 2011. Genome duplication. New York:
Garland Science.
Galbraith D. 2010. Flow cytometry and fluorescence-activated cell sorting
in plants: the past, present, and future. Biomédica 30, 65–70.
Galbraith DW, Harkins KR, Maddox JM, Ayres NM, Sharma DP,
Firoozabady E. 1983. Rapid flow cytometric analysis of the cell cycle in
intact plant tissues. Science 220, 1049–1051.
Gilbert DM, Takebayashi SI, Ryba T, Lu J, Pope BD, Wilson KA,
Hiratani I. 2010. Space and time in the nucleus: developmental control
of replication timing and chromosome architecture. Cold Spring Harbor
Symposium in Quantitative Biology 75, 143–153.
Greer E, Martin AC, Pendle A, Colas I, Jones AME, Moore G, Shaw
P. 2012. The Ph1 locus auppresses Cdk2-type activity during premeiosis
and meiosis in wheat. The Plant Cell 24, 152–162.
Hayashi K, Hasegawa J, Matsunaga S. 2013. The boundary of the
meristematic and elongation zones in roots: endoreduplication precedes
rapid cell expansion. Science Reporter 3, 2723.
Heckmann S, Schroeder-Reiter E, Kumke K, Ma L, Nagaki K,
Murata M, Wanner G, Houben A. 2011. Holocentric chromosomes of
Luzula elegans are characterized by a longitudinal centromere groove,
chromosome bending, and a terminal nucleolus organizer region.
Cytogenetic and Genome Research 134, 220–228.
2756 | Bass et al.
Higgins J. 2013. Analyzing meiosis in barley. In: Pawlowski WP, Grelon M,
Armstrong S, editors, Plant meiosis, vol. 990. New York: Humana Press,
pp 135–144.
Howe ES, Murphy SP, Bass HW. 2013. Three-dimensional acrylamide
fluorescence in situ hybridization for plant cells. In: Pawlowski WP, Grelon
M, Armstrong SA, editors, Plant meiosis , vol. 990. New York: Humana
Press, pp 53–66.
Ilina EL, Logachov AA, Laplaze L, Demchenko NP, Pawlowski
K, Demchenko KN. 2012. Composite Cucurbita pepo plants with
transgenic roots as a tool to study root development. Annals of Botany
110, 479–489.
Ishida T, Adachi S, Yoshimura M, Shimizu K, Umeda M,
Sugimoto K. 2010. Auxin modulates the transition from the mitotic
cycle to the endocycle in Arabidopsis. Development 137, 63–71.
Jacob Y, Stroud H, Leblanc C, Feng S, Zhuo L, Caro E, Hassel
C, Gutierrez C, Michaels SD, Jacobsen SE. 2010. Regulation
of heterochromatic DNA replication by histone H3 lysine 27
methyltransferases. Nature 466, 987–991.
Joubes J, Chevalier C. 2000. Endoreduplication in higher plants. Plant
Molecular Biology 43, 735–745.
Kelliher T, Walbot V. 2011. Emergence and patterning of the five cell
types of the Zea mays anther locule. Developmental Biology 350, 32–49.
Klemme S, Banaei-Moghaddam AM, Macas J, Wicker T, Novak P,
Houben A. 2013. High-copy sequences reveal distinct evolution of the rye
B chromosome. New Phytology 199, 550–558.
Koumbaris GL, Bass HW. 2003. A new single-locus cytogenetic
mapping system for maize (Zea mays L.): overcoming FISH detection limits
with marker-selected sorghum (S. propinquum L.) BAC clones. The Plant
Journal 35, 647–659.
Kotogány E, Dudits D, Horvath GV, Ayaydin F. 2010. A rapid and
robust assay for detection of S-phase cell cycle progression in plant cells
and tissues by using ethynyl deoxyuridine. Plant Methods 6, 5.
Kuznetsova MA, Sheval EV. 2013. Detection of replication sites in plant
cell nuclei by using semithin sections. Cell and Tissue Biology 7, 375–378.
Lander ES, Linton LM, Birren B, et al. 2001. Initial sequencing and
analysis of the human genome. Nature 409, 860–921.
Lee HO, Davidson JM, Duronio RJ. 2009. Endoreplication: polyploidy
with purpose. Genes and Development 23, 2461–2477.
Lee TJ, Pascuzzi PE, Settlage SB, et al. 2010. Arabidopsis thaliana
chromosome 4 replicates in two phases that correlate with chromatin
state. Plos Genetics 6.
Nordman J, Orr-Weaver TL. 2012. Regulation of DNA replication during
development. Development 139, 455–464.
Ogawa A, Taguchi N, Miyoshi K. 2010. Variations in endopolyploidy
level during the short period of the early growing stage in the roots and
leaves of maize (Zea mays) seedlings. Plant Biotechnology Reports 4,
117–123.
Poethig RS. 2001. Life with 25,000 genes. Genome Research 11,
313–316.
Pope BD, Gilbert DM. 2013. The replication domain model: regulating
replicon firing in the context of large-scale chromosome architecture.
Journal of Molecular Biology, 23, 4690–4695.
Pryor A, Faulkner K, Rhoades MM, Peacock WJ. 1980. Asynchronous
replication of heterochromatin in maize. Proceedings of the National
Academy of Sciences, USA 77, 6705–6709.
Salic A, Mitchison TJ. 2008. A chemical method for fast and sensitive
detection of DNA synthesis in vivo. Proceedings of the National Academy
of Sciences, USA 105, 2415–2420.
Sabelli PA, Larkins BA. 2009. The development of endosperm in
grasses. Plant Physiology 14, 14–26.
Schnable PS, Ware D, Fulton RS, et al. 2009. The B73 maize genome:
complexity, diversity, and dynamics. Science 326, 1112–1115.
Schubert I, Schubert V, Fuchs J. 2011. No evidence for ‘break-induced
replication’ in a higher plant—but break-induced conversion may occur.
Frontiers in Plant Science 2, 8.
Sedat J, Manuelidis L. 1978. A direct approach to the structure of
eukaryotic chromosomes. Cold Spring Harbor Symposium in Quantitative
Biology 42, 331–350.
She W, Grimanelli D, Rutowicz K, Whitehead MW, Puzio M, Kotlinski
M, Jerzmanowski A, Baroux C. 2013. Chromatin reprogramming during
the somatic-to-reproductive cell fate transition in plants. Development 140,
4008–4019.
Shultz RW, Tatineni VM, Hanley-Bowdoin L, Thompson WF. 2007.
Genome-wide analysis of the core DNA replication machinery in the higher
plants Arabidopsis and rice. Plant Physiology 144, 1697–1714.
Swedlow JR, Agard DA, Sedat JW. 1993. Chromosome structure inside
the nucleus. Current Opinion in Cell Biology 5, 412–416.
Tornøe CW, Christensen C, Meldal M. 2002. Peptidotriazoles on solid
phase: [1,2,3]-triazoles by regiospecific copper(i)-catalyzed 1,3-dipolar
cycloadditions of terminal alkynes to azides. Journal of Organic Chemistry
67, 3057–3064.
Tresch S, Schmotz J, Grossmann K. 2011. Probing mode of action in
plant cell cycle by the herbicide endothall, a protein phosphatase inhibitor.
Pesticide Biochemistry and Physiology 99, 86–95.
Van’t Hof J, Kuniyuki A, Bjerknes CA. 1978. The size and number
of replicon families of chromosomal DNA of Arabidopis thaliana.
Chromosoma 68, 269–285.
Vanstraelen M, Baloban M, Da Ines O, Cultrone A, Lammens T,
Boudolf V, Brown SC, De Veylder L, Mergaert P, Kondorosi E.
2009. APC/C-CCS52A complexes control meristem maintenance in the
Arabidopsis root. Proceedings of the National Academy of Sciences, USA
106, 11806–11811.
Waterston RH, Lindblad-Toh K, Birney E, et al. 2002. Initial
sequencing and comparative analysis of the mouse genome. Nature 420,
520–562.
Woodfine K, Fiegler H, Beare DM, Collins JE, McCann OT,
Young BD, Debernardi S, Mott R, Dunham I, Carter NP. 2004.
Replication timing of the human genome. Human Molecular Genetics
13, 191–202.
Xiong Y, McCormack M, Li L, Hall Q, Xiang CB, Sheen J. 2013.
Glucose-TOR signalling reprograms the transcriptome and activates
meristems. Nature 496, 181–187.
Xu DY, Huang WH, Li Y, Wang H, Huang H, Cui XF. 2012. Elongator
complex is critical for cell cycle progression and leaf patterning in
Arabidopsis. The Plant Journal 69, 792–808.
Xu P, Yuan D, Liu M, Li C, Liu Y, Zhang S, Yao N, Yang C. 2013.
AtMMS21, an SMC5/6 complex subunit, is involved in stem cell niche
maintenance and DNA damage responses in Arabidopsis roots. Plant
Physiology 161, 1755–1768.
Rhind N, Gilbert DM. 2013. DNA replication timing. Cold Spring Harbor
Perspectives in Biology 5, a010132.
Zhou W, Wei L, Xu J, et al. 2010. Arabidopsis tyrosylprotein
sulfotransferase acts in the auxin/PLETHORA pathway in regulating
postembryonic maintenance of the root stem cell niche. The Plant Cell 22,
3692–3709.
Rostovtsev VV, Green LG, Fokin VV, Sharpless KB. 2002. A stepwise
huisgen cycloaddition process: copper(I)-catalyzed regioselective ‘ligation’
of azides and terminal alkynes. Angewandte Chemie International Edition
41, 2596–2599.
Zhu Y, Weng MJ, Yang Y, Zhang C, Li Z, Shen WH, Dong AW. 2011.
Arabidopsis homologues of the histone chaperone ASF1 are crucial for
chromatin replication and cell proliferation in plant development. The Plant
Journal 66, 443–455.