Visualization of early stress responses in plant leaves
Laury Chaerlea, Martin vande Venb, Roland Valcke*b and Dominique Van Der Straeten§a
a
Department of Molecular Genetics, Ghent University, Ghent, Belgium
b
Limburgs Universitair Centrum, Department S.B.G., Diepenbeek, Belgium
ABSTRACT
Plant leaves possess 'microscopic valves', called stomata, that enable control of transpirational water loss. In case of
water shortage, stomata close, resulting in decreased transpirational cooling. The ensuing temperature increase is readily
visualized by thermography. Salicylic acid, a central compound in the defense of plants against pathogens, also closes
stomata in several species. In previous work, thermography permitted to monitor an increase in temperature after
infection of resistant tobacco by tobacco mosaic virus, before visual symptoms appeared. Furthermore, cell death was
visualized with high contrast in both tobacco and Arabidopsis. In addition to transpiration, photosynthetic assimilation
is a key physiological parameter. If the amount of light absorbed by chlorophyll exceeds the capacity of the
photosynthetic chain, the surplus is dissipated as light of longer wavelength. This phenomenon is known as chlorophyll
fluorescence. If a plant leaf is affected by stress, photosynthesis is impaired resulting in a bigger share of non-utilized
light energy emitted as fluorescence. The potential of an automated imaging setup combining thermal and fluorescence
imaging was shown by monitoring spontaneous cell death in tobacco. This represents a first step to multispectral
characterization of a wide range of emerging stresses, which likely affect one or both key physiological parameters.
Key-words: thermography, visible-light-induced chlorophyll fluorescence imaging, transpiration, photosynthesis,
incompatible plant pathogen interaction, salicylic acid, spontaneous plant cell death
1. INTRODUCTION
Plants control the transpiration stream that enables uptake of water and nutrients from the soil by regulating stomatal
aperture. Upon drought, levels of the hormone abscisic acid (ABA) increase and cause stomatal closure1. Thermal
imaging was used to screen for barley mutants unable to respond to ABA2. Recently, the interest for such screenings
increased again3. Historically, thermography was first used for plant research in a remote-sensing context4. In the mid
seventies, a thermal imaging system was used at lab scale to study the necrotising damage caused by tobacco mosaic
virus5.
Plants of some so-called thermogenic species increase the temperature of their flowers by increasing respiration. The
kinetics of these processes have been recorded by thermal imaging6. Salicylic acid (SA) was discovered to trigger the
metabolic upregulation resulting in the observed heat-production. Thermal imaging indicated that SA increased the
temperature of leaves when applied by spraying on tobacco leaves7. In addition, SA was shown to close stomata in other
species, including common bean8,9. These data were obtained by microscopical observation and measurement of the
increase in humidity of air circulated through cuvettes clamped on leaves (direct transpiration measurements).
Importantly, salicylic acid is a signaling compound in the defense reaction of plants to pathogens10. After infection of
resistant tobacco by tobacco mosaic virus (TMV), the model system for incompatible plant-pathogen interaction, SA
rises to high levels before any visual necrotic symptoms become apparent11. During this plant-pathogen interaction,
presymptomatic increases in temperature were visualized which coincided with the later occurring necrotic cell death12.
Thermal imaging permitted to discern the increase in temperature 8 hours before the emergence of pinpoint necrotic
lesions. More important, 2 days after infection, the area with increased temperature reached its maximal extension,
corresponding to the final necrotic lesion formed 4 days later. Direct transpiration measurements proved the major share
of the temperature increase was due to stomatal closure. In support of this finding, plant leaves have a high surface to
volume ratio, making it impossible to accumulate metabolic heat13. The process of cell death in response to invading
pathogens is mimicked in spontaneous cell death mutant or transgenic plants14,15. Thermal imaging monitored a
Correspondence: § [email protected]; Tel.: +32 9 264 50 95; Fax.: +32 9 264 53 33; http://www.plantgenetics.rug.ac.be
* [email protected]; Tel.: +32 11 26 83 81; Fax.: +32 11 26 83 01; http://www.luc.ac.be
temperature increase before the appearance of the spontaneous cell death16. In addition, the cell death process is
visualized with high contrast, since the dying tissue evaporates water from disrupted cells, and appears in dark shades in
thermograms. Thermography thus clearly visualizes pending damage to plant leaves. This would enable early
monitoring of disease outbreaks or screening for increased disease-resistance.
Chlorophyll fluorescence imaging was shown to be a second straightforward way for detection of stress17. Different
fluorescence imaging systems (FIS) have been developed, and applied to study biotic or abiotic stress18-20. Light energy
absorbed by the chlorophyll pigment in plant leaves is used for photosynthetic carbon assimilation. Energy captured in
excess of the capacity of the photosynthetic system is dissipated as emission of light or heat21. When stress - either
biotic or abiotic - causes the photosynthetic capacity to decrease, the emission of light (and heat) increases for equal
illumination conditions. Chlorophyll re-emits light at specific wavelengths. When excited with visible light,
fluorescence is emitted at red (690nm) and far-red (740nm) wavelengths. By equipping the visible light source with a
cut-off low-pass blue filter and placing a cut-off high-pass red B+W 092 filter in front of the CCD camera chlorophyll
fluorescence can be recorded without interference from excitation light. When using UV-excitation, fluorescence from
other leaf constituents can be recorded in addition to chlorophyll fluorescence22.
By combining two imaging techniques each monitoring a major plant physiological parameter, it is expected that many
stress situations will be detected at early stages, enabling timely control measures. In addition, monitoring of differences
in the kinetics of stress responses among populations of plants could be useful in breeding programs for selection of
plants more resistant to stress23,24.
2. METHODOLOGY
2.1.
Imaging systems
A custom-designed measuring chamber was built, in which a constant temperature within 0.1ºC limits is obtained.
Inside this chamber a Cartesian positioning system was installed (See fig.1). The thermal imaging system (FLIR-Agema
THV-900 LW - Stirling cooled) and a color CCD-camera were fitted one next to the other on a supporting structure at
the basis of the Z-axis. Recently a chlorophyll fluorescence imaging system (FIS), consisting of a black and white
CCD-camera and a circular high-intensity illuminating system consisting of small halogen bulbs was mounted on the
system. Imaging positions are entered into the system by a teach-in procedure, using the thermal or video camera for
visualization. The PLC controlling the robot can be set to automatically generate imaging positions for the color-CCD
camera and FIS system (see Fig. 2). The resulting images captured from the 3 cameras are thus co-localized. If
necessary, these images can be further aligned by using image registration procedures. After teach-in programming, the
robot is made to run the programmed image capture cyclus at a fixed time interval (typically 15minutes, 30 minutes or 1
hour). The images from the three cameras are captured following a continuous numbering scheme. After each capture
cycle, images are transferred to a central storage disk. A workstation PC subsequently renumbers the files to reflect
Figure 1. Model of the portal robot - the enclosure of the
measuring room was omitted for clarity. Photographic views of
the
measuring
chamber
can
be
accessed
at
www.plantgenetics.rug.ac.be/PlantIR/Setup.
The system is build of extruded aluminum profiles with
integrated belts. Stepper-motors are mounted on each of the 3
axis. The working volume is X x Y x Z: 2m x 1.2m x 0.5m.
Repetitive accuracy was specified and measured to be 0.1mm.
For stability the aluminum X-axis are mounted at 1.2m height on
an iron chassis, built into the measuring cabinet. The control
display of the robot and displays for the cameras are located
outside the climate room. To avoid damage to plants when
moving the cameras, the system was programmed to lift the Zaxis to its highest position before moving laterally.
capture position and capture cycle. Next the infrared FLIR Agema files are converted to bitmap. Subsequently, thermal,
chlorophyll fluorescence and video images from capture positions belonging to a single leaf or plant are combined in a
concatenated file. These files make it possible to visualize on-line changes in thermal patterns of the monitored plants
and ultimately to guide sampling of presymptomatic leaf tissue for molecular and biochemical analysis. The overview
images are automatically combined into an MPEG animation file, either incrementally after each capture cycle, or a
separate file is created each 24h. After completion of the experiment, the MPEG-animation files permit an efficient
visual selection of the gathered data. Selected leaves or plants are subsequently analyzed for local leaf temperature, area
of temperature increase and area of visual symptom extension.
Figure 2. Flowchart of the image capture, processing and visualization method. This setup permits to organize and visualize thermal,
fluorescence and video sequences in an efficient way. On selected leaf regions, nearly real-time presymptomatic sampling can be
carried out, or temperatures and areas can be measured off-line.
2.2.
Plant growth conditions
Tobacco plants (Nicotiana tabacum cv. Xanthi NN, Xanthi nc) are grown in a conventional plant growth chamber at
21±1ºC and 60±10%RH. At the 6 to 8 leaf stage plants are transferred to the measuring room, conditioned at 21±0.1ºC.
Since measurements can span several weeks, plants have a constant supply of water to avoid effects of water
availability on leaf transpiration and temperature. Fresh tobacco mosaic virus suspension was obtained by grinding
infected tissue from Nicotiana tabacum cv. SR1. Pumice was used as an abrasive to permit TMV to infect leaf
epidermal cells.
3. RESULTS
3.1.
Salicylic acid increases leaf surface temperature
Salicylic acid was previously shown to increase leaf surface temperature of tobacco leaves after spraying7. A
concentration of 3mM resulted in maximal leaf-surface temperature increase. Higher concentrations were phytotoxic
and resulted in necrosis. In order to determine the minimum concentration that results in a measurable temperature
increase, single droplets of SA-solutions were applied at the leaf surface. Comparing the reaction of different-leaves
from different plants proved difficult due to heterogeneous response at the leaf surface. The result from the application
as droplets of a dilution series of salicylic acid on a tobacco leaf surface is consistent with the data obtained by
spraying. In both experiments, 3mM resulted in a clearly identifiable temperature increase.
Figure 3. Droplet application of a dilution-series of SA in phosphate buffer
pH7 solutions on an attached leaf of a tobacco Xanthi nc plant. The leaf
veins delimit application zones. Above the main vein from right to left:
control buffer, 10mM, 1mM, 10-2 mM and 10-4 mM. Under the mid vein
from right to left: 5mM, 3mM, 0.1mM, 10-3 mM and 10-4mM. The upper
panel shows thermal (left) and visual reflectance (right) images, just after
the application of the 20 µl droplets. The dark droplets have a minimum
temperature of 19ºC, the untreated leaf blade is at 20 to 20.3ºC depending
on the position and the main vein reaches 21.5ºC. The pictures of the lower
panel were taken 26 hours later. Treatments from 3 to 10mM result in clear
symptoms in the thermogram. The maximum temperature at the treated loci
was 21.3ºC, whereas the untreated tissue in-between was at 20.9ºC. One
and 0.1mM cause very little local increase in temperature.
3.2.
Thermography aided presymptomatic sampling of tobacco-mosaic virus infected leaf tissue
In previous work, a correlation in the kinetics of temperature increase, transpiration decrease and salicylic acid
concentration increase was observed12. Sampling for SA-determinations can be done on isolated infection spots (see
Fig. 5), or areas bounded by leaf veins can be homogeneously infected (see Fig. 4). Sampling is obviously much faster
with the latter method, although some spatial variation is introduced due to uneven infection of the whole region.
Figure 4. TMV infection on resistant tobacco leaf sections for
thermography-aided presymptomatic sampling. Infection was carried out
by dusting the leaf area with pumice, adding TMV-suspension and gentle
abrasion with a bent glass-rod. The two pictures from each panel are
composed of 2 overlapping images. The attached tobacco leaves are kept
horizontal by supporting nylon wires.
The upper panel was taken 28 hours after infection. Neither on the thermal
(left) or visual reflectance (right) images symptoms can be seen. The
temperature of the leaf blade was on average 20.2ºC. At 41 hours after
infection the increase in temperature has reached its maximum (middle
panel). The maximum temperature of the warmer, lighter colored regions
was 20.6ºC. No visual symptoms were yet detectable at this time point. As
a control, the leaf regions in-between the TMV-infected regions were
inoculated by the same method, without TMV. These regions have a
temperature of 20.2ºC, corresponding with the temperature of the leaf blade
in the upper panel. The spread of the infection over the delimiting veins
into the control region proved difficult to prevent. The lower panel shows a
strong decrease of temperature in the infected areas due to cell death. The
dark regions at the infected sites have a minimum temperature of 19.4ºC.
The control-infected areas now have a slightly higher temperature of 20.320.4ºC, likely due to transport of SA from the infected areas. At this time,
visual symptoms are clearly discernable as drying, light-reflecting regions.
3.3.
Tobacco-mosaic virus infection induces local presymptomatic leaf surface temperature increase
In a previous publication the exact spatial colocalization of the local thermal effect with the subsequently formed patch
of dead tissue was proven12. Figure 5 illustrates thermal imaging at high resolution of the previously published
presymptomatic increase in leaf surface temperature at the local infection site.
Figure 5. Close-up of a TMV-infection site. The region to be
infected was encircled with a black felt marker. Pumice was
deposited in the center, followed by a 20µl droplet of TMVsuspension. A fine glass rod was used to infect locally by gentle
friction. The images of panel a were captured 1 day after infection.
The average temperature inside the marked region of the thermal
image (left) was 21.4ºC (min. 21.3ºC and max. 20.5ºC). Half a day
later (panel b) a thermal effect is visible, although no visual
symptoms can be observed. The temperature has increased by
0.2ºC (avg. 21.6ºC, min. 21.5ºC, max. 21.7ºC). Panel c (2d post
infection) shows necrosis at the site of infection. Thermally, these
regions have a low temperature (21ºC, dark blue). Visually, the
necrotic fleck has a brownish appearance. Two and a half day after
infection, the necrotic lesion has dried. Evaporation has completely
ceased, causing the region to appear much warmer (max. 22.3ºC)
than the surrounding tissue. The width of the images is 2.75 cm.
3.4.
Combination of thermography and chlorophyll fluorescence imaging
Spontaneous cell death is preceded by a temperature increase,
as was observed previously16. We wanted to find out if
chlorophyll fluorescence imaging would also permit
presymptomatous detection of pending cell death. From figure
6 it can be concluded that, as is the case with thermal imaging,
regions with different color can be discerned in
asymptomatous tissue. However, these regions are not of
higher contrast or extent compared to the thermal effects.
From these preliminary images, it can be seen that cell death
is visualized with very high contrast compared to the visual
color reflection images.
Figure 6. Parallel thermal and chlorophyll fluorescence imaging of
tobacco bO spontaneous cell death mutant showing spreading
necrosis. The temperature of the thermal images spans 1ºC, from
21ºC (dark blue spots) to 22ºC (yellow regions). The unaffected
portions of the leaf blade have a temperature of 21.5ºC
In the thermal images (left column) emerging cell death is visible as
dark blue spots. In panel a cell death is visible at the bottom of the
image. One day later (panel b), cell death has expanded along the
major and side veins. A faintly warmer front is discernable at the left
of the cell death zone. A more detailed pattern of cell death is visible
from the chlorophyll fluorescence image (middle column), which
provides a much higher contrast than the visual images (right
column). A thermal front 0.4ºC higher in temperature has expanded
during the next 12h. New initiation of cell death is visible in panel d
(2d after first frame). The spread of cell death via the side veins is
clearly visible in the chlorophyll fluorescence image. In the last panel
(3d post infection), the thermal effect is more diffuse. Th dark
regions in the chlorophyll fluorescence image correspond to patches
of dead cells where chlorophyll is completely degraded and thus
fluorescence emission is absent.
4. CONCLUSIONS
Imaging technology clearly has a lot of potential to assess the physiological status of plants. In addition to the here
illustrated thermography and chlorophyll fluorescence imaging, which respectively monitor transpiration and
photosynthetic efficiency, hyperspectral reflectance imaging can provide additional information on changes in leaf
composition, structure or orientation25,26. Information on accumulation or degradation of a particular compound can be
obtained from images taken from specific narrow spectral regions (of the order of 10nm). For each particular stresssituation, multiple images captured in the visual or near-infrared spectrum can be analyzed for deviations. In table 1 a
concluding overview of the three mentioned imaging techniques is given along with their pros and cons. By combining
hyperspectral reflectance imaging with thermography and chlorophyll fluorescence imaging, a multispectral imaging
setup would become available to realize a so called 'stress-catalogue', enabling timely determination of emerging
stresses. This would extend the application of such a multispectral monitoring system beyond the point of simply
visualizing an emerging stress condition.
Technology
Thermography
Chlorophyll Fluorescence
imaging
Reflectance imaging
Parameter monitored
Transpiration
Photosynthetic efficiency
Changes in leaf
constituents
Advantages
Passive
Permits quantification of
photosynthetic parameters
Disadvantages
Sensitive to environmental
changes
Illumination needed
hyperspectral imaging is
powerful but needs
expensive systems
Table 1: Overview of imaging techniques that could be combined into a versatile multispectral imaging system.
ACKNOWLEDGEMENTS
This research was supported by grants from the Fund for Scientific Research (Flanders) (G0068.90, G0021.92,
G.0023.95N, 1.5.514.98 and G0015-01). Laury Chaerle is a postdoctoral research assistant of the Fund for Scientific
Research (Flanders). We thank Wim and Frank De Boever for the generation of the model in Figure 1. Bacterio-opsin
(bO) transgenic plants of tobacco were a generous gift from E. Lam (AgBiotech Centre, Rutgers - The State University
of New Jersey, USA).
REFERENCES
1.
2.
3.
4.
5.
6.
7.
A. M. Hetherington, "Plant physiology: Spreading a drought warning," Curr. Biol. 8, pp. R911-R913, 1998.
I. Raskin and J. A. R. Ladyman, "Isolation and characterization of a barley mutant with abscisic-acid-insensitive
stomata," Planta 173, pp. 73-78, 1988.
S. Merlot, F. Gosti, D. Guerrier, A. Vavasseur, J. Giraudat, "The ABI1 and ABI2 protein phosphatases 2C act in
a negative feedback regulatory loop of the abscisic acid signalling pathway," Plant J. 25, pp. 295-303, 2001.
H. E. Nilsson, "Remote sensing and image analysis in plant pathology," Annu. Rev. Phytopathol. 33, pp. 489527, 1995.
C. De Carolis, G. G. Conti and G. M. Lechi, "Investigations on thermal behaviour of plants affected by virus and
fungus diseases," Proc. 10th Int. Symp. Rem. Sens. of Environm., Ann. Arbor 21, pp. 659-672, 1975.
H. Skubatz, T. A. Nelson, A. M. Dong, B. J. D. Meeuse and A. J. Bendich, "Infrared thermography of Arum lily
inflorescences," Planta 182, pp. 432-436, 1990.
D. Van Der Straeten, L. Chaerle, G. Sharkov, H. Lambers and M. Van Montagu, "Salicylic acid enhances the
activity of the alternative pathway of respiration in tobacco leaves and induces thermogenicity," Planta 196, pp.
412-419, 1995.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
B. Manthe, M. Schulz and H. Schnabl, "Effects of salicylic acid on growth and stomatal movements of Vicia
faba L.: Evidence for salicylic acid metabolization," Journal of Chemical Ecology 18, pp. 1525-1539, 1992.
A. Larqué-Saavedra, "Stomatal closure in response to acetylsalicylic acid treatment," Zeitschrift fuer
Pflanzenphysiologie 93, pp. 371-375, 1979.
T. Delaney, S. Uknes, B. Vernooij, L. Friedrich, K. Weymann, D. Negrotto, T. Gaffney, M. Gutrella, H.
Kessmann, E. Ward and J. Ryals, "A central role of salicylic acid in plant disease resistance," Science 266, pp.
1247-1250, 1994.
A. J. Enyedi, N. Yalpani, P. Silverman and I. Raskin, "Localization, conjugation, and function of salicylic acid in
tobacco during the hypersensitive reaction to tobacco mosaic virus," Proc. Natl. Acad. Sci. USA 89, pp. 24802484, 1992.
L. Chaerle, W. Van Caeneghem, E. Messens, H. Lambers, M. Van Montagu and D. Van Der Straeten,
"Presymptomatic visualization of plant-virus interactions by thermography," Nature Biotechnol. 17, pp. 813-816,
1999.
R. W. Breidenbach, M. J. Saxton, L. D. Hansen and R. S. Criddle, "Heat generation and dissipation in plants: can
the alternative oxidative phosphorylation pathway serve a thermoregulatory role in plant tissues other than
specialized organs?," Plant Physiol. 114, pp. 1137-1140, 1997.
R. Dietrich, T. Delaney, S. Uknes, E. Ward, J. Ryals and J. Dangl, "Arabidopsis mutants simulating disease
resistance response," Cell 77, pp. 565-577, 1994.
R. Mittler, V. Shulaev and E. Lam, "Coordinated activation of programmed cell death and defense mechanisms
in transgenic tobacco plants expressing a bacterial proton pump," Plant Cell 7, pp. 29-42, 1995.
L. Chaerle, F. De Boever, M. Van Montagu and D. Van der Straeten, "Thermographic visualization of cell death
in tobacco and Arabidopsis," Plant Cell and Environment 24, pp. 15-25, 2001.
H. K. Lichtenthaler and J. A. Miehé, "Fluorescence imaging as a diagnostic tool for plant stress," Trends Plant
Sci. 2, pp. 316-320, 1997.
C. B. Osmond, P. F. Daley, M. R. Badger and U. Lüttge, "Chlorophyll fluorescence quenching during
photosynthetic induction in leaves of Abutilon striatum Dicks. Infected with Abutilon mosaic virus, observed
with a field-portable imaging system," Bot. Acta 111, pp. 390-397, 1998.
R. Valcke, M. Ciscato, F. Heisel, J. Miehé and M. Sowinska, "Analysis of heavy metal stressed plants by
fluorescence imaging," in Laser Radar Technology and Applications IV, G.W. Kamerman and C. Werner,
Editors, Proc. SPIE 3707, pp. 82-90, 1999.
J. D. Scholes and S. A. Rolfe, "Photosynthesis in localised regions of oat leaves infected with crown rust
(Puccinia coronata): quantitative imaging of chlorophyll fluorescence," Planta 199, pp. 573-582, 1996.
G. H. Krause and E. Weis, "Chlorophyll fluorescence and photosynthesis: The basics," Annu. Rev. Plant.
Physiol. Plant Mol. Biol. 42, pp. 313-349, 1991.
C. Buschmann and H. K. Lichtenthaler, "Principles and characteristics of multi-colour fluorescence imaging of
plants," J. Plant Physiol. 152, pp. 297-314, 1998.
L. Chaerle and D. Van Der Straeten, "Imaging techniques and the early detection of plant stress," Trends Plant
Sci. 5, pp. 495-501, 2000.
L. Chaerle and D. Van Der Straeten, "Seeing is believing: imaging techniques to monitor plant health," Biochim.
Biophys. Acta - Gene Struct. Expression 1519, pp. 153-166, 2001.
H. G. Jones, J. M. Anderson and R. Casa, "Multispectral and multiangular remote sensing of crop canopy stress,"
Aspects of Applied Biology 60, p155-162, 2000.
J. Peñuelas and I. Filella, "Visible and near-infrared reflectance techniques for diagnosing plant physiological
status," Trends Plant Sci. 3, pp. 151-156, 1998.