Journal of Experimental Botany, Vol. 49, No. 320, pp. 535–546, March 1998
Consequences of chlorophyll deficiency for leaf
carotenoid composition in tobacco synthesizing
glutamate 1-semialdehyde aminotransferase antisense
RNA: dependency on developmental age and growth light
Heiko Härtel1,2,3 and Bernhard Grimm1
1 Institut für Pflanzengenetik und Kulturpflanzenforschung, Corrensstrasse 3, D-06466 Gatersleben, Germany
2 Humboldt Universität zu Berlin, Institut für Biologie/Pflanzenphysiologie, Philippstrasse 13, D-10115 Berlin,
Germany
Received 26 June 1997; Accepted 5 November 1997
Abstract
Transgenic tobacco (Nicotiana tabacum), with a
reduced chlorophyll content of up to less than 10% of
the wild-type level due to a different expression of
antisense RNA coding for glutamate 1-semialdehyde
aminotransferase, were used to study the relationship
between chlorophyll accumulation and changes in
carotenoid composition in developing and mature
leaves grown either under low (30 mmol photons
m−2 s−1) or high light (300 mmol photons m−2 s−1).
Regardless of the extent to which chlorophyll synthesis
was reduced, under low light the ratios of total chlorophyll to carotenoids remained constant. In contrast,
under high light the content of carotenoids was elevated relative to chlorophyll and increased further with
progressive inhibition in chlorophyll synthesis. The
xanthophyll-cycle pigment pool was most strongly
increased (up to 18-fold) upon supression of chlorophyll synthesis. Concurrently to the higher pool sizes
a higher extent of violaxanthin was converted into
antheraxanthin and zeaxanthin and this was found to
be correlated with a decrease in the quantum yield of
photosystem II photochemistry. While lutein increased
(up to 3-fold) with decreasing chlorophyll contents in
high light transformants, neoxanthin remained rather
constant in all plants analysed. Based on the present
results, two different levels for the regulation of carotenoid synthesis are proposed depending on (i) the
chlorophyll synthesizing capacity, and (ii) the photosynthetic light utilization efficiency. The first point
suggests a co-regulation between carotenoid and
chlorophyll synthesis; the second emphasizes the
special role of carotenoids for protection against
light stress.
Key words: Light acclimation, photoprotection, quantum
yield, lutein, xanthophyll cycle.
Introduction
Like chlorophylls (Chls), carotenoids of leaves are ubiquitous structural components of the photosynthetic
apparatus. In higher plants, Chls and carotenoids are
bound to specific proteins to form either reaction centre
pigment–protein complexes or light-harvesting pigment–
protein complexes (LHCs) of photosystem (PS) I and
PSII (LHCII ). Attempts to ascertain the distribution of
carotenoids within the multiple pigment–protein complexes by detergent fractionation yielded varying results
between research groups, probably due to distinct isolation procedures and differences in both the growth conditions and plant species. Nevertheless, b-carotene was
consistently found in the reaction centre/core complexes
of both PSI and PSII, whereas xanthophylls mainly bind
to LHCs ( Thayer and Björkman, 1992; Bassi et al., 1993;
Ruban et al., 1994; Lee and Thornber, 1995). Lutein is
normally the predominant xanthophyll, accounting for
up to 50% of total carotenoids in most plants. 80% and
more of all the lutein and neoxanthin were found to be
associated with the peripheral trimeric LHCII, consisting
3 To whom correspondence should be addressed at Gatersleben: Fax: +49 39482 5139. E-mail: [email protected]
© Oxford University Press 1998
536
Härtel and Grimm
of the major polypeptides Lhcb1 and Lhcb2 ( Thayer and
Björkman, 1992; Bassi et al., 1993; Ruban et al., 1994).
However, low amounts of lutein were frequently found
to be associated with the reaction centre/core complex as
well (Bassi et al., 1993; Lee and Thornber, 1995). Whilst
lutein is considered to bind within the protein backbone
of LHCs, neoxanthin seems to bind along the periphery
(Plumley and Schmidt, 1987; Kühlbrandt et al., 1994).
Violaxanthin ( V ) and its de-epoxidation products antheraxanthin (A) and zeaxanthin (Z ) are enriched in the
inner LHCII antenna proteins Lhcb4 to Lhcb6 (Bassi
et al., 1993).
Carotenoids have important functional roles in the
photosynthetic apparatus of higher plants. They serve as
accessory pigments by supplementing Chl in light harvesting and thereby enable plants to utilize light efficiently
over a broad wavelength range. Furthermore, carotenoids
have long been associated with a pivotal role in the
photoprotection of the photosynthetic apparatus by limiting the destructive reactions of 1O via direct quenching
2
of 1O or of the Chl triplet state that sensitizes formation
2
of reactive oxygen species (Cogdell et al., 1992; Telfer
et al., 1994; for reviews see Siefermann-Harms, 1987;
Bartley and Scolnik, 1995). Because of their supposed role
in deactivating excess excitation energy of light (DemmigAdams et al., 1996), the regulation of the synthesis of the
xanthophyll-cycle pigments has attracted much interest.
It is thought that the carotenoids are involved in the
structural reorganization of the individual PS components
in response to varying light conditions. Variations in
carotenoid composition and energy dissipation among
different species grown under various field conditions
have been described previously ( Thayer and Björkman,
1990; Demmig-Adams and Adams, 1992; Johnson et al.,
1993; Brugnoli et al., 1994; Demmig-Adams et al., 1995).
Moreover, the abundance of individual carotenoids is
differently affected in mutant plants with altered antenna
composition (Höfer et al., 1987; Knoetzel and Simpson,
1991; Schindler et al., 1994; Falbel et al., 1994; Plumley
and Schmidt, 1995; Härtel et al., 1996) and in transgenic
plants with a genetically lowered photosynthetic capacity
as well (Bilger et al., 1995).
However, the regulation of carotenoid synthesis and
the precise function of carotenoids in photosynthesis and
the assembly of the photosynthetic apparatus have not
been completely understood. The use of mutants affected
in their ability to synthesize specific carotenoids is one
important tool to assign their function in vivo (Rock
et al., 1992; Pogson et al., 1996). Another promising
approach to gain new information on the functional and
regulatory significance of carotenoids is the manipulation
of Chl synthesis. It seems well established that both Chl
and carotenoid biosynthesis exclusively take place in
chloroplasts. Carotenoids and Chls are characterized by
their well-defined location and orientation within the
molecular structure of the thylakoid membrane, and their
chemical and physical properties are strongly influenced
by other molecules in their vicinity, such as proteins and
membrane lipids (Plumley and Schmidt, 1987;
Kühlbrandt et al., 1994; for a review see Britton, 1995).
Because the pigment concentration of leaves per unit area
can vary considerably in wild-type plants dependent on
the growth conditions (Björkman, 1981), no consistent
correlation between the photosynthetic pigment accumulation and the growth light intensity was found until now.
In this respect the recently described glutamate
1-semialdehyde aminotransferase (GSA-AT ) antisense
RNA plants of tobacco provide a novel tool. These plants
are characterized by different degrees of lowered GSA-AT
activities leading to a reduced 5-aminolevulinic acidsynthesizing capacity and in consequence to a lowered
Chl content relative to the leaf area (Höfgen et al., 1994;
Härtel et al., 1997). The plants show a wide variety of
inheritable Chl variegation pattern in tobacco GSA-AT
antisense RNA plants (Höfgen et al., 1994; Härtel et al.,
1997). The spectrum of Chl abundance ranged from wild
type-like to uniformly bleached yellow leaves. Other
enzymes of the tetrapyrrole pathway seem not to be
influenced by the antisense gene effect, which largely rules
out the possibility that potentially dangerous tetrapyrrole
intermediates accumulate (Härtel et al., 1997). Therefore,
these GSA-AT antisense plants offer a novel opportunity
to gain greater insight into how the carotenoid biosynthetic pathway is regulated in plants upon Chl deprivation. Since carotenoids are of crucial importance for the
photoprotection of the photosynthetic apparatus, studies
of these plants improves the understanding of the strategies developed to cope with excess light energy. The
results demonstrate a distinctly different response of individual carotenoid components under Chl shortage. It is
suggested that the carotenoid pattern in leaves depends
on both the Chl turnover and the photosynthetic light
utilization efficiency.
Materials and methods
Plant material and growth conditions
Wild-type tobacco (Nicotiana tabacum var. Samsun NN ) plants
and GSA-AT antisense RNA transformants were cultivated in
soil with a 12/12 h light/dark cycle at 65% relative humidity in
a growth chamber under two different photon flux densities
(PFD, 400–700 nm) constant throughout the light period: High
light (300 mmol photons m−2 s−1, HL) and low light (30 mmol
photons m−2 s−1, LL) were both provided by an array of
halogen lamps (SON-T AGRO 400, Phillips, Eindhoven,
Netherlands). PFDs incident upon the top of plants were
determined using a quantum sensor (LI-189A, Li-Cor, Lincoln,
Nebraska, USA). The day/night temperature regime was
controlled at 25/20 °C. For all measurements, 7–10-week-old
plants were usually used. The construction of the GSA-AT
antisense gene transformants is described in Höfgen et al. (1994).
Carotenoids in chlorophyll-deficient tobacco 537
Pigment analysis
Samples were harvested 4–5 h after the beginning of light
exposure. Leaves were rapidly frozen in liquid nitrogen for
subsequent pigment determination. Pigments were extracted
and analysed by high-performance liquid chromatography
applying the method of Thayer and Björkman (1990) modified
as described (Härtel et al., 1996). All data shown represent the
mean of three to five replicates on different plants of the same
line. The de-epoxidation state of xanthophyll-cycle pigments
(DPS) was calculated as DPS=(A+Z)/( V+A+Z ).
Chl a fluorescence measurements
Measurements of room temperature Chl a fluorescence were
performed in the growth chamber with a portable PAM-2000
Chl fluorometer in conjunction with a leaf clip holder ( Walz,
Effeltrich, Germany) immediately before the same samples were
harvested for carotenoid determinations. The PSII quantum
yield (W ) was calculated according to Genty et al. (1989). In
PSII
order to ensure the coverage of representative leaf sections,
fluorescence was measured in triplicate for light-exposed plants
at three different locations on the leaf. The same leaf sections
were used for pigment determinations.
Results
Effect of GSA-AT antisense RNA synthesis on Chl
accumulation in dependence on developmental age and
growth light
GSA-AT tobacco transformants nos. 57, 25 and 42 were
selected for these investigations. They have been previously subjected to an extensive genetical and biochemical
characterization (Höfgen et al., 1994; Härtel et al., 1997).
These transformants cover a wide range of reduced Chl
content per leaf tissue, which correspond to the reduction
of GSA-AT activity ( Fig. 1). The leaves of the most
chlorotic transformant no. 42 are uniformly bleached,
while those of other transformants are slightly variegated
in the pigment pattern. To get further information about
the mechanism which controls carotenoid accumulation,
plants were grown under two distinct different PFDs,
30 mmol m−2 s−1 (LL) and 300 mmol m−2 s−1 (HL). No
significant macroscopic differences of the phenotypes were
observed in plants grown under LL or HL conditions.
For pigment determinations either whole leaves
(number 2 and 4) or large leaf areas (9 cm2) covering
representative leaf sections were used. Figure 2 depicts
Chl contents measured in total leaf extracts of wild type
and transformants dependent on leaf age and growth
light. In LL-exposed plants of wild type and transformant
no. 57, maximal amounts of Chl were determined in
leaves 6 or 8 numbered from the plant top downwards
( Fig. 2A), while in chlorotic transformants nos. 25 and
42 the Chl content was only marginally changed throughout the entire leaf development. Chl accumulated during
the development of HL-exposed plants similar as observed
for LL plant, even though the differences in Chl content,
Fig. 1. Typical leaves of wild-type tobacco (SNN ) and the GSA-AT
antisense RNA transformants nos 57, 25 and 42 of plants that were
grown for 7 weeks under HL at a PFD of 300 mmol photons m−2 s−1.
in particular between the wild type and the most chlorotic
transformant no. 42, further increased (Fig. 2B). A minor
shift in the maximum Chl contents towards leaves 8 and
10 occurred in HL-grown wild type and transformant
no. 57.
The Chl a/b ratios were higher in HL leaves of each
line compared to those of LL leaves of the same line,
except for the oldest leaves in transformant no. 42
( Fig. 2C, D). The Chl a/b ratio strongly altered throughout leaf development in transformant no. 42 with higher
ratios in young leaves and lower ratios in old leaves in
comparison to both the HL transformants nos. 57 and
25 and the wild type. Remarkably, in all LL lines, the
Chl a/b ratios were very similar, with the exeption of the
very young leaves of no. 42, despite the fact that the Chl
content was reduced up to a factor of seven in transformants versus the wild type ( Fig. 2A, B). Unchanged
Chl a/b ratios upon Chl-deficiency have not been previously reported for pigment mutants ( Knoetzel and
Simpson, 1991; Falbel et al., 1994) or intermittent-light
grown plants (Höfer et al., 1987; Härtel et al., 1996),
indicating the distinct different response of the GSA-AT
transformants to the Chl shortage.
538
Härtel and Grimm
Fig. 2. Chl content on a unit fresh weight basis (A, B) and Chl a/b ratio (C, D) in tobacco wild type (%) and the GSA-AT transformants no. 57
(( ), no. 25 (#) and no. 42 (6) grown either under LL (closed symbols) or HL (open symbols) dependent on developmental leaf age. The standard
deviations were always less than 22% (A, B) and 14% (C, D), respectively.
The carotenoid content and composition of GSA-AT
transformants depends on developmental age and growth
light
Figure 3 shows that the reduction in the total carotenoid
content parallels the deprivation of Chl synthesis in
transformants. Like total Chls, GSA-AT transformants
grown under LL possessed considerably lower levels of
carotenoids per unit fresh weight compared to wild-type
plants. As a consequence, except for the very young leaves
2 and 4 of transformant no. 42, the relative proportions
between Chls and carotenoids did not vary between wild
type and transformants (Fig. 3A). However, a different
response was observed for plants grown under HL. More
carotenoids accumulated in the most Chl-deficient transformants nos. 25 and 42, yielding higher levels of carotenoids per Chl relative to those of wild type ( Fig. 3B). In
all HL plants maximal carotenoid contents were present
in leaf 8. A quantitative inspection of the amounts of
individual carotenoids revealed that the higher carotenoid
contents in HL-grown transformants were mainly due to
higher amounts of lutein and the xanthophyll-cycle pigments V+A+Z (Fig. 3H, K ), whereas neoxanthin and
to a lesser extent b-carotene yielded similar ratios to Chl
in all plants during leaf development ( Fig. 3D, F ). The
decline in the carotenoid content in older leaves, which
is most prominent in chlorotic transformants nos. 25 and
42, may be due to the increasing distance from the light
source and shading from upper leaves.
Interestingly, in leaf 2 the ratio of all individual carotenoid components to Chl was very similar when the same
lines were compared independent of whether plants were
grown under LL or HL conditions. This may indicate
that changes in carotenoid content and composition at
this early stage of leaf development are mainly controlled
by endogenous factors associated with chloroplast differentation (e.g. Chl and protein synthesis), whereas changes
in carotenoids relative to Chl upon further maturation of
leaves are additionally influenced by the light regime (as
indicated by the increase in the carotenoid to Chl ratio
in HL transformants nos. 25 and 42). The higher carotenoid to Chl ratio in the premature leaves of transformant
no. 42 is likely due to the steadily synthesized low amount
of carotenoids that is usually found in dark-grown plants
independent of the presence of Chl ( Frosch and Mohr,
1980).
During assembly of light-harvesting antenna complexes, antenna protein and pigment biosynthesis are
tightly linked regulated processes (Härtel et al., 1997).
Chl b is exclusively associated with LHCs, in particular
the peripheral LHCII. Hence, the amount of Chl b is a
rough indicator of the amount of LHCs. Furthermore,
lutein and neoxanthin are considered to be predominantly
associated with the LHCs. Thus, alterations in these
pigments should be indicative for changes in the lightharvesting antenna organization. In LL plants, the
neoxanthin/Chl b and lutein/Chl b ratios were indistinguishable, with the exception of the young leaves 2 and
4 of no. 42, which contained higher levels of neoxanthin
and lutein compared to Chl b (Fig. 4A, C ). This constant
stoichiometry between these carotenoids and Chl b
throughout an increase in the Chl b content of nearly one
order of magnitude from transformant no. 42 to the wild
type (Fig. 2A) indicate that the synthesis of these pigments is closely coupled under LL conditions. On the
other hand, a pronounced increase, in particular in the
lutein/Chl b ratio but also the neoxanthin/Chl ratio, was
noticed in the HL-exposed transformants nos 42 and 25
Carotenoids in chlorophyll-deficient tobacco 539
Fig. 3. Contents of total carotenoids (A, B), b-carotene (C, D), neoxanthin ( E, F ), lutein (G, H ) and xanthophyll-cycle pigments (I, K ) relative to
the Chl content in tobacco wild type (%) and the GSA-AT transformants no. 57 (( ), no. 25 (#) and no. 42 (6) grown either under LL (closed
symbols) or HL (open symbols) dependent on developmental leaf age. The standard deviations were always less than 10%. V, A, Z denote the
contents of violaxanthin, antheraxanthin and zeaxanthin, respectively.
in comparison both with their LL counterparts and the
HL wild type ( Fig. 4D). Virtually no differences were
found in the neoxanthin/Chl b and lutein/Chl b ratios
between HL and LL variants of wild type and transformant no. 57 (Fig. 4B, D).
The pigment content and composition is distinctly different
in mature leaves of tobacco wild type and GSA-AT
transformants
Since both carotenoid and Chl maximally accumulated
in leaf 8 further analyses were focused on the relations
between carotenoid and Chl content in this particular
stage of leaf development. Figure 5 shows the relationship
between the content of total carotenoids and total Chls
in leaves of wild type and GSA-AT transformants, each
expressed on a unit area basis. A strict linear relation
between both pigment groups was found (r2=0.952, solid
line) irrespective of the PFD that plants experienced
during growth and the degree of suppression of Chl
synthesis. The correlation improved further (r2 ≥0.99)
when LL and HL variants were plotted separately (see
dashed lines). Whilst the LL plot extrapolates to the
origin, the HL plot does not, which is due to higher
540
Härtel and Grimm
Fig. 4. Ratios of neoxanthin/Chl b (A, B) and lutein/Chl b (C, D) in tobacco wild type (%) and the GSA-AT transformants no. 57 (( ), no. 25
(#) and no. 42 (6) grown either under LL (closed symbols) or HL (open symbols) dependent on developmental leaf age. The standard deviations
were always less than 8%.
Fig. 5. Relationship between total leaf carotenoid and Chl content per
unit area in mature leaves of tobacco wild type (%) and the GSA-AT
transformants no. 57 (( ), no. 25 (#) and no. 42 (6) grown either
under LL (closed symbols) or HL (open symbols). The standard
deviations were always less than 10%. The solid line correspond to the
first-order regression for both the LL and HL data. In addition separate
regression lines for LL or HL data are fitted (dashed lines).
amounts of carotenoids in particular in the chlorotic
transformants nos 25 and 42.
The Chl content increased per unit area in the
HL-grown wild type to 142% of the LL content, and in
the HL-grown transformant no. 57 to 147%, remained
rather constant in no. 25 (99%) and displayed a decrease
in no. 42 (58%) ( Table 1). Consequently, the most severe
Chl-deficient transformant no. 42 contained 16% (LL)
and 7% (HL) of the respective wild-type Chl content.
Neoxanthin and to a lesser extent also b-carotene contents
followed the same pattern. In contrast, lutein content
increased 1.5-fold in leaves of HL plants relative to LL
plants. Of the carotenoids, xanthophyll-cycle pigments
responded most strongly to the acclimation to HL versus
LL. When HL were compared with LL plants, the
V+A+Z pool sizes increased 2.0-fold (wild type),
3.0-fold (no. 57), 3.1-fold (no. 25) and 10.4-fold (no. 42),
respectively. Due to the different light response of Chl
synthesis (Fig. 2), the differences in V+A+Z contents
on a Chl basis between wild type and transformants were
more pronounced under HL exposure yielding 1.4-fold
(wild type) to 18.3-fold (no. 42) higher pool sizes than in
LL plants (Fig. 6D). On a Chl basis, HL-grown no. 42
contained a 12.8-fold higher V+A+Z pool size than the
HL-grown wild type. The higher V+A+Z pool sizes did
not simply occur at the expense of b-carotene, which is a
precursor in the synthesis of xanthophyll-cycle pigments.
The b-carotene contents were equal or even higher upon
HL exposure in all lines ( Fig. 6A). An intriguing result
is the dramatic increase in lutein contents throughout leaf
maturation (up to 300%) in HL transformants no. 25 and
42, while virtually no changes were found in the wild type
and transformant no. 57 (Fig. 6B). On the other hand,
neoxanthin accumulates independent on changes in Chl
content and PFD (Fig. 6C ).
The carotenoid composition correlates with the quantum
efficiency of PSII photochemistry
Figure 7 illustrates the relationship between the levels of
the various carotenoid components (as a percentage of
total carotenoids) and the relative quantum yield of PSII
photochemistry, W , which reflects the degree of
PSII
absorbed light utilized in photochemistry under the
respective light-adapted state. In LL plants, virtually no
differences in the relative distribution were found among
all individual carotenoids. In particular, HL-grown Chldeficient transformants showed dramatic changes in the
Carotenoids in chlorophyll-deficient tobacco 541
Table 1. Pigment contents in mature leaves of wild-typpe tobacco (SNN) and the GSA-AT transformants nos 57, 25 and 42 grown
under low light (LL) or high light (HL) conditions on a leaf area basis (in mmol m−2)
All values represent the means of at least seven different plants. The standard deviations were always less than 15%.
Lines
SNN
57
25
42
LL
HL
Chl
b-carotene
V+A+Z
Neoxanthin
Lutein
Chl
b-carotene
V+A+Z
Neoxanthin
Lutein
360.2
263.7
163.3
58.6
32.0
25.2
18.0
4.8
10.2
8.0
6.0
1.7
14.9
11.0
7.3
2.5
42.6
34.5
19.1
6.7
511.4
387.7
161.0
33.9
52.7
40.8
20.4
4.8
20.7
24.1
18.4
17.6
19.8
17.8
6.8
1.6
61.9
50.0
25.5
10.0
Fig. 6. Relationship between individual carotenoids normalized to Chl and the Chl content per unit area in mature leaves of tobacco wild type (%)
and the GSA-AT transformants no. 57 (( ), no. 25 (#) and no. 42 (6) grown either under LL (closed symbols) or HL (open symbols). The
standard deviations were always less than 8%.
carotenoid pattern. The strongest alteration was found
for xanthophyll-cycle pigments, which increased from
about 10% in LL plants to more than 50% in
HL-transformant no. 42. While the W
values were
PSII
similar for all LL plants (solid symbols), they showed a
decrease with reduced Chl content in HL transformants
(open symbols), indicative for a higher extent of excess
light in HL transformants. There is a positive correlation
between these changes in the W values and the fractions
PSII
of b-carotene (r2=0.896, Fig. 7A), lutein (r2=0.860,
Fig. 7B) and neoxanthin (r2=0.885, Fig. 7C ) and a negative correlation between W and the fraction of xanthoPSII
phyll-cycle pigments (r2=0.935, Fig. 7D).
Decreases in W
were linearly related to increases in
PSII
the DPS ( Fig. 8A). All values closely fit to the same
regression line (r2=0.959), despite the large differences
in both the absolute amounts of A and Z present as well
as in the Chl content ( Fig. 2; Table 1). Since both W
PSII
and DPS are mainly controlled by the acidification of the
thylakoid lumen space, these results point to an increasing
excitation pressure on PSII in parallel to the reduction in
Chl content in HL transformants. A close relationship
(r2=0.977) was also apparent when the V+A+Z pool
size was plotted against the DPS ( Fig. 8B). No difference
was found in the DPS levels in plants grown under LL.
Virtually all violaxanthin remained epoxidized. Leaves of
HL plants did not only reveal higher pool sizes of
xanthophyll-cycle pigments, but also showed a higher
DPS than LL plants. 85% of the total xanthophyll-cycle
pigments were present as A and Z in transformant no.
42 upon HL conditions.
The high differences in the pool sizes of V+A+Z (on
a Chl basis) between the wild type and transformants
under HL conditions rule out that the PFD to which the
plants were exposed during growth is the sole determinator. In fact, there appears to be a close link between the
V+A+Z pool size and the depression of W , i.e. the
PSII
increase in the excitation energy pressure on PSII. To
support this idea, the amounts of xanthophyll-cycle pigments, DPS and W were correlated in leaves of different
PSII
542
Härtel and Grimm
Fig. 7. Relationship between b-carotene (A), lutein (B) and neoxanthin (C ) and xanthophyll-cycle pigments each expressed as a percentage of total
carotenoids and W
in mature leaves of tobacco wild type (%) and the GSA-AT transformants no. 57 (( ), no. 25 (#) and no. 42 (6) grown
PSII
either under LL (closed symbols) or HL (open symbols). The standard deviations were always less than 8%. The dashed lines correspond to the
first-order regression to the data.
Fig. 8. Relationship between the DPS and either W (A) or the pool size of xanthophyll-cycle pigments (B) in mature leaves of tobacco wild type
PSII
(%) and the GSA-AT transformants no. 57 (( ), no. 25 (#) and no. 42 (6) grown either under LL (closed symbols) or HL (open symbols). The
standard deviations were always less than 14%. The dashed lines correspond to the first-order regression to the data.
age of transformant no. 42 grown either under LL or HL
conditions. No. 42 has the greatest difference in the
V+A+Z pool size between LL and HL-grown plants
and among the leaves of different age of HL plants.
Figure 9A shows the plot of the DPS versus the corresponding V+A+Z pool size. As apparent, a curvilinear
relationship was obtained between both parameters. This
is mainly caused by the fact that the DPS did not respond
to the altered V+A+Z pool size in LL transformants
(see also Fig. 8A), while in HL transformants a strong
linear correlation exists. A similar relationship was
obtained when W
was plotted against the pool size of
PSII
xanthophyll-cycle pigments (Fig. 9B). However, W linPSII
early correlated with the DPS along the developmental
leaf gradient of both LL and HL-grown transformant no.
42 (r2=0.962; Fig. 9C ).
Discussion
The present studies were conducted to understand better
the regulatory mechanisms determining carotenoid synthesis and accumulation in plants. For this purpose,
transgenic tobacco plants with differently reduced Chl
contents were grown under two different PFDs. The
alterations in the Chl content were achieved by expressing
antisense RNA for GSA-AT in tobacco plants ( Höfgen
et al., 1994; Härtel et al., 1997). GSA-AT catalyses the
last step in the formation of 5-aminolevulinic acid, which
Carotenoids in chlorophyll-deficient tobacco 543
Fig. 9. Relationship between the DPS (A) or W
(B) and the pool size of xanthophyll-cycle pigments, and between W
and the DPS (C ) in
PSII
PSII
developing leaves of the GSA-AT transformant no. 42 (6) grown either under LL (closed symbols) or HL (open symbols). Values obtained with
leaves 2 to 8 are shown. The standard deviations were always less than 14%. The dashed line in (C ) corresponds to the first-order regression to the data.
is rate limiting for the entire Chl synthesis (Beale and
Weinstein, 1990). This approach enables us to follow
changes in the carotenoid content from two different
points of view. First, the large range of Chl reduction of
selected transgenic plants allows a more detailed study
on the interrelations between Chl and carotenoid synthesis
in developing and mature leaves. Second, it was possible
to study the effects of light dosis on the carotenoid pattern
in plants of a single species, but with genetically defined
reductions in Chl content under highly standardized
growth conditions in the laboratory. So far, most studies
on carotenoid changes under different PFDs were carry
out with plants grown under field conditions, which inhere
the difficulties to exclude secondary effects by other
environmental factors.
Diminished GSA-AT activity results in dramatic
changes not only in the Chl but also in the carotenoid
contents ( Figs 2, 3). Independent of the extent of Chl
deficiency, the proportions between total Chl and carotenoids appreciably correlated under the LL regime. This
commensurate decline of carotenoids with Chls is certainly related to the changes in the amount of individual
pigment-protein complexes which bind carotenoids, but
which obligatorily require Chl for stabilization (Härtel
et al., 1997, and references cited therein). Assuming that
the carotenoid contents measured mainly reflect rates of
synthesis and are not merely due to higher turnover rates,
this finding suggests that Chl and carotenoid synthesis
are co-regulated in plants. In fact, both the Chl and
carotenoid pathway share the precursor geranyl geranyl
pyrophosphate (Beale and Weinstein, 1990; Bartley and
Scolnik, 1995). Thus, co-ordination of the synthesis of
Chls and carotenoids could take place at the level of
geranyl geranyl pyrophosphate availability. A largely
proportional decrease in Chl and carotenoid contents was
recently also reported for spinach leaves that were exposed
to nitrogen stress ( Verhoeven et al., 1997), which may
corroborate the assumption of a tight control in the
synthesis of both pigment groups.
However, under HL conditions the gradual reduction
of total Chl in GSA-AT transformants quite differently
affects the abundance of individual carotenoids. In all
lines higher carotenoid contents were found relative to
Chl (Fig. 5). Xanthophyll-cycle pigments responded most
strongly to the change of the light regime. Consistent
with results obtained with numerous different field-grown
plant species ( Thayer and Björkman, 1990; DemmigAdams and Adams, 1992; Johnson et al., 1993; Brugnoli
et al., 1994; Demmig-Adams and Adams, 1996)
HL-grown plants contained elevated pool sizes of xanthophyll-cycle pigments compared with LL plants ( Figs 3, 6;
Table 1). However, large differences in the enhancement
of the V+A+Z pool size between wild type and the pale
GSA-AT transformants were found. Depending on
544
Härtel and Grimm
whether Chl or leaf area was used as reference unit the
increase varied from 1.4- or 2.0-fold, respectively, in the
wild type to 18.3- or 10.4-fold, respectively, in transformant no. 42. This implies that HL exposure to plants
does not lead to a proportionate increase in the pool size
of xanthophyll-cycle pigments and that a clear trend exists
for the V+A+Z content to increase with decreasing
Chl content.
One may argue that the lack of a direct correlation
between total V+A+Z and the PFD incident on transgenic plants could simply be explained by the variations
in the amount of Chl, since this was the reference unit
for expressing the pool sizes. First of all, differences in
the pool sizes were not found in Chl-deficient GSA-AT
transformants grown under LL conditions when expressed
on a Chl basis. Apart from this, it has to be emphasized
that in HL plants the amount of xanthophyll-cycle pigments drastically more increased than does the Chl content declined ( Fig. 2; Table 1). Moreover, the progressive
increase in V+A+Z contents at consistent low levels of
Chl in expanding leaves of the most chlorotic transformant no. 42 (Fig. 3K ) is clearly indicative of a regulatory response. On the other hand, due to the drastic
decline of the Chl content in GSA-AT transformants, a
progressive decline in the proportion of incident light that
is absorbed by leaves can be predicted. This implies that
the changes in the V+A+Z pool size under both LL
and HL conditions are not a direct consequence of the
photon receipt by leaves and that another intrinsic factor
must control the carotenoid synthesis. Owing to the
standardized growth conditions, the possibility can be
excluded that differences in light quality and temperature
may have affected the V+A+Z pool size. Importantly,
the V+A+Z pool size correlated with the PSII quantum
efficiency. The transformants with the lowest W values,
PSII
i.e. with the lowest fraction of absorbed light that is
utilized in PSII photochemistry, had much higher
V+A+Z pool sizes than the wild type and those transformants with higher W values (Figs 8, 9). This suggests
PSII
that the excitation energy pressure on the photosynthetic
apparatus seems to determine the V+A+Z pool size and
not the PFD during growth. Given the fact that the
degree to which the xanthophyll cycle was converted to
A+Z tended to correlate with the degree to which the
quantum yield of PSII photochemistry was depressed in
HL transformants, the lowered W values could be due
PSII
a higher extent of non-radiative excitation energy dissipation in these plants (Demmig-Adams et al., 1995;
Verhoeven et al., 1997).
The content of the two other xanthophylls, neoxanthin
and lutein, was quite differently affected in transformants
by the different growth PFD. In agreement with previous
reports on several field-growing sun and shade species
( Thayer and Björkman, 1990; Demmig-Adams and
Adams, 1992, 1996), but in contrast to Brugnoli et al.
(1994), leaves of plants grown under HL showed significantly higher lutein contents (up to 3-fold in line no. 42;
Figs 3H, 6C ) than those grown under LL. On the other
hand, no appreciable differences were found in neoxanthin
contents in LL and HL grown plants ( Figs 3F, 6C ).
Under the assumption that (i) all carotenoids are also
bound to proteins in transgenic lines, and (ii) the individual pigment-binding proteins possess specific binding
sites for the different pigments, one might in fact expect,
due to the higher Chl a/b ratios in transformants, a
decrease in the levels of both neoxanthin and lutein that
are thought to be associated mostly with the peripheral
LHCII. While the neoxanthin/Chl b ratios showed, except
for the younger leaves of no. 42, minor differences
( Fig. 4A, B), the higher lutein/Chl b ratios in
HL-transformants no. 42 and 25 (Fig. 4D) may indicate
that either the proportion of lutein associated with the
inner PSII antenna increased and/or large amounts of
lutein must bind to the reaction centre/core complexes
under Chl-deficiency. In favour of a regulatory response
is the progressive increase in the lutein content throughout
leaf development in HL transformants no. 42 and 25
( Fig. 3H ), which seems to parallel the changes in the
V+A+Z contents (Fig. 3K ). This increase of lutein
contents would be difficult to understand, if large proportions of lutein are present as free (unbound to protein
complexes) pigments in transformants.
The changes in the amounts of xanthophyll-cycle pigments in GSA-AT transformants are fully consistent with
their proposed role in photoprotection. Evidence has been
accumulated that dissipation of potentially harmful excess
excitation energy occurs directly within the LHCs of PSI
and PSII and that the xanthophyll cycle mediates this
process (for recent reviews see Demmig-Adams et al.,
1996; Gilmore, 1997). Recent work has indicated the
importance of the minor LHCs of PSII in the process of
excess excitation energy dissipation. They are enriched in
xanthophyll-cycle pigments (Bassi et al., 1993; Ruban
et al., 1994) and they seem to be the main site for the
formation of the quenching species which mediate heat
dissipation (Crofts and Yerkes, 1994; Härtel and Lokstein,
1995; Gilmore, 1997). Since lutein is a structural isomer
of zeaxanthin, differing only in the location of one double
bound, it is tempting to assume that it also plays a role
in excess excitation energy dissipation, that could also
provide a reasonable explanation for its increase in
HL-exposed transformants. The strong differences in the
V+A+Z pool sizes and in lutein contents make the
GSA-AT transformants an interesting tool with which to
correlate the pigment alterations in leaves with their
capacity to dissipate excess light energy non-radiatively.
The Chl a/b ratio is very sensitive to the amounts of
LHCs per PS unit and, hence, is a good indicator not
only for changes in the antenna organization, but also
the PSII antenna size. Under LL conditions the ratio of
Carotenoids in chlorophyll-deficient tobacco 545
Chl a/b was virtually the same for the wild type and all
transformants, indicating that the antenna organization
is not affected by the strongly restricted Chl synthesizing
capacity. This finding confirms our previous study, that
the Chl deficiency in GSA-AT transformants does not
reduce the antenna size in tobacco grown under low
PFDs (Härtel et al., 1997). As expected, HL plants
showed higher Chl a/b ratios than plants grown under
LL conditions, which is consistent with a decrease of the
peripheral PSII antenna size relative to the reaction
centre/core complex, and an increase in the proportion
of non-appressed thylakoid membranes (Anderson,
1986). The Chl a/b ratio of HL-transformants concomitantly increased with lowering Chl content (Fig. 2D). This,
together with the changes in carotenoids, points to a
different adaptive response in antenna organization
during LL/HL acclimation under Chl deficiency.
Presumably, the acclimation to HL conditions is associated with a deprivation of the peripheral antenna size.
However, the distribution of Chls and carotenoids among
the pigment–protein complexes and the amount of these
complexes remains to be elucidated.
HL-grown wild type and transformant no. 57 accumulated more Chl than under LL conditions, whereas line
no. 25 accumulated similar amounts and line no. 42 less
Chl (Fig. 2, Table 1). An explanation for the decreasing
Chl content could be an additional inactivation of Chl
synthesis in HL-grown pale transformants. It is well
established that synthesis of 5-aminolevulinic acid is controlled by light ( Kannangara and Gough, 1979). At
present, it can not be excluded that, additionally to the
antisense RNA effect, synthesis of 5-aminolevulinic acid
is down-regulated in most Chl-deficient HL transformants. Alternatively, photo-oxidative processes could
be responsible for the additional decrease in Chl content.
Photo-oxidative processes are known to be characterized
by a preferential degradation of reaction centre Chl a due
to the high efficiency of energy transfer from Chl b and
a loss of carotenoids, which are primary targets for
activated oxygen species (Härtel et al., 1993). Because
higher Chl a/b ratios and carotenoid contents were found
in the HL-grown transformants as compared to the wild
type, photo-oxidative damage can hardly account for the
additional Chl decrease.
Based on studies with Chl-deficient wheat and barley
chlorina mutants, Falbel et al. (1996) recently hypothesized that the rate of Chl synthesis controls to a large
extent the assembly of the photosynthetic apparatus.
According to Falbel et al. (1996), Chl b is synthesized
from Chl a that is not bound by the reaction centre core
complexes and that therefore remains available as substrate for Chl b synthesis. Considering the strong reduction in total Chl without any effect on the Chl a/b ratio
in GSA-AT transformants grown under LL conditions
these results do not agree with this suggestion and imply
a high regulatory complexity between Chl synthesis and
assembly of photosystems.
In summary, the commensurate accumulation of carotenoids with Chls and their dependency on the light regime
provide evidence for the high flexibility of their synthesis
and/or turnover to adapt to the respective environmental
conditions. Based on the results presented at least two
factors seem to control the carotenoid pattern of plants:
(i) the Chl-synthesizing capacity, and (ii) the photosynthetic light utilization efficiency. Although our results do
not explain the molecular mechanism, they are consistent
with the idea that a co-regulation exists between Chl and
carotenoid synthesis, which in turn is governed by a
selective accumulation of individual carotenoid components. It is suggested that the ratio of absorbed to photochemically utilized light is an important determining
factor in the signalling pathway leading to this selective
accumulation. It remains a challenging task to find out
the mechanism for interactive control of Chl and carotenoid synthesis.
Acknowledgements
This work was supported by the Deutsche Forschungsgemeinschaft (grants no. Gr 936/4 and 936/5 and Ho 1757/1)
and by a grant from the Innovationskolleg ‘Zellspezialisierung’
(Deutsche Forschungsgemeinschaft) to HH Barbara Hickel is
acknowledged for expert technical assistance.
References
Anderson JM. 1986. Photoregulation of the composition,
function, and structure of thylakoid membranes. Annual
Review of Plant Physiology 37, 93–136.
Bassi R, Pineau B, Dainese P, Marquardt J. 1993. Carotenoidbinding proteins of photosystem II. European Journal of
Biochemistry 212, 297–303.
Bartley GE, Scolnik PA. 1995. Plant carotenoids: pigments for
photoprotection, visual attraction, and human health. The
Plant Cell 7, 1027–38.
Beale SI, Weinstein JD. 1990. Tetrapyrrole metabolism in
photosynthetic organisms. In: Dailey HA, ed. Biosynthesis of
heme and chlorophyll. New York: McGraw-Hill, 287–391.
Bilger W, Fisahn J, Brummert W, Kossmann J, Willmitzer L.
1995. Violaxanthin cycle pigment contents in potato and
tobacco plants with genetically reduced photosynthetic capacity. Plant Physiology 108, 1479–86.
Björkman O. 1981. Responses to different quantum flux
densities. In: Lange OL, Nobel PS, Osmond CB, Ziegler
H, eds. Physiological plant ecology. Berlin: Springer Verlag,
57–107.
Britton G. 1995. Structure and properties of carotenoids in
relation to function. The FASEB Journal 9, 1551–8.
Brugnoli E, Cona A, Lauteri M. 1994. Xanthophyll cycle
components and capacity for non-radiative energy dissipation
in sun and shade leaves of Ligustrum ovalifolium exposed to
conditions limiting photosynthesis. Photosynthesis Research
41, 451–63.
546
Härtel and Grimm
Cogdell RJ, Andersson PO, Gillbro T. 1992. Carotenoid singlet
states and their involvement in photosynthetic light-harvesting
pigments. Journal of Photochemistry and Photobiology, B:
Biology 15, 105–12.
Crofts AR, Yerkes CT. 1994. A molecular mechanism for
qE-quenching. FEBS Letters 352, 265–70.
Demmig-Adams B, Adams WW III. 1992. Carotenoid composition in sun and shade leaves of plants with different life
forms. Plant, Cell and Environment 15, 411–19.
Demmig-Adams B, Adams WW III. 1996. Chlorophyll and
carotenoid composition in leaves of Euonymus kiautschovicus
acclimated to different degrees of light stress in the field.
Australian Journal of Plant Physiology 23, 649–59.
Demmig-Adams B, Adams WW III, Logan BA, Verhoeven AS.
1995. Xanthophyll cycle-dependent energy dissipation and
flexible photosystem II efficiency in plants acclimated to light
stress. Australian Journal of Plant Physiology 22, 249–60.
Demmig-Adams B, Gilmore AM, Adams WW III. 1996. In vivo
functions of carotenoids in plants. The FASEB Journal
10, 403–12.
Falbel TG, Staehelin A, Adams WW III. 1994. Analysis of
xanthophyll cycle carotenoids and chlorophyll fluorescence
in light intensity-dependent chlorophyll-deficient mutants of
wheat and barley. Photosynthesis Research 42, 191–202.
Falbel TG, Meehl JB, Staehelin A. 1996. Severity of mutant
phenotype in a series of chlorophyll-deficient wheat mutants
depends on light intensity and the severity of the block in
chlorophyll synthesis. Plant Physiology 112, 821–32.
Frosch S, Mohr H. 1980. Analysis of light-controlled accumulation of carotenoids in mustard (Sinapis alba L.) seedlings.
Planta 148, 279–86.
Genty B, Briantais JM, Baker NR. 1989. The relationship
between the quantum yield of photosynthetic electron
transport and quenching of chlorophyll fluorescence.
Biochimica et Biophysica Acta 990, 87–92.
Gilmore AM. 1997. Mechanistic aspects of xanthophyll cycledependent photoprotection in higher plant chloroplasts and
leaves. Physiologia Plantarum 99, 197–209.
Härtel H, Kruse E, Grimm B. 1997. Restriction of chlorophyll
synthesis due to expression of glutamate-1 semialdehyde
aminotransferase antisense RNA does not reduce the light
harvesting antenna size in tobacco. Plant Physiology 113,
1113–24.
Härtel H, Lokstein H. 1995. Relationship between quenching
of maximum and dark-level chlorophyll fluorescence in vivo:
dependence on photosystem II antenna size. Biochimica et
Biophysica Acta 1228, 91–4.
Härtel H, Lokstein H, Grimm B, Rank B. 1996. Kinetic studies
on the xanthophyll cycle in barley leaves. Influence of antenna
size and relations to non-photochemical chlorophyll fluorescence quenching. Plant Physiology 110, 471–82.
Härtel H, Walter G, Hanke T. 1993. Photoinduced damage in
leaf segments of wheat (Triticum aestivum L.) and lettuce
(Lactuca sativa L.) treated with 5-aminolevulinic acid. I.
Effect on structural components of the photosynthetic
apparatus. Journal of Plant Physiology 142, 230–6.
Höfer M, Walter G, Meister A, Hoffmann P. 1987. Pigment
accumulation and energy flow between carotenoids and
chlorophyll during greening of etiolated wheat seedlings
under continuous and intermittent irradiation. 1. Pigment
biogenesis. Photosynthetica 21, 124–30.
Höfgen R, Axelsen KA, Kannangara CG, Schüttke I, Pohlenz
H-D, Willmitzer L, Grimm B, von Wettstein D. 1994. A visible
marker for antisense mRNA expression in plants: inhibition
of chlorophyll synthesis with a glutamate 1-semialdehyde
aminotransferase antisense gene. Proceedings of the National
Academy of Sciences, USA 91, 1726–30.
Johnson GN, Scholes JD, Horton P, Young AJ. 1993.
Relationships between carotenoid composition and growth
habit in British plant species. Plant, Cell and Environment
16, 681–6.
Kannangara CG, Gough SP. 1979. Biosynthesis of d-aminolevulinate in greening barley leaves. II. Induction of enzyme
synthesis by light. Carlsberg Research Communications 44,
11–20.
Knoetzel J, Simpson D. 1991. Expression and organisation of
antenna proteins in the ight- and temperature-sensitive barley
mutant chlorina-104. Planta 185, 111–23.
Kühlbrandt W, Wang DN, Fujiyoshi Y. 1994. Atomic model of
plant light-harvesting complex by electron crystallography.
Nature 367, 614–21.
Lee ALC, Thornber JP. 1995. Analysis of the pigment
stoichiometry of pigment-protein complexes from barley
(Hordeum vulgare). The xanthophyll cycle intermediates occur
mainly in the light-harvesting complexes of photosystem I
and photosystem II. Plant Physiology 107, 565–74.
Plumley FG, Schmidt GW. 1987. Reconstitution of chlorophyll
a/b light-harvesting complexes: xanthophyll-dependent assembly and energy transfer. Proceedings of the National Academy
of Sciences, USA 84, 146–50.
Plumley FG, Schmidt GW. 1995. Light-harvesting chlorophyll
a/b complexes: interdependent pigment synthesis and protein
assembly. The Plant Cell 7, 689–704.
Pogson B, McDonald KA, Truong M, Britton G, DellaPenna D.
1996. Arabidopsis carotenoid mutants demonstrate that lutein
is not essential for photosynthesis in higher plants. The Plant
Cell 8, 1627–39.
Rock CD, Bowlby NR, Hoffmann-Benning S, Zeevaart JAD.
1992. The aba mutant of Arabidopsis thaliana (L.) Heynh.
has reduced chlorophyll fluorescence yields and reduced
thylakoid stacking. Plant Physiology 100, 1796–1801.
Ruban AV, Young A, Pascal AA, Horton P. 1994. The effects of
illumination on the xanthophyll composition of the photosystem II light-harvesting complexes of spinach thylakoid
membranes. Plant Physiology 104, 227–34.
Schindler C, Reith P, Lichtenthaler HK. 1994. Differential levels
of carotenoids and decrease of zeaxanthin cycle performance
during leaf development in a green and an aurea variety of
tobacco. Journal of Plant Physiology 143, 500–7.
Siefermann-Harms D. 1987. The light-harvesting and protective
functions of carotenoids in photosynthetic membranes.
Physiologia Plantarum 69, 561–8.
Telfer A, Dhami S, Bishop SM, Phillips D, Barber J. 1994. bCarotene quenches singlet oxygen formed by isolated photosystem II reaction centers. Biochemistry 33, 14469–74.
Thayer SS, Björkman O. 1990. Leaf xanthophyll content and
composition in sun and shade determined by HPLC.
Photosynthesis Research 23, 331–43.
Thayer SS, Björkman O. 1992. Carotenoid distribution and
de-epoxidation in thylakoid pigment-protein complexes from
cotton leaves and bundle sheath cells of maize. Photosynthesis
Research 33, 213–26.
Verhoeven AS, Demmig-Adams B, Adams WW. 1997. Enhanced
employment of the xanthophyll cycle and thermal energy
dissipation in spinach exposed to high light and N stress.
Plant Physiology 113, 817–24.