ARTICLE
DOI: 10.1038/s41467-018-03231-x

OPEN

Photosystem II Subunit S overexpression increases
the efﬁciency of water use in a ﬁeld-grown crop

1234567890():,;

Katarzyna Głowacka1,2, Johannes Kromdijk1, Katherine Kucera1, Jiayang Xie1,3, Amanda P. Cavanagh1,
Lauriebeth Leonelli4, Andrew D.B. Leakey1,3,5, Donald R. Ort1,6, Krishna K. Niyogi4,7 & Stephen P. Long

1,8

Insufﬁcient water availability for crop production is a mounting barrier to achieving the 70%
increase in food production that will be needed by 2050. One solution is to develop crops
that require less water per unit mass of production. Water vapor transpires from leaves
through stomata, which also facilitate the inﬂux of CO2 during photosynthetic assimilation.
Here, we hypothesize that Photosystem II Subunit S (PsbS) expression affects a chloroplastderived signal for stomatal opening in response to light, which can be used to improve wateruse efﬁciency. Transgenic tobacco plants with a range of PsbS expression, from undetectable
to 3.7 times wild-type are generated. Plants with increased PsbS expression show less stomatal opening in response to light, resulting in a 25% reduction in water loss per CO2
assimilated under ﬁeld conditions. Since the role of PsbS is universal across higher plants, this
manipulation should be effective across all crops.

1 Carl R. Woese Institute for Genomic Biology, University of Illinois, Urbana, IL 61801, USA. 2 Institute of Plant Genetics, Polish Academy of Sciences, 60-479
Poznań Poland. 3 Department of Crop Sciences, University of Illinois, Urbana, IL 61801, USA. 4 Howard Hughes Medical Institute, Department of Plant and
Microbial Biology, University of California Berkeley, Berkeley, CA 94720, USA. 5 Department of Plant Biology, University of Illinois, Urbana, IL 61801, USA.
6 Photosynthesis Research Unit, US Department of Agriculture-Agricultural Research Service, University of Illinois, Urbana, IL 61801, USA. 7 Molecular
Biophysics and Integrated Bioimaging Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA. 8 Lancaster Environment Centre,
University of Lancaster, Lancaster LA1 1YX, UK. These authors contributed equallly: Katarzyna Głowacka and Johannes Kromdijk. Correspondence and
requests for materials should be addressed to K.K.N. (email: niyogi@berkeley.edu) or to S.P.L. (email: slong@illinois.edu)

NATURE COMMUNICATIONS   (2018)9:868

  DOI: 10.1038/s41467-018-03231-x   www.nature.com/naturecommunications

1

 ARTICLE

NATURE COMMUNICATIONS   DOI: 10.1038/s41467-018-03231-x

D

emand for primary foodstuffs, that is, grains and seeds of
our major crops, may increase by 70–100% by 20501,2.
One major barrier to meeting this large demand will be
availability of water for crop production. Crop productivity
strongly depends on having a sufﬁcient supply of freshwater, and
agriculture consumes 90% of total global freshwater3. A large
proportion of global food crops depend on irrigation, which is
depleting global groundwater, and putting the sustainability of
global food production at risk4. To capture atmospheric CO2
during photosynthesis, stomatal pores need to stay open to allow
CO2 diffusion into the leaf. However, stomatal opening causes
most of the water absorbed by plant roots to be lost via transpiration. Transpiration is proportional to the water vapor pressure deﬁcit (VPD) from leaf to air, which represents the gradient
between the humidity in leaf internal airspaces and drier air
surrounding the leaf. With the global rise in air and surface
temperatures, VPD has been increasing, thus increasing demand
for irrigation5,6.
Because stomatal opening controls both the CO2 inﬂux and the
water vapor efﬂux, stomata have to respond to many different
cues to balance the ﬂuxes7. Progress has been made in unraveling
the molecular basis of the response of stomata to intercellular
CO2 concentration8 and blue light9, but much less is known
about stomatal response to light quantity. Stomatal opening in
response to light is typically much less pronounced in detached
epidermal layers, but can be restored when the connection with
mesophyll cells is restored10–12. Therefore, although some control
resides in the guard cells, stomatal responses to light intensity
seem to rely strongly on a signal derived from the underlying
mesophyll tissue. The rate of photosynthetic CO2 assimilation at
high light intensity is usually limited by the restriction of CO2
inﬂux imposed by stomata. Thus, control of stomatal opening by
a signal derived directly from photosynthesis could provide a
feedback loop to match the light energy processed by the photosynthetic light-dependent reactions with sufﬁcient supply of
CO2. However, several mutants deﬁcient in speciﬁc components
of the photosynthetic light-dependent or carbon reactions typically show vast decreases in the rate of CO2 assimilation without
corresponding changes in stomatal conductance. For example, in
tobacco plants containing reduced amounts of cytochrome b6/f13,
ribulose-1,5-bisphosphate carboxylase-oxygenase (Rubisco)13,14,
glyceraldehyde 3-phosphate dehydrogenase15, or sedoheptulosebisphosphatase16 net assimilation rate was substantially reduced,
but stomatal conductance often remained relatively unaltered
compared to the wild-type (WT). These results show clearly that
the stomatal opening signal does not scale directly with the rate of
CO2 uptake. However, the interpretation of these results is
severely complicated by the strong decrease of net CO2 assimilation rate associated with these transgenic alterations, which
greatly increases the CO2 concentration in the intercellular airspaces within the leaf (Ci), providing a potent signal for stomatal
closing8.
A recent analysis suggested the redox state of chloroplastic
quinone A (QA) as an early signal for stomatal opening in
response to light17, with a more reduced QA pool corresponding
to increased stomatal opening. QA is the primary electron
acceptor downstream of photosystem II and its oxidation state
reﬂects the balance between excitation energy at photosystem II
and the rate of the Calvin–Benson cycle. This predicts that
decreasing the excitation pressure at photosystem II should
directly affect stomatal opening in response to light by keeping
QA more oxidized. This prediction is tested here by altering
expression of Photosystem II Subunit S (PsbS). PsbS expression
directly affects the rate at which excitation energy absorbed by the
antenna complex of photosystem II is used to reduce QA, because
of its role in non-photochemical quenching (NPQ). NPQ protects
2

NATURE COMMUNICATIONS   (2018)9:868

the photosynthetic machinery under excessive light conditions via
controlled dissipation of absorbed light energy as heat18,19. PsbS
expression strongly stimulates NPQ and promotes photoprotection under high light or rapidly ﬂuctuating conditions20, but
typically does not affect steady-state rates of net CO2 assimilation21, thus keeping control of stomatal movements via Ci relatively unaltered. Manipulation of PsbS expression thus provides
an ideal test case to verify if QA redox state is indeed an early
signal for light-induced stomatal opening. Since PsbS stimulates
the thermal dissipation of excitation energy, we predicted that
increased expression of this protein would keep the redox state of
QA more oxidized, decrease stomatal opening in response to light,
and decrease water loss at the leaf level. To test this hypothesis,
Nicotiana tabacum lines with both increased and decreased PsbS
expression were generated and analyzed under controlled and
ﬁeld conditions. N. tabacum cv “Petite Havana” was transformed
with the coding sequence of N. benthamiana PsbS fused to the
cauliﬂower mosaic virus 35S promoter for constitutive strong
expression (Supplementary Fig. 1). Four independent, single-copy
transformation events with increased NPQ amplitude (PSBS-28,
PSBS-34, PSBS-43, and PSBS-46) were selected and selfed to
obtain progeny homozygous for the transgene. Additionally, two
events exhibiting spontaneous partial silencing of PsbS expression
(psbs-4 and psbs-50) were selected for further analysis. Our
results show that the light response of stomatal conductance is
clearly affected by PsbS expression. Plants overexpressing PsbS
show an average 25% reduction in water loss per CO2 assimilated
under ﬁeld conditions.
Results
PsbS expression under controlled conditions. PSBS-28, PSBS43, psbs-4, and WT N. tabacum plants were grown in a
controlled-environment cabinet and PsbS transcript and protein
levels were measured in samples from the youngest fully expanded leaves. PSBS-28 and PSBS-43 samples showed 4.2-fold and
3.5-fold increases in total (transgenic and native) PsbS transcript
relative to WT (Fig. 1a), whereas transcript levels in psbs-4 were
10-fold less than WT. PsbS protein expression, normalized to the
large subunit of the oxygen-evolving complex (PsbO) as a relative
measure for the abundance of photosystem II, was 2.7-fold higher
in PSBS-43 and 3.5-fold higher in PSBS-28 relative to WT while
virtually absent in psbs-4 (Fig. 1b, c).
PsbS expression affects intrinsic water-use efﬁciency. Net CO2
uptake (An) increased in response to light intensity until
approximately 800 μmol m−2 s−1 and was not signiﬁcantly
affected by PsbS expression (P = 0.6, analysis of variance
(ANOVA); Fig. 1d). The maximum capacity for carboxylation of
ribulose-bisphosphate (Vcmax) and the maximum rate of wholechain electron transport (Jmax) showed weak positive trends with
PsbS content (Supplementary Fig. 2a–d). Rubisco content was
similar between lines (Supplementary Fig. 2e), but Rubisco activation state was slightly lower in psbs-4 (85%), compared to WT
(95%), whereas the overexpressing lines were similar to wild-type
(PSBS-28, 93%) or slightly higher (PSBS-43, 102%; Supplementary Fig. 2f). Stomatal limitation to net CO2 assimilation also
signiﬁcantly differed with PsbS content (Supplementary Fig. 2g);
however, the aforementioned changes in photosynthetic capacity
and Rubisco biochemistry counteracted these differences, leaving
An unchanged between all lines.
As expected, differences in PsbS protein expression led to
pronounced differences in NPQ (Fig. 1e). At high light, NPQ was
signiﬁcantly higher in PSBS-28 and PSBS-43 relative to WT (P ≤
0.05, Dunnett’s two-way test) and signiﬁcantly lower in psbs-4
(P ≤ 0.02, Dunnett’s two-way test). In concert with these
  DOI: 10.1038/s41467-018-03231-x   www.nature.com/naturecommunications

 ARTICLE
c

*

37 kDa

0.02
PsbO

25 kDa
psbs-4

PSBS-43

WT

PSBS-43

PSBS-28

h

P = 0.60

PSBS-28

PsbS

*

2
R = 0.98; P < 0.0001

20
psbs-4

0.4
gs (mol H2O m–2 s–1)

WT

10

PSBS-28
PSBS-43

0

e

P < 0.0001

*

3
NPQ

P < 0.0001

*

0.00

PSBS-43

psbs-4

d

PSBS-28

*

b

WT

400

0.04

psbs-4

*
*

0

An (µmol m–2 s–1)

P < 0.0001

PsbS protein content
(normalized to PsbO)

a
800

WT

PsbS mRNA normalized
to actin and tubulin

NATURE COMMUNICATIONS   DOI: 10.1038/s41467-018-03231-x

0.2

2
1

0.8

0.0

P = 0.0001

0.4

*

0.0
gs (mol H2O m–2 s–1)

*
*

f

An / gs 800–2,000 (µmol CO2 mol–1 H2O)

QA redox state (1-qL)

0

g

P = 0.73

0.25

0.00
0

500

1,000

1,500

2,000

PFD (µmol m–2 s–1)

i

0.4
QA redox state (1-qL)

0.8

2
R = 0.92; P = 0.03

70

65

60

0.000

0.025
PsbS protein content (normalized to PsbO)

Fig. 1 Photosynthesis and water-use efﬁciency in Nicotiana tabacum plants with modiﬁed PsbS levels. a PsbS mRNA levels normalized to actin and tubulin
sampled from fully expanded leaves of psbs-4, PSBS-28, PSBS-43 and wild-type (WT) N. tabacum plants. b PsbS protein levels normalized to the large
subunit of oxygen-evolving complex of photosystem II (PsbO), determined from densitometry on immunoblots. c Representative immunoblot for PsbS and
PsbO. d Net CO2 ﬁxation rate (An), e NPQ, f quinone A (QA) redox state, and g stomatal conductance (gs) as a function of incident light intensity in fully
expanded leaves. h Linear correlation between QA redox state and gs. Broken lines indicate measurements at highest light intensity. i Linear correlation
between PsbS protein levels and intrinsic water-use efﬁciency (An/gs) at light intensity above 600 μmol m−2 s−1. Asterisks/lines show signiﬁcant
differences from WT (black for silencing, red for overexpressing lines; Dunnett’s two-way test; α = 0.05). Error bars indicate s.e.m. (n = from 6 to 10
biological replicates)

differences, the redox state of QA was signiﬁcantly more oxidized
in PSBS-28 and PSBS-43 relative to WT (P ≤ 0.02, Dunnett’s twoway test; Fig. 1f) and more reduced in psbs-4 (P ≤ 0.002,
Dunnett’s two-way test; Fig. 1f). These differences in QA redox
state at high light were reﬂected in differences in stomatal
conductance (gs; Fig. 1g). The change in gs was consistent with
altered regulation of stomatal opening rather than any changes in
stomatal or epidermal anatomy, that is, pore dimensions or
stomatal density. Stomatal density was 21% (abaxial) to 23%
(adaxial) lower in psbs-4 and 18% (abaxial) lower in PSBS-43,
relative to WT (P = 0.006 for abaxial and P = 0.02 for adaxial,
ANOVA; Supplementary Fig. 3a), but unchanged in PSBS-28.
NATURE COMMUNICATIONS   (2018)9:868

Stomatal pore dimensions on both abaxial and adaxial leaf
surfaces were very similar between all lines (Supplementary
Fig. 3b–d). In addition, all measurements of QA redox state and gs
in all lines could be described by a single highly signiﬁcant
positive correlation (P < 0.0001, ANOVA; Fig. 1h), consistent
with a role for QA redox state as an early determinant for stomatal
opening in response to light intensity17. Furthermore, the
differences in gs associated with PsbS expression in combination
with unchanged An resulted in a strong correlation between
intrinsic water-use efﬁciency (An/gs, WUEi) and PsbS expression
(R2 = 0.92, P = 0.03, ANOVA; Fig. 1i).

  DOI: 10.1038/s41467-018-03231-x   www.nature.com/naturecommunications

3

 ARTICLE

NATURE COMMUNICATIONS   DOI: 10.1038/s41467-018-03231-x

37 DAE

a

b

41 DAE

45 DAE

c

d

1000

0

psbs-4

0

WT

500

0.10

i

PSBS-28

e

f

g

h

PSBS-34

NPQ

PSBS-43
PSBS-46

0.05

PSBS-50
psbs-4
psbs-50
WT
PSBS-28
PSBS-34
PSBS-43
PSBS-46

psbs-4
psbs-50
WT
PSBS-28
PSBS-34
PSBS-43
PSBS-46

5
psbs-4
psbs-50
WT
PSBS-28
PSBS-34
PSBS-43
PSBS-46

0.00

psbs-4
psbs-50
WT
PSBS-28
PSBS-34
PSBS-43
PSBS-46

PsbS protein content PsbS mRNA normalized
(normalized to PsbO)
to actin and tubulin

34 DAE
1500

Fig. 2 PsbS expression and photoprotection in Nicotiana tabacum plants grown under ﬁeld conditions. a–d PsbS mRNA levels and e–h PsbS protein levels in
several tobacco genotypes with modiﬁed PsbS expression levels as well as wild-type (WT) tobacco at four time points during the ﬁeld experiment (DAE =
days after emergence). i Representative levels of NPQ for each genotype determined on leaf discs. All genotype means were signiﬁcantly different from
WT (Dunnett’s two-way test; P ≤ 0.008 for mRNA; P ≤ 0.001 for protein). Error bars indicate s.e.m. (n = 4 biological replicates)

WUE and productivity under ﬁeld conditions. The differences
in WUEi observed under controlled conditions were subsequently
tested under ﬁeld conditions. A ﬁeld experiment was conducted
with transformants with both increased (PSBS-28, PSBS-34,
PSBS-43, and PSBS-46) and decreased PsbS expression (psbs-4
and psbs-50) in an incomplete block design (Supplementary
Fig. 4). Western blotting conﬁrmed differences in PsbS expression
in the youngest fully expanded leaf for each genotype at 34, 37,
41, and 45 days after emergence (DAE) (Fig. 2a–h). As predicted,
transformants showed a broad range of PsbS expression from
almost none to 3.7-fold higher than WT, which were directly
reﬂected in levels of NPQ measured on leaf discs (Fig. 2i). Critically, as under controlled conditions, net CO2 assimilation rate
did not differ among the transformants and WT (Fig. 3a, Supplementary Fig. 5a), but gs correlated negatively to PsbS expression (P = 0.0001, ANOVA; Fig. 3b and Supplementary Fig. 5b).
The reduction in gs due to PsbS overexpression varied between 4
and 30%, whereas gs was increased by 46% due to decreased PsbS
expression (Fig. 3b). Once again, WUEi was signiﬁcantly affected
by genotype (P = 0.007, ANOVA; Fig. 3c and Supplementary
Fig. 5d) and correlated positively with PsbS expression (R2 = 0.94,
P = 0.004, ANOVA; Fig. 3d). Increased PsbS expression resulted
in 25–33% increased WUEi, whereas decreased PsbS expression
led to 14% reduction in WUEi. Final size and dry weight were
determined at the end of the ﬁeld experiment. All biomass productivity traits were signiﬁcantly affected by PsbS expression (P ≤
0.008, ANOVA; Fig. 3e–g). Decreased PsbS expression signiﬁcantly reduced dry weight (22%, P ≤ 0.002, Dunnett’s two-way
test; Fig. 3e, Supplementary Fig. 6), leaf area (15%, signiﬁcant
only in psbs-50, P = 0.03, Dunnett’s two-way test; Fig. 3f) and
plant height (15%, P < 0.0001, Dunnett’s two-way test; Fig. 3g).
The productivity measures in transformants with increased PsbS
expression did not show a consistent response. PSBS-28
showed signiﬁcant decreases in dry weight (−18%, P = 0.008,
Dunnett’s two-way test; Fig. 3e), leaf area (−19%, P = 0.005,
Dunnett’s two-way test; Fig. 3f), and plant height was also signiﬁcantly smaller in PSBS-28 and PSBS-34 (−9 and −8%, P ≤
0.01, Dunnett’s two-way test; Fig. 3g), whereas the same productivity measures were not signiﬁcantly affected in PSBS-43 and
PSBS-46, relative to WT.
4

NATURE COMMUNICATIONS   (2018)9:868

Discussion
Our results provide direct proof, through genetic manipulation,
that increasing PsbS expression suppresses stomatal opening with
little effect on CO2 uptake and so increases WUE. We showed a
strong dependence of gs on PsbS expression (Figs. 1g, 3b) and that
overexpression of PsbS signiﬁcantly improved WUEi (Fig. 3c),
representing a strong decrease (averaging 25%) in the amount of
water used for each molecule of CO2 assimilated at leaf level by an
irrigated ﬁeld crop. Novel bioengineering strategies to improve
crop WUE such as exempliﬁed here are urgently needed, especially considering the long timelines for developing new crop
varieties22. Although this test of concept was performed on
tobacco, the role of PsbS in NPQ is universal across higher
plants23, so this manipulation can be expected to be effective
across all crops.
Here a large improvement of leaf level WUEi is shown (Figs. 1i,
3c), which can be expected to conserve soil moisture and may
result in increased productivity if the crop becomes water-limited.
However, many feedbacks could lessen this improvement at the
whole crop level. Open-air elevation of CO2 has provided a direct
test of the signiﬁcance of such feedbacks. When stomatal conductance in a mature soybean canopy was reduced by 10% due to
open-air elevation of [CO2], this led to a corresponding decrease
in the measured ecosystem evapotranspiration by 8.6%24.
Our data for the ﬁrst time show that modulation of QA redox
state via PsbS expression affects stomatal conductance and leaf
WUE. Interestingly, manipulation of PsbS expression did not
directly impact steady-state photosynthesis, which is consistent
with previous ﬁndings21 and thus allowed for speciﬁc targeting of
the putative QA-redox signal without strongly affecting the CO2
control over stomatal conductance via Ci (Fig. 4a). Although Ci
was still somewhat affected (Supplementary Fig. 5c), there was a
strong linear correlation of QA redox state with gs (Fig. 1h), in
contrast to several previous manipulations, where blockage of
electron transfer components downstream of QA13,25 or impairment of Calvin–Benson cycle function14–16,26,27 strongly affected
stomatal control by both QA redox state and Ci in opposing ways,
resulting in ambiguous effects on stomatal conductance. Furthermore, when electron transfer is blocked upstream of QA, both
Ci and QA-redox signals stimulate stomatal closure, which matches observations of stomatal conductance in plants with reduced
  DOI: 10.1038/s41467-018-03231-x   www.nature.com/naturecommunications

 ARTICLE

NATURE COMMUNICATIONS   DOI: 10.1038/s41467-018-03231-x

a

c

An / gs

Y Data

gs

45

P = 0.0001

0

*

d

PSBS-43

PSBS-28

psbs-4

PSBS-43

PSBS-34

psbs-4

PSBS-28

PSBS-43

PSBS-28

psbs-4

60

P = 0.007

*

PSBS-34

P = 0.96

–45

An / gs 400–2,000
(µmol CO2 mol–1 H2O)

b

An

PSBS-34

Relative difference
with WT (%)

90

R 2 = 0.94; P = 0.004

45
psbs-4
WT
PSBS-28

30

PSBS-34
PSBS-43

2D Graph 2
0.000

0.025

0.050

0.075

PsbS protein content (normalized to PsbO)

e

Dry weight

f

Leaf area

g

Plant height

0

–20
P = 0.0001

*

*
P = 0.008

*
* *
P < 0.0001

psbs-4
psbs-50
PSBS-28
PSBS-34
PSBS-43
PSBS-46

*

*

psbs-4
psbs-50
PSBS-28
PSBS-34
PSBS-43
PSBS-46

* *
–40

psbs-4
psbs-50
PSBS-28
PSBS-34
PSBS-43
PSBS-46

Relative difference
with WT (%)

20

Fig. 3 Photosynthetic water-use efﬁciency and productivity of ﬁeld-grown
Nicotiana tabacum plants with modiﬁed PsbS levels. a Net CO2 ﬁxation rate
(An), b stomatal conductance (gs), c intrinsic water-use efﬁciency (An/gs)
in tobacco genotypes with modiﬁed PsbS expression levels relative to wildtype tobacco. d Linear correlation between PsbS protein levels and An/gs at
light intensity above 300 μmol m−2 s−1. e Total dry weight, f Leaf area, and
g plant height. Error bars indicate s.e.m. (a–d N = 4 biological replicates;
e–g n = 6 blocks for transgenic and n = 12 blocks for WT), and asterisks
indicate signiﬁcant differences between transgenic lines and WT
(Dunnett’s two-way test; α = 0.05), P-values indicate signiﬁcance of line
effect in ANOVA (a–c and e–g) or signiﬁcance of regression (d)

levels of PsbO28 or treated with DCMU (3-(3,4-dichlorophenyl)1,1-dimethylurea)29. Thus, when the stomatal control by QA
redox and Ci are both accounted for, as shown by the schematic
in Fig. 4b, our results complement previous observations and for
the ﬁrst time show experimental evidence for molecular regulation of stomatal response to light intensity. Although a constitutive promoter was used here, and thus guard cell expression
of PsbS was also affected, it seems most likely that the QA-redox
signal to open stomata in response to light is perceived in the
mesophyll, especially considering that the ﬂuorescence signal
used to estimate QA-redox state is primarily originating from the
chloroplast-rich mesophyll tissue, consistent with several previous
investigations7.
Since water supply was not limiting for growth in our ﬁeld
experiment, the increased WUE associated with PsbS
NATURE COMMUNICATIONS   (2018)9:868

overexpression was not advantageous for biomass productivity,
but instead led to slight decreases in ﬁnal plant size and dry
weight (Fig. 3e–g and Supplementary Fig. 6). We hypothesize that
the increase in PsbS and associated increased levels of NPQ may
have adversely affected the light-use efﬁciency of CO2 assimilation under ﬂuctuating light as previously shown in rice30. We
have previously demonstrated that this efﬁciency is an important
determinant of biomass productivity of tobacco under similar
ﬁeld conditions31 and therefore may explain these ﬁndings.
Since gs directly affects the supply of CO2 to photosynthesis,
decreases in gs often result in decreased An32. Interestingly, the
effects of PsbS on stomatal conductance did not translate into
differences in An, even though CO2 supply to photosynthesis was
slightly affected (Supplementary Fig. 2g). Instead, maximum
RuBP carboxylation capacity (Vcmax) and the maximal rate of
linear electron transport (Jmax) showed weak positive relationships with the amount of PsbS (Supplementary Fig. 2c, d), consistent with previous ﬁndings in rice with altered PsbS levels30
and Rubisco activation state also showed a weak positive trend
with PsbS content (Supplementary Fig. 2f). These results indicate
that plants may be able to compensate for the effects of a decrease
in stomatal conductance on CO2 uptake by increasing photosynthetic capacity, thereby limiting the negative feedback on
biomass productivity.
Methods

Plant material. WT N. tabacum cv. “Petit Havana” seeds carrying TMV resistance
(NN) were a gift from Professor Spencer Whitney. Lines exhibiting increased or
reduced expression of PsbS were generated within this study as described below.
Recombinant DNA and transformation. The N. benthamiana PsbS gene coding
sequence (www.uniprot.org, Q2LAH0_NICBE) was cloned in between the cauliﬂower mosaic virus 35S and octopine synthase terminator in the pEARLYGATE
100 binary vector. The resulting binary vector pEG100-NbPsbS conferred microbial resistance to kanamycin and bialaphos resistance in planta (Supplementary
Fig. 1). Nicotiana tabacum cv. “Petite Havana” was transformed with pEG100NbPsbS using the Agrobacterium tumefaciens-mediated protocol33. Copy number
and homozygosity were assessed using digital droplet PCR34. Results shown are for
homozygous offspring unless otherwise described.

Transcription and protein expression. Five leaf discs (total 2.9 cm2) were from
the youngest fully expanded leaf of ﬁve plants per genotype (controlled conditions)
or four plants per genotype (ﬁeld). Samples were taken 2 h after the start of the
photoperiod. Protein and mRNA were extracted from the same leaf sample
(NucleoSpin RNA/Protein kit, REF740933, Macherey-Nagel GmbH & Co., Düren,
Germany). Extracted mRNA was treated by DNase (Turbo DNA-free kit; AM1907,
Thermo Fisher Scientiﬁc, Waltham, MA, USA) and transcribed to cDNA using
Superscript III First-Strand Synthesis System for RT-PCR (18089-051, Thermo
Fisher Scientiﬁc, Waltham, MA, USA). Quantitative reverse transcription PCR was
used to quantify PsbS transcripts (5′-GGCACAGCTGAATCTTGAAAC-3′ and 5′CAGGGACAGGGTCATCAATAAA-3′) relative to NtActin (5′-CCTCACAGAAGCTCCTCTTAATC-3′ and 5′-ACAGCCTGAATGGCGATATAC-3′) and
NtTubulin (5′-GTACATGGCCTGTTGTTTGATG-3′ and 5′CTGGATGGTCCTCTTTGTCTTT-3′).
Total protein concentration was quantiﬁed using a protein quantiﬁcation assay
(ref. 740967.50, Macherey-Nagel GmbH & Co., Düren, Germany). Samples
containing 1 µg total protein were separated by sodium dodecyl sulfatepolyacrylamide gel electrophoresis electrophoresis, blotted to membrane
(Immobilon-P, IPVH00010, Millipore, Tullagreen, Carrigtwohill, Ireland) using
semi-dry blotting (Trans-Blot SD, Bio-Rad, Hercules, CA, USA), and sequentially
immuno-labeled with primary antibodies raised against AtPsbS (1:2,000 dilution;
AS09533, Agrisera, Vännäs, Sweden) and AtPsbO (1:20,000 dilution; AS06142-33,
Agrisera, Vännäs, Sweden) followed by incubation with secondary antibodies
(1:2,500 dilution; W401B, Promega, Madison, WI, USA). The sequential use of the
two primary antibodies was veriﬁed empirically against blots where only one
antibody was used and dilution series were used to establish the quantiﬁable range.
Chemiluminescence was detected using a scanner (ImageQuant LAS-4010, GE
Healthcare Life Sciences, Pittsburgh, PA, USA). A protein ladder (Precision Plus
Protein Kaleidoscope Prestained Protein Standards, #1610375, Bio-Rad, Hercules,
CA, USA) was used as a size indicator on each gel. Protein bands were quantiﬁed
using densitometry with ImageQuant TL software (version 7.0 GE Healthcare Life
Sciences, Pittsburgh, PA, USA) (Data Set 2 and 11 in Data Repository; https://data.
mendeley.com/datasets/nsbjps9rkg/draft?a = 10508d31-685a-4a62-8f96cb591c569e97). PsbS expression was normalized based on PsbO bands.

  DOI: 10.1038/s41467-018-03231-x   www.nature.com/naturecommunications

5

 ARTICLE

NATURE COMMUNICATIONS   DOI: 10.1038/s41467-018-03231-x

a

b

Stomatal pore

CO2

H2O

Ci

PsbO silencing28
DCMU treatment29

Blocked electron
transport
upstream of Q A

Silencing
13
of Cyt b6 /f
or FNR25

Blocked electron
transport
downstream of
QA

Silencing
14
of Rubisco ,
26
16
PRK , SBPase ,
27
FBP aldolase ,
15
or GAPDH

Impaired CalvinBenson cycle
function

Q A signal: Open

Silencing
of PsbS

Increased
excitation
pressure at PSII

Q A signal: Open

Overexpression
of PsbS

Decreased
excitation
pressure at PSII

Q A signal: Close

Calvin–Benson
cycle

CO2
RuBP

NADPH
ATP
G3P

Thylakoid membrane

ATP

ADP

NADP
ADP

+

Q A redox signal
to stomata
ATP
e

Fd

Pheo

PSI

QA
QB

PSII

Cyt b6f

NPQ

PC
e–

+

H O
2

H2O

NADPH

–

e–
PsbS

Decrease gs

Q A signal: Open
C i signal: Close

Ambiguous

3PGA

ADP

+

C i signal: Close

Stomatal
cavity

C i signal
to stomata

H

Q A signal: Close

OE

C

FNR

C i signal: Close

C i signal: No
change

Ambiguous

Increase g s

e–

+

H
+
NADP

e–

LHCI

LHCII

C i signal: No
change

Decrease g s

Fig. 4 Interactions between light and CO2 control over stomatal movements. a Schematic representation of processes in the chloroplast thylakoid
membrane, Calvin–Benson cycle in the chloroplast stroma, and at the interchange between intercellular airspace and atmosphere through the stomatal
pore. Indicated are where the CO2 signal (Ci signal) and the proposed QA-redox signal originate. b Chain of events showing the direct effects on the
stomatal control signals from Ci and QA redox state of the manipulation of PsbS expression in the current work and several previously published
manipulations, and the subsequent direction of change in stomatal conductance (gs). Superscripted numbers indicate corresponding literature references.
Abbreviations: 3PGA – 3-Phosphoglycerate; ADP – adenosine diphosphate; ATP – adenosine triphosphate; Cyt b6f – Cytochrome b6f; DCMU – (3-(3,4dichlorophenyl)-1,1-dimethylurea); FBP aldolase – Fructose-bisphosphate aldolase; Fd – Ferredoxin; FNR – Ferredoxin NADP(+) reductase; G3P –
Glyceraldehyde-3-phosphate; GAPDH – Glyceraldehyde 3-phosphate dehydrogenase; LHCI or II – light-harvesting complex I or II; NADP+ / NADPH –
nicotinamide adenine dinucleotide phosphate (oxidized/reduced); NPQ – non-photochemical quenching; OEC – Oxygen evolving complex; PC –
Plastocyanin; Pheo – pheophytin; PRK – Phosphoribulokinase; PsbO – subunit of the oxygen evolving complex (OEC); PsbS – Photosystem II subunit S; PSI
or II – photosystem I or II; QA – Plastoquinone A; QB – Plastoquinone B; Rubisco – Ribulose-1,5-bisphosphate carboxylase-oxygenase; RuBP – Ribulose 1,5bisphosphate, SBPase – Sedoheptulose-bisphosphatase
Photosynthetic gas exchange under controlled conditions. Seedlings of psbs-4,
PSBS-43, PSBS-28, and WT were germinated on growing medium (LC1 Sunshine
mix, Sun Gro Horticulture, Agawam, MA, USA) in a controlled-environment walkin growing chamber (Environmental Growth Chambers, Chagrin Falls, OH, USA)
with photoperiod set to 12 h and temperature controlled at 23/18 °C (day/night).
Five days after germination, psbs-4 seedlings with low NPQ were identiﬁed
through chlorophyll ﬂuorescence imaging and together with PSBS-28, PSBS-43,
and WT seedlings transplanted to 3.8-L pots and randomly positioned in a
controlled-environment chamber (PGC20, Conviron, Winnipeg, MB, Canada)
with photoperiod set to 16 h and air temperature controlled at 20/25 °C (night/
day). Light intensity at leaf-level was controlled at 500 µmol m−2 s−1. Plants were
watered and plant positions were repositioned at random every 2 days until the
ﬁfth leaf was fully expanded. Gas exchange measurements were performed using an
open gas exchange system (LI6400XT, LI-COR, Lincoln, NE, USA) equipped with
a 2-cm2 leaf chamber and integrated modulated ﬂuorometer. All chlorophyll
ﬂuorescence measurements were performed using the multiphase ﬂash routine35.
To determine the light response of An and whole-chain photosynthetic electron
transport, gas exchange and pulse amplitude-modulated chlorophyll ﬂuorescence
were measured at a range of light intensities. Block temperature was controlled at
25 °C, [CO2] inside the cuvette was maintained at 380 µmol mol−1 and leaf-to-air
water VPD was controlled to 1.1–1.4 kPa. Leaves were clamped in the leaf cuvette
and dark-adapted for 1 h, after which minimal (Fo) and maximal ﬂuorescence (Fm)
were measured to determine maximal efﬁciency of whole-chain electron transport36 (Fv/Fm, Eq. 1)
Fv =Fm

¼ ðFm   Fo Þ=Fm :

ð1Þ

Subsequently, light intensity (100% red LEDs, λpeak 630 nm) was slowly
increased from 0 to 50, 80, 110, 140, 170, 200, 300, 400, 500, 600, 800, 1,000, 1,500,
and 2,000 µmol m−2 s−1. When steady state was reached, An, gs, and Ci were
logged, and F′ and Fm′ were measured to estimate the operating efﬁciency of
6

NATURE COMMUNICATIONS   (2018)9:868

whole-chain electron transport36 (Fq′/Fm′, Eq. 2). Since stomatal movements can
include very long-term diurnal components37,38, our routine was aimed at
measuring only relatively short-term stomatal responses to changes in light
intensity, and steady-state waiting times were kept between 10 and 20 min per step.
NPQ of chlorophyll ﬂuorescence was determined according to Eq. 3 assuming a
Stern–Volmer quenching model39. Minimal ﬂuorescence without dark adaptation
(Fo′) was also determined (using a short far-red pulse to fully oxidize QA). The
ﬂuorescence parameter qL (Eq. 4) was used to estimate the fraction of QA in its
oxidized state (and correspondingly, QA redox state as 1 – qL). The derivation of
this parameter is assuming a “lake” model for photosynthetic antenna complexes
(i.e., antennae are shared between reaction centers)40:
Fq ′ =Fm ′
NPQ
qL

¼

¼ ðFm ′   F ′ Þ=Fm ′ ;

ð2Þ

¼ Fm =Fm ′   1;

ð3Þ

ð1=F ′   1=Fm ′ Þ=ð1=Fo ′   1=Fm ′ Þ :

ð4Þ

To evaluate the CO2 response of An, leaves were allowed to reach steady state at a
light intensity of 2,000 µmol m−2 s−1 (100% red LEDs, λpeak = 630 nm), with block
temperature controlled at 25 °C and [CO2] in the airstream set to 400 µmol mol−1.
Subsequently, [CO2] was varied from 400 to 300, 200, 100, 75, 400, 400, 500, 600,
700, 800, 1,000, 1,200, and 1,500 µmol mol−1. When steady state was attained,
An, gs, and Ci were logged. Vcmax was determined from the response of An to
chloroplastic CO2 concentration (Cc) by ﬁtting a biochemical model41 with
temperature corrections42 to measurements. Cc required an estimate of mesophyll
conductance to CO2 transfer (gm). This was estimated independently for each point
in the CO2 response curve from parallel chlorophyll ﬂuorescence measurements
according to the variable J method43. Jmax was determined by ﬁtting a nonrectangular hyperbola to light response curves of linear electron transport
estimated from chlorophyll ﬂuorescence44. Stomatal limitation of An was computed

  DOI: 10.1038/s41467-018-03231-x   www.nature.com/naturecommunications

 ARTICLE

NATURE COMMUNICATIONS   DOI: 10.1038/s41467-018-03231-x
using measurements at ambient CO2 (Ca = 380 μmol mol−1) and saturating light
intensity, and predicted values of An when stomata are not limiting (i.e., Ci would
equal Ca)44.
Rubisco activation state and content. Plants were grown under controlled
conditions as described above. Youngest fully expanded leaves were clamped in the
cuvette of an open gas exchange system (LI6400XT with 2 × 3 LED light source),
with light intensity set to 1800 μmol m−2 s−1, CO2 concentration set to 400 μmol
mol−1, and block temperature set to 25 °C. After steady-state gas exchange was
reached, leaves were rapidly removed and a disc of 0.55 cm2 from the center of the
portion of the leaf that had been enclosed in the cuvette was snap frozen in liquid
N. Rubisco activity was determined by the incorporation of 14CO2 into acid-stable
products at 25 °C following an existing protocol45. Samples were ground in tenbroek glass homogenizers with ~2 mL cm−2 CO2-free extraction buffer containing
100 mM Hepes-KOH (pH 7.5), 2 mM Na2ethylenediaminetetraacetic acid (EDTA),
20 mM MgCl2, 5 mM dithiothreitol (DTT), 5 mg mL−1 polyvinyl pyrrolidine, 15
mM amino-n-caproic acid and 3.5 mM benzamidine, and 5% v/v protease inhibitor
cocktail (P9599, Sigma, St. Louis, MO, USA). Within 30 s of extraction, samples
were assayed for initial Rubisco activity in a buffer containing 100 mM BicineNaOH (pH 8.2), 1 mM Na2EDTA, 20 mM MgCl2, 5 mM DTT, 1 mM ATP, 0.5 mM
ribulose-1,5-bisphosphate, and 12.8 mM NaH14CO3 (15 Bq nmol−1, Vitrax, Placentia, CA, USA). Assays were run for 30 s and terminated with the addition of
300 μL 5 N formic acid. The radioactivity of acid-stable products was determined
by liquid scintillation counting (Packard Tri-Carb 1900 TR, Canberra Packard
Instruments Co., Downers Grove, IL, USA). After determining initial activity, the
extract was incubated with 10 mM NaHCO3 and 20 mM MgCl2 for 20 min at room
temperature, and the total activity of the extract was assayed as above. Unless stated
otherwise, all other reagents were purchased from Sigma (St. Louis, MO, USA).
Puriﬁed RuBP was used in both initial and total activity assays to avoid underestimating the activation state46. The activation state of Rubisco is determined by
the ratio of initial to total activity. Rubisco content was determined from carbamylated samples extracted as above using a [14C]carboxy-arabinitol bisphosphatebinding assay47 with a speciﬁc activity of 583 Bq nmol−1 Rubisco, assuming eight
binding sites per Rubisco45.
Stomatal density and stomatal complex dimension. Plants were grown under
controlled conditions as described above. Fresh leaf samples were taken from the
youngest fully expanded leaf and mounted onto a microscope slide using doublesided tape. Topographies of the adaxial and abaxial surfaces were measured using a
μsurf explorer optical topometer (Nanofocus, Oberhausen, Germany). The 20×/
0.60 objective lens (image size 0.8 × 0.8 mm2) and the 50×/0.80 objective lens
(image size 0.32 × 0.32 mm2) were used, respectively, for stomatal density quantiﬁcation and measurements of stomatal complex dimensions. Based on a prior
bootstrap analysis, 8 and 10 images were analyzed for each of four biological
replicates for stomatal density quantiﬁcation and measurements of stomatal
complex dimensions, respectively. For stomatal density quantiﬁcation, raw images
were ﬁrst optimized for light contrast in μsoft analysis software (Nanofocus,
Oberhausen, Germany) and then exported into TIF format. Each stoma was labeled
and counted manually using multi-point function in ImageJ (ImageJ 1.51k, NIH,
Rockville, MD, USA). Stomatal density was derived by dividing the number of
stomata in each image by the image size (0.64 mm2). To measure the length and
width of the stomatal complex, the distance measurement function in μsoft analysis
software was applied on each stoma which could be completely observed in the
image. Lines were drawn manually from end to end of the stomatal complex ellipse
to measure stomatal length and width.
Seedling propagation for ﬁeld experiment. Homozygous T2 (psbs-50 and PSBS28) or T3 (PSBS-34, PSBS-43, and PSBS-46) seeds as well as segregating T1 seed
from psbs-4 and WT seed from the same harvest date were sown in the greenhouse
on May 16, 2016. Five days after germination, seedlings exhibiting severely reduced
NPQ were identiﬁed by chlorophyll ﬂuorescence imaging of the psbs-4 T1 progeny.
These low NPQ seedlings as well as seedlings from all other lines and WT were
propagated hydroponically for 2 weeks in ﬂoating trays (Transplant Tray GP009
6 × 12 cells, Speedling Inc., Ruskin, FL, USA) ﬁlled with specialized growing
medium for hydroponics (Pro-mix PGX, Premier Tech, Quakertown, PA, USA).
The concentration of total dissolved solids in the solution was measured every
2 days with a handheld TDS meter (COM-100, HM Digital Inc., Culver City, CA,
USA) and adjusted to 100 ppm by the addition of 20-10-20 water-soluble fertilizer
(Jack’s Professional, JR Peters Inc., Allentown, PA, USA). Five days after the
transplant to trays, Etridiazole fungicide (Terramaster 4EC to a ﬁnal concentration
of 78 μl L−1, Crompton Manufacturing Company Inc., Middlebury, CT, USA) was
added to the solution to protect the plants against root fungus disease in the ﬁeld.
Two applications of Mancozeb (Dithane Rainshield Fungicide at 1 g L−1, Dow
AgroSciences Canada Inc., Calgary, AB, Canada) were applied 6 and 9 days after
transplant to prevent foliar fungus disease. On the same days, seedlings were
sprayed with fermentation solids and solubles from Bacillus thuringiensis, subsp
israelensis, strain AM65-52 (Gnatrol WDG Biological larvicide at 1 mL L−1, Valent
Biosciences Corp., Libertyville, IL, USA) to reduce the greenhouse population of
fungus gnats.
NATURE COMMUNICATIONS   (2018)9:868

Field experimental design. Seedlings were transplanted to an experimental ﬁeld
site at the University of Illinois Energy Farm (40.11°N, 88.21°W) on June 9, 2016.
The ﬁeld was prepared 2 weeks prior to transplant by rototilling, cultivation, and
harrowing. At this time, chlorpyrifos (1.5 g m−2 Lorsban 15 G Insecticide, Dow
AgroSciences Canada Inc., Calgary, AB, Canada) was worked into the soil to
suppress cutworm damage, sulfentrazone (29 μL m−2 Spartan 4 F preemergence
herbicide, FMC Agricultural Solutions, Philadelphia, PA, USA) was applied to
reduce the emergence of weeds and slow-release fertilizer (30.8 g m−2 ESN Smart
Nitrogen, Agrium US Inc., Denver, CO, USA) was put down. After transplant, all
seedlings were sprayed with thiamethoxam (7 mg/plant Platinum 75 SG insecticide,
Syngenta Crop Protection LLC, Greensboro, NC, USA) to prevent damage from
insect herbivory, and 12 days after the ﬁeld transplant, all plants were sprayed with
fermentation solids, spores, and insecticidal toxins from Bacillus thuringiensis,
subsp. kurstaki, strain ABTS-351 (2.6 mL L−1 DiPel Pro dry ﬂowable biological
insecticide, Valent Biosciences Corp.) to suppress tobacco hornworm. The ﬁeld
experiment was set up as an incomplete randomized block design with 12 blocks of
6 × 6 plants spaced 30 cm apart (Supplementary Fig. 4). Each block contained four
rows of four plants per genotype in north–south (N–S) orientation, surrounded by
one border row of WT. WT was present in all blocks (n = 12), whereas the four
PSBS overexpression and two psbs knock-down lines were randomly assigned to
six blocks (n = 6). The blocks were positioned in a 3 (N–S) × 4 (E–W) rectangle
with 75 cm spacing between blocks. The entire experiment was surrounded by two
border rows of WT.
Light intensity (LI-190R quantum sensor, LI-COR, Lincoln, NE, USA) and air
temperature (Model 109 temperature probe, Campbell Scientiﬁc, USA) were
measured nearby on the same ﬁeld site and half-hourly averages were logged using
a datalogger (CR1000, Campbell Scientiﬁc, USA). Precipitation was measured at
two locations close to the ﬁeld using precipitation gauges (NOAH IV Precipitation
Gauge, ETI Instrument Systems Inc., Fort Collins, CO, USA) (Supplementary
Fig. 7). Watering to restore ﬁeld capacity was provided daily when needed through
parallel drip irrigation lines with emitters every 30 cm (17 mm PC Drip Line
#DL077, The Drip Store, Vista, CA, USA) spanning the whole experiment in E–W
orientation and spaced 30 cm apart in N–S direction. To improve soil drainage
after watering and precipitation events, two trenches with a depth of approximately
10 cm were dug in N–S direction between the blocks and connected on the south
side of the experiment to a 15 cm deep E–W trench. Photosynthesis measurements
were performed on the youngest fully expanded leaf 22 days after transplanting.
Plants were harvested on July 7, 2016. At ﬁnal harvest, stem length and the number
of leaves were determined, and leaf area was measured with a conveyor-belt
scanner (LI-3100C Area meter, LI-COR, Lincoln NE, USA). Leaf, stem, and root
fractions were dried to constant weight at 60 °C in a custom-built drying oven
equipped with condenser to further dry the recirculated air, after which the dry
weights were determined.
Non-photochemical quenching in ﬁeld-grown plants. Leaf discs were sampled
pre-dawn from ﬁeld-grown plants of psbs-4, psbs-50, PSBS-28, PSBS-34, PSBS-43,
PSBS-46, and WT control and stored in darkness in glass vials for up to 4 h until
measurement. Humidity in the vials was maintained fully saturated by placing a
piece of wet ﬁlter paper in each vial. Dark-adapted leaf discs were positioned on a
piece of wet ﬁlter paper in a chlorophyll ﬂuorescence imager (CFimager, Technologica, Colchester, UK) to determine maximal ﬂuorescence (Fm). Subsequently,
leaf discs were exposed to 15 min of 1000 µmol m−2 s−1, after which maximal
ﬂuorescence without dark adaptation was determined (Fm'). NPQ was then
determined according to Eq. 3.
Photosynthetic gas exchange in ﬁeld. The response of photosynthetic gas
exchange to light intensity was measured on the youngest fully expanded leaf of
four plants of psbs-4, PSBS-28, PSBS-34, PSBS-43, and WT control in the S–W
blocks. Measurements were performed in four complete sets to account for random
effects of N–S position of plants, and time of day. Leaves were clamped in the
cuvette of an open gas exchange system (LI6400XT, LI-COR, Lincoln, NE, USA)
and allowed to reach steady-state gas exchange at saturating light intensity of 2000
µmol m−2 s−1, with block temperature set to 30 °C and [CO2] in the airstream
controlled to 400 µmol mol−1 and vapor pressure deﬁcit between air and leaf kept
below 1.5 kPa. Subsequently, light intensity was varied from 2,000 to 1,500, 1,000,
800, 600, 400, 300, 200, 170, 140, 110, 80, and 50 µmol m2 s−1. Due to the limited
window suitable for measuring gas exchange in ﬁeld trials, waiting time for steady
state was kept between 5 and 10 min for these measurements. When steady state
was reached, net assimilation rate, stomatal conductance, and intercellular [CO2]
were logged. After gas exchange measurements were performed, leaf absorptance
was determined using an integrating sphere (LI1800, LI-COR, Lincoln, NE, USA)
connected to a spectrometer (USB-2000, Ocean Optics Inc., Dunedin, FL, USA).
Statistical analysis. All statistical analysis was performed with SAS (version 9.3,
SAS Institute Inc., Cary, NC, USA). Data were tested with the Brown–Forsythe test
for homogeneity of variance and the Shapiro–Wilk test for normality. One-way
analysis of variance was applied to transcription levels, protein expression, gas
exchange data, Rubisco content and activation state, stomatal density, and
dimension data. Measurement set was included as a random effect in analysis of the

  DOI: 10.1038/s41467-018-03231-x   www.nature.com/naturecommunications

7

 ARTICLE

NATURE COMMUNICATIONS   DOI: 10.1038/s41467-018-03231-x

ﬁeld gas exchange data to account for variation caused by N–S plant position and
time of day. Biomass, leaf area, and plant height data were analyzed with a linear
mixed model accounting for block and genotype effects with Welch–Satterthwaite
adjustment of degrees of freedom to account for the different replication rate of
WT (PROC MIXED). Signiﬁcant genotype effects in ANOVA (α = 0.05) were
followed by testing of genotype means against WT control (α = 0.05), using
Dunnett’s multiple comparison correction. Correlations between QA redox state
with gs and protein levels and with An/gs were evaluated using Pearson’s correlation
coefﬁcient.
Data availability. All relevant data and plant materials are available from the
authors upon request. Raw data corresponding to the ﬁgures and results described
in this manuscript have been deposited at: https://data.mendeley.com/datasets/
nsbjps9rkg/draft?a=10508d31-685a-4a62-8f96-cb591c569e97.

Received: 28 July 2017 Accepted: 26 January 2018

References
1.

2.

3.

4.
5.
6.
7.

8.
9.
10.
11.
12.
13.

14.

15.

16.

17.

18.
19.

20.

8

Long, S.P., Marshall-Colon, A. & Zhu, X.G. Meeting the global food demand
of the future by engineering crop photosynthesis and yield potential. Cell 161,
56–66 (2015).
Tilman, D., Balzer, C., Hill, J. & Befort, B. L. Global food demand and the
sustainable intensiﬁcation of agriculture. Proc. Natl. Acad. Sci. USA 108,
20260–20264 (2011).
Scanlon, B. R. et al. Groundwater depletion and sustainability of irrigation in
the US High Plains and Central Valley. Proc. Natl. Acad. Sci. USA 109,
9320–9325 (2012).
Dalin, C., Wada, Y., Kastner, T. & Puma, M. J. Groundwater depletion
embedded in international food trade. Nat. Commun. 543, 700–704 (2017).
Lobell, D. B. et al. Greater sensitivity to drought accompanies maize yield
increase in the U.S. Midwest. Science 344, 516–519 (2014).
Ort, D. R. & Long, S. P. Limits on yields in the corn belt. Science 344, 484–485
(2014).
Lawson, T., Simkin, A. J., Kelly, G. & Granot, D. Mesophyll photosynthesis
and guard cell metabolism impacts on stomatal behaviour. New Phytol. 203,
1064–1081 (2014).
Hu, H. et al. Carbonic anhydrases are upstream regulators in guard cells of
CO2-controlled stomatal movements. Nat. Cell Biol. 12, 87–118 (2010).
Shimazaki, K.-i, Doi, M., Assmann, S. M. & Kinoshita, T. Light regulation of
stomatal movement. Annu. Rev. Plant Biol. 58, 219–247 (2007).
Lee, J.-S. & Bowling, D. J. F. Inﬂuence of the mesophyll on stomatal opening.
Aust. J. Plant Physiol. 22, 357–363 (1995).
Lawson, T. Guard cell photosynthesis and stomatal function. New Phytol. 181,
13–34 (2009).
Mott, E. A. Guard cell photosynthesis and stomatal function. Plant Cell
Environ. 32, 1479–1486 (2009).
Baroli, I., Price, G.D., Badger, M.R. & von Caemmerer, S. The contribution of
photosynthesis to the red light response of stomatal conductance. Plant
Physiol. 146, 737–747 (2008).
von Caemmerer, S. et al. Stomatal conductance does not correlate with
photosynthetic capacity in transgenic tobacco with reduced amounts of
Rubisco. J. Exp. Bot. 55, 1157–1166 (2004).
Price, G. D., Evans, J. R., von Caemmerer, S., Yu, J.-W. & Badger, M. R.
Speciﬁc reduction of chloroplast glyceraldehyde-3-phosphate dehydrogenase
activity by antisense RNA reduces CO2 assimilation via a reduction in ribulose
bisphosphate regeneration in transgenic tobacco plants. Planta 195, 369–378
(1995).
Lawson, T., Lefebvre, S., Baker, N. R., Morison, J. I. L. & Raines, C. A.
Reductions in mesophyll and guard cell photosynthesis impact on the control
of stomatal responses to light and CO2. J. Exp. Bot. 59, 3609–3619 (2008).
Busch, F. A. Opinion: the red-light response of stomatal movement is sensed
by the redox state of the photosynthetic electron transport chain. Photosynth.
Res. 119, 131–140 (2014).
Müller, P., Li, X.-P. & Niyogi, K. K. Non-photochemical quenching. A
response to excess light energy. Plant Physiol. 125, 1558–1566 (2001).
Ruban, A. V. Nonphotochemical chlorophyll ﬂuorescence quenching:
mechanism and effectiveness in protecting plants from photodamage. Plant
Physiol. 170, 1903–1916 (2016).
Li, X.-P., Müller-Moule, P., Gilmore, A. M. & Niyogi, K. K. PsbS-dependent
enhancement of feedback de-excitation protects photosystem II from
photoinhibition. Proc. Natl. Acad. Sci. USA 99, 15222–15227 (2002).

NATURE COMMUNICATIONS   (2018)9:868

21. Logan, B. A., Terry, S. G. & Niyogiy, K. K. Arabidopsis genotypes with
differing levels of PsbS expression differ in photosystem II quantum yield,
xanthophyll cycle pool size, and aboveground growth. Int. J. Plant Sci. 169,
597–604 (2008).
22. Kromdijk, J. & Long, S. P. One crop breeding cycle from starvation? How
engineering crop photosynthesis for rising CO2 and temperature could be one
important route to alleviation. Proc. R. Soc. Lond. Ser. B 283, 1–8 (2016).
23. Brooks, M., Jansson, S. & Niyogi, K. in Non-Photochemical Quenching and
Energy Dissipation in Plants, Algae and Cyanobacteria Advances in
Photosynthesis and Respiration Vol. 40 (eds Govindjee & Sharkey, T. D.)
297–314 (Springer, Berlin, 2014).
24. Bernacchi, C. J., Kimball, B. A., Quarles, D. R., Long, S. P. & Ort, D. R.
Decreases in stomatal conductance of soybean under open-air elevation of
[CO2] are closely coupled with decreases in ecosystem evapotranspiration.
Plant Physiol. 143, 134–144 (2007).
25. Hajirezaei, M.-R. et al. Small chanages in the activity of chloroplastic
NADP+-dependent ferredoxin oxidoreductase lead to impaired plant growth
and restrict photosynthetic activity of transgenic tabacco plants. Plant J. 29,
281–293 (2002).
26. Haake, V., Zrenner, R., Sonnewald, U. & Stitt, M. A moderate decrease
of plastid aldolase activity inhibits photosynthesis, alters the levels of
sugars and starch, and inhibits growth of potato plants. Plant J. 14, 147–157
(1998).
27. Paul, M. J. et al. Reduction in phosphoribulokinase activity by antisense RNA
in transgenic tobacco: effect on CO2 assimilation and growth in low
irradiance. Plant J. 7, 535–542 (1995).
28. Dwyer, S. A. et al. Antisense reductions in the PsbO protein of photosystem II
leads to decreased quantum yield but similar maximal photosynthetic rates. J.
Exp. Bot. 63, 4781–4795 (2012).
29. Messinger, S., Buckley, T. & Mott, K. Evidence for involvement of
photosynthetic processes in the stomatal response to CO2. Plant Physiol. 140,
771–778 (2006).
30. Hubbart, S., Ajigboye, O. O., Horton, P. & Murchie, E. H. The photoprotective
protein PsbS exerts control over CO2 assimilation rate in ﬂuctuating light in
rice. Plant J. 71, 402–412 (2012).
31. Kromdijk, J. et al. Improving photosynthesis and crop productivity by
accelerating recovery from photoprotection. Science 354, 857–861 (2016).
32. Farquhar, D. & Sharkey, T. Stomatal conductance and photosynthesis. Annu.
Rev. Plant Physiol. 33, 317–345 (1982).
33. Clemente, T. in Agrobacterium Protocols Vol. 343 Methods in Molecular
Biology (ed. Wang, K.) Ch. 12, 143–154 (Humana Press, Totowa, 2006).
34. Głowacka, K. et al. An evaluation of new and established methods to
determine T-DNA copy number and homozygosity in transgenic plants. Plant
Cell Environ. 39, 908–917 (2016).
35. Loriaux, S. D. et al. Closing in on maximum yield of chlorophyll ﬂuorescence
using a single multiphase ﬂash of sub-saturating intensity. Plant Cell Environ.
36, 1755–1770 (2013).
36. Genty, B., Briantais, J.-M. & Baker, N. R. The relationship between quantum
yield of photosynthetic electron transport and quenching of chlorophyll
ﬂuorescence. Biochim. Biophys. Acta 990, 87–92 (1989).
37. Vialet-Chabrand, S., Dreyer, E. & Brendel, O. Performance of a new dynamic
model for predicting diurnal time courses of stomatal conductance at the leaf
level. Plant Cell Environ. 36, 1529–1546 (2013).
38. Matthews, J. S. A., Vialet-Chabrand, S. & Lawson, T. Diurnal variation in gas
exchange: the balance between carbon ﬁxation and water loss. Plant Physiol.
174, 614–623 (2017).
39. Bilger, W. & Björkman, O. Relationships among violaxanthin deepoxidation,
thylakoid membrane conformation, and non-photochemical chlorophyll
ﬂuorescence quenching in leaves of cotton (Gossypium hirsutum L.). Planta
193, 238–246 (1994).
40. Kramer, D. M., Johnson, G., Kiirats, O. & Edwards, G. E. New ﬂuorescence
parameters for the determination of QA redox state and excitation energy
ﬂuxes. Photosynth. Res. 79, 209–218 (2004).
41. Farquhar, G. D., Von Caemmerer, S. & Berry, J. A. A biochemical model of
photosynthetic CO2 assimilation in leaves of C3 species. Planta 149, 78–90
(1980).
42. Sharkey, T. D., Bernacchi, C. J., Farquhar, G. D. & Singsaas, E. L. Fitting
photosynthetic carbon dioxide response curves for C3 leaves. Plant Cell
Environ. 30, 1035–1040 (2007).
43. Harley, P., Thomas, R., Reynolds, J. & Strain, B. Modelling photosynthesis of
cotton grown in elevated CO2. Plant Cell Environ. 15, 271–282 (1992).
44. Long, S. P. & Bernacchi, C. J. Gas exchange measurements, what can they tell
us about the underlying limitations to photosynthesis? Procedures and sources
of error. J. Exp. Bot. 54, 2393–2401 (2003).

  DOI: 10.1038/s41467-018-03231-x   www.nature.com/naturecommunications

 ARTICLE

NATURE COMMUNICATIONS   DOI: 10.1038/s41467-018-03231-x

45. Kubien, D., Brown, C. & Kane, H. in Quantifying the Amount and Activity
of Rubisco in Leaves Vol. 684 Methods in Molecular Biology (ed. Carpentier,
R.) Ch. 27, 349–362 (Springer Science+Business Media, Berlin, 2011).
46. Sharwood, R. E., Sonawane, B. V., Ghannoum, O. & Whitney, S. M. Improved
analysis of C4 and C3 photosynthesis via reﬁned in vitro assays of their carbon
ﬁxation biochemistry. J. Exp. Bot. 67, 3137–3148 (2016).
47. Ruuska, S. et al. The interplay between limiting processes in C3 photosynthesis
studied by rapid-response gas exchange using transgenic tobacco impaired in
photosynthesis. Aust. J. Plant Physiol. 25, 859–870 (1998).

Additional information
Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467018-03231-x.
Competing interests: The authors declare no competing ﬁnancial interests.
Reprints and permission information is available online at http://npg.nature.com/
reprintsandpermissions/
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional afﬁliations.

Acknowledgements
We thank David Drag and Ben Harbaugh for plant management in greenhouse and ﬁeld
studies, Madeline Steiner and Sydney Gillespie for general assistance during lab work and
ﬁeldwork. This research was supported by the Bill and Melinda Gates Foundation
[OPP1060461] titled “RIPE—Realizing Increased Photosynthetic Efﬁciency for Sustainable Increases in Crop Yield”. K.K.N. is an investigator of the Howard Hughes Medical
Institute and the Gordon and Betty Moore Foundation (through Grant GBMF3070). A.
D.B.L. was supported by the National Science Foundation Directorate of Biological
Sciences (through Grant PGR–1238030).

Author contributions
K.G., J.K., K.K.N., and S.P.L. designed experiments. L.L. prepared plasmid. K.G. and J.K.
generated transgenic tobacco lines. K.G., K.K., and J.K. performed experiments. A.P.C.
and D.R.O. measured Rubisco content and activation state. J.X. and A.D.B.L. measured
stomatal density and dimensions of stomatal complex. K.G. and J.K. analyzed experiments. K.G., J.K., K.K., L.L., K.K.N., and S.P.L. wrote the manuscript.

NATURE COMMUNICATIONS   (2018)9:868

Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative
Commons license, and indicate if changes were made. The images or other third party
material in this article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not included in the
article’s Creative Commons license and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this license, visit http://creativecommons.org/
licenses/by/4.0/.
© The Author(s) 2018

  DOI: 10.1038/s41467-018-03231-x   www.nature.com/naturecommunications

9