| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
Plant Physiol, April 2000, Vol. 122, pp. 1399-1416 The Effects of Elevated CO2 Concentrations on Cell Division Rates, Growth Patterns, and Blade Anatomy in Young Wheat Plants Are Modulated by Factors Related to Leaf Position, Vernalization, and GenotypeResearch School of Biological Sciences, Institute of Advanced Studies, Australian National University, Canberra, Australian Capital Territory 2601, Australia
This study demonstrates that elevated [CO2] has profound effects on cell division and expansion in developing wheat (Triticum aestivum L.) leaves and on the quantitative integration of these processes in whole-leaf growth kinetics, anatomy, and carbon content. The expression of these effects, however, is modified by intrinsic factors related to genetic makeup and leaf position, and also by exposure to low vernalizing temperatures at germination. Beyond these interactions, leaf developmental responses to elevated [CO2] in wheat share several remarkable features that were conserved across all leaves examined. Most significantly: (a) the contribution of [CO2] effects on meristem size and activity in driving differences in whole-blade growth kinetics and final dimensions; (b) an anisotropy in cellular growth responses to elevated [CO2], with final cell length and expansion in the paradermal plane being highly conserved, even when the rates and duration of cell elongation were modified, while cell cross-sectional areas were increased; (c) tissue-specific effects of elevated [CO2], with significant modifications of mesophyll anatomy, including an increased extension of intercellular air spaces and the formation of, on average, one extra cell layer, while epidermal anatomy was mostly unaltered. Our results indicate complex developmental regulations of sugar effects in expanding leaves that are subjected to genetic variation and influenced by environmental cues important in the promotion of floral initiation. They also provide insights into apparently contradictory and inconsistent conclusions of published CO2 enrichment studies in wheat.
The consequences for plant growth and morphogenesis of variations in photosynthesis depend on the efficiency of the conversion of triose-P into Suc and of phloem loading at the sites of Suc production and unloading in growing sinks. Ultimately, however, they depend on the sensitivity to sugar supply of a suite of developmental processes involved in meristem initiation, cell division, expansion, and differentiation, and of the mechanisms that regulate the integration of these processes in the formation of organs with a certain shape, size, and structure. The aim of the present study was to investigate this latter area using atmospheric [CO2] as a tool for manipulating sugar supply to expanding organs and wheat (Triticum aestivum L.) as a model experimental system. Based on experiments at the whole-plant or whole-leaf level, it has
been reported that, relative to other species, wheat is not very
responsive to elevated [CO2], especially at
early stages. It has been argued (Nicolas et al., 1993 The aim of the work presented here was to examine the effects of
elevated [CO2] on the spatial and temporal
patterns of cell division and cell expansion in developing wheat leaves
and on their translation into variations of whole-leaf growth kinetics, anatomical features, and carbon content. The analysis was conducted in
two genotypes. Vernalization was used as a way of shifting the timing
of floral initiation and associated changes in carbon allocation within
the plant (e.g. Griffiths and Lyndon, 1985
Growth Conditions Wheat (Triticum aestivum L.) plants were grown from
seed in two adjacent, well-ventilated greenhouses matched for
temperature (23°C ± 0.1°C SE day and
night) and relative humidity (62% ± 1.0% during the day and 52% ± 1.2% at night), but providing contrasted atmospheric
CO2 concentrations. One greenhouse was run at
present ambient CO2 concentrations, i.e. on a
typical day, 350 ± 10 ppm during the day and 420 ± 19 ppm at night, while the other greenhouse was run at an elevated
CO2 concentration of 900 ± 12 ppm, day and
night. Day length increased slowly over the duration of the experiment
(4 weeks) from 11 to 12 h according to seasonal variations in
Canberra, Australia at that time of the year (August to September). The
incident radiation varied from day to day according to outside conditions around a daily average of 718 ± 40 µmol quanta
m Two wheat cultivars of contrasting genetic background, morphology, and
growth habit were used: cv Hartog, a selection from the CIMMYT (Centro
Internacional de Mejoramiento de Maiz y Trigo) variety Pavon 76, carrier of the Rht2 dwarf gene and classified as a spring wheat, and cv
Birch 75, a true winter wheat derived from crosses between English and
CIMMYT lines. cv Birch and cv Hartog seeds were germinated in Petri
dishes on wet filter paper and vernalized at 2°C to 3°C in the dark
in a cold room for 7 weeks. By then, seminal roots were 30 to 40 mm
long, and the first leaf 20 to 40 mm long. Five days before sowing,
another batch of seeds was germinated in similar conditions and kept in
the dark at 23°C to obtain non-vernalized "control" seedlings of
similar size at sowing as the vernalized ones. All seedlings were
transplanted on the same day to pots filled with a 1:2 sand:perlite mix
saturated with nutrient solution (Hewitt and Smith, 1975 Destructive Growth Measurements Plants of cv Birch and cv Hartog were harvested from each
greenhouse on d 1, 3, 8, d 15 to 17 (date of sampling for detailed kinematic analysis of leaf elongation, see below), and finally on d 23 (cv Birch) or d 27 (cv Hartog). On d 3 and 8, only non-vernalized plants were sampled; on the other dates, both vernalized and
non-vernalized plants were harvested. Roots were cut at the crown
level; individual tillers were identified according to their position
on the plant (Masle, 1984 Kinematic Analysis of Leaf Elongation Leaf elongation rates were analyzed non-destructively through
daily measurements of the lengths of all visible leaves to the nearest
0.5 mm. The underlying cellular responses were investigated using the
kinematic approach pioneered by Goodwin and Stepka (1945) Leaf Elongation Rate, Cell Length Profile, and Size of the Growth Zone Leaf 6 was selected for this analysis. Following blade emergence, leaf length was measured every 3 h over a full day/night cycle. A regression line was fitted through these measurements over time (r2 > 0.999 in all cases), and the slope of that line was taken as a measurement of the leaf linear rate of elongation (E). Twenty-four hours after blade emergence, the leaf was quickly dissected from the plant under the microscope as close as possible to its insertion on the apex. The leaf was then immediately immersed in boiling methanol until all chlorophyll had been removed and then cleared in lactic acid. Five leaves were analyzed for each treatment (2 vernalization levels × 2 CO2 concentrations × 2 genotypes). Because of [CO2] and vernalization effects on the rate of leaf development (see "Results"), leaves 6 were not all synchronized and were, depending on treatment, sampled over 2 d (d 15-17). The cleared leaf was mounted on a light microscope (axioscope, Zeiss, Jena, Germany) fitted with a video camera (model WV-CL 702E, Panasonic, Tokyo). A file of epidermal cells of the same type for all leaves (file of sister cells adjacent to a stomatal row) was selected and individual cell lengths were measured throughout the growth zone. This was done from video images using the morphometric program MTV (Datacrunch, San Clemente, CA). Note was taken of the position of thin transverse cell walls, which are indicative of recent cell divisions. Such walls were typically visible only in the basal 4 to 7 mm of the growth zone. The position of the most distal fresh wall was taken as defining the transition between the division and the elongation-only zone, where cells had lost the ability to divide and were starting to undergo rapid expansion. In many leaves, the last divisions were "asymmetrical," yielding two daughter cells of very unequal length. The shortest daughter cells could easily be traced through the elongation zone, where they hardly expanded and started to differentiate into trichomes. The positions from the base of the leaf (x0) of the last symmetrical and asymmetrical divisions were denoted as xsd and xad, respectively. Individual cell lengths were also measured along a 10-mm segment of mature blade and their average was taken as an estimate of final cell length, lf. Individual cell lengths, l(x), and individual elemental lengths, l(x)* (the length of a cell and its associated trichome), were plotted against position (x) along the growth zone. In the elongation-only zone, cell lengths were fitted by a Richards function as in Morris and Silk (1992) for cells and Beemster
et al. (1996) for elements. The distal end of the growth zone was
defined as the location (xel) where
the fitted cell length reached 95% of
lf. For the meristem, where cell
length does not bear any direct relationship to position, cell length
data were smoothed by calculating moving averages over 11 cell
intervals around x. The number of meristematic cells along a
file in the leaf growth zone were denoted as
Nsd and
Nad for the zones of symmetrical and
asymmetrical divisions, respectively, and the number of elongating-only
cells was denoted as Nel.
Growth Velocities and Strain Rates The velocities of displacement v(x) and relative cell elongation rates r(x), the latter often referred to as strain rates in the kinematic literature, can be calculated as a function of position x. Velocities in the elongation-only zone are given by equation 11 in Morris and Silk (1992):
 (x) is the number of newly formed
cross-walls between the base of the meristem and position x
as a proportion of such walls from the base to
xsd (Beemster et al., 1996).
The relative elongation rates r(x) are the
derivatives of velocities with respect to position (equation 2 in
Morris and Silk, 1992):
).
Cell Partitioning Rates The average cell partitioning rate, , in
the zone of symmetrical division and its inverse, the average cell
cycling time,  c, time
elapsed between two successive divisions, were calculated as in Green
(1976):
as the average
partitioning rate over intervals (i) of m cells
around the cell at location x, using:
(i,x) denotes the number of newly
formed cross walls found in interval i, as a proportion of
the total number of such walls from xo
to xsd. Calculations were done on
intervals of 20 cells.
Quantitative Analysis of Mature Blade Anatomy At final harvest, mature blades of leaf 6 were sampled for detailed anatomical observations. Four contiguous short segments, each about 3 mm long (denoted segment 1-4 below), were taken mid-length along the blade. Morphometric Analysis of Cleared Mature Epidermis Segment 1 was cleared in methanol and lactic acid as described above. The numbers of epidermal files between veins 1 and 2, 2 and 3, and 3 and 4 were counted (veins numbered from the mid-rib). An 800-µm-long field of view delimited by veins 2 and 3, equivalent to 0.2 to 0.3 mm2 of leaf tissue, was selected on the abaxial side of the blade for determining the densities and relative proportions of the various types of cells constituting the mature epidermis (stomata and interstomatal cells, sister cells in adjacent rows, elongated non-specialized cells, and trichomes). The length and maximum width of 10 contiguous cells of each type were measured, excluding stomata and trichomes. This was replicated three times per leaf.Morphometric Analysis of the Mesophyll Tissue on Leaf Sections Segments 2 and 3 were immediately fixed in 2% (v/v) glutaraldehyde in 50 mM 1,4-piperazinediethanesulfonic acid (PIPES) buffer for 3 h (1 h of which was under vacuum), post-fixed in 1% (v/v) osmium tetroxide in 25 mM phosphate buffer, pH 7.0, slowly dehydrated in ethanol, and embedded in Spurr's resin. Three-micrometer-thick cross-sections (segment 2) and longitudinal sections (segment 3) were cut, stained with toluidine blue, and mounted on a microscope fitted with a video camera for morphometric measurements. Cross-sections were used to determine the number of mesophyll cell layers and accurately measure the thicknesses of the whole blade, mesophyll tissue, and each epidermis, and the cross-sectional areas of individual mesophyll and epidermal cells. These parameters were measured at three positions across the blade centered mid-way between veins 1 and 2, 2 and 3, and 3 and 4. At each position the individual cross-sectional areas (aj) of all mesophyll cells comprised within a small rectangle of total area A were measured. The difference 1  j=1paj/A was
calculated as an estimate of the proportion of tissue cross-sectional
area occupied by air spaces. Depending on treatment, the number of
mesophyll cells, p, measured at each location varied from 12 to 20. Four sections were analyzed at each of three locations across
the embedded segment, and there were four replicated leaves per
treatment. Mesophyll cell lengths were measured on 3-µm longitudinal sections cut parallel to veins. In wheat, mesophyll cells are elongated
with several deep lobes (e.g. Parker and Ford, 1982), so mesophyll cell
lengths were only measured for cells whose end walls were clearly
visible and abutting the adjacent cells in the file.
Estimation of Mesophyll Cell Numbers Segment 4 (segment symmetrical to segment 3 with respect to mid-rib) was used to estimate mesophyll cell numbers. The slice of fresh tissue was fixed in 3.5% (v/v) glutaraldehyde for 1 h at room temperature, then transferred into 0.1 M EDTA (pH 9.0), incubated for 3 h at 60°C, and stored at 4°C. The slice of tissue was later macerated in 5% (w/w) chromium trioxide at 4°C for 24 h until cells could be easily teased apart without damage. Mesophyll cell counts were made a few days later using a 0.1-mm-depth hemocytometer under a microscope (Zeiss) on five loadings per leaf and four leaves per treatment.Sugar Analyses Soluble sugars and starch were extracted twice from dried blade and root powder in 80% (v/v) ethanol. Color pigments were removed by the addition of activated charcoal (Norit SA3, Aldrich Company, Milwaukee, WI) to the extract, followed by centrifugation at 12,000 rpm for 10 min. The supernatant was dried down and resuspended in water. Glc, Fru, and Suc concentrations were determined enzymatically in three steps after the addition of a formulation of hexokinase and Glc-6-P dehydrogenase (Glc [HK] diagnostics reagent, Sigma-Aldrich, St. Louis) and of invertase and isomerase. Glc moieties were measured spectrophotometrically at 340 nm following each reaction. Statistics Treatment effects on kinematic and morphometric parameters were assessed by analysis of variance using General Linear Models algorithms (statistical package GLIM, version 3.77, 1985, Royal Statistical Society, London). In addition, differences in the spatial distributions of strain rates and partitioning rates were compared by the test of Kolomgorov-Smirnov.
[CO2] Effects on Whole-Plant Growth, Carbon Accumulation, and Leaf Expansion Plants under 900 ppm [CO2] grew significantly more than those under 350 ppm (Table I), showing a 52% to 93% increase in total dry weight at the end of the experiment and a 39% to 82% increase in leaf area. All tillers contributed to that response (Fig. 1), including the main tiller, where a growth stimulation by elevated [CO2] was detectable within 4 to 15 d after sowing (Fig. 2) and led to a 25% to 40% increase in dry weight at final harvest (d 23-27). Elevated [CO2] accelerated the emergence rate of secondary and tertiary tillers (data not shown), leading to plants with a greater number of concurrently expanding sinks at a given time. For those tillers, the size differences shown in Figure 1 therefore reflect the compounded effects of earlier appearance and faster growth rate thereafter. Although elevated [CO2] led to significant increases in the soluble sugar content in leaves (Fig. 3), this effect was insufficient to fully account for the increased dry weights shown in Table I and Figures 1 and 2.
Figure 4 describes the kinetics of elongation of successive leaves of the main stem, from emergence of the blade to completion of elongation. In the two genotypes examined, and whether seeds had been vernalized or not, increased atmospheric [CO2] stimulated the expansion growth of individual leaves. Remarkably, however, this stimulatory effect only became visible and significant beyond leaf 2 (vernalized plants) or even leaf 4 (non-vernalized plants). Moreover, although in later leaves elevated [CO2] systematically resulted in longer mature blades, the underlying reasons varied depending on genotype and vernalization treatment. In vernalized cv Hartog plants, for example, elevated [CO2] caused earlier blade emergence (see in Fig. 4 the 1-1.5 d lag between the two CO2 levels for leaves 4-8) and extended growth duration. In the non-vernalized seedlings of the two genotypes, however, these two parameters were not significantly affected by exposure to elevated [CO2], but the elongation rate of emerged blades was enhanced. In vernalized cv Birch plants, blades emerged earlier and elongated faster thereafter.
Kinematic Analysis of Leaf Elongation Cell Flux and Mature Cell Length An analysis of the cellular responses underlying the effects of elevated [CO2] on leaf elongation was undertaken in leaf 6. Table II gives the average blade elongation rate (E) over the 24-h period following tip blade emergence. For each individual leaf, E was estimated by the regression line fitted to the three hourly measurements of blade lengths taken during that period (see "Materials and Methods"). Consistent with data shown in Figure 4, E was always greater under 900 ppm than 350 ppm. Vernalization had no significant effect except in cv Birch under high [CO2]. Variations in E can be analyzed using Equation 2 in relation to variations in F, the flux of cells moving out of the growth zone per unit of time, and in lf*, the length of the newly fully expanded elements. Table II shows that elevated [CO2] caused a 15% to 20% increase in cell flux (P < 5%) irrespective of genotype and vernalization treatment. In cv Birch, F was also affected by vernalization, although to a lesser extent, being 12% greater in non-vernalized than in vernalized seedlings. In contrast, final elemental lengths and final cell lengths were remarkably stable across growth [CO2] levels, vernalization treatments, and genotype (see Table II; actual overall mean cell length 200 ± 10 µm; predicted mean from the analysis of variance = 199.6 ± 2.8 µm).
Meristem Size and Cell Partitioning Rates By definition during steady growth the flux of cells moving out of the growth zone is equal to the number of cells displaced out of the meristem into the elongation-only zone during the same period. It therefore integrates variations in both the number of meristematic cells and their partitioning rate (Eq. 9). In cv Birch, the average interval between two successive divisions,c (calculated using Eq. 9), was 3 to 4 h shorter under 900 compared with 350 ppm
CO2, equivalent to an approximately 13%
reduction (Table III), and the number of
symmetrically dividing cells per file
(Nsd) was increased, although
relatively less so (+8%; P < 0.05) (Table III). In cv
Hartog, elevated [CO2] had, qualitatively,
similar effects; their relative importance, however, depended on
vernalization treatment (Table III). In vernalized cv Hartog seedlings,
faster cycling rates were the dominant response (14% decrease in
c versus an 8%
increase in Nsd), as was also seen in
cv Birch, while in the absence of vernalization, elevated
[CO2] had only a weak effect on
c (5%, non-significant decrease), but caused a 20% increase in meristematic cell number. Even
when individually non-statistically significant, the combined effects
of elevated [CO2] on
c and
Nsd always gave rise to highly
significant differences in cell fluxes between low- and
high-[CO2]-grown plants (Table II).
Size of the Leaf Growth Zone: Kinetics of Cell Elongation Elongation in the Meristem. Elevated [CO2] had no significant effect on the distributions of meristematic cell lengths in either genotype or vernalization treatment (Fig. 6, insets). Variation in meristem length was therefore highly correlated to variations in meristematic cell number. Consistent with the variations of Nsd described earlier, elevated [CO2] systematically increased the longitudinal extension of the division zone (P < 5%), especially in the non-vernalized seedlings of cv Hartog (Table III, Lsd values). Elevated [CO2] also enhanced local elongation rates in the meristem; however, as for partitioning rates, this effect was mostly confined to the basal 2 to 3 mm of the meristem (Fig. 5b).
Anatomy of Mature Blades The above data described growth kinetics of epidermal sister cells. Morphometric analysis of cleared epidermis and blade sections allowed examination of the final dimensions of other cell types and of blade anatomy. As found for sister cells, the lengths of other epidermal cell types (interstomatal cells and non-specialized elongated cells) did not differ significantly between low- and high-[CO2]-grown blades (data not shown). The densities of these three cell types were also similar (Fig. 8), implying that their surface area and width were also insensitive to ambient [CO2]. Stomatal indices or densities were also not affected (Fig. 8). The only significant effect of elevated [CO2] was a reduction of trichome frequency in the non-vernalized leaves of cv Hartog (trichome index reduced from 9.8%-3.9%). Interestingly, vernalization also had little effect on epidermis anatomy (Fig. 8).
Contrary to our expectations based on observations of cleared epidermis, the examination of blade sections revealed several striking anatomical differences between leaf tissue generated under 350 or 900 ppm CO2, which also depended on both genotype and vernalization. While consistently having no effect on mesophyll cell length or cell projected area (Table IV), elevated [CO2] caused a significant increase in the cross-sectional areas of those cells except in the cv Hartog non-vernalized seedlings (Fig. 9, top panel). No such increase was observed in epidermal cells except in the non-vernalized leaves of cv Birch, where the cross-sectional area of epidermal cells of all types was also greater in high-[CO2]-grown leaves (Fig. 9, bottom panel). Remarkably, elevated [CO2] consistently caused an increase in air space volume in the mesophyll tissue. This latter effect was more marked in cv Hartog, and significant at any position across the blade (Fig. 10). In cv Birch, it was closest to the mid-vein while disappearing toward the blade margins. In addition, elevated [CO2] caused an increase in the number of mesophyll cell layers (on average, one out of three to five layers in total, data not shown) with the exception, again, of the non-vernalized seedlings of cv Hartog. Overall, these various effects resulted in an increase in blade thickness and structural carbon content per unit leaf area in high-[CO2]-grown leaves (Fig. 10); in the non-vernalized leaves of cv Hartog, however, increased leaf thickness was confined to the mid-rib region. Because ambient [CO2] had no effect on the size of major or minor veins (vein diameters of 60-80 µm regardless of [CO2] and vernalization, data not shown), the ratio of mesophyll tissue volume to vein volume was greater in high-[CO2]-grown leaves.
Elevated [CO2] Caused an Early Stimulation of Plant Growth and Modified the Development of Individual Leaves Contrary to earlier conclusions in the literature, wheat proved to
be an excellent model system to analyze [CO2]
effects on leaf development. In the two genotypes examined, vegetative
growth was increased under 900 ppm CO2. The
stimulatory effect of elevated [CO2] was
detectable early, in the first week after sowing, on leaf area
expansion and biomass accumulation, at a time that did not coincide
with any specific phenological stage nor necessarily with the beginning
of tillering (Fig. 4). Depending on genotype and vernalization, a
positive growth response to elevated [CO2] was
detected before the appearance of the first tiller (cv Hartog, vernalized leaves), at approximately the same time (e.g. cv Birch, vernalized leaves), or significantly later (e.g. Birch non-vernalized leaves) (Fig. 4). One can therefore conclude that there is no direct
causal link between the two events, as has been suggested (e.g. Nicolas
et al., 1993 Within a few days after the effects of high
[CO2] on leaf growth were detected,
high-[CO2]-grown plants were characterized by
greater ratios of carbon (total or structural) laid down in the whole
plant to leaf area expanded. These differences mostly reflected an
increase in C content per unit leaf area, a feature observed in other
CO2 enrichment studies with wheat or other
species and, more generally, in response to increased
CO2 assimilation rates (e.g. Thomas and Harvey,
1983 Our data leave no doubt that there are intrinsic developmental differences between low- and high-[CO2]-grown leaves, which overall lead to the deposition of more structural carbon per unit leaf area in the latter. Elevated [CO2] affected cell division and expansion rates (Figs. 5 and 7) and also leaf histogenesis, final dimensions, and anatomy (Figs. 9 and 10). Furthermore, for the first time to our knowledge, it is shown that these quantitative and qualitative effects of elevated [CO2] on leaf growth are subject to significant intraspecific genetic variation and, most unexpectedly, are modified by seed vernalization, a major developmental cue for the switch between vegetative and reproductive development in cereals. The effects of atmospheric [CO2] that are reported here were primarily due to increased photoassimilate supply (see in Fig. 3 the increased sugar contents and the 25% increase in rate of leaf photosynthesis per unit leaf area measured in a preliminary experiment and similar growth conditions). Elevated [CO2] caused a reduction in stomatal aperture, but, due to the well-ventilated conditions of the greenhouses, there was no detectable difference in leaf temperature between low- and high-[CO2]-grown plants. Although one cannot totally exclude some role for improved leaf water status under elevated [CO2], this effect was most likely minor. The soil was maintained very wet and, more importantly, quantitatively similar positive responses to elevated [CO2] as those described in Figures 1 and 4 were observed in a parallel experiment where plants were grown in hydroponics and under higher air humidity (6-7 mbar leaf-to-air vapor pressure difference against 11-14 mbar in this experiment [J. Masle, unpublished data]). Cellular Bases of [CO2] Effects on Leaf Growth and Carbon Deposition per Unit Leaf Area The cellular responses underlying whole-leaf responses to elevated
[CO2] were examined using a combination of
microscopy techniques and the theoretical framework developed earlier
for the kinematic analysis of "steady-state" growth zones,
characterized by constant cell length profiles (see "Materials and
Methods"). All leaves analyzed in the present study were sampled at
similar developmental stages during the period of linear elongation
that followed blade emergence. The ligule meristem was then just
initiated and at 0.5 mm at the most from the base of the leaf, so that
the whole growth zone could be treated as one continuous zone for the
derivation of kinematic parameters (Kemp, 1980 Elevated [CO2] Enhanced (Epidermal) Cell Division Rates A consistent effect of elevated [CO2] was to reduce the time interval between successive divisions (Table III). This is the first time that such an effect has been demonstrated in expanding leaves. Earlier studies gave evidence for a direct role of Suc in mitosis in both buds (Ballard and Wildman, 1964) and root
meristems (Webster and Henry, 1987), but being based on comparisons of
mitotic indices, these data did not allow the separation of the
respective contributions of changes in the cell cycle per se versus
changes in the number of cycling cells. The present finding of faster cycling rates in leaves expanding under elevated
[CO2] is consistent with the recent report by
Kinsman et al. (1997) of enhanced cell division rates in the shoot apex
of Dactylis under 700 ppm CO2 compared
with 350 ppm. Furthermore, although determined by different methods,
their estimates of cell cycling times are comparable to ours, however,
with a more pronounced CO2 effect (20% change for a doubling in atmospheric [CO2]).
Elevated [CO2] Affects Epidermal Cell Elongation Rate But Not Cell Length at Partitioning Nor Final Cell Length Elevated [CO2] also affected cellular growth in the meristem and beyond. This study reveals several new features about these effects. Growth [CO2] influenced local rates of cell elongation but, remarkably, had no detectable effect on cell length at partitioning (lsd), i.e. upon cell entry into the elongation-only zone (Fig. 6). Nor did it affect mature cell length (lf, Table II) or width (data not shown). The elongation rates of meristematic cells were consistently increased by elevated [CO2]. In non-vernalized leaves this stimulatory effect of elevated [CO2] on wall elongation was maintained after cell migration out of the meristem (greater rmax, Fig. 7). In those leaves non-meristematic cells elongated faster under elevated [CO2] but for a shorter time (Table III), with the net result being no change in lf. In vernalized leaves, however, none of these effects of elevated [CO2] was detectable. The maximum rate of cell elongation and the cell residence time in the elongation-only zone were insensitive to [CO2] (Fig. 7; Table III), hence the conserved final cell length. There are two possible hypotheses for the constancy of lsd and lf. The first hypothesis is that for cell partitioning to occur, meristematic cells have to reach a set threshold size and, once they have lost the ability to divide, cells elongate to a fixed length. Under that interpretation there is no direct sugar effect on cell cycling time or duration of cell elongation. Variations inc and
Tel with growth
[CO2] follow from [CO2]
effects on cell elongation rates. In the absence of such effects, as in vernalized leaves (Fig. 7), Tel is
unchanged. This first interpretation would require mechanisms by which
cells can measure their size. While there is some evidence for that in
yeast (e.g. Nurse and Fantes, 1981), it is totally unknown whether such
mechanisms operate in higher plants (Francis and Halford, 1995). Data
from our other experiments (Beemster et al., 1996) have shown that
there is no absolute size threshold for cell division in wheat, and no
stable relationship between the partitioning rate and the elongation rate in meristematic cells.
The second hypothesis is that cell partitioning and cessation of growth
are determined by the cell metabolic status and/or state of
differentiation rather than by size per se. Under that interpretation,
the concurrent decrease in
c (or
Tel) and increase in
rsd (or
rel) observed under elevated
[CO2] are caused by the increased metabolic
activity associated with increased photoassimilate supply. The
constancy of lsd and
lf seen in this experiment would then
simply mean that the rate of cell elongation and the rate at which a
cell reaches the "critical" state, conditioning partitioning
and cessation of growth, are increased by a similar proportion when
cellular sugar concentrations increase. However, there is no reason to
exclude the possibility that in other species and under different
growth conditions, these two sets of parameters may show a differential
sensitivity to sugars, leading to variations in
lsd or
lf as reported, for example, by Ferris
et al. (1996) in perennial rye-grass under elevated
[CO2] or by Beemster et al. (1996) in wheat
under root stress. This second interpretation for the constancy of
lf in this experiment appeals to us in
that it provides a unified explanation for the present data as well as
others where lsd and/or
lf have been found to vary.
Furthermore, while not requiring any role of cell size per se it does
not exclude it.
Whether direct or indirect a tight link between
[CO2] effects on cell division and expansion
processes is shown by the correlation between the spatial patterns of
local rates of cell partitioning and cell elongation (Fig. 5).
Furthermore, the fact that [CO2] effects on
cell elongation are most important at the base of the meristem (Fig. 5)
and smaller in the elongation zone indicates that these effects are
modulated by factors related to cell position and/or meristematic
status. Increased cellular Suc contents in the basal part of the growth
zone may be one of these factors, as suggested by Schnyder and Nelson
(1987) data in Festuca.
Elevated [CO2] Modifies Growth Anisotropy Morphometric analysis of mature blade sections demonstrates that elevated [CO2] has more complex effects on cellular growth than revealed by the examination of cell lengths and does modify cell properties. Thus in cv Birch, while not affecting final cell lengths (Tables II and IV), elevated [CO2] caused a significant increase in the cross-sectional area of mesophyll cells and, in non-vernalized leaves, also of epidermal cells (Fig. 9). In cv Hartog, changes in mesophyll cell cross-sectional area were only observed in vernalized leaves. These observations lead to the suggestion that elevated [CO2] modifies the anisotropy of cell growth, and that the underlying control mechanisms are genetically variable and sensitive to factors influenced by exposure to low vernalizing temperatures. Moreover, the absence of a stable correlation between dimensional changes observed in epidermal and mesophyll cells (see Fig. 9) indicates that [CO2] effects on growth anisotropy may be in part tissue or even cell type specific.From Cells to Whole Leaf (Regulation of Whole-Leaf Elongation Rate) The Enhancement of Leaf Growth Rate under Elevated [CO2]: a Crucial Role of Effects in the Meristem The kinematic methods of growth analysis provide a conceptual and mathematical framework for a quantitative analysis of the relationships between cell and whole-organ responses (Erickson and Sax, 1956; Green,
1976; Gandar, 1983a, 1983b) from two points of view:
First, whole leaf elongation, E, can be expressed as the
integral of local strain rates, r(x), over the
length of the whole growth zone, with no explicit role of partitioning
rate nor of the number of cells contributing to growth. This approach
(Fig. 7) indicates that differences in E with growth
[CO2] arose in part from differences in the
spatial regulation of cell wall elongation along the growth zone, with
elevated [CO2] either up-regulating r at a given position without affecting the size of the
growth zone (non-vernalized leaves), or displacing the peak of maximum cellular activity toward more distal positions along a longer growth
zone (vernalized leaves).
Second, leaf elongation may also be seen as the integrals of growth
activities of a certain number of cells, which sequentially move away
from the meristem. Attention is now focused on the growth trajectories
of individual material particles (cells here) with respect to time in
considering that the endowment of a cell for elongation may be
specified at the outset (e.g. see Van't Hof, 1973, and discussion in
previous section) and that the size of the growth zone may be related
to cell number rather than being physically fixed. This second
approach, encapsulated in Equation 2, points to the importance in the
present experiment of meristem size and activity in determining
variations in leaf growth and anatomy with growth
[CO2]. With
lsd and
lf being highly conserved, variations
in the overall elongations generated in the division and
elongation-only zones were directly proportional to variations in cell
flux, F. Furthermore, variations in final leaf length were
proportional to those in total number of cells per file, i.e. in the
number of proliferative divisions.
The importance of cell flux, often referred to as the cell production
rate, in driving environmental effects on leaf expansion has been
emphasized in several recent studies on roots (e.g. Muller et al.,
1998), all of which concluded with the dominant role of changes in the
number of cycling cells (Nsd) rather
than changes in cell cycling rate
(tc). This cannot be generalized to
the [CO2]-induced increase in epidermal cell
fluxes observed in this study. The present data demonstrate that
elevated [CO2] may affect both
Nsd and
c and that its relative
impact on each parameter varies depending on genotype and vernalization treatment.
A Relative Insensitivity of Leaf Area Expansion to Elevated [CO2]: CO2 Effects on Leaf Anatomy In all cases, at the leaf level, elevated [CO2] had relatively less effect on processes involved in expansive growth than on those determining growth in mass, leading to increased structural carbon content per unit leaf area (Fig. 10). This increase was not necessarily associated with increased blade thickness (e.g. in cv Hartog, non-vernalized leaves), a common characteristic of high-[CO2]-grown leaves (e.g. Downton et al., 1980; Thomas and Harvey, 1983). Increased carbon
content was the overall result of, first, a large insensitivity of the
final structure of the epidermis a tissue thought to place major
constraints on expansive growth in leaves (Kutschera et al., 1987)to
sugar supply. This insensitivity is manifest in the highly conserved
epidermis anatomy, including stomatal density and cell dimensions
in the paradermal leaf plane. Increased carbon content was also
due to complex histological changes in the mesophyll tissue affecting
both mesophyll cell initiation and development.
While the formation of an extra mesophyll cell layer under elevated
[CO2] or an increase in mesophyll cell
enlargement have been noted in other C3 species
(soybean, Thomas and Harvey, 1983; Phaseolus, Radoglou and
Jarvis, 1992), the systematic increase in intercellular air spaces in
high-[CO2]-grown leaves is an intriguing new
finding. The formation of air spaces in leaves is generally thought to
be developmentally regulated and to occur in a predictable manner
involving localized lysis of cell wall components and local cell wall
thickening (Knox, 1992; Raven, 1996). Its sensitivity to atmospheric
[CO2] shown here is, however, consistent with
experimental evidence and theoretical prediction that cell separation
forces increase with cell turgor and cell diameter (Jarvis, 1998).
Indeed, Suc contents were significantly greater in the
high-[CO2]-grown leaves (Fig. 3), most likely
resulting in increased mesophyll cell osmotic pressure and
turgor; furthermore, in both cv Birch and cv Hartog, mesophyll
cell diameter was increased.
While the formation of more numerous mesophyll cell layers under
elevated [CO2] would allow increased carbon
deposition per unit leaf area, the greater extension of intercellular
spaces plays in the opposite direction. The fact that leaf
C/m2 was increased even when this latter effect
was the largest or even the only one to be significant, as in cv Hartog
(non-vernalized leaves), implies that in these leaves, elevated
[CO2] either increased carbon/unit cell wall
area and/or increased the mesophyll cell area to volume ratio, through
changes in cell lobing for example.
The significance of the anatomical changes induced in wheat leaves by
elevated [CO2] for leaf function needs to
be considered. Effects that contribute to increased leaf
carbon/m2 constitute limitations to the
magnitude of long-term whole-plant growth enhancement by elevated
[CO2] (see equation 1 in Masle et al., 1990).
On the other hand, more developed air spaces and increased cell area to
volume ratio should facilitate CO2 diffusion in
the mesophyll tissue (Parkhurst, 1994), resulting in a smaller drop of
CO2 partial pressure between the substomatal
cavity and sites of carboxylation, and thereby in higher effective
assimilation rate at a given stomatal conductance.
[CO2] Effects on Leaf Growth Are Modified by Factors Related to Leaf Position, Vernalization, and Genotype Genotypic Effects There have been several recent reports of interspecific variation in growth responses to elevated [CO2] within a genus (Populus, Gardner et al., 1995; Dactylis,
Kinsman et al., 1997). This study demonstrates intraspecific variation
among cultivars of an intensively bred species. Genetic variation is
shown in both the magnitude of [CO2] effects on
a range of developmental processes and in the relative contributions of
these processes to variations in leaf growth rate and anatomy. Our
data, however, also reveal some highly conserved attributes, such as
final cell length or number of epidermal cell files, which constrained
leaf growth responses to elevated [CO2] in the
two genotypes examined. These two groups of attributes define targets
for an effective genetic manipulation of growth responses to
[CO2] by classical breeding programs using natural genetic variation and by more directed genetic engineering.
Vernalization Effects Aside from modifying the effects of atmospheric [CO2] on the size of the growth zone and on the kinetics of elongation within it, vernalizing temperatures had profound effects per se on overall blade elongation rate and mature blade anatomy. Unexpectedly, this was the case even in a genotype such as cv Hartog, classified as a spring type based on its vernalization requirements for flowering. In many respects, however, the effects of seed vernalization differed between cv Hartog and cv Birch, especially in the meristem. In the winter vernalized cv Birch leaves were characterized by a smaller meristem, comprising fewer epidermal cells with a faster partitioning rate than in non-vernalized seedlings (Table III) and smaller mesophyll cells (Table IV). In the spring cv Hartog, the opposite effects were observed, i.e. a longer meristem with more numerous epidermal cells that cycled more slowly and thicker mesophyll cells. In the two genotypes, however, the two sets of effects resulted in only a small change in cell flux of similar direction (10% greater flux in non-vernalized seedlings). Vernalization systematically increased the residence time of cells in the elongation zone, especially under elevated [CO2] (Table III), but caused a decrease in cell elongation rates (Fig. 7) so that final cell length remained mostly unchanged (Table II). Overall, these data reveal that early exposure to low vernalizing temperatures affects the expression of a number of genes involved in leaf development and that these effects are modified by the plant carbohydrate status. Whether there is a direct link between the observed effects of low temperatures on leaf development and on vernalization per se, i.e. on the promotion of flowering, or whether these are independent effects is an open question. Several decades ago, Purvis and Hatcher (1959) observed in several cereals, including wheat,
that vernalized seedlings had a shorter coleoptile and first leaf than
non-vernalized ones. Since then, several genes controlling
vernalization requirements in wheat have been identified (vrn genes),
but studies of their expression have been restricted to a few genotypes
with respect to flowering response and apical development only, with no
attention being paid to possible pleiotropic effects on leaf development.
Leaf Position Effect A feature common to all [CO2] growth responses analyzed in this experiment is that significant [CO2] effects on leaf elongation could not be detected in the first two to four leaves. A similar pattern was noted by Williams and Williams (1968) in their analysis of expansive growth
under different light levels, and is also present in the data of Friend
et al. (1962), which also showed variation in irradiance. How can this
be explained? While it may be argued, following Williams (1960), that
leaves 1 and 2 derive most of their carbon from seed reserves rather
than from current photosynthesis, this is not the case for subsequent
leaves. By the time leaf 3 emerges, seed reserves are depleted.
Furthermore, as discussed earlier, the data presented in Figure 4 also
rule out a direct causal relationship between the onset of tillering, which causes a sharp increase in carbon demand by axillary meristems and the appearance of a carbon limitation in the expanding leaves of
the main tiller.
A third, non-exclusive interpretation is that the sensitivity of leaf
growth to carbohydrates is developmentally regulated and is confined to
early stages in leaf ontogeny. Recent molecular studies provide several
examples of genes that are differentially regulated by sugars depending
on the physiological/developmental context of the leaf (Kovtun and
Daie, 1995). In the present experiments, leaves 1 to 5 all started to
develop before sowing, i.e. before first expo |
TOP
ABSTRACT
INTRODUCTION
2 s
) and 900 ppm [CO2] (
); bars across symbols
represent 2 SE. Note that the y axis is
common to the two genotypes and is in log-scale.
5 mg/g).
c, for meristematic cells or
residence time in the growth zone (Tel) for
elongating-only cells. Values followed by a different letter within a
column were statistically significantly different (P 
