Journal of Experimental Botany, Vol. 58, No. 4, pp. 743–756, 2007
Imaging Stress Responses in Plants Special Issue
doi:10.1093/jxb/erl157 Advance Access publication 14 December, 2006
SPECIAL ISSUE PAPER
Intact plant MRI for the study of cell water relations,
membrane permeability, cell-to-cell and long-distance
water transport
Henk Van As*
Laboratory of Biophysics and Wageningen NMR Centre, Wageningen University, Dreijenlaan 3,
6703 HA Wageningen, The Netherlands
Received 15 May 2006; Accepted 17 August 2006
Abstract
Water content and hydraulic conductivity, including
transport within cells, over membranes, cell-to-cell, and
long-distance xylem and phloem transport, are strongly
affected by plant water stress. By being able to measure
these transport processes non-invasely in the intact
plant situation in relation to the plant (cell) water balance,
it will be possible explicitly or implicitly to examine many
aspects of plant function, plant performance, and stress
responses. Nuclear magnetic resonance imaging (MRI)
techniques are now available that allow studying plant
hydraulics on different length scales within intact plants.
The information within MRI images can be manipulated in
such a way that cell compartment size, water membrane
permeability, water cell-to-cell transport, and xylem and
phloem flow hydraulics are obtained in addition to anatomical information. These techniques are non-destructive
and non-invasive and can be used to study the dynamics
of plant water relations and water transport, for example,
as a function of environmental (stress) conditions. An
overview of NMR and MRI methods to measure such
information is presented and hardware solutions for
minimal invasive intact plant MRI are discussed.
Key words: Cell compartments, diffusion, flow conducting
area, hydraulic conductivity, phloem, stress imaging, T2
relaxation, xylem.
Introduction
Long-term growth and crop yield are considerably reduced compared with maximum attainable yield due to
water stress. This topic is turning into an enormous social
and environmental problem due to the potential impacts of
climate change on rainfall patterns and temperature extremes. Several abiotic stresses are united by the fact that
at least part of their detrimental effects on plant performance is caused by disruption of plant water status
(Verslues et al., 2006). Water content and hydraulic conductivity, including transport within cells, over membranes, from cell-to-cell, and long-distance xylem and
phloem transport, is therefore key information in studying
the effect of water stress. On the cell and tissue level, aquaporins, water channel-forming proteins, are of eminent
importance in defining hydraulic conductivity (Verkman,
2000; Javot and Maurel, 2002; Tyerman et al., 2002;
Tournaire-Roux et al., 2003; Lee et al., 2005). In xylem
and phloem the flow conducting area and the resistance
within the vessel or tracheid connections determine
hydraulics (Sperry et al., 2006).
By being able to measure transport processes in relation
to the plant (cell) water balance it will be possible explicitly or implicitly to examine many aspects of plant
function and plant performance. This is key information to
validate biophysical functional-structural plant models
based on integrated carbon and water allocation and functions. Such models are in use to address water stressinduced effects and growth limitations (Daudet et al.,
2002; Tardieu, 2005). In addition, such models are used to
count for the contribution of plant evapotranspiration and
carbon exchange within global atmospheric circulation
models (Sellers et al., 1997).
Transport within intact plants is difficult to measure
because only a few techniques are suitable. In the phloem,
but also in the xylem and the surrounding tissue, complex
* E-mail: [email protected]
ª The Author [2006]. Published by Oxford University Press [on behalf of the Society for Experimental Biology]. All rights reserved.
For Permissions, please e-mail: [email protected]
744 Van As
and fragile gradients in pressure and osmotic potential
exist that are easily disturbed by invasive experimentation
(Verkman, 2000; Steudle, 2001; Koch et al., 2004). On
the cell level, the cell pressure probe has been proven to
be very valuable to measure water (and solute) membrane
permeability, either diffusional (Pd) or under hydrostatic
or osmotic pressure gradients (Pf) (Tomos and Leigh,
1999; Verkman, 2000). On the tissue level, the pressure
bomb and root pressure probe techniques allow water
potential and tissue hydraulics to be measured (Henzler
et al., 1999). These techniques have been applied to excised roots, leaves, and other pieces of a plant and are
clearly very informative but destructive. Hydraulic conductivity has been studied by the use of the high pressure
flow meter (Tyree et al., 1995). That method can only be
applied to excised parts as well.
For several decades heat tracer methods were used to
measure the mass flow in xylem. Up to now they have
been used in several ecophysiological investigations, and
have led to acceptable results if precautions against potential sources of errors are taken (Smith and Allen, 1996). A
general problem is that heat dissipation probes do not
truly integrate velocity along the probe length. Furthermore, the placement of the sensor itself is a source of
errors, particularly for the heat-pulse method (Clearwater
et al., 1999). Calculations of mass flow rates from sap
velocities obtained by heat pulse techniques require a
reliable estimate of sap-conducting surface area. Measurement of the active surface, in general, is very difficult. In
addition, it is known that shrinking and swelling of the
stem occur periodically (Sevanto et al., 2002), and that
day/night rhythm may be accompanied by changes in sapconducting surface area (Windt et al., 2006). All these
factors can result in substantial underestimation of the
actual sap flow.
A very promising and attractive method for providing
detailed, non-invasive, and quantitative information on
water transport and water balance in intact plants is Magnetic Resonance Imaging (MRI). MRI was considered to
include NMR Microscopy or MRM, the high spatial resolution analogue of MRI. MRI is, in essence, spatially
resolved Nuclear Magnetic Resonance (NMR). NMR is
a non-invasive and non-destructive technique that has a
large number of anatomical and physiological applications
to plant cells, plant organs, and living plants (Walter
et al., 1992; Ratcliffe, 1994; Shachar-Hill and Pfeffer,
1996; Chudek and Hunter, 1997; Ishida et al., 2000;
Köckenberger, 2001; Ratcliffe et al., 2001). 1H NMR has
long been used for the characterization of the physical
state of water in plant tissue and for the non-invasive
measurement of molecular displacement such as flow and
diffusion in plants (MacFall and Van As, 1996).
Although well known from (bio-)medical applications,
MRI in plant research is still far from being a routine tool.
Dedicated hardware is required in order to image intact
plants or even trees. MRI systems that are used for plant
studies have often been developed for other purposes,
such as medical imaging. As a result, many machines
have been used with horizontal bore superconducting
magnets with cylindrical geometry of the magnetic field
gradient coils. In such systems plants have to be placed
horizontally instead of vertically. Fitting the shoot or roots
of a plant through the narrow cylindrical bore of a gradient
set can be stressful and damaging for the plant. Better
solutions for minimal invasive studies on intact plants,
with a (potted) root system and extended shoot (leaves),
require special hardware: open access magnets, open access gradient and detector coil systems, and climate
control. Dedicated MRI equipment and methods are now
available to study cell water balance, cell-to-cell, phloem
and xylem transport in (large) potted plants. An overview
of NMR and MRI methods and hardware to measure such
information is presented here.
NMR and MRI basics related to plant (cell)
structure
Some basics of NMR and MRI and the information
available by NMR directly related to water in plant
(cell) structures is first briefly introduced. For more extended descriptions of the NMR technique and MRI strategies see one of the many excellent recent textbooks
(Callaghan, 1993; Levitt, 2001) and the reviews cited in
the Introduction.
Water molecules contain two protons (1H), which are
spin-bearing nuclei. In a strong magnetic field (created by
a magnet) aligning the magnetic moments of such nuclei
results in a weak sample magnetization, which can be
manipulated by applying time-dependent magnetic field
pulses at the proper frequency (radio frequency or rf
pulses). The component of the sample magnetization
perpendicular to the main magnetic field induces a weak
induction voltage in a detector coil placed around the
sample. The time-dependent induction signal can be
analysed into frequency components by Fourier transformation, resulting in a frequency spectrum. In a homogeneous main magnetic field B0, equal spins (e.g. protons
of the water molecules) have identical Larmor precession
frequency or resonance frequency, and a single resonance
line in the frequency spectrum is observed. When a welldefined constant magnetic field gradient G, G¼@B0/@r,
is created within the magnet, identical spins at different
positions along this gradient have different resonance
frequencies, because the resonance frequency is proportional to the local magnetic field experienced by the spins.
G can be created in three independent directions x, y, or z,
or combinations thereof. In this way spins can be uniquely
spatially encoded. This is the basis for NMR imaging.
Position labelling by magnetic field gradients can be
performed in a variety of ways (Callaghan, 1993).
MRI and water transport in plants 745
Depending on the actual method used, the process of
position labelling will take some time and acquisition of
the signal occurs at a certain time TE (echo-time) after the
excitation of the spin system. During that time, the
observed signal will decay according to the transverse or
T2 relaxation process: S(TE)¼A0 exp(–TE/T2). A0 is the
signal amplitude directly after excitation, and is a direct
measure of the amount of spins under observation in the
detector coil. In imaging it is primarily a function of the
spin density. In order to obtain a full two-dimensional
image of N3N pixels, the sequence has to be repeated N
times for position encoding in the second direction. If the
repetition time TR is long enough, the spin system has
restored equilibrium along the magnetic field direction:
TR >3T1, where T1 is the signal decay time in the longitudinal or main magnetic field direction. If TR <3T1, the
effective signal amplitude, Aeff, does not represent the spin
density in each pixel uniquely, but depends on a combination of the spin density and the relaxation time T1: Aeff¼
A0(1–exp(–TR/T1)). As a result, NMR image intensity
usually depends on a combination of these parameters,
reflecting spin density, T1, and T2. In addition to these
parameters, diffusion behaviour of the molecules can also
contribute to the contrast (see below) (Duce et al., 1992;
Callaghan et al., 1994a; Xia, 1995; Edzes et al., 1998).
Methods are available by means of which quantitative
images are obtained that represent each of these parameters separately. Multiple Spin-Echo (MSE) MRI (Edzes
et al., 1998) and Inversion recovery (IR) MSE MRI
(Donker et al., 1996) are examples of such sequences. In
MSE a series of images is obtained during the decay of
the signal after the first excitation rf pulse. Single parameter images can now be processed from the MSEexperiment by assuming a mono-exponential relaxation
decay as a function of n3TE in each picture element or
pixel. Here n is the image number acquired during the
signal decay. The resulting images are: signal amplitude
(A0), T2, and T1 (the latter in the case of Inversion
Recovery (IR-)MSE MRI). In addition, images representing the diffusion coefficient, D, and flow can be obtained
(see below). In this way a combination of anatomical and
physiological (functional) information is obtained, referred
to as functional imaging. Some examples of single parameter images are presented in Fig. 1.
NMR relaxation times and compartments
A variety of interactions between the magnetic moments
of the observed spins and the surrounding nuclei and
Fig. 1. (A) Single parameter images of a 3 mm slice through the trunk of a poplar tree (1.8 m total length), obtained at 0.7 T. These images have
been calculated based on a mono-exponential fit per pixel of the signal decay as a function of decay time n3TE. a: Signal amplitude or spin density
image, b: T21 image, c: T2 image. (B) Single parameter images from a slice of the stem of a Ricinus pant. a: Signal amplitude or spin density image,
b: T1 image, c: T2 image. In-plane resolution is around 1003100 lm2. Both examples demonstrate the sensitivity of T2 to tissue structure or cell size.
Amplitude images show water content times tissue density. Images by courtesy of C Windt.
746 Van As
electrons contribute to the relaxation times T1 and T2.
These interactions make it possible to probe the physicochemical properties of the spin environment using NMR
relaxation measurements. The protons in water molecules
experience an intramolecular dipolar interaction between
the two proton spins within one and the same water
molecule, as well as an intermolecular interaction with
protons of neighbouring water molecules. Both interactions fluctuate when the molecules rotate or translate.
When the rotation correlation time of the molecules is
short, as is the case for free water molecules (sc~1012 s),
both T1 and T2 are equal and relatively long (~2 s). Water
close to macromolecules or to solid surfaces generally
have slower tumbling rates (sc~1012–1010 s), which
leads to a reduction in both relaxation times. In addition,
exchange of protons between water and other molecules,
such as sugars, proteins, and other macro-molecules, also
influences (shortens) the relaxation times (Hills and Duce,
1990).
The signal from water in plant tissue, containing among
others vacuoles, cytoplasm, cell walls, and extracellular
spaces, decays with different relaxation times. In biological systems multi-exponential relaxation is therefore
normally observed and the different decay times (relaxation times) observed could be used to obtain information
on the relative proportions of water in different environments or compartments (Belton and Ratcliffe, 1985). In
plant tissue the different relaxation times can be assigned
more or less uniquely to either water in the vacuole
(longest T1 and T2), cytoplasm (T1 >T2, both shorter than
vacuolar T1 and T2), or cell wall/extracellular space (T2
depends strongly on the water content in this compartment, and ranges from about 5 ms up to hundreds of ms),
respectively (Snaar and Van As, 1992; Van Dusschoten
et al., 1995). In leaves, water in chloroplasts can be discriminated as well (McCain, 1995). Within these compartments diffusive exchange results in single exponential
behaviour.
Proton exchange over the plasmalemma and the
tonoplast membrane affects the observed relaxation times.
Due to the effect of exchange, which depends on the
difference in the relaxation times of water in the exchanging compartments, T1 and T2 results are, in general,
different, even for the number of observed exponentials.
Differences in T1 for the different compartments are
relatively small, and exchange between the compartments
results in averaging over the compartments. Therefore T1
values relate to water content more directly. Differences in
T2 are more pronounced, and exchange over membranes
only results in partial averaging (depending on the size of
the compartments and membrane water permeability). Cell
compartments, therefore, can best be discriminated based
on T2 values (Snaar and Van As, 1992; Van Dusschoten
et al., 1995).
Because of the relatively poor spatial resolution,
most pixels within an image will contain information
that originates from different subcellular compartments
or even different cells. This is called the partial
volume problem. In general, the low S/N per pixel in
imaging (typically in the order of 10–50) does not allow
multi-exponential fitting, which might provide a way to
resolve subcellular information. Such information is
available from non-spatially resolved NMR measurements, with a much higher S/N (1000 or higher).
Therefore the information presented in images mostly is
called ‘apparent’: e.g. apparent T2, T2,app, or apparent D,
Dapp. In images, the S/N can be enhanced by summing
up the signal of pixels containing identical information
(same tissue). In this way subcellular information of that
tissue can be obtained (Scheenen et al., 2002).
T2, cell size, and (tonoplast) membrane
permeability
Membrane permeability in plant cells (as well as in many
other systems, for example, red blood cells (Regan and
Kuchel, 2000, 2002)) has been determined using the
NMR relaxation times of intracellular water protons based
on the Conlon–Outhred technique (Conlon and Outhred,
1972; Stout et al., 1978; Ratković and Bačić, 1980; Snaar
and Van As, 1992; Zhang and Jones, 1996; Donker et al.,
1999). The disadvantage of this technique is that it needs
the introduction of paramagnetic ions (such as Mn2+,
which can pass membranes, or other MRI contrast agents
that cannot pass membranes) in high, non-physiological
concentrations. A non-invasive approach is based on the
principle that the observed transverse relaxation time T2 of
water in a confined compartment such as a vacuole can be
described as a function of the bulk T2, T2,bulk, of the water
and the probability that water molecules reach the membrane and lose magnetization at the membrane, either
by a direct interaction with the membrane (acting as a sink
for relaxation) or by passing the membrane and entering
a compartment with a (much) shorter relaxation time
(Brownstein and Tarr, 1979; van der Weerd et al., 2001).
The probability to reach the membrane is defined by the
diffusion time and thus directly related to the compartment radii. No evidence has been found that membranes
themselves act as a relaxation sink (McCain, 1995; van
der Weerd et al., 2002a). The net loss of magnetization
in a vacuole therefore depends on the membrane water
permeability of the tonoplast and the effective relaxation
in the cytoplasm (van der Weerd et al., 2002b; L van der
Weerd, JEM Snaar, FJ Vergelt, H Van As, unpublished
data). As a result the observed relaxation time depends,
in addition to T2,bulk, to the radii of the compartment
along the x, y, and z directions (Rx,y,z) and the net
loss of magnetism at the compartment boundary, the
MRI and water transport in plants 747
so-called magnetization sink strength (H) (van der Weerd
et al., 2001):
1
T2;obs
¼
H
1
þ
ðRx þ Ry þ Rz Þ T2;bulk
ð1Þ
H is linearly related to the actual membrane permeability (van der Weerd et al., 2002b; L van der Weerd, JEM
Snaar, FJ Vergelt, H Van As, unpublished data). A simple
common-sense approach is sufficient to explain the effect
of compartment properties on relaxation. The intermembrane distances (R) and the bulk diffusion coefficients (D)
determine the average diffusion time of a water molecule
to cross a compartment (tdif¼R2/(2D)). The relaxation rate
1
T2;bulk
in that compartment determines the chance that the
molecule still bears magnetization once it reaches the
1
membrane. If that is the case (2D/R2)< T2;bulk
, the membrane permeability determines the exchange rate to the
next compartment, and thereby influences the relaxation
rate; if the above condition does not apply, the membrane
permeability is of no consequence for the NMR signal and
equation 1 does not hold.
An example of this relation based on MRI results is
shown in Fig. 2 for cells in the apex zone of the stem of
intact maize and pearl millet plants. In intact pearl millet
plants, H in cells in this zone has been shown to change
during osmotic stress experiments, in contrast to the same
cells in maize plants where no changes were observed
(van der Weerd et al., 2001, 2002a), pointing to a response
to stress most probably related to changes in aquaporin
functioning.
Some care must be taken into account by using T2,obs
from images. The observed T2 value in images with
respect to its value in non-imaging NMR depends on
Fig. 2. Relation between the relaxation time T2 as observed by MRI
and the cell dimensions for maize (filled triangles) and pearl millet
(filled diamonds) cells of different internodes of the apex zone of the
stem. Following the MRI measurements on the intact plants, the same
plants were used for microscopic sections to determine the cell
dimensions. After van der Weerd et al., 2001 (van der Weerd L,
Claessens MMAE, Efdé C, Van As H. 2002. Nuclear Magnetic
Resonance imaging of membrane permeability changes in plants during
osmotic stress. Plant, Cell and Environment 25, 1538–1549) and
reproduced by kind permission of Blackwell Publishing.
a number of contributions that relates to details of the
image experiment and plant tissue characteristics (Rofe
et al., 1995; Edzes et al., 1998):
1
T2;obs
¼
1
1
1
þ i þ ii
T2 T2 T2
ð2Þ
Here T2 represents the original T2 of the liquid in the
tissue, which is also observed in non-imaging mode experiments. T2i and T2ii originate from diffusion effects.
In plants, especially in leaves and woody tissue, small
air spaces in the order of a few lm up to 100 lm are
present. Due to the difference in the magnetic permeability of air and tissue/water, local magnetic field gradients,
gz, originate. Displacement through these local gradients
due to diffusion results in a reduction of the observed
T2 value:
1
}c2 Æg2z æTE2 D
T2i
ð3Þ
Here c is the gyromagnetic ratio, which is a constant for
each type of nuclear spin. D is the (self-)diffusion coefficient of water in the tissue. The strength of these field
inhomogeneities is proportional to the applied magnetic
field (Lüdecke et al., 1985). Minimizing this effect can be
achieved by imaging at relatively low magnetic field
strength, B0, and at short TE values (Donker et al., 1996,
1997, 1999). In non-spatially resolved T2 measurements
this TE value can be chosen to be very short (in the order
of 200–400 ls) whereas in imaging mode TE is in the
order of a few ms or longer.
In addition to the effect of local field gradients,
diffusion through the repeatedly applied position encoding
magnetic field gradients contributes to the observed
relaxation time:
1
4d
2 2 2
D
ð4Þ
¼c G d 1
3TE
T2ii
Here 2d is the duration of the applied gradient. In
practice, G and d are dictated by the choice of pixel
resolution, image size, and actual TE. To image relaxation
times in the order of 1 s, as a rule of thumb a lower limit
of 100 lm for the pixel size for objects of 1–2 cm
diameter can be used (Edzes et al., 1998). At smaller pixel
sizes the diffusive attenuation in the position-encoding
gradient becomes the leading contribution to the observed
transverse relaxation time, and the information available
from the actual T2 is lost and mixed up with (restricted)
diffusion information and equation 1 no longer holds.
From equation 1 it is clear that, for a proper interpretation of T2 measurements in terms of membrane
permeability, it is necessary to know the cell dimensions,
or more precisely the dimensions of the vacuole. This
information can be obtained by NMR as well by
748 Van As
(restricted) diffusion measurements, which, in addition,
gives access to membrane permeability of water over
longer distances, resulting in cell-to-cell transport.
Diffusion, restricted diffusion, cell (compartment)
dimension and cell-to-cell transport
All molecules in a fluid are subject to Brownian motion.
The extent of this motion depends on the temperature and
the viscosity of the fluid, which are incorporated in the
bulk diffusion coefficient of the fluid. When an ensemble
of molecules is followed in time, the root mean square
distance travelled, r, increases with time as long as no
boundaries are encountered, according to the Einstein
relation:
pffiffiffiffiffiffiffiffi
r ¼ 2Dt
ð5Þ
D can be measured by NMR using a so-called pulsed field
gradient (PFG) experiment. In this experiment a sequence
of two magnetic field gradient pulses of duration d and
equal magnitude G but opposite sign (or equal sign but
separated by an inverting rf pulse) label the protons as
a function of their position. If the spins remain at exactly
the same position the effect of the gradient pulses compensate each other. However, as soon as translational
(displacement) motion occurs, the gradients do not exactly
compensate each other any more, resulting in attenuation
of the signal amplitude. The amount of this attenuation is
determined by d and G of the gradient pulses, and by the
mean translational distance travelled during the interval D
between the two pulses. The NMR-signal amplitude S(G)
normalized to the signal amplitude S(0) at G¼0 for free
diffusion at a given D is given by (Stejskal and Tanner,
1965; Norris, 2001):
SðG; DÞ
d
2 2 2
¼ exp c d G D D ð6Þ
Sð0; DÞ
3
By measuring S(G, D) as a function of G, D can be
obtained. The distance travelled depends on the bulk
diffusion coefficient of the fluid in the compartment and D
(equation 5 with t¼D). If water experiences a barrier to
diffusion, for example, a cell membrane, the cell dimensions determine the maximum displacement (in case of an
impermeable barrier) or to what extent it is hindered (semipermeable membrane). For restricted diffusion equation 6
no longer holds and becomes dependent on the geometry
and size of the compartment and the permeability of the
surrounding membrane (Callaghan et al., 1999; Norris,
2001; van der Weerd et al., 2002b; Regan and Kuchel,
2000, 2002).
To extract the information on size and membrane permeability, diffusion coefficients have to be measured as
a function of the diffusion labelling time, or observation
time, D, the time between the two gradient pulses (for
a tutorial on this subject see Sen, 2004). By varying D the
distance over which the spins can diffuse freely and to
what extend they can pass over the membrane, which restricts the diffusion process, are observed. The probability
to pass the membrane is a direct measure of the water
permeability of the surrounding membrane (Regan and
Kuchel, 2000). At longer values of D, water molecules
can even travel from cell to cell. An example of
a theoretical curve of D versus D for diffusion in a confined geometry is presented in Fig. 3 for a number of
different values of the water membrane permeability. In
the first region, at short diffusion times, free diffusion is
observed. In the next region the diffusion becomes restricted, but the averaging of local properties over a large
enough distance does not yet occur. In the last region, at
long diffusion times, hindered diffusion is observed,
which is defined by the permeability of the system, and
for plants reflects cell-to-cell transport. The effective permeability P can now be estimated. For diffusion through
a geometry consisting of a series of semi-permeable membranes (thin walls) separated by a distance d, Crick (1970)
obtained:
1
1
D1
inf ¼ D0 þ ðPdÞ
ð7Þ
or
P¼
ðDinf D0 Þ
ðD0 Dinf Þd
ð8Þ
Dinf is the D value in the limit of D to infinity. If D0,
Dinf, and d are known, P can be obtained. This is easily
done for D(D) information as presented in Fig. 3, where
Dinf is reached. In practice, D(D) is obtained over a smaller
Fig. 3. Simulated behaviour of D(D) as a function of D in a system
with a series of compartments separated by semi-permeable membranes
(thin walls) at a distance of 10 lm, for different values of the
permeability coefficient P (in m s1) of the membranes. For P¼0 fully
restricted diffusion is observed. At increasing P values Dinf becomes
higher. Plot by courtesy of T Sibgatullin.
MRI and water transport in plants 749
range of D values and a procedure is needed to extract this
information based on a limited range of D(D) values.
In the limit of short D, D depends linearly on the square
root of the diffusion time. The slope of this dependence is
determined by the surface-to-volume ratio (S/V) of the
water-containing compartment, irrespective of whether
these compartments are connected or disconnected (Sen,
2004), according to the equation
DðDÞ
S 4 pffiffiffiffiffiffiffiffiffi
pffiffiffi D0 D
¼1
D0
V9 p
ð9Þ
For a sphere S/V¼3/R, where R is the radius of the
sphere.
In the limit of long diffusion time (D >>R2/D0) and fully
restricted diffusion (impermeable wall) the apparent
diffusion coefficient varies inversely with D according to
Einstein’s equation
DðDÞ ¼
R2
2D
ð10Þ
R can thus be obtained from the slope, the maximum
value of r to be travelled in the confinement. However,
semi-permeable membranes will result in deviations from
this dependence (Fig. 3). If semi-permeable membranes
are present D versus D curves can be described by the
following equation (Anisimov et al., 1998; Valiullin and
Skirda, 2001)
D0 Dinf
DðDÞ ¼ Deff1 ðDÞ
þ Dinf
ð11Þ
Deff1 ðDÞ þ D0
where
Deff1 ðDÞ ¼
D0 Deff2 ðDÞ
D0 Deff2 ðDÞ
ð12Þ
and
DðDÞ Dinf
Deff2 ðDÞ ¼
D0 Dinf
ð13Þ
An example of experimental data is given in Fig. 4.
This figure also illustrates that the experimental range of D
values is limited by the relaxation times T2 and T1 and
Dinf is hard to obtain experimentally. Deff2(D) (equation
13) behaves in much the same way as D(D) for the case of
a geometry enclosed by an impermeable membrane (cf.
equation 10). The effect of the permeable wall is removed
and it results in a good estimate of D0 at short D values
(Fig. 4). However, due to the limited range of td the
dependence D(D)~D1 is not yet observed (Fig. 4, circles). Rescaling according to equation 12 (Deff1) enlarges
the range of diffusion time (to the shorter D values) where
the dependence according to equation 10 is observed. It is
very useful when the long diffusion times are unavailable
Fig. 4. Experimental D(D) as a function of D as observed in root
segments of maize plants. (Filled squares with open centre): Experimental D values; filled triangles, Deff1 (equation 12), from which R is
obtained (dotted line); filled circles, Deff2 (equation 13), resulting in D0.
A fit to the experimental data according to equation 11 (solid line)
results in Dinf, in addition to D0 and R. This information is used to
calculate P (equation 8). Plot by courtesy of T Sibgatullin.
experimentally. As a result parameter R can be determined
(cf. equation 10), and can be used to fit the experimental
data D(D) versus D according to equation 11 resulting in
D0 and Dinf. Alternatively, D0 can be estimated from
equation 9 at short D values, if experimentally available.
P can now be calculated from D0, Dinf, and R (¼d, see
equation 8).
P will include the permeability of the tonoplast,
plasmalemma, walls and plasmodesmata of neighbouring
cells. At long observation times cell-to-cell transport becomes visible. Therefore P is not identical to H as
obtained from T2 measurements (equation 1). The value
of d as a result of the diffusion measurements can directly
be used to obtain H from the T2 measurements. A complication is the effect of the geometry of the confined compartment. Equation 8 assumes plan-parallel membranes,
which might be a good assumption for an array of cells in
one dimension (the magnetic field gradient direction). The
result of H from equation 1 depends on the actual
geometry of the compartment: for a sphere, the first term
in equation 1 becomes 3H/R, for a cylindrical geometry it
becomes 2H/R. This problem can be overcome by
measuring D in different directions, which is possible by
applying magnetic field gradients in different directions.
Some first results are now available. By combining the
results of T2 and D measurements on water in apple
parenchyma (Granny Smith) H was found to be around
13105 m s1 (T2,obs¼1.25 s and R¼86 lm). Under the
assumption of parallel planes P¼2.93106 m s1 (TA
Sibgatullin, PA de Jager, FJ Vergeldt, AV Anisimov,
E Gerkema, H Van As, unpublished results). For Cox
apple parenchyma cells a tonoplast water membrane
permeability of Pd¼2.443105 m s1 was reported (Snaar
and Van As, 1992) based on the Conlon–Outhred method.
750 Van As
In maize roots, Anisimov et al. (1998) found a higher
value of P: around 53105 m s1. Recently values were
obtained of P in excised roots of normal and osmotically
stressed maize and pearl millet plants: P was found to be
around 33105 m s1 for both normal and stressed maize,
whereas P was around 93105 m s1 for normal pearl
millet plants and 33105 m s1 for stressed plants (TA
Sibgatullin and H Van As, unpublished results). Ionenko
et al. (2006) recently reported P values for water in roots
of maize seedlings of 33105 m s1, which decreased by
a factor of 1.7 due to water stress or HgCl2 treatment. The
latter clearly demonstrates the contribution of aquaporin
functioning in P. Up to now H was found to be higher
than P, as expected.
By time-dependent diffusion coefficient measurements
combined with MRI the size and the membrane permeability of, for example, the vacuoles in vacuolated plant
tissue have been measured, even spatially resolved within
single pixels of an image (TA Sibgatullin, FJ Vergeldt,
E Gerkema, H Van As, unpublished results).
As stated above, in imaging, Dapp is normally observed.
In tissue with (large) vacuolated cells Dapp will mainly
reflect D of vacuolar water. In other tissue it can become
more complex (van der Toorn et al., 2000). Correlated
diffusion-T2 measurements have been developed that
allow unambiguously to relate D and T2 values of
different water-containing cell compartments, even within
single pixels in images.
Correlated diffusion-T2 measurements
Water in different cell compartments can best be discriminated on the basis of the differences in relaxation
behaviour (T2) and (restricted) diffusion behaviour. By
combined relaxation and diffusion measurements, together
with a recently developed efficient and stable twodimensional fitting procedure based on a Fast Laplace
Inversion algorithm (Venkataramanan et al., 2002;
Hürlimann et al., 2002), two-dimensional correlation plots
between D and T2 can now be generated, which greatly
enhance the discrimination of different water pools in
subcellular compartments. In this way, an unambiguous
correlation between relaxation time and compartment
size can be obtained, resulting in a general approach to
quantify water in the different cell compartments. This
approach is very promising in non-spatially resolved measurements (Qiao et al., 2005), and has been shown to
obtain sub-pixel information in images as well (Van
Dusschoten et al., 1996; for recent plant applications H
Van As, FJ Vergeldt, CW Windt, unpublished data).
Flow, xylem and phloem hydraulics
The method that has been most successful in providing
detailed, non-invasive information on the characteristics of
water transport in the xylem, as well as in the phloem of
intact plants, is MRI [for overviews see MacFall and Van
As (1996) and Köckenberger (2001)]. Both non-imaging
and imaging methods have been developed and applied to
plants. Several groups have used different MRI methods.
Most of them are based on (modified) Pulsed Field
Gradient (PFG) methods, either by using a limited number
of PFG steps or by (difference) propagator approaches
(Callaghan et al., 1994b, see below). Also flow measurements based on uptake and transport of (paramagnetic)
tracers have been used (Link and Seelig, 1990; Clearwater
and Clark, 2003).
The first (non-imaging) method to measure xylem water
transport in plants was presented some 20 years ago (Van
As and Schaafsma, 1984; Reinders et al., 1988a, b;
Schaafsma et al., 1992). Based on that method a (trans)
portable NMR bioflowmeter has been developed with
a U-shaped permanent magnet (open access from one
side) and an openable, hinged, rf coil. It has been used in
greenhouse situations on intact plants (Van As et al.,
1994). By the applied method averaged linear flow
velocity and flux (volume flow) are obtained, and the ratio
results in the effective flow conducting area. However, to
interpret the data in terms of averaged flow velocity,
calibration is needed: the results depend on the actual flow
profile. Flow in a single xylem vessel, in general, is
assumed to be laminar, but velocities within the different
xylem vessels in a stem will be different, depending on
vessel diameter. So the flow profile within the total crosssection of a stem is therefore not known a priori. In
addition, xylem and phloem flow can not be discriminated
by that method, the sum of both is observed.
To quantify xylem and phloem flow accurately, one
needs to determine both the direction of the flow, and the
actual flow profile, from which the flow velocity, the flux,
and the flow conducting area can be obtained. At the same
time, diffusing water molecules (stationary water in cells)
and flow have to be discriminated. These goals can best
be obtained by the use of PFG techniques. To quantify the
unknown displacement-behaviour of an observed ensemble of spins correctly, one has to measure the NMR-signal
S(G) as a function of G. By applying a Fourier Transform
on S(G) as a function of G, the propagator P(R, D) is
obtained (cf. Fig. 5). P(R, D) presents the probability that
a spin at any initial position is displaced by a distance R in
time D. For flow, dividing the displacement axis R by D
results in the flow profile P(v). This type of measurement
is referred to as displacement or propagator imaging and
has been applied to measure flow in many porous systems,
including plants (for some introductions and reviews see
Callaghan et al., 1999; Fukushima, 1999; Mantle and
Sedesman, 2003; Stapf and Han, 2005).
The propagator for free, unhindered, diffusing water
has a Gaussian shape, centred around R¼0 (Fig. 5). The
root mean square displacement, r, of diffusing protons,
MRI and water transport in plants 751
Fig. 5. Simulated propagators for a system consisting of 75% stationary water and 25% flowing water, with a laminar flow profile. D is 2.23109 m2
s1. The mean linear average flow velocity is 0.2 mm s1 (typical for phloem). (a) D¼15 ms (r¼8.1 lm, r¼3.0 lm); (b) D¼100 ms (r¼21 lm,
r¼20 lm); (c) D¼1000 ms (r¼66 lm, r¼200 lm). The arrow indicates the signal of the flowing water, that emerges from the Gaussian shaped
signal centred on R¼0 of diffusing water. Reprinted from Scheenen TWJ, Vergeldt FJ, Windt CW, de Jager PA, Van As H. Microscopic imaging of
slow flow and diffusion: a pulsed field gradient stimulated echo sequence combined with turbo spin echo imaging. Journal of Magnetic Resonance
151, 94–100, copyright 2001, with kind permission from Elsevier.
observed by NMR, is proportional to the root of D times
D (equation 5). r is directly related to the width of the
Gaussian distribution (Fig. 5). By contrast, the mean
displacement r of flowing protons is linearly proportional
to D itself:
r ¼ mav D
ð14Þ
where vav is the average flow velocity of the flowing
protons. The labelling time between the two PFGs has to
be long (in the order of 150 ms or longer) to discriminate
between slow (phloem) flow and diffusion (D in the order
of 23109 m2 s1) (Scheenen et al., 2001). At lower
velocities D has to be increased further (Fig. 5).
A crucial step in the quantification of the flow is to
discriminate stationary and flowing water. The fact that
the propagator for stationary water is symmetrical around
zero is used to separate the stationary from the flowing
water. The signal in the non-flow direction is mirrored
around zero displacement and subtracted from the signal
in the flow direction, to produce the displacement distribution of the flowing and the stationary water. Because
the signal amplitude is proportional to the density of the
mobile protons, the integral of the propagator provides
a measure for the amount of water. The average velocity
of the flowing water is then calculated by taking the
amplitude weighted average of the velocity distribution.
The volume flow rate or flux is calculated by taking the
integral of amplitude times displacement of the velocity
distribution. Using this approach, for each pixel in the
image the following flow characteristics are extracted in
a model-free fashion, as described by Scheenen et al.
(2000): total amount of water, amount of stationary water,
amount of flowing water (or flow conducting area),
average linear velocity (including the direction of flow),
and volume flow per pixel. A propagator flow imaging
method was developed that allowed the flow profile of
every pixel in an image to be recorded quantitatively, with
a relatively high spatial resolution, while keeping measurement times down to 15–30 min (Scheenen et al., 2000,
2001, 2002).
Phloem transport in plants is particularly difficult to
measure quantitatively. The slow flow velocities and the
very small flowing volumes in the presence of large
amounts of stationary water make it difficult to distinguish
the slowly flowing phloem sap from freely diffusing
water. Windt et al. (2006) further optimized the propagatorfast imaging method to measure quantitatively, for the
first time, detailed flow profiles of phloem flow in large
and fully developed plants. An example of xylem and
phloem flow imaging in stem of a Ricinus plant is shown
in Fig. 6. In this way the dynamics in phloem and xylem
flow and flow conducting area were studied. The observed
differences for day and night in flow conducting area,
which directly relate to xylem and phloem hydraulics, are
one of the most striking observations (Windt et al., 2006),
which demonstrates the potential of the method to study
hydraulics in intact plants under normal and stress conditions. In addition, the phloem-to-xylem flux ratio reflects
the fraction of xylem water that is used for phloem
transport, also known as Münch’s counter flow. This ratio
was surprisingly large at night (hardly any evaporation)
for poplar (0.19), castor bean (0.37), and tobacco (0.55),
but low in tomato (0.04) (Windt et al., 2006).
Hardware considerations
In many cases dedicated hardware is required in order to
image intact plants. Normally, only small parts of the
plant (e.g. stem, a leaf, petioles, seed pots, fruit stalk, etc)
will be chosen for study. If so, an optimal signal-to-noise
ratio, S/N, is obtained by optimizing the radius of the rf
coil, r, with respect to that part of the plant to be measured. The smaller r, the higher S/N. The best approach is
to construct rf detector coils that closely fit that part of the
752 Van As
Fig. 6. Quantitative NMR flow images of xylem water moving upwards through the hypocotyl of Ricinus communis (a)–(d), and phloem water
moving downwards to the roots (e)–(h). Shown are the volume of stationary water per pixel (a, e), the flow conducting area per pixel (b, f), the
average linear velocity per pixel (c, g), and the average volume flow per pixel (d, h). The xylem flow images were calculated from a single flow
imaging measurement; the phloem flow images were constructed from seven consecutive individual flow imaging measurements. The differences in
the amounts of water per pixel shown in (a) and (e) are due to differences in image matrix size (1283128 versus 64364), slice thickness (3 mm
versus 6 mm) and scaling. After Peuke et al., 2006 (Peuke AD, Windt C, Van As H. 2006. Effects of cold-girdling on flows in the transport phloem
in Ricinus communis: is mass flow inhibited? Plant, Cell and Environment 25, 15–25) and reproduced by kind permission of Blackwell Publishing.
plant or tree to be imaged (Scheenen et al., 2002, Windt
et al., 2006).
Also, when the distance between the magnetic field
gradient coils is made smaller, higher gradient strengths
can be obtained. This is important to obtain a high spatial
resolution, fast switching of gradients, and to measure low
flow velocities: small displacements require a high G
value to be detected.
Another way to increase S/N is to use higher field
strength, B0. However, for plant tissues with extracellular
air spaces, this results in increased susceptibility artefacts,
as discussed above. These artefacts can be overcome by
increasing maximum imaging gradients, but this would
result in a decrease in S/N and loss of T2 information (see
equations 3 and 4). At higher B0 the effective (and bulk)
T2 is (much) shorter than at lower field strength, limiting
the number of measurable images during the signal decay,
and resulting in a lower sensitivity to extract the permeability information form T2 measurements (equation 1).
Some hardware solutions for intact plant NMR and
MRI are presented in Figs 7 and 8. In Fig. 7a and b a low
field (0.7 T) imaging system based on an open access
electromagnet is shown. A comparable (permanent)
magnet has been used by Utsuzawa et al. (2005) to study
xylem cavitation due to pine wilt disease. On top of the
magnet a climate chamber for environmental control is
placed (Fig. 7b). Maximal access into the centre of the
magnet is obtained by use of plan parallel gradient plates
Fig. 7. Magnet (a) and climate control unit on top of magnet (b) of
a 0.7 T imager based on an electromagnet. The two plan parallel
gradient coils plates on top of the poles of the magnet (one is seen as
a dark plate in the centre of the magnet, see arrow) results in maximum
access to the centre of the magnet (a).
(Fig. 7a). To maximize the filling factor a (solenoid) rf
coil is wrapped directly around the stem of the plants. A
solenoid coil results in about a factor 2 to 3 more
sensitivity than Helmholtz or birdcage coils of the same
diameter. This configuration allowed large plants, up to a
size of two metres, to be placed upright easily in the NMR
MRI and water transport in plants 753
Fig. 8. (a) 50 cm diameter vertical free bore of the superconducting-magnet of an intact plant 3 T MRI system. Within this bore air temperature,
humidity and gas composition and light intensity can be varied and controlled. (b, c) Gradient coil system consisting of four half-cylindrical parts that
can be opened from one side (b), allowing the insertion of a plant. The closed gradient system containing the plant and placed on top of a support
system (c) has to be inserted in the 50 cm bore of the magnet (a).
magnet. Placement of plants can be undertaken without
causing much stress to the subject, other than the stress
that is caused by moving and handling it. Potentially
stressful actions like mounting an rf coil or, when necessary, the removal of a branch or leaf, can be undertaken
well in advance.
In Fig. 8 part of a dedicated intact plant 3 T MRI system is shown. The superconducting magnet has a 50 cm
vertical free bore (Fig. 8a). One of the gradient coils is
able to be opened. It consists of four parts and can easily
be mounted around the stem of a plant or trunk of a tree
(Fig. 8b, c). An rf coil (4 cm inner diameter) that consists
of two parts completes the fully openable construction.
Inside the bore of the magnet the climate is controlled by
use of a remote climate control unit. This 3 T MRI system
provides an excellent and unique infrastructure for MRI
on intact plants and trees.
Future of functional plant MRI
The combination of dedicated NMR/MRI equipment and
T2, diffusion and propagator flow MRI methods is now
available to measure and quantify cell, tissue, phloem and
xylem hydraulics routinely in intact plants up to a size of
several metres and over periods of weeks. Plant responses
in hydraulics and water content of the different tissues can
now be studied as a function of changes in environmental
conditions. The method has already been applied to study
day–night rhythms in flow and flow conducting area in the
stem of large potted plants (Windt et al., 2006), dynamics
in phloem in response to stem cold girdling (Peuke et al.,
2006), root cooling (inducing xylem air embolism, follow
refilling and functionality of xylem; Scheenen, 2001), root
anoxia and long dark periods, and the flow (and changes
therein) in the stalk of a tomato truss during a 5-week
period of fruit development (CW Windt et al., unpublished results). Flow MRI has been proven to be
quantitative in terms of volume flow. The additional
information on effective flow conductive area is totally
new. First results demonstrate surprising, unexpected dynamics in this parameter (Windt et al., 2006) which may
result in a better understanding of the mechanism and
regulation of long-distance transport. This information is
very difficult to measure, or cannot be measured at all, in
intact plants using other techniques. Flow MRI can be of
help in understanding the results obtained by heat pulse
methods, which are routinely used in field situations. First
comparisons between flow MRI and (modified) heat pulse
methods have very recently been made in our laboratory.
Clearly, flow MRI teaches us some striking lessons about
sap flow of importance for the interpretation of other
methods, improving their reliability.
By contrast to the measurement of long-distance transport (xylem and phloem), MRI methods to study hydraulics on the membrane and tissue level based on diffusion
measurements have not yet been demonstrated to be quantitative. First results indicate that MRI can at least be used
to monitor changes in hydraulics within a single plant and
that these MRI methods are sensitive for aquaporin
functioning. All tissues, independent of position, are accessible by MRI, in contrast to, for example, the pressure
probe technique. At the moment it has to be demonstrated
that both Pd and Pf can be obtained by MRI. In the next
step it would be necessary to corroborate MRI results by
results obtained by more established and quantitative
methods for hydraulic conductivity, for example, by highpressure flow meter or by cell pressure probe. For this
purpose, the development of MRI image guided cell
pressure probe measurements will be of great interest, as
well as combining pressure bomb or high pressure flow
meter and MRI.
Functional intact plant imaging by MRI offers exciting
new possibilities for the non-invasive physiological
mapping of intact plants. Functional MRI is expected to
754 Van As
become a powerful tool to characterize functionality of
aquaporins to be studied. In this way it will contribute to
resolve the role of aquaporins in plant water balance,
hydraulics and stress tolerance (Verkman, 2000; Javot and
Maurel, 2002; Tyerman et al., 2002; Tournaire-Roux
et al., 2003; Lee et al., 2005). Over-expression plants,
knock-out plants or anti-sense plants show different
aquaporin gene expression patterns compared with wildtype plants and may be of great help in understanding the
relationship between aquaporin gene expression and
function. NMR imaging can elucidate in vivo gene functionality in tissue hydraulics at different levels and thereby
bridge the gap between gene expression and function.
Currently MRI in plant science is still far from being
a routine tool. Some reasons for this could be the high costs
of MRI equipment, the difficult theoretical fundamentals
on which the technique is based, the horizontal orientation
of most of the standard imaging set-ups, the limited accessibility (for large plants) of the magnetic field’s iso-centre
of most set-ups, and the specific (climatic) requirements
which have to be met when measuring intact plants. It is to
be expected, however, that relatively cheap imaging setups based on permanent magnet systems will soon become
available (Rokitta et al., 2000; Haishi et al., 2001).
For NMR flow and hydraulic measurements to be
applicable in situ (greenhouses, field situations), quantitative non-spatially resolved (non-imaging) methods with
specifically designed magnets are being developed. The
basic principles and components are now available to develop (reasonably priced) NMR plant flow-meters, as are
specially designed magnets (Raich and Blümler, 2004).
We can also think about small NMR or MRI detectors by
making use of rather small magnets, as presented by
Blümich and others (Blümich et al., 2002).
Acknowledgements
The research described in part herein was supported by the Dutch
Technology Foundation (STW), Applied Science Division of the
Netherlands Organization for Scientific Research (NWO) (projects
WAW07.0068, WBI 3493, and WBI 4803). The funding of the
dedicated intact plant 3 T MRI facility of Wageningen NMR Centre
by NWO, the Dutch Ministry of Agriculture, Nature and Food
Quality (LNV), and the Wageningen University and Research
Centre is greatly acknowledged. I acknowledge the contributions
and many stimulating discussions on the research topics presented
in this overview by my colleagues and coworkers Carel Windt,
Timur Sibgatullin, Frank Vergeldt, Edo Gerkema, Tom Scheenen,
and Louise van der Weerd. Financial support of the EU Transnational Access to Research Infrastructure programs starting from
1994 has stimulated Trans-European co-operation on this research.
References
Anisimov AV, Sorokina NY, Dautova NR. 1998. Water diffusion
in biological porous media: a NMR approach. Magnetic Resonance Imaging 16, 565–568.
Belton PS, Ratcliffe RG. 1985. NMR and compartmentation in
biological tissues. Progress in Nuclear Magnetic Resonance
Spectroscopy 17, 241–279.
Blümich B, Anferov V, Anferova S, Klein M, Fechete R,
Adams M, Casanova F. 2002. Simple NMRMOUSE with a bar
magnet. Concepts in Magnetic Resonance B 15, 255–261.
Brownstein KR, Tarr CE. 1979. Importance of classical diffusion
in NMR studies of water in biological cells. Physical Review A
19, 2446–2453.
Callaghan PT. 1993. Principles of nuclear magnetic resonance
microscopy. Oxford: Clarendon Press.
Callaghan PT, Clark CJ, Forde LC. 1994a. Use of static and
dynamic NMR microscopy to investigate the origins of contrast
in images of biological tissues. Biophysical Chemistry 50,
225–235.
Callaghan PT, Codd SL, Seymour JD. 1999. Spatial coherence
phenomena arising from translational spin motion in gradient spin
echo experiments. Concepts in Magnetic Resonance 11, 181–202.
Callaghan PT, Köckenberger W, Pope JM. 1994b. Use of
difference propagators of imaging of capillary flow in the
presence of stationary fluid. Journal of Magnetic Resonance B
104, 183–188.
Chudek JA, Hunter G. 1997. Magnetic resonance imaging of
plants. Progress in Nuclear Magnetic Resonance Spectroscopy
31, 43–62.
Clearwater MJ, Clark CJ. 2003. In vivo magnetic resonance
imaging of xylem vessel contents in woody lianas. Plant, Cell
and Environment 26, 1205–1214.
Clearwater MJ, Meinzer FC, Andrade JL, Goldstein G,
Holbrook M. 1999. Potential errors in measurement of nonuniform sap flow using heat dissipation probes. Tree Physiology
19, 681–687.
Conlon T, Outhred R. 1972. Water diffusion permeability of
erythrocytes using an NMR technique. Biochimical et Biophysical
Acta 288, 354–361.
Crick F. 1970. Diffusion in embryogenesis. Nature 225, 420–422.
Daudet FA, Lacointe A, Gaudillère JP, Cruiziat P. 2002.
Generalized Münch coupling between sugar and water fluxes for
modeling carbon allocation as affected by water status. Journal of
Theoretical Biology 214, 481–498.
Donker HCW, Van As H, Edzes HT, Jans AWH. 1996. NMR
imaging of white button mushroom (Agaricus bisporus) at
various magnetic fields. Magnetic Resonance Imaging 14,
1205–1215.
Donker HCW, Van As H, Snijder HJ, Edzes HT. 1997.
Quantitative 1H-NMR imaging of water in white button mushrooms (Agaricus bisporus). Magnetic Resonance Imaging 15,
113–121.
Donker HCW, Van As H. 1999. Cell water balance of white
button mushrooms (Agaricus bisporus) during its post-harvest
lifetime studied by quantitative magnetic resonance imaging.
Biochimica et Biophysica Acta 1427, 287–297.
Duce SL, Carpenter TA, Hall LD, Hills BP. 1992. An investigation of the origin of contrast in NMR spin echo im ages of
plant tissue. Magnetic Resonance Imaging 10, 289–297.
Edzes HT, van Dusschoten D, Van As H. 1998. Quantitative T2
imaging of plant tissues by means of multi-echo MRI microscopy. Magnetic Resonance Imaging 16, 185–196.
Fukushima E. 1999. Nuclear magnetic resonance as a tool to study
flow. Annual Review of Fluid Mechanics 31, 95–123.
Haishi T, Uematsu T, Matsuda Y, Kose K. 2001. Development of
a 1.0 T MR microscope using a Nd-Fe-B permanent magnet.
Magnetic Resonance Imaging 19, 875–880.
Henzler T, Waterhouse RN, Smyth AJ, Carvajal M, Cooke DT,
Schäffner AR, Steudle E, Clarkson DT. 1999. Diurnal
MRI and water transport in plants 755
variations in hydraulic conductivity and root pressure can be
correlated with the expression of putative aquaporins in the roots
of Lotus japonicus. Planta 210, 50–60.
Hills BP, Duce SL. 1990. The influence of chemical and diffusive
exchange on water proton transverse relaxation in plant tissues.
Magnetic Resonance Imaging 8, 321–332.
Hürlimann MD, Venkataramanan L, Flaum C. 2002. The
diffusion-spin relaxation time distribution as an experimental
probe to characterize fluid mixtures in porous media. Journal of
Chemical Physics 117, 10223–10232.
Ionenko IF, Anisimov AV, Karimova FG. 2006. Water transport
in maize roots under the influence of mercuric chloride and water
stress: a role of water channels. Biologia Plantarum 50, 74–80.
Ishida N, Koizumi M, Kano H. 2000. The NMR microscope,
a unique and promising tool for plant science. Annals of Botany
86, 259–278.
Javot H, Maurel C. 2002. The role of aquaporins in root water
uptake. Journal of Experimental Botany 54, 2035–2043.
Koch GW, Sillet SC, Jennings GM, Davis SD. 2004. The limits to
tree height. Nature 428, 851–854.
Köckenberger W. 2001. Functional imaging of plants by magnetic
resonance experiments. Trends in Plant Science 6, 286–292.
Lee SH, Chung GC, Steudle E. 2005. Low temperature and
mechanical stresses differently gate aquaporins of root cortical
cells of chilling-sensitive cucumber and -resistant figleaf gourd.
Plant, Cell and Environment 28, 1191–1202.
Levitt M. 2001. Spin dynamics: basics of nuclear magnetic
resonance. Chichester: Wiley.
Link J, Seelig J. 1990. Comparison of deuterium NMR imaging
methods and applications to plants. Journal of Magnetic
Resonance 89, 310–330.
Lüdecke KM, Röschmann P, Tischler R. 1985. Susceptibility
artefacts in NMR imaging. Magnetic Resonance Imaging 3,
329–343.
MacFall JJ, Van As H. 1996. Magnetic resonance imaging of
plants. In: Shachar-Hill Y, Pfeffer PE, eds. Nuclear magnetic
resonance in plant biology. Rockville: The American Society of
Plant Physiologists, 33–76.
Mantle MD, Sederman AJ. 2003. Dynamic MRI in chemical
process and reaction engineering. Progress in Nuclear Magnetic
Resonance Spectroscopy 43, 3–60.
McCain D. 1995. Nuclear magnetic resonance study of spin
relaxation and magnetic field gradients in maple leaves. Biophysical Journal 69, 1111–1116.
Norris DG. 2001. The effects of microscopic tissue parameters on
the diffusion weighted magnetic resonance imaging experiment.
NMR in Biomedicine 14, 77–93.
Peuke AD, Windt C, Van As H. 2006. Effects of cold-girdling on
flows in the transport phloem in Ricinus communis: is mass flow
inhibited? Plant, Cell and Environment 29, 15–25.
Qiao Y, Galvosas P, Callaghan PT. 2005. Diffusion correlation
NMR spectroscopic study of anisotropic diffusion of water in
plant tissue. Biophysical Journal 89, 2899–2905.
Raich H, Blümler P. 2004. Design and construction of a dipolar
Halbach array with a homogeneous field from identical bar
magnets: NMR Mandhalas. Progress in Nuclear Magnetic
Resonance Spectroscopy B 23, 16–25.
Ratcliffe RG. 1994. In vivo NMR studies of higher plants and
algae. Advances in Botanical Research 20, 42–123.
Ratcliffe RG, Roscher A, Shachar-Hill Y. 2001. Plant NMR
spectroscopy. Progress in Nuclear Magnetic Resonance Spectroscopy 39, 267–300.
Ratković S, Bačić G. 1980. Water exchange in Nitella cells: a 1H
NMR study in the presence of paramagnetic Mn2+ ions.
Bioelectrochemistry and Bioenergetics 7, 405–412.
Regan DG, Kuchel PW. 2000. Mean residence time of molecules
diffusing in a cell bounded by a semi-permeable membrane:
Monte Carlo simulations and an expression relating membrane
transition probability to permeability. European Biophysical
Journal 29, 221–227.
Regan DG, Kuchel PW. 2002. Simulations of molecular diffusion
in lattices of cells: insight for NMR of red blood cells.
Biophysical Journal 83, 161–171.
Reinders JEA, Van As H, Schaafsma TJ, de Jager PA, Sheriff
DW. 1988a. Water balance in Cucumis plants, measured by
nuclear magnetic resonance. I. Journal of Experimental Botany
39, 1199–1210.
Reinders JEA, Van As H, Schaafsma TJ, Sheriff DW. 1988b.
Water balance in Cucumis plants, measured by nuclear magnetic resonance. II. Journal of Experimental Botany 39,
1211–1220.
Rokitta M, Rommel E, Zimmermann U, Haase A. 2000.
Portable nuclear magnetic resonance imaging system. Review of
Scientific Instruments 71, 4257–4262.
Rofe CJ, van Noort J, Back PJ, Callaghan PT. 1995. NMR
microscopy using large pulsed magnetic field gradients. Journal
of Magnetic Resonance 108, 125–136.
Schaafsma TJ, Van As H, Palstra WD, Snaar JE, de Jager PA.
1992. Quantitative measurement and imaging of transport processes in plants and porous media by 1H NMR. Magnetic
Resonance Imaging 10, 827–836.
Scheenen TWJ. 2001. Nuclear magnetic resonance imaging of
water motion in plants. PhD thesis (dissertation nr. 3032),
Wageningen University.
Scheenen TWJ, van Dusschoten D, de Jager PA, Van As H.
2000. Quantification of water transport in plants with NMR
imaging. Journal of Experimental Botany 51, 1751–1759.
Scheenen TWJ, Heemskerk AM, de Jager PA, Vergeldt FJ,
Van As H. 2002. Functional imaging of plants: a Nuclear Magnetic Resonance study of a cucumber plant. Biophysical Journal
82, 481–492.
Scheenen TWJ, Vergeldt FJ, Windt CW, de Jager PA, Van As
H. 2001. Microscopic imaging of slow flow and diffusion:
a pulsed field gradient stimulated echo sequence combined with
turbo spin echo imaging. Journal of Magnetic Resonance 151,
94–100.
Sellers PJ, Dickinson RE, Randall DA, et al. 1997. Modeling the
exchanges of energy, water, and carbon between continents and
the atmosphere. Science 275, 502–509.
Sen PN. 2004. Time-dependent diffusion coefficient as a probe of
geometry. Concepts in Magnetic Resonance A 23, 1–21.
Sevanto S, Vesala SJ, Peramaki M, Nikinmaa E. 2002. Time
lags for xylem and stem diameter variations in Scots pine tree.
Plant, Cell and Environment 25, 1071–1077.
Shachar-Hill Y, Pfeffer PE (eds). 1996. Nuclear magnetic
resonance in plant biology. Rockville: American Society of Plant
Physiologists.
Smith DM, Allen SJ. 1996. Measurement of sap flow in plant
stems. Journal of Experimental Botany 47, 1833–1844.
Snaar JEM, Van As H. 1992. Probing water compartment and
membrane permeability in plant cells by proton NMR relaxation
measurements. Biophysical Journal 63, 1654–1658.
Sperry JS, Stiller V, Hacke UW. 2006. Xylem hydraulics and the
soil–plant–atmosphere continuum: opportunities and unresolved
issues. Agronomy Journal 95, 1362–1370.
Stapf S, Han S-I (eds). 2005. NMR imaging in chemical
engineering. Weinheim: Wiley.
Stejskal EO, Tanner JE. 1965. Spin diffusion measurements: spin
echoes in the presence of a time-dependent field gradient. Journal
of Chemical Physics 42, 288–292.
756 Van As
Steudle E. 2001. The cohesion–tension mechanism and the
aquisition of water by plant roots. Annual Reviews in Plant
Physiology and Plant Molecular Biology 52, 847–875.
Stout DG, Steponkus PL, Cotts M. 1978. Nuclear magnetic
resonance relaxation-times and plasmalemma water exchange in
ivy bark. Plant Physiology 62, 636–642.
Tardieu F. 2005. Plant tolerance to water deficit: physical limits
and possibilities for progress. Comptes Rendus Geoscience 337,
57–67.
Tomos AD, Leigh RA. 1999. The pressure probe: a versatile tool in
plant cell physiology. Annual Review of Plant Physiology and
Plant Molecular Biology 50, 447–472.
Tournaire-Roux C, Sutka M, Javot H, Gout E, Gerbeau P,
Luu DT, Bligny R, Maurel C. 2003. Cytosolic pH regulates
root water transport during anoxic stress through gating of aquaporins. Nature 425, 393–397.
Tyerman SD, Niemietz CM, Bramley H. 2002. Plant aquaporins:
multifunctional water and solute channels with expanding roles.
Plant, Cell and Environment 25, 173–194.
Tyree MT, Patino S, Bennink J, Alexander J. 1995. Dynamic
measurements of root hydraulic conductance using a highpressure flow meter in the laboratory and field. Journal of
Experimental Botany 46, 83–94.
Utsuzawa S, Fukuda K, Sakaue D. 2005. Use of magnetic
resonance microscopy for the non-destructive observation of
xylem cavitation caused by pine wilt disease. Phytopathology 95,
737–743.
Valiullin R, Skirda V. 2001. Time-dependent self-diffusion coefficient of molecules in porous media. Journal of Chemical
Physics 114, 452–458.
Van As H, Reinders JEA, de Jager PA, van der Sanden PACM,
Schaafsma TJ. 1994. In situ plant water-balance studies using
a portable NMR spectrometer. Journal of Experimental Botany
45, 61–67.
Van As H, Schaafsma TJ. 1984. Noninvasive measurement of
plant water-flow by Nuclear Magnetic Resonance. Biophysical
Journal 45, 496–472.
van der Toorn A, Zemah H, Van As H, Bendel P, Kamenetsky R.
2000. Developmental changes and water status in tulip bulbs
during storage: visualization by NMR imaging. Journal of
Experimental Botany 51, 1277–1287.
van der Weerd L, Claessens MMAE, Efdé C, Van As H. 2002a.
Nuclear Magnetic Resonance imaging of membrane permeability
changes in plants during osmotic stress. Plant, Cell and
Environment 25, 1538–1549.
van der Weerd L, Claessens MMAE, Ruttink T, Vergeldt FJ,
Schaafsma TJ, Van As H. 2001. Quantitative NMR microscopy
of osmotic stress responses in maize and pearl millet. Journal of
Experimental Botany 52, 2333–2343.
van der Weerd L, Melnikov SM, Vergeldt FJ, Novikov EG, Van
As H. 2002b. Modelling of self-diffusion and relaxation time
NMR in multi-compartment systems with cylindrical geometry.
Journal of Magnetic Resonance 156, 213–221.
van Dusschoten D, de Jager PA, Van As H. 1995. Extracting
diffusion constants from echo-time-dependent PFG NMR data
using relaxation-time information. Journal of Magnetic Resonance A 116, 22–28.
van Dusschoten D, Moonen CTW, de Jager PA, Van As H.
2006. Unraveling diffusion constants in biological tissue by combining Carr–Purcell–Meiboom–Gill imaging and Pulsed Field
Gradient NMR. Magnetic Resonance in Medicine 36, 907–913.
Venkataramanan L, Song YQ, Hurlimann MD. 2002. Solving
Fredholm integrals of the first kind with tensor product structure
in 2 and 2.5 dimensions. IEEE Transaction on Signal Procession
50, 1017–1026.
Verkman AS. 2000. Water permeability measurement in living
cells and complex tissues. Journal of Membrane Biology 173,
73–87.
Verslues PE, Agarwal M, Katiyar-Agarwal S, Zhu JH, Zhu JK.
2006. Methods and concepts in quantifying resistance to drought,
salt and freezing, abiotic stresses that affect plant water status.
The Plant Journal 45, 523–539.
Walter L, Callies R, Altenburger R. 1992. Studies of plant
systems by in vivo NMR spectroscopy. In: de Certaines JD,
Bovée WMMJ, Podo F, eds. Magnetic resonance spectroscopy in
biology and medicine. Oxford: Pergamon Press, 573–610.
Windt CW, Vergeldt FJ, de Jager PA, Van As H. 2006. MRI
of long-distance water transport: a comparison of the phloem
and xylem flow characteristics and dynamics in poplar, castor
bean, tomato, and tobacco. Plant, Cell and Environment 29,
1715–1729.
Xia Y. 1995. Contrast in NMR imaging and microscopy. Concepts
in Magnetic Resonance 8, 205–225.
Zhang WH, Jones GP. 1996. Water permeability in wheat root
protoplasts determined from nuclear magnetic resonance relaxation times. Plant Science 118, 97–106.