Journal of Experimental Botany, Vol. 54, No. 388, pp. 1743±1752, July 2003
DOI: 10.1093/jxb/erg186
RESEARCH PAPER
The responses of guard and mesophyll cell
photosynthesis to CO2, O2, light, and water stress in a
range of species are similar
Tracy Lawson, Kevin Oxborough, James I. L. Morison* and Neil R. Baker
Department of Biological Sciences, University of Essex, Colchester CO4 3SQ, UK
Received 20 December 2002; Accepted 7 April 2003
Abstract
High resolution chlorophyll a ¯uorescence imaging
was used to compare the photosynthetic ef®ciency of
PSII electron transport (estimated by Fq¢/Fm¢) in guard
cell chloroplasts and the underlying mesophyll in
intact leaves of six different species: Commelina
communis, Vicia faba, Amaranthus caudatus,
Polypodium vulgare, Nicotiana tabacum, and
Tradescantia albifora. While photosynthetic ef®ciency
varied between the species, the ef®ciencies of guard
cells and mesophyll cells were always closely
matched. As measurement light intensity was
increased, guard cells from the lower leaf surfaces of
C. communis and V. faba showed larger reductions
in photosynthetic ef®ciency than those from the
upper surfaces. In these two species, guard cell
photosynthetic ef®ciency responded similarly to that
of the mesophyll when either light intensity or CO2
concentration during either measurement or growth
was changed. In all six species, reducing the O2 concentration from 21% to 2% reduced guard cell photosynthetic ef®ciency, even for the C4 species A.
caudatus, although the mesophyll of the C4 species
did not show any O2 modulation of photosynthetic
ef®ciency. This suggests that Rubisco activity is signi®cant in the guard cells of these six species. When
C. communis plants were water-stressed, the guard
cell photosynthetic ef®ciency declined in parallel
with that of the mesophyll. It was concluded that
the photosynthetic ef®ciency in guard cells is
determined by the same factors that determine it in
the mesophyll.
Key words: Chlorophyll ¯uorescence, guard cell, mesophyll,
photosynthesis, stomata, water stress.
Introduction
The guard cells that form the stomatal pore control the ¯ux
of CO2, H2O and other gases between the plant and the
atmosphere and are regulated by both internal and external
factors. Stomatal movements are due to the loss or
accumulation of ions, which require energy (Willmer and
Fricker, 1996). In the majority of species examined guard
cells contain well-developed chloroplasts, unlike the other
epidermal cells from which they are formed. The role of
these chloroplasts is still not clear, although they are not
always essential to stomatal function since achlorophyllous guard cells do open and close (in Paphiopedilum sp.,
Nelson and Mayo, 1975). Guard cell chloroplasts have
been proposed as signi®cant energy sources for H+
extrusion and ion transport (Wu and Assmann, 1993;
Tominaga et al., 2001), and are involved in several
different light transduction pathways (Zeiger et al., 2002).
Although many studies have shown that guard cells
contain several of the main Calvin cycle enzymes
(Shimazaki and Zeiger, 1985; Zemel and Gepstein, 1985;
Gotow et al., 1988; Shimazaki et al., 1989), few have
presented conclusive evidence for signi®cant Calvin cycle
activity within these cells (Outlaw, 1989; Reckmann et al.,
* To whom correspondence should be addressed. Fax: +44 (0)1206 873416. E-mail: [email protected]
Abbreviations: Ca, ambient CO2 concentration; F¢, chlorophyll ¯uorescence in the light-adapted state; Fm¢, chlorophyll ¯uorescence when PSII centres are
maximally closed in the light-adapted state; Fo¢, chlorophyll ¯uorescence when PSII centres are maximally open in the light-adapted state; Fq¢, difference
between F¢ and Fm¢; Fq¢/Fm¢, ¯uorescence parameter that provides an estimate of the operating ef®ciency of PSII photochemistry; Fq¢/Fv¢, factor relating the
operating and maximum ef®ciencies of PSII photochemistry; Fv¢, variable chlorophyll ¯uorescence (Fm¢±Fo¢); Fv¢/Fm¢, ¯uorescence parameter that provides an
estimate of the maximum ef®ciency of PSII photochemistry (when all PSII centres are open); PPFD, photosynthetic photon ¯ux density; Rubisco, ribulose
1,5-bisphosphate carboxylase oxygenase; VPD, vapour pressure de®cit.
Journal of Experimental Botany, Vol. 54, No. 388, ã Society for Experimental Biology 2003; all rights reserved
1744 Lawson et al.
1990), and there is continuing debate about the role and
nature of guard cell photosynthesis (Zeiger et al., 2002).
Chlorophyll ¯uorescence is a powerful, non-invasive
technique to investigate photosynthetic metabolism in
guard cells. The majority of chlorophyll ¯uorescence
measurements from guard cells have been restricted to
epidermal peels (Melis and Zeiger, 1982; Ogawa et al.,
1982) or guard cell protoplasts (Goh et al., 1997, 1999,
2001) or have used variegated leaves; they have mainly
been restricted to ¯uorescence induction curves or
responses of the steady-state ¯uorescence signal (F¢).
However, using high spatial resolution chlorophyll a
¯uorescence imaging in intact green leaves of
Commelina communis, it has recently been shown that
guard cell quantum ef®ciency for PSII photochemistry
(Fq¢/Fm¢=(Fm¢±F¢)/Fm¢, `photosynthetic ef®ciency') was
approximately 70±80% of that of the mesophyll cells
(Baker et al., 2001; Lawson et al., 2002) across a wide
range of light intensities. It has also been shown that
photosynthetic ef®ciency in both guard and mesophyll
cells of C. communis was similarly altered by O2
concentration at low, but not at high CO2 concentration,
indicating that photorespiration is a major sink for ATP
and NADPH produced through electron transport in guard
cells, and that Rubisco activity is signi®cant (Lawson et al.,
2002).
Some of the previous disagreements over the role of
chloroplast photosynthetic activity in guard cells could be
due to differences in plant material, as a range of species
are widely used in stomatal physiological studies from
very different taxa with different stomatal anatomy and
often grown in different conditions. As Zeiger et al. (2002)
have recently emphasized, guard cell chloroplasts show
remarkable functional plasticity. The aims of this study
were (1) to compare guard and mesophyll cell photosynthetic ef®ciencies in six species previously used in
stomatal physiology, and (2) to investigate the response
to different growth and measurement conditions, in
particular light, CO2 and O2 concentrations, and water
stress. The choice of species was largely dictated by
whether the stomata were large enough to be imaged
clearly, but included a fern (Polypodium vulgare) and a C4
species (Amaranthus caudatus) to compare with the C3
species usually used in stomatal physiology.
Materials and methods
Plant material
Seeds of Commelina communis L., Nicotiana tabacum L. and Vicia
faba L. were sown in a peat and loam based compost (F2, Levington
Horticulture Ltd, Ipswich, UK) in a controlled environment chamber
(SGC066 Fitotron, Sanyo Gallenkamp, Leicester, UK). After 3
weeks, seedlings were transplanted into 100 mm diameter pots and
used 6±7 weeks after sowing in the case of C. communis and
N. tabacum and 4±5 weeks in V. faba. Cuttings of the variegated
plant Tradescantia albi¯ora Kunth. were grown in the same compost
and environment chamber. The chamber air temperature was
maintained at 18 °C at night and 22 °C through the day. Light was
provided by halogen quartz iodide lamps (Neutralweiss, Germany)
from 06.00±21.00 h, at a constant PPFD of 530 mmol m±2 s±1 at plant
height. Relative humidity was maintained at 70% through the day
and 65% at night. Amaranthus caudatus L. and Polypodium vulgare
L. were grown from seed in the same compost and maintained in the
glasshouse where supplementary lighting was provided by highpressure sodium lamps and the temperature was maintained between
18 °C and 30 °C. All plants were kept well watered using capillary
matting, except those in the drought stress experiment.
Growth treatments
Drought stress: Plants of C. communis were grown in the controlled
environment chamber until about 6-weeks-old. Water was then
withheld from half the plants for 12 d, while control plants were kept
well watered. Leaf water potential was measured using a pressure
chamber (SKPM1400, Skye Instruments, Powys, UK). After 12 d,
stressed plants were rewatered, resulting in full recovery of the water
potential after 2 d.
Elevated CO2: C. communis and V. faba were grown in two
controlled environment chambers as above, but CO2 concentration
was maintained at either 360 or 700 mmol mol±1 using an injection
system operated by an infrared gas analyser (WMA-2, PP Systems).
Growth PPFD: Plants of C. communis and V. faba were grown in the
above controlled environment chambers, but half of the chamber
was shaded, using neutral density tissue paper to give PPFDs at the
height of the plants of approximately 640 and 260 mmol m±2 s±1.
The microscope imaging system
The optical part of the imaging microscope used in these experiments is essentially the same as that described previously
(Oxborough and Baker, 1997) with the modi®cation of the lower
light source and a purpose-designed microscope cuvette attached to
an infrared gas analyser to control CO2, O2 and VPD as described by
Lawson et al. (2002). CO2 of known concentration was supplied
through the gas analyser system (CIRAS1; PP Systems, Hitchin,
UK), and known O2 concentration was supplied from external gas
bottles (BOC, Surrey, UK) attached to the air inlet on the gas
analyser. Unless otherwise stated, conditions in the microscope
cuvette were 23±25°C with 21% O2, a Ca of 360 mmol mol±1 and a
VPD of approximately 0.6 kPa. Chlorophyll ¯uorescence was
de®ned by a 680 nm bandpass ®lter (Coherent, Watford, UK). Fo¢
and Fm¢ de®ne the minimal and maximal ¯uorescence levels from
leaves in the light, respectively. F¢ is the ¯uorescence level at any
point between Fo¢ and Fm¢. For the construction of parameterized
images, the speci®c term Fq¢ was recently introduced (Oxborough
and Baker, 1997; Oxborough et al., 2000) which denotes the
difference between Fm¢ and F¢ measured immediately before
application of a saturating pulse to measure Fm¢. Under these
conditions, Fq¢/ Fm¢ equates to the operating quantum ef®ciency of
PSII photochemistry. Images of Fv¢/Fm¢ and Fq¢/Fv¢ were generated
from images of Fo, Fm and Fm¢ as described previously (Lawson
et al., 2002). There was no attempt to estimate rates of electron
transport from Fq¢/ Fm¢ because there are uncertainties concerning
the exact light absorption and contribution of PSI ¯uorescence for
the guard and mesophyll cell chloroplasts. All images were taken
from the abaxial surface of leaves (except where stated) using a 403
objective, which provided images of 3103205 mm with a pixel size
of 534 nm2. Replicates are individual stomatal complexes on leaves
of different plants. The mesophyll areas used for comparison were
those immediately adjacent to the guard cells. Chloroplasts within
guard cell pairs were isolated from images using the ends-in search
Guard cell photosynthesis 1745
and other editing tools described in Oxborough and Baker (1997)
and Oxborough et al. (2000).
Statistical analysis
Mean values of chloroplast photosynthetic ef®ciencies (Fq¢/Fm¢)
were calculated from the images, and differences between species,
cell type (mesophyll or guard cell, abaxial or adaxial) or treatments
(light, CO2 or O2 concentration) were compared using ANOVAs
with mixed `between subjects' (e.g. species) and `within subjects'
(e.g. cell type, CO2 or O2 concentration) designs as appropriate. The
data in the regression in Fig. 6b were examined for the effect of any
treatments using analysis of covariance. All statistical analyses were
carried out with SPSS v. 10, or Systat v. 5.
Results
Fluorescence images and guard cell chloroplast
appearance
Images of steady-state ¯uorescence (F¢) from guard cells
in the six species examined are shown in Fig. 1, and
demonstrate the variation in stomatal pore and guard cell
size in the different species and the differences in
chloroplast number and orientation within individual
guard cells. For example, in N. tabacum (Fig. 1d) the
chloroplasts are located on the outer wall of the guard cells,
Fig. 1. Steady-state chlorophyll ¯uorescence images (F¢) obtained
under the microscope from intact leaves showing abaxial stomatalguard cells with chloroplasts of (a) Amaranthus caudatus; (b)
Commelina communis; (c) Polypodium vulgare; (d) Nicotiana
tabacum; (e) Tradescantia albi¯ora, and (f) Vicia faba.
whereas in T. albi¯ora (Fig. 1e) chloroplasts appear to be
distributed evenly throughout the cells around the vacuole.
Guard cells of the fern P. vulgare (Fig. 1c) contain many
chloroplasts, as do the other epidermal cells, as previously
noted (Willmer and Fricker, 1996). Despite these differences, all the chlorophyll ¯uorescence parameters were
readily measured and hence photosynthetic electron
transport ef®ciency (estimated by Fq¢/Fm¢) could be
calculated for both guard cells and underlying mesophyll
cells in all of the species.
Photosynthetic ef®ciencies in guard and mesophyll
cells in different species
There were signi®cant differences between species in
Fq¢/Fm¢ of guard and underlying mesophyll cells during
steady-state photosynthesis under identical measurement
conditions (species difference P <0.001; Fig. 2a). The fern
P. vulgare exhibited the lowest ef®ciencies, with guard
cell and mesophyll Fq¢/Fm¢ values of 0.24 and 0.31,
respectively, while the maximum values of 0.62 and 0.65,
respectively, were found in V. faba. The low values of
Fq¢/Fm¢ observed in P. vulgare are notable, particularly as
the large number of chloroplasts present might have been
expected to result in a high degree of self-shading, which
would reduce the average incident-light intensity and,
consequently, result in a higher photosynthetic ef®ciency.
One explanation might be the higher light level during
growth reducing photosynthetic ef®ciency as these plants
were glasshouse-grown, but A. caudatus was grown in a
similar environment and yet showed higher ef®ciencies.
For P. vulgare, N. tabacum and T. albi¯ora guard cell
Fq¢/Fm¢ values were 79%, 92% and 87% of the values of
the adjacent mesophyll cells, respectively, although no
signi®cant differences were observed for A. caudatus,
V. faba and C. communis (overall cell type difference
P <0.001; species3cell type interaction P=0.049), and
there was a close linear correlation between mesophyll and
guard cell Fq¢/Fm¢ (Fig. 2b).
Effect of growth and measurement PPFD
Photosynthetic ef®ciency is dependent both on measurement PPFD and on growth conditions which affect the
development of the photosynthetic apparatus, as exempli®ed by the large differences observed for sun and shade
leaves (Pearcy, 1998). In species that are amphistomatous,
stomatal guard cells provide an interesting system to
examine the effect of growth PPFD on photosynthetic
behaviour, as guard cells in the upper surface are exposed
to much higher PPFD than those on the lower surface. In
C. communis and V. faba, plants grown at moderate light
intensity (530 mmol m±2 s±1) guard cell Fq¢/Fm¢ in both
upper (adaxial) and lower (abaxial) leaf surfaces decreased
with increasing incident PPFD (Fig. 3; P <0.001). The two
species differed in the response of the photosynthetic
ef®ciency of the upper and lower guard cells to PPFD
1746 Lawson et al.
Fig. 2. (a) Comparison of guard cell (open bars) and mesophyll cell
(®lled bars) photosynthetic ef®ciency (estimated by Fq¢/Fm¢) in intact
leaves of six species. Measurements were taken after 10 min
stabilization under the following conditions Ca 360 mmol mol±1, 21%
O2, PPFD 114 mmol m±2 s±1, 24 °C and a VPD of 0.6 kPa. Data are
the means of three replicates 6SE. (b) Relationship between Fq¢/Fm¢
of guard cells and mesophyll cells for individual stomata of six
species under the conditions detailed above. The broken line
represents the y=x relationship. (Amar.=Amaranthus caudatus (®lled
diamonds);
Comm.=Commelina
communis
(®lled
squares);
Polyp.=Polypodium vulgare (®lled triangles); Nicot.=Nicotiana
tabacum (open squares); Trades.=Tradescantia albi¯ora (open
triangles); and Vicia=Vicia faba (®lled circles).
(P=0.003). In C. communis the decrease of Fq¢/Fm¢ was
substantially larger in guard cells in the lower surface
(Fig. 3a). In V. faba at the lowest measurement PPFD of 93
mmol m±2 s±1 guard cells in the lower surface operated at a
higher ef®ciency than those in the upper surface (Fig. 3b).
However, as PPFD increased there were large decreases in
ef®ciency in guard cells in both surfaces and no signi®cant
differences were observed between the guard cells on the
lower and upper surfaces (Fig. 3b). Fq¢/Fm¢ was lower in
guard cells of V. faba than those from C. communis at the
higher PPFD levels, particularly in the upper surface.
The in¯uence of growth PPFD on the lower surface
stomata and any possible acclimation to light was studied
by growing C. communis and V. faba at PPFD of 260 and
640 mmol m±2 s±1 and measuring photosynthetic ef®ciency
at PPFDs of 93 and 428 mmol m±2 s±1 (Fig. 4). While
Fq¢/Fm¢ declined in higher measurement PPFD in both cell
Fig. 3. Effect of measurement PPFD on guard and mesophyll cell
photosynthetic ef®ciency (estimated by Fq¢/Fm¢) from the lower (open
bars) and upper surface (®lled bars) in (a) Commelina communis and
(b) Vicia faba. Data are the means of three replicates 6SE.
types in both species (P <0.001), overall the photosynthetic
ef®ciencies measured in V. faba were about 0.10 lower
than those of C. communis (overall species difference
P=0.026), and there was a larger relative response of
Fq¢/Fm¢ to measurement PPFD in V. faba (P=0.012). There
was also a small (but signi®cant, P=0.018) difference
between species in the response to growth PPFD, with C.
communis plants grown at 640 mmol m±2 s±1 having a
higher Fq¢/Fm¢ for both guard and mesophyll cells,
regardless of the measurement PPFD (Fig. 4a). In contrast,
V. faba showed no differences in Fq¢/Fm¢ between plants
grown under the two different light intensities (Fig. 4b).
While there were no signi®cant differences between Fq¢/
Fm¢ of guard and mesophyll cells in each combination of
growth and measurement PPFD treatments for C. communis, guard cells in V. faba had signi®cantly lower Fq¢/
Fm¢ values (approximately 8%) than the mesophyll cells,
for all growth and measurement PPFDs (overall cell
type3species interaction P=0.018). Both the short-term
responses of Fq¢/Fm¢ to measurement PPFD and the longer
term acclimation to growth PPFD were very similar for
both guard and mesophyll cells (Fig. 4c).
Effect of CO2 and O2 concentration
The effects of CO2 and O2 concentration on the
photosynthetic ef®ciencies of the guard and mesophyll
cells were determined for the six species (Table 1). As in
Guard cell photosynthesis 1747
was a signi®cant overall CO23O2 interaction, P=0.036).
The reduction in Fq¢/Fm¢ in low [O2] or low [CO2] was
almost entirely due to a decrease in Fq¢/Fv¢, with little
change being observed in Fv¢/Fm¢ (data not shown),
indicating that the reduced availability of terminal electron
acceptors for electron transport from PSII was primarily
responsible for the decrease in photosynthetic ef®ciency. In
the C4 species no effects of [O2] on Fq¢/Fm¢ of the
mesophyll at either [CO2] were observed. Clearly, in low
[CO2] in the mesophyll of C3, but not C4 species, O2 is
acting as a sink for the products of photosynthetic electron
transport due to Rubisco oxygenase activity and photorespiratory metabolism. However, low [O2] decreased Fq¢/
Fm¢ of the guard cells in all six species (range from 7±24%)
at both CO2 concentrations (O2 effect P <0.001, no
CO23O2 interaction, P=0.301), indicating that Rubisco is
active in the guard cells of the C4 species as well as of the
C3 species.
Effect of growth CO2 concentration
Fig. 4. Effect of growth and measurement PPFD on guard and
mesophyll cell photosynthetic ef®ciency (estimated by Fq¢/Fm¢) for (a)
Commelina communis and (b) Vicia faba. Plants were grown at two
PPFD (260 and 640 mmol m±2 s±1) and measured at 93 and 428 mmol
m±2 s±1. Data are the means of three replicates 6SE. (c) Relationship
between Fq¢/Fm¢ of guard cells and mesophyll cells for individual
measurements on C. communis and V. faba at the different
measurement and growth PPFD. The broken line represents the y=x
relationship.
the previous experiments, Fq¢/Fm¢ of the guard cells were
lower than those of the adjacent mesophyll in some, but not
all, species (signi®cant cell type3species interaction
P=0.003). Overall, there was a more pronounced increase
of Fq¢/Fm¢ with an increase of [CO2] from 150 to 360 mmol
mol±1 for the C3 species than for the C4 species Amaranthus
caudatus (overall CO23species interaction P=0.036; if
data from A. caudatus are excluded there is no signi®cant
interaction, P=0.274). In mesophyll cells in the ®ve C3
species, decreases of [O2] from 21% to 2% decreased
Fq¢/Fm¢ substantially at the lower [CO2] (7±26% reduction), but not in higher [CO2] (there were no signi®cant
species differences in response to CO2, P=0.232, but there
The effects of growth [CO2] (360 or 700 mmol mol±1) on
Fq¢/Fm¢ was examined for C. communis and V. faba to
determine if photosynthetic acclimation resulted in different behaviour of guard cells compared with mesophyll
cells (Fig. 5). Fq¢/Fm¢ of guard cells was slightly (2±9%)
lower than that of mesophyll cells, except for V. faba
grown in high [CO2] where there was no difference
(overall cell type effect, P=0.002). Fq¢/Fm¢ in both guard
and mesophyll cells increased by approximately 10% with
an increase in measurement [CO2] from 150 to 360 mmol
mol±1, but not in higher [CO2]. The effect of growth [CO2]
on Fq¢/Fm¢ of both guard and mesophyll cells differed
between the two species (P=0.014). In C. communis,
growth in high [CO2] substantially reduced Fq¢/Fm¢ of both
guard and mesophyll cells in all CO2 measurement
concentrations (Fig. 5a; P <0.001), whereas V. faba
showed no signi®cant effect of growth CO2 concentration
(Fig. 5b). The decrease in Fq¢/Fm¢ in C. communis was
attributable to a decrease in Fq¢/Fv¢ since Fv¢/Fm¢ did not
change signi®cantly (data not shown); again this indicates
that the decrease in photosynthetic ef®ciency is due to a
decrease in the ability to use the products of electron
transport.
Effect of water stress on the relationship between
guard and mesophyll cells
Close correlations were found between the values of
Fq¢/Fm¢ of guard and adjacent mesophyll cells for different
species and in differing measurement and growth conditions (Figs 2c, 4c, 5c). To examine if such correlations
between the guard and mesophyll cell photosynthetic
ef®ciencies are conserved during water stress, C. communis
plants were subjected to mild water stress by withholding
water for 12 d. Leaf water potentials decreased to
approximately ±0.75 MPa in the drought-stressed plants,
1748 Lawson et al.
Table 1. Response of guard and mesophyll cell Fq¢/Fm¢ in intact leaves of six different species to reductions in O2 from 21% to
2% at high (360 mmol mol±1) and low (150 mmol mol±1) CO2 concentration
PPFD was approximately 240 mmol m±2 s±1. Data are the means of three replicates, and the pooled
Species
150 mmol mol
±1
Amaranthus caudatus
Vicia faba
Nicotiana tabacum
Polypodium vulgare
Tradescantia albi¯ora
Commelina communis
Pooled SE
appropriate for each column is shown.
360 mmol mol±1 CO2
CO2
Guard cell
SE
Mesophyll
Guard cell
Mesophyll
21% O2
2% O2
21% O2
2% O2
21% O2
2% O2
21% O2
2% O2
0.528
0.547
0.453
0.139
0.263
0.537
0.019
0.507
0.489
0.422
0.108
0.200
0.495
0.017
0.538
0.606
0.533
0.219
0.340
0.544
0.025
0.555
0.553
0.483
0.181
0.253
0.508
0.025
0.562
0.640
0.557
0.244
0.387
0.587
0.020
0.529
0.598
0.532
0.219
0.330
0.570
0.018
0.566
0.672
0.608
0.309
0.443
0.592
0.028
0.568
0.655
0.584
0.324
0.420
0.586
0.028
while well-watered controls remained at c. ±0.05 MPa
throughout. Stomatal aperture showed a steady decline as
water was withheld, but recovered within 48 h after
rewatering (data not shown). Both guard and mesophyll
cell Fq¢/Fm¢ declined during the stress, but recovered along
with stomatal aperture 48 h after rewatering (Fig. 6a). This
decrease in Fq¢/Fm¢ was again largely due to a decrease in
Fq¢/Fv¢ not in Fv¢/Fm¢ (data not shown), again indicating
the limitation on photosynthetic ef®ciency being the ability
to use the products of electron transport. However, there
was a close linear relationship between guard and
mesophyll cell Fq¢/Fm¢ as the values changed through the
drying cycle which was not distinguishable from that of the
control plants (Fig. 6b, pooled regression for all treatments
r2=0.902). Increasing measurement Ca from 360 to
700 mmol mol±1 resulted in a substantial increase in
Fq¢/Fm¢ of both guard and mesophyll cells in the waterstressed plants (mean increase at the end of the cycle was
from 0.23 to 0.37 for mesophyll cells), but only a small
increase was observed in control plants (Fig. 6b).
However, even with increased CO2 concentration Fq¢/Fm¢
of drought-stressed plants did not increase to that of the
control plants. This suggests that as water stress increased,
stomata adjusted to reduce water loss, and restricted CO2
diffusion into the leaf, resulting in a decrease in Fq¢/Fm¢
through a reduction in the capacity for the consumption
of the products of electron transport. The relationship
between Fq¢/Fm¢ of the guard and mesophyll cells was not
affected by CO2 or water stress (although there is a
statistically signi®cant difference of the intercept for
the water stress treatment (P <0.001) at 0.028 it is
negligible).
Discussion
Much of the previous work investigating guard cell
photosynthesis has been con®ned to species such as
C. communis or V. faba due to the ease with which the
epidermis can be removed from the leaf and the ability to
obtain uncontaminated guard cell protoplasts (Willmer and
Fricker, 1996). In this study, photosynthetic ef®ciency in
guard cells during steady-state photosynthesis has been
measured in intact leaves from a number of species.
Taxonomically, these species are very diverse, including
one fern and ®ve angiosperms, and the angiosperms
include two monocots (C. communis and T. albi¯ora, both
in the Commelinaceae) and three dicots, each from
families in different subclasses (V. faba: Fabaceae,
subclass Rosidae; A. caudatus: Amaranthaceae, Caryophillidae; and N. tabacum: Solanaceae, Asteridae).
Consequently, it is believed that these conclusions on the
responses of the photosynthetic ef®ciency of guard cells to
environmental factors have considerable general application. In leaves of all species the values of photosynthetic
ef®ciency for guard cells were either indistinguishable
from or only slightly lower (minimum of 79%) than those
of the underlying, spongy mesophyll cells. In all species
examined the responses of guard and mesophyll photosynthetic ef®ciency to changes in light, [CO2] and [O2]
were similar, although there were demonstrable differences between species in the actual values of guard cells
and mesophyll, and their responses to growing conditions
(Figs 2±5; Table 1). Some of these differences may be due
to the plants being grown under different environmental
conditions, as it has been demonstrated that photosynthetic
ef®ciency can vary with the growth environment (Figs 4±
6). However, the lower values of Fq¢/Fm¢ for T. albi¯ora
than for four other species from the chambers is consistent
with previous work (Lawson et al., 2002). The lower
ef®ciencies observed in T. albi¯ora and P. vulgare are also
re¯ected in lower photosynthetic rates compared with the
other species, indicating lower intrinsic light use ef®ciencies for photosynthesis. These two species also tend to
have a greater number of chloroplasts in the guard cells
than other species, which may result in increased light
absorption and a reduced ef®ciency at any given incident
PPFD.
Guard cell photosynthesis 1749
Fig. 5. Effect of growth and measurement CO2 on guard and
mesophyll cells for (a) Commelina communis and (b) Vicia faba.
Plants were grown at CO2 concentrations of 360 and 700 mmol mol±1
and measured in 150, 360 and 700 mmol mol±1 CO2. Data are the
means of three replicates 6SE. (c) Relationship between Fq¢/Fm¢ of
guard cells and mesophyll cells for individual measurements on C.
communis and V. faba at the different measurement and growth CO2
concentrations. The broken line represents the y=x relationship.
Effect of light
Some differences were observed in the values and
responses of photosynthetic ef®ciency between the guard
cells on the adaxial and abaxial surfaces, and the way it
was affected by PPFD (Fig. 3). In V. faba a difference was
observed only at low PPFD, agreeing with the report of
similar photosynthetic rates per unit chlorophyll for
adaxial and abaxial guard cell protoplasts from the same
species by Goh et al. (1997). In C. communis the different
light responses of the guard cells in adaxial and abaxial
surfaces were similar to those seen in leaf photosynthetic
responses for sun and shade leaves, respectively (Pearcy,
1998). In addition, photosynthetic ef®ciency of C. communis
Fig. 6. (a) Time-course of guard (squares) and mesophyll cell
(triangles) photosynthetic ef®ciency (estimated by Fq¢/Fm¢) for
Commelina communis for either drought-stressed plants (open
symbols) or well-watered plants (solid symbols). Arrow indicates
time of rewatering. Data are the means of three replicates 6SE.
(b) Relationship between guard cell and mesophyll cells Fq¢/Fm¢ for
well-watered (closed symbols) and drought-stressed plants (open
symbols). Ambient CO2 concentration was maintained at 360 or 700
mmol mol±1. The broken line represents the y=x relationship, and the
solid line a linear regression with the equation y=0.914x, r2=0.902.
guard and mesophyll cells was affected by the PPFD
during growth, while this was not the case in V. faba. The
contrast between the two species may be due to the growth
habit. Leaves of C. communis are more horizontal,
consequently the adaxial surface is better illuminated
than in V. faba, and the abaxial surface only receives the
light that is transmitted though the leaf. On the other hand,
leaves of V. faba are less rigid in their angle to the stems,
and are frequently twisted, exposing both surfaces to
similar, but reduced, PPFD. Such differences in photosynthetic ef®ciency may lie behind the higher sensitivity of
abaxial stomata to light (Travis and Mans®eld, 1981), and
may be associated with differences in guard cell pigment
content (Lu et al., 1993; Goh et al., 1997).
Effect of changes in CO2 and O2
In the C3 species, the photosynthetic ef®ciencies of the
guard and mesophyll cells were reduced at low CO2
1750 Lawson et al.
concentration (Fig. 4; Table 1). Similar reductions were
observed in Fq¢/Fm¢ of both the guard and the mesophyll
cells in ®ve C3 species when the ambient O2 was decreased
from 21% to 2% at low CO2 concentrations (Table 1). We
argue, as others have done (Cardon and Berry, 1992), that
such effects of CO2 and O2 on photosynthesis indicate that
Rubisco is a major sink for the products of photosynthetic
electron transport in guard cells, and con®rm and extend
earlier evidence for this from T. albi¯ora (Lawson et al.,
2002). The argument is strengthened by the lack of
response of Fq¢/Fm¢ to [O2] in the mesophyll in
Amaranthus caudatus, a C4 species, which is as expected
as there is no Rubisco activity in mesophyll cells of C4
leaves. However, there was an O2 response in the guard
cells (Table 1) which agrees with recent immunogold
labelling studies on Amaranthus viridis which showed
substantial amounts of Rubisco in guard cells, but no
labelling in mesophyll cells (Ueno, 2001). Consequently, it
appears that although the mesophyll cells of the C4 leaves
lack Rubisco and do not operate a carbon reduction cycle
as expected, the guard cells do exhibit Rubisco and Calvin
cycle activity.
Furthermore, the guard cells in C. communis showed the
same reduction in Fq¢/Fm¢ that occurred in the mesophyll
cells in response to growth in high [CO2] (Fig. 5). The
current explanations for photosynthetic acclimation during
growth in high CO2 are either that increased carbohydrate
supply is not matched by sink demand, thus inhibiting
Calvin cycle activity either by limiting RuBP regeneration
or by causing the down-regulation of Rubisco or that the
amount of Rubisco is reduced due to a change in N
allocation (Drake et al., 1997). Therefore, the matching
reduction of photosynthetic ef®ciency in the guard and
mesophyll cells in high CO2 also argues for comparable
Calvin cycle activity in guard and mesophyll cells. Such
Rubisco regulation in both guard and mesophyll cells may
be the mechanism behind the parallel acclimation of
stomatal conductance and mesophyll photosynthesis to
high [CO2] sometimes observed (Morison, 1998;
Assmann, 1999; Lodge et al., 2001).
Effect of water stress
During slowly imposed water stress, there were parallel
declines in the photosynthetic ef®ciencies of the guard and
mesophyll cells over a time-course of days (Fig. 6). Goh
et al. (2001) have described declines in photosynthetic
ef®ciency in the guard and mesophyll cell protoplasts
under hypertonic osmotic stress, but this was accompanied
by declines in photochemical and non-photochemical
quenching. In their experiment there were differences
between the guard and mesophyll cells, but they imposed
rather severe osmotic stress on cells without walls. Some
of the decline in photosynthetic ef®ciency in this study was
due to reductions in CO2 supply since doubling Ca
increased Fq¢/Fm¢ and reductions of Ca below ambient
reduced Fq¢/Fm¢ markedly, as shown in Fig. 5 (see also
Lawson et al., 2002). However, it is also likely that some
of the decline in Fq¢/Fm¢ was due to other water stress
effects causing `impaired photosynthetic metabolism' (see
review by Lawlor, 2002) because increased Ca did not
completely offset the decline in Fq¢/Fm¢. Previous work
indicates that Ca, and by inference intercellular [CO2],
would have to be very low to cause the declines observed
here (Lawson et al., 2002; Fig. 5b). While these results
cannot distinguish which of the possible mechanisms are
involved in this photosynthetic inhibition, an important
result is that both the guard and mesophyll cells are
similarly affected. This suggests that if guard cell
photosynthetic electron transport or Calvin cycle activity
is important to aperture maintenance, then there must be an
additional positive feedback of reduced photosynthesis in
guard cells during water stress, contributing to reductions
in stomatal aperture.
Photosynthesis in guard cells
There is a long-standing controversy over the role and
activity of the Calvin cycle in guard cells (see recent
review by Zeiger et al., 2002). A possible explanation for
the discrepancy between results indicating substantial
Calvin cycle activity and those concluding that there is
little activity is that photosynthetic regulation in guard
cells re¯ects the pretreatment of leaves and the measurement conditions (Zeiger et al., 2002). For example, it has
been proposed that the accumulation of sucrose near the
guard cells in the apoplastic phloem loader V. faba may
suppress Calvin cycle enzymes (Lu et al., 1997) so that
studies of guard cell photosynthetic activities using intact
leaves may give different results to those with peels.
Secondly, Talbott and Zeiger (1996) observed changes in
the osmotic regulation of guard cells of V. faba during the
day, with K+ being the main osmoticum early in the day,
but replaced by sucrose later. Thirdly, the light regulation
of stomatal function is complex: both blue and red light
stimulate photosynthesis and sucrose accumulation in
guard cells (Tallman and Zeiger, 1988; Talbott and
Zeiger, 1993), but the relative proportion of red and blue
light appears to change the balance between the starch±
sugar and the K+±malate osmotic mechanisms (Talbott and
Zeiger, 1993). The present work provides strong evidence
for Calvin cycle activity in guard cells of six species,
including V. faba, during steady-state photosynthesis in
intact leaves, using moderate to high ¯uence rates of blue
light when sucrose supply in the apoplast should have been
substantial. Such moderate and high ¯uence rates will have
driven normal photosynthetic activity in the guard cells
(Wu and Assmann, 1993) and this probably contributed to
stomatal opening in addition to any low intensity, chlorophyll-independent blue light response (Tallman and
Zeiger, 1988; Taylor and Assmann, 2001). Furthermore,
these measurements of guard cell photosynthetic ef®ciency
Guard cell photosynthesis 1751
agree well with those of Goh et al. (1999, 2001) using
chlorophyll ¯uorescence in the guard cell protoplasts.
An important result emerging from these studies is that
photosynthetic ef®ciency of guard cells in intact leaves
responded quantitatively to light, CO2, O2 and water stress
in a similar way to adjacent mesophyll cells. As highlighted by Wong et al. (1979), there is often a close
positive correlation at the leaf scale between stomatal
conductance and mesophyll CO2 assimilation rate across a
range of environmental conditions. This close relationship
has been attributed to the in¯uence of internal CO2
concentration (Raschke, 1976), but there have also been
suggestions that there is another signal transmitted from
the mesophyll cells to the guard cells such that mesophyll
photosynthesis controls the degree of stomatal opening
(Heath and Russell, 1954; Wong et al., 1979; Lee and
Bowling, 1995). However, the nature of any such
messengers is not clear. Sucrose movement within the
transpiration stream has been suggested recently by
Outlaw and colleagues, as this has been shown to be a
major source of organic carbon for the guard cells, and can
also exert an osmotic effect by accumulation in the cell
apoplast (Lu et al., 1997; Outlaw and De Vlieghere-He,
2001). Alternatively, photosynthetic metabolism in the
guard cells may be behind the co-ordination of the stomata
and the mesophyll (Farquhar and Wong, 1984; Jarvis and
Davies, 1998). Figures 2c, 4c, 5c, and 6b show that a close,
linear correlation between the guard cell and mesophyll
photosynthetic activity exists at the cell level. This
suggests that the guard cell photosynthetic activity may
provide the sensing mechanism linking stomatal movement to mesophyll photosynthetic rate.
Acknowledgement
This work was funded by the BBSRC (grant 84/P10409).
References
Assmann SM. 1999. The cellular basis of guard cell sensing of
rising CO2. Plant, Cell and Environment 22, 629±637.
Baker NR, Oxborough K, Lawson T, Morison JIL. 2001. High
resolution imaging of photosynthetic activities of tissues, cells
and chloroplasts in leaves. Journal of Experimental Botany 52,
615±621.
Cardon ZG, Berry J. 1992. Effects of O2 and CO2 concentration
on the steady-state ¯uorescence yield of single guard cell pairs in
intact leaf discs of Tradescantia albi¯ora. Plant Physiology 99,
1238±1244.
Drake BG, Gonzalez-Meler MA, Long SP. 1997. More ef®cient
plants: a consequence of rising atmospheric CO2? Annual Review
of Plant Physiology and Plant Molecular Biology 48, 609±639.
Farquhar GD, Wong SC. 1984. An empirical model of stomatal
conductance. Australian Journal of Plant Physiology 11, 191±210.
Goh C-H, Hedrich R, Schreiber U. 2001. Osmotic stress induces
inactivation of photosynthesis in guard cell protoplasts of Vicia
leaves. Plant and Cell Physiology 42, 1186±1191.
Goh C-H, Oku T, Shimazaki K. 1997. Photosynthetic properties
of adaxial guard cells from Vicia leaves. Plant Science 127, 149±
159.
Goh C-H, Schreiber U, Hedrich R. 1999. New approach of
monitoring changes in chlorophyll a ¯uorescence of single guard
cells and protoplasts in response to physiological stimuli. Plant,
Cell and Environment 22, 1057±1070.
Gotow K, Taylor S, Zeiger E. 1988. Photosynthetic carbon
®xation in guard cell protoplasts of Vicia faba L. Plant
Physiology 86, 700±705.
Heath OVS, Russell J. 1954. Studies in stomatal behaviour. VI. An
investigation of the light responses of wheat stomata with the
attempted elimination of control by the mesophyll. Part 2.
Interactions with external carbon dioxide and general discussion.
Journal of Experimental Botany 5, 269±292.
Jarvis AJ, Davies WJ. 1998. The coupled response of stomatal
conductance to photosynthesis and transpiration. Journal of
Experimental Botany 49, 399±406.
Lawlor DW. 2002. Carbon and nitrogen assimilation in relation to
yield: mechanisms are the key to understanding production
systems. Journal of Experimental Botany 53, 773±787.
Lawson T, Oxborough K, Morison JIL, Baker NR. 2002.
Responses of photosynthetic electron transport in stomatal guard
cells and mesophyll cells in intact leaves to light, CO2 and
humidity. Plant Physiology 128, 52±62.
Lee J, Bowling DJF. 1995. In¯uence of the mesophyll on stomatal
opening. Australian Journal of Plant Physiology 22, 357±363.
Lodge RJ, Dijkstra P, Drake BG, Morison JIL. 2001. Stomatal
acclimation to increased CO2 concentration in a Florida scrub oak
species Quercus myrtifolia Willd. Plant, Cell and Environment
24, 77±88.
Lu P, Outlaw WH Jr, Smith BG, Freed GA. 1997. A new
mechanism for the regulation of stomatal aperture size in intact
leaves. Accumulation of mesophyll derived sucrose in the guard
cell wall of Vicia faba L. Plant Physiology 114, 109±114.
Lu Z, Quinones MA, Zeiger E. 1993. Abaxial and adaxial stomata
from Pima cotton (Gossypium barbadense L.) differ in their
pigment content and sensitivity to light quality. Plant, Cell and
Environment 16, 851±858.
Melis A, Zeiger E. 1982. Chlorophyll a ¯uorescence transients in
mesophyll and guard cells. Plant Physiology 69, 642±647.
Morison JIL. 1998. Stomatal response to increased CO2
concentration. Journal of Experimental Botany 49, 443±453.
Nelson SP, Mayo JM. 1975. The occurrence of functional nonchlorophyllous guard cells in Paphiopedilum spp. Canadian
Journal of Botany 53, 1±7.
Ogawa T, Grantz D, Boyer J, Govindjee. 1982. Effects of cations
and abscisic acid on chlorophyll a ¯uorescence in guard cells of
Vicia faba. Plant Physiology 69, 1140±1144
Outlaw Jr WH. 1989. Critical examination of the quantitative
evidence for and against photosynthetic CO2 ®xation by guard
cells. Physiologia Plantarum 77, 275±281.
Outlaw Jr WH, De Vlieghere-He X. 2001. Transpiration rate: an
important factor controlling the sucrose content of the guard cell
apoplast of broad bean. Plant Physiology 126, 1716±1720.
Oxborough K, Baker NR. 1997. An instrument capable of imaging
chlorophyll a ¯uorescence from intact leaves at very low
irradiance and at cellular and subcellular levels of organization.
Plant, Cell and Environment 20, 1473±1483.
Oxborough K, Hanlon ARM, Underwood GJC, Baker NR.
2000. In vivo estimation of the photosystem II photochemical
ef®ciency of individual microphytobenthic cells using highresolution imaging of chlorophyll a ¯uorescence. Limnology and
Oceanography 45, 1420±1425.
Pearcy RW. 1998. Acclimation to sun and shade. In: Raghavendra
AS, ed. Photosynthesis: a comprehensive treatise. Cambridge:
Cambridge University Press, 250±263.
1752 Lawson et al.
Raschke K. 1976. Stomatal action. Annual Review of Plant
Physiology 26, 309±340.
Reckmann U, Scheibe R, Raschke K. 1990. Rubisco activity in
guard cells compared with the solute requirement for stomatal
opening. Plant Physiology 92, 246±253.
Shimazaki K, Terada J, Tanaka K, Kondo N. 1989. Calvin±
Benson cycle enzymes in guard cell chloroplasts from Vicia faba
L. Plant Physiology 90, 1057±1064.
Shimazaki K-I, Zeiger E. 1985. Cyclic and non-cyclic
photophosphorylation in isolated guard cell chloroplasts from
Vicia faba L. Plant Physiology 78, 211±214.
Tallman G, Zeiger E. 1988. Light quality and osmoregulation in
Vicia faba guard cells: evidence for involvement of three
metabolic pathways. Plant Physiology 88, 887±895.
Talbott LD, Zeiger E. 1993. Sugar and organic acid accumulation
in guard cells of Vicia faba in response to red and blue light.
Plant Physiology 102, 1163±1169.
Talbott LD, Zeiger E. 1996. Central roles for potassium and
sucrose in guard cell osmoregulation. Plant Physiology 111,
1051±1057.
Taylor AR, Assmann SM. 2001. Apparent absence of a redox
requirement for blue light activation of pump current in broad
bean guard cells. Plant Physiology 125, 329±338.
Tominaga M, Kinoshita T, Shimazaki K. 2001. Guard cell
chloroplasts provide ATP required for H+ pumping in the plasma
membrane and stomatal opening. Plant and Cell Physiology 42,
795±802.
Travis AJ, Mans®eld TA. 1981. Light saturation of stomatal
opening on the adaxial and abaxial epidermis of Commelina
communis. Journal of Experimental Botany 32, 1169±1179.
Ueno O. 2001. Ultrastructural localization of photosynthetic and
photorespiratory enzymes in epidermal, mesophyll, bundle
sheath, and vascular bundle cells of the C4 dicot Amaranthus
viridis. Journal of Experimental Botany 52, 1003- 1013.
Willmer C, Fricker M. 1996. Stomata, 2nd edn. London: Chapman
and Hall.
Wong SC, Cowan IR, Farquhar GD. 1979. Stomatal
conductance correlates with photosynthetic capacity. Nature
282, 424±426.
Wu W, Assman SM. 1993. Photosynthesis by guard cell
chloroplasts of Vicia faba L.: effects of factors associated with
stomatal movement. Plant and Cell Physiology 34, 1015±
1022.
Zemel E, Gepstein S. 1985. Immunological evidence for the
presence of ribulose bisphosphate carboxylase in guard cell
chloroplasts. Plant Physiology 78, 586±590.
Zeiger E, Talbott LD, Frechilla S, Srivastava A, Zhu J. 2002.
The guard cell chloroplast: a perspective for the twenty-®rst
century. New Phytologist 153, 415±424.