Systemic signaling during abiotic stress combinationin plantsSara I. Zandalinasa,b, Yosef Fichmana,b

, Amith R. Devireddya,b, Soham Senguptac, Rajeev K. Azadc,d,and Ron Mittlera,b,e,1

aDivision of Plant Sciences, College of Agriculture, Food and Natural Resources, Christopher S. Bond Life Sciences Center, University of Missouri, Columbia,MO 65201; bInterdisciplinary Plant Group, Christopher S. Bond Life Sciences Center, University of Missouri, Columbia, MO 65201; cDepartment of BiologicalSciences, College of Science, University of North Texas, Denton, TX 76203-5017; dDepartment of Mathematics, College of Science, University of North Texas,Denton, TX 76203-5017; and eThe Department of Surgery, University of Missouri School of Medicine, Christopher S. Bond Life Sciences Center, University ofMissouri, Columbia, MO 65201

Edited by Donald R. Ort, University of Illinois at Urbana Champaign, Urbana, IL, and approved April 21, 2020 (received for review March 17, 2020)

Extreme environmental conditions, such as heat, salinity, and de-creased water availability, can have a devastating impact on plantgrowth and productivity, potentially resulting in the collapse of entireecosystems. Stress-induced systemic signaling and systemic acquiredacclimation play canonical roles in plant survival during episodes ofenvironmental stress. Recent studies revealed that in response to asingle abiotic stress, applied to a single leaf, plants mount a compre-hensive stress-specific systemic response that includes the accumula-tion of many different stress-specific transcripts and metabolites, aswell as a coordinated stress-specific whole-plant stomatal response.However, in nature plants are routinely subjected to a combination oftwo or more different abiotic stresses, each potentially triggering itsown stress-specific systemic response, highlighting a new fundamen-tal question in plant biology: are plants capable of integrating twodifferent systemic signals simultaneously generated during conditionsof stress combination? Here we show that plants can integrate twodifferent systemic signals simultaneously generated during stresscombination, and that the manner in which plants sense the differentstresses that trigger these signals (i.e., at the same or different partsof the plant) makes a significant difference in how fast and efficientthey induce systemic reactive oxygen species (ROS) signals; transcrip-tomic, hormonal, and stomatal responses; as well as plant acclimation.Our results shed light on how plants acclimate to their environmentand survive a combination of different abiotic stresses. In addition,they highlight a key role for systemic ROS signals in coordinating theresponse of different leaves to stress.

abiotic stress | reactive oxygen species | stress combination | systemicacquired acclimation | systemic signaling

Abiotic stress conditions, such as heat, salinity, and decreasedwater availability, can have a devastating impact on plant

growth and productivity, potentially resulting in extensive yieldlosses to agriculture, as well as the collapse of entire ecosystems(1, 2). To withstand harsh environmental conditions, plants evolvedsophisticated perception, signaling, and acclimation mechanismsthat allow them to survive different stress conditions, even at thecost of reduced growth and yield (2). However, successful accli-mation of plants to stress conditions requires an efficient, timely,and coordinated response that spans most, if not all, parts and tis-sues of the plant (3). To achieve such as a coordinated response,plants evolved multiple systemic signaling pathways that allow themto communicate different stress signals from a particular part of theplant, that initially sensed the stress (i.e., local tissue), to the entireplant (i.e., systemic tissue), within minutes (3–15). Once these sys-temic signals are perceived in the systemic tissues of plants, theyinduce an acclimation process, termed “systemic acquired accli-mation” (SAA) (16), that enables these tissues to withstand thestress even if they did not sense or experience it yet (5, 6, 15).Among the different systemic signals found to propagate from

a stressed local tissue to the entire plant within minutes areelectric, calcium, reactive oxygen species (ROS), and hydraulicwaves (3–15, 17), as well as changes in the levels of the plant

hormones jasmonic acid (JA), abscisic acid, ethylene, and dif-ferent metabolites (18). Rapid and systemic whole-plant trans-mission of electric, calcium, and ROS waves was recently shownto be required for plant acclimation to heat or light stresses, aswell as for systemic wound responses (4–7, 15). Although thesesystemic signals were demonstrated to function in response to asingle stress stimuli affecting a particular leaf or a root tip of theplant (4–7, 15), in nature plants frequently encounter more thanone environmental stress condition at a time, resulting in acondition termed “stress combination” (19–22). Acclimation to astate of stress combination (e.g., a combination of drought andheat) has been shown to involve integrating responses to each ofthe individual stresses that simultaneously impact the plant (e.g.,drought or heat), as well as the induction of a new type of re-sponse, sometimes involving thousands of transcripts, that isunique to the state of stress combination (19–22).The different abiotic stresses that simultaneously impact a plant

during stress combination may be sensed by the same or differentparts (or tissues) of the plant (3). In addition, it was found thateach different abiotic stress sensed by the plant (e.g., wounding,high light or heat stress) will trigger its own abiotic stress-specificsystemic signaling and acclimation responses that include the ac-cumulation of many different stress-specific transcripts andmetabolites,

Significance

Environmental stresses such as heat, drought, and salinity,especially in combination with intense light conditions, canhave devastating economical and sociological impacts. Althoughour knowledge of how each of these stresses affects plantswhen applied individually is vast, we know very little about howplants acclimate to a combination of different stresses. Here wereveal that plants can integrate different local and systemicsignals generated during conditions of stress combination. Wefurther show that the specific part at which plants sense the twoco-occurring stresses makes a significant difference in how fastand efficient they acclimate. Our results shed light on how plantsacclimate to and survive a combination of different stresses.

Author contributions: S.I.Z., Y.F., and R.M. designed research; S.I.Z., Y.F., and A.R.D. per-formed research; S.I.Z., Y.F., A.R.D., S.S., R.K.A., and R.M. analyzed data; and S.I.Z., Y.F.,R.K.A., and R.M. wrote the paper.

The authors declare no competing interest.

This article is a PNAS Direct Submission.

This open access article is distributed under Creative Commons Attribution License 4.0(CC BY).

Data deposition: Raw and processed RNA-seq data files have been deposited in the GeneExpression Omnibus (GEO) database, https://www.ncbi.nlm.nih.gov/geo (accession no.GSE138196).1To whom correspondence may be addressed. Email: mittlerr@missouri.edu.

This article contains supporting information online at https://www.pnas.org/lookup/suppl/doi:10.1073/pnas.2005077117/-/DCSupplemental.

First published May 29, 2020.

13810–13820 | PNAS | June 16, 2020 | vol. 117 | no. 24 www.pnas.org/cgi/doi/10.1073/pnas.2005077117

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

11, 2

020

as well as a coordinated stress-specific canopy-wide stomatal re-sponse (5, 23). The possible coactivation of different systemicsignals in the same plant during stress combination raises a newfundamental question in plant biology: are plants capable ofintegrating different systemic signals that simultaneously orig-inate at the same or different parts of the plant during stresscombination?To address this question, we studied the local and systemic

response of the flowering plant Arabidopsis thaliana to a com-bination heat and light stress applied to the same or two differentleaves of the same plant.

ResultsSystemic Response of Plants to a Combination of High Light and HeatStress Applied to a Single Leaf. To study systemic signal integrationduring stress combination, we subjected a single leaf of Arabi-dopsis to a local treatment of high light (HL), heat stress (HS), ora combination of light and heat stresses (applied to the same

leaf), and studied local and systemic responses (Fig. 1, Table 1,SI Appendix, Fig. S1, and Datasets S1–S9). A combination of HLand HS is common to field-grown plants during midday at tem-perate and tropical regions worldwide and was shown to have anadverse impact on photosynthesis, plant growth, and plant survivalcompared with each of its different components (HL or HS) ap-plied individually (19–22). Applying a combination of the twocomponents to the same leaf (HL+HS) resulted in a local re-sponse that included transcripts specific to light or heat stress, aswell as transcripts unique to the stress combination (Fig. 1A). Thisstate of stress combination generated a systemic signal(s) thatresulted in a systemic response to stress combination in systemicleaves (Fig. 1A). Although the systemic response of plants to a localstate of stress combination (HL+HS) included transcripts uniqueto each individual stress, as well as to the stress combination, itscomposition was nonetheless different from that of local leaves,representing an overall overlap of ∼25 to 30% with just 51 tran-scripts common between the stress combination-unique transcripts

15

20

25

30

% o

verla

p w

ith lo

cal

HL

HS

HL+HS

593

1833

404

6811859

924

742(4206) (4838)

(3780)

HL+HSHS

HL

HS

HL+HS2126

485

1196 276345

649 217(4316)

(3477)

(2964)Systemic HL+HS

Local HL+HS

51

225

(276)

1782

(1833)

367153

A B

C

Fig. 1. Plants can integrate two different systemic signals generated simultaneously at the same leaf during stress combination. (A) Overlap between thetranscriptomic response of plants to a local treatment of HL, HS, or a combination of heat and light stresses applied to the same leaf (HL+HS). Venn diagramsfor the overlap between the different responses are shown at the bottom for local leaves and at the top for systemic leaves. Black arrows in HL+HS plantsrepresent the number of HL-specific (solid) or HS-specific (dashed) transcripts common between local and systemic leaves. (B) Venn diagram showing theoverlap between stress combination-specific transcripts in local and systemic leaves. (C) Bar graph showing the percent overlap between the systemic HL+HSresponse and local HL, HS, or HL+HS responses. Local leaves were subjected to HL, HS, or a combination of HL+HS, and local and systemic leaves were sampledat 2 and 8 min following stress application. All experiments were repeated at least three times with 40 plants per biological repeat. All transcripts shown weresignificantly different from controls at P < 0.05 (negative binomial Wald test followed by Benjamini–Hochberg correction).

Zandalinas et al. PNAS | June 16, 2020 | vol. 117 | no. 24 | 13811

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

11, 2

020

of local and systemic leaves (Fig. 1 B and C). Furthermore, althoughthe response of local and systemic leaves of plants individuallyand locally subjected to HL or HS included many hydrogen per-oxide (H2O2)-, HL-, HS-, and salicylic acid (SA)-response tran-scripts, these transcripts were not highly represented in thesystemic response of plants simultaneously subjected to HL+HS(Table 1). These results demonstrate that plants are capable ofintegrating two different systemic signals (one for heat and onefor high light), that the response of plants to stress combinationin systemic leaves differs from that of local leaves, and thatcompared with the systemic response of plants to HL or HSindividually applied to a single leaf, the systemic response ofplants to the two stresses applied simultaneously to the sameleaf (HL+HS) is less comprehensive, at least when it comesto HL-, HS-, SA-, or H2O2-response transcripts (Fig. 1 andTable 1).

Systemic Response of Plants to a Combination of High Light and HeatStress Applied to Two Different Leaves of the Same Plant. We nextexamined the systemic response of plants to a combination ofheat and light stress applied to two different leaves (Fig. 2, Ta-ble 1, and Datasets S1–S9). Applying a combination of light andheat stress to two different leaves of the same plant (HL&HS)resulted in local responses in the two different leaves that in-cluded transcripts specific to light or heat stress, as well as tran-scripts unique to the stress combination (Fig. 2A). This state ofstress combination generated at least two different systemic signalsthat resulted in a systemic response to stress combination in sys-temic leaves (Fig. 2A). Although the systemic response of plants toa local state of stress combination (HL&HS) included transcriptsunique to each individual stress, as well as to the stress combi-nation, its composition also differed from that of local leaves,representing an overall overlap of ∼30 to 35%, with only 47 or 69transcripts common between the stress combination-unique tran-scripts of local and systemic leaves (Figs. 1 and 2B). Interestingly,compared with the systemic response of plants to two differentstresses applied to the same leaf (HL+HS), the response of sys-temic leaves of plants subjected to two different stresses, eachapplied to a different leaf (HL&HS), included a higher proportion

of H2O2-, HL-, HS-, and SA-response transcripts (Table 1). Theseresults demonstrate that plants are capable of integrating twodifferent systemic signals (one for heat and one for high light),generated at two different leaves, that the response of plants tostress combination in systemic leaves is different from that of localleaves, and that the systemic response of plants to two differentstresses simultaneously applied to two different local leaves(HL&HS) could potentially be different and more comprehensivecompared with the systemic response of plants to two differentstresses simultaneously applied to the same leaf (HL+HS) (Figs. 1and 2 and Table 1).

Comparing the Systemic Response of Plants to Stress CombinationApplied to the Same or Two Different Leaves. The results pre-sented in Figs. 1 and 2 and Table 1 suggest that the systemic re-sponse of plants to abiotic stress combination applied to the same(HL+HS) or two different (HL&HS) leaves could be different.Therefore, we conducted a more extensive analysis of the differ-ences between these two systemic responses. As shown in Fig. 3A,the systemic response of plants to two different stresses appliedsimultaneously to two different leaves (HL&HS) was different andmore extensive than the response to two different stresses appliedto the same leaf (HL+HS). Although both systemic responsesincluded transcripts unique to light or heat stress, as well as uniqueto the stress combination, the systemic response to two differentstresses applied to two different leaves (HL&HS) was morecomprehensive, with ∼1,200 additional transcripts (Fig. 3A). Inaddition, although an overlap of 2,477 transcripts was found be-tween the two systemic responses (HL&HS and HL+HS) (Fig. 3),there were many more systemic transcripts unique to the state oftwo different stresses applied to two different leaves (HL&HSstress combination: 1,692 [Fig. 3A], as well as 540 transcriptsunique to the stress combination [Fig. 2A]), compared with that oftwo different stresses applied to the same leaf (HL+HS stresscombination: 487 [Fig. 3A], with only 276 transcripts unique tostress combination [Fig. 1A]), with a higher percent overlap be-tween local and systemic responses under conditions of HL&HS(Figs. 1C and 2C).

Table 1. Presence of hormone-, ROS-, HL-, and HS-response transcripts in the different groups of transcripts significantly up-regulatedin local and systemic leaves of plants subjected to HL and/or HS simultaneously applied to the same or two different leaves of thesame plant

ABA, abscisic acid; ACC, 1‐aminocyclopropane‐1‐carboxylic acid; BL, brassinolide; CK, cytokinin; GA, gibberellin; IAA, auxin. All transcripts included weresignificantly different from controls at P < 0.05 (negative binomial Wald test followed by a Benjamini–Hochberg correction). Shading intensity is proportionalto percent representation level of each hormone-, ROS-, or treatment-specific transcripts in the different tissues.

13812 | www.pnas.org/cgi/doi/10.1073/pnas.2005077117 Zandalinas et al.

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

11, 2

020

In addition, compared with the systemic response of plants totwo different stresses simultaneously applied to the same leaf(HL+HS), the representation of many hormone-, HL-, HS-, andROS-response transcripts was higher in the systemic response ofplants to two different stresses simultaneously applied to twodifferent leaves (HL&HS) (Table 1). Analysis of different groupsof transcription factors (TFs) involved in the integration of dif-ferent abiotic stress-specific signals during stress combination(24) further revealed that, in contrast to systemic leaves of plantssubjected to a local treatment of HL, HS, or HL and HS si-multaneously applied to two different leaves (HL&HS), many ofthese TFs were not expressed in systemic leaves of plants sub-jected to a local treatment of two different stresses simulta-neously applied to the same leaf (HL+HS) (Fig. 3B). Theseincluded, for example, Rap2.12, CBF1, ERF1, EREBF11,MYB16, and MYB77. Therefore, the systemic response of plantsto two different stresses simultaneously applied to two differentleaves is different and more extensive than the response to two

different stresses simultaneously applied to the same leaf (Figs.1–3 and Table 1).

Leaf-To-Leaf Communication in Plants Subjected to Stress CombinationApplied to Two Different Leaves. The differences observed betweenthe response of plants to two different stresses applied to the sameor two different leaves (Figs. 1–3 and Table 1) could result fromthe two different local leaves subjected to the two differentstresses exchanging signals with each other (3, 15, 23). Indeed,heat-specific transcripts could be found in local leaves subjected tolight stress- and light-specific transcripts could be found in localleaves subjected to heat stress in plants simultaneously subjectedto light and heat stress on two different leaves (HL&HS; Fig. 4A).In addition, as shown in Fig. 4B, comparing the expression patternof TFs involved in the integration of different abiotic stress-specific signals during stress combination (24) between localleaves subjected to HL, HS, HL(HL&HS), and HS(HL&HS)reveals that several HL-response TFs are expressed in local

HL

HS

HL(HL&HS)

675

1362

322

6931784

999

730(4206) (4198)

(3780)

HS

HL

HS

HL&HS2466

411

669 540872

309 291(4316)

(3477)

(4169)

A

B

HL

HL

HS

HS(HL&HS)

534

1984

463

6291874

909

794(4206) (5115)

(3780)

Systemic HL&HS

Local HL(HL&HS)

47

493

(540)

1315

(1362)

Systemic HL&HS

Local HS(HL&HS)

69

471

(540)

1915

(1984)

HLHS

25

30

35

40

% o

verla

p w

ith lo

cal

Systemic HL&HS

C

467

176

Fig. 2. Plants can integrate two different systemic signals generated simultaneously at two different leaves of the same plant during stress combination. (A)Overlap between the systemic transcriptomic response of plants to a local application of HL, HS, or a combination of heat and light stress simultaneouslyapplied to two different leaves (HL&HS). Venn diagrams for the overlap between the different responses are shown at the bottom for local leaves and at thetop for systemic leaves. Black arrows in HL&HS plants represent the number of HL-specific (solid), or HS-specific (dashed) transcripts common between localand systemic leaves. (B) Venn diagrams showing the overlap between stress combination-specific transcripts from local (HL or HS) and systemic leaves. (C) Bargraph showing the percent overlap between the systemic HL&HS response and local HL, HS, HL(HL&HS), or HS(HL&HS) responses. All experiments wererepeated at least three times with 40 plants per biological repeat. All transcripts included in the figure were significantly different from controls at P < 0.05(negative binomial Wald test followed by Benjamini–Hochberg correction). HL(HL&HS) and HS(HL&HS) denote a local HL- or HS- treated leaf of a plantsubjected to HS or HL on another local leaf, respectively.

Zandalinas et al. PNAS | June 16, 2020 | vol. 117 | no. 24 | 13813

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

11, 2

020

HS(HL&HS) leaves (e.g., SNZ, MYB4), and several-HS responseTFs are expressed in local HL(HL&HS) leaves (e.g., At1g21910,At4g28140, MYB29). These findings, as well as the overlap inoverall heat and light stress responses of the two different local

leaves (Fig. 2) and the similarity in hormone- and ROS-responsetranscript representation between the two different local leaves[HL(HL&HS) and HS(HL&HS); Table 1], demonstrate that inaddition to generating two different systemic signals that integrate

Systemic

HL+HS HL&HS

2477487 1692

(2964) (4169)

HS(HL&HS)

HL(HL&HS)

HL+HS

353

819

881 3122643

192531

(4838) (5115)

(4198)

HL+HS HLHS

467

176367153

HL HS HL+HS HL&HS1.176 1.186 n.s. n.s. MYB30 AT3G28910n.s. 1.058 n.s. n.s. MYB31 AT1G74650n.s. 1.045 n.s. n.s. MYB34 AT5G60890

1.025 n.s. n.s. n.s. MYB3R1 AT4G32730n.s. n.s. n.s. 1.755 MYB51 AT1G18570n.s. n.s. n.s. 2.186 MYB95 AT1G74430

1.562 1.658 n.s. 3.382 MYB77 AT3G500601.120 1.168 n.s. 1.261 MYB16 AT5G153101.531 1.310 1.043 1.447 MYB3R4 AT5G115101.415 n.s. n.s. 1.453 MYB96 AT5G62470n.s. n.s. 1.139 n.s. MYB70 AT2G23290n.s. n.s. 1.631 n.s. MYB10 AT3G12820

1.138 1.202 1.394 1.226 MYB6 AT4G094601.162 1.105 1.234 1.147 MYB29 AT5G076901.357 1.271 1.664 1.774 MYB44 AT5G67300

Systemic

n.s. n.s. n.s. 8.034 HSFA4A AT2G26150n.s. n.s. n.s. 1.321 HSFA7A AT4G18880n.s. n.s. n.s. 1.778 HSFA7B AT5G62020n.s. n.s. n.s. 3.186 HSFA8 AT3G51910n.s. n.s. n.s. 2.079 HSFB2A AT4G11660

HL HS HL+HS HL&HS1.219 1.132 n.s. 1.093 AP2-like AT1G254701.280 1.100 n.s. n.s. TINY-like AT1G012504.623 n.s. n.s. n.s. WIN1 AT1G153601.036 n.s. n.s. n.s. AP2-lik AT3G258901.391 1.441 n.s. n.s. WRI1 AT3G543201.585 1.485 1.457 1.323 TINY-like AT4G328001.359 n.s. 1.534 1.777 DRE-binding AT1G21910n.s. 0.173 n.s. n.s. Rap2.6 AT1G43160

1.226 n.s. n.s. 1.296 AP2-like AT4G237502.065 1.665 n.s. 2.079 Rap2.12 AT4G34410

11.150 18.810 n.s. 32.653 CBF1 AT1G126102.176 2.560 n.s. 3.897 ERF1 AT4G175002.921 5.572 n.s. 6.188 EREBF11 AT1G283703.461 13.087 n.s. 14.136 AP2-like AT1G74930n.s. 4.296 n.s. 3.630 AP2-like AT1G19210n.s. 1.151 n.s. 1.112 EREBF3 AT1G50640n.s. 2.175 n.s. 1.896 ERF2 AT5G47220

1.120 1.217 1.394 2.079 DREB1A AT1G630302.937 2.543 3.821 8.605 ERF6 AT4G17490n.s. 1.198 n.s. 1.621 Rap2.4 AT1G22190n.s. 1.658 n.s. 3.382 TINY-like AT1G33760

10.689 19.467 n.s. 61.592 CBF2 AT4G254701.225 1.392 n.s. 2.098 ERF4 AT3G152104.297 7.559 n.s. 17.462 ERF13 AT2G448401.710 2.560 n.s. 4.939 ERF5 AT5G47230n.s. n.s. n.s. 1.233 Putative AT1G78080n.s. n.s. n.s. 9.962 CBF1 AT4G25490n.s. n.s. n.s. 3.013 DREB2A AT5G05410n.s. n.s. n.s. 5.584 Rap2.10 AT5G52020

1.613 1.767 n.s. 6.267 EREBF8 AT5G61600n.s. 1.040 1.052 n.s. Putative AT1G53910

2.224 3.005 3.969 2.753 CBF3 AT4G25480n.s. 1.072 1.244 n.s. AP2-like AT4G36920

1.351 1.309 1.622 1.340 SHINE2 AT5G253902.179 2.143 2.486 1.847 EREBF-like AT5G615900.794 n.s. n.s. n.s. AP2-like AT1G448300.934 n.s. n.s. 1.014 AP2-liken.s. 1.406 n.s. n.s. SNZ AT2G39250n.s. 1.401 n.s. n.s. AP2-like AT4G31060n.s. 1.022 n.s. n.s. AP2-like AT5G64750 0 15<

A B

Fig. 3. The systemic response of plants to two different stresses simultaneously applied to two different local leaves is more extensive than the response totwo different stresses simultaneously applied to the same local leaf. (A) Overlap between the systemic transcriptomic responses of plants to a combination ofheat and light stress applied to the same (HL+HS) or different (HL&HS) local leaves. Venn diagrams for the overlap between the different responses are shownat the bottom for local leaves and at the top for systemic leaves. Black arrows represent the number of HL-specific (solid) or HS-specific (dashed) transcriptscommon between local and systemic leaves. (B) Heat maps showing the expression pattern of TFs belonging to the ethylene response (AP2-EREBPs), MYB, andheat shock factor (HSF) families (24) in systemic tissues of plants subjected to a local treatment of HL, HS, HL+HS, or HL&HS. All transcriptomics experimentswere repeated at least three times with 40 plants per biological repeat (P < 0.05, negative binomial Wald test followed by Benjamini–Hochberg correction).

HL HS(HL&HS)

26681538 2447(4206) (5115)

HL

HS HL(HL&HS)

21061674 2092(3780) (4198)

794

463HL HS

HS

HL HS HL+HS HL(HL&HS)

HS(HL&HS)

0.783 n.s. n.s. n.s. 0.967 MYB20 AT1G66230n.s. 1.243 1.282 1.229 1.231 MYB29 AT5G07690

4.261 9.494 4.879 7.507 7.856 MYB15 AT3G232501.534 2.488 1.821 2.257 2.508 MYB51 AT1G18570

12.645 18.492 15.852 18.042 23.167 MYB77 AT3G500601.368 1.917 1.579 1.501 2.025 MYB96 AT5G624700.955 1.146 0.990 0.933 1.117 MYB16 AT5G153101.316 1.827 n.s. n.s. 1.791 MYB30 AT3G28910n.s. 1.360 0.853 0.865 1.230 MYB3 AT1G22640

1.294 1.616 n.s. n.s. 1.289 MYB6 AT4G094602.743 5.696 3.332 2.381 3.809 MYB73 AT4G37260n.s. 1.695 n.s. n.s. n.s. MYB70 AT2G23290n.s. 1.040 n.s. n.s. n.s. MYB75 AT1G56650n.s. 1.517 1.271 n.s. n.s. MYB23 AT5G40330n.s. n.s. 0.853 n.s. n.s. MYB124 AT1G14350n.s. n.s. 0.954 n.s. n.s. MYB3R4 AT5G11510

1.353 n.s. n.s. 1.175 1.468 MYB4 AT4G38620n.s. n.s. n.s. 1.353 1.660 MYB43 AT5G16600n.s. n.s. n.s. 1.093 1.140 MYB76 AT5G07700n.s. n.s. 0.999 1.190 1.092 MYB34 AT5G60890n.s. n.s. n.s. 6.257 n.s. MYB83 AT3G08500

10.180 n.s. n.s. 64.878 n.s. MYB103 AT1G63910n.s. 0.802 n.s. n.s. n.s. MYB59 AT5G59780

2.741 1.646 4.162 3.385 2.824 MYB7 AT2G16720n.s. n.s. 1.338 n.s. n.s. MYB32 AT4G34990

3.485 3.527 4.207 3.546 3.757 MYB44 AT5G67300n.s. n.s. n.s. n.s. 0.914 MYB47 AT1G18710

1.561 1.473 1.800 n.s. n.s. MYB95 AT1G744307.798 2.910 n.s. n.s. n.s. MYB61 AT1G095402.285 n.s. n.s. n.s. n.s. MYB86 AT5G266607.163 n.s. n.s. n.s. n.s. MYB31 AT1G74650

1.806 1.442 n.s. 1.341 n.s. HSFA2 AT1G679702.009 3.148 1.865 1.760 1.985 HSFA7A AT4G18880n.s. 1.218 n.s. n.s. n.s. HSFA4C AT5G45710

40.535 16.906 n.s. n.s. 86.341 HSFA4A AT2G261502.919 3.037 5.571 4.652 5.633 HSFA7B AT5G620201.487 n.s. 3.332 3.385 2.244 HSFB1 AT3G633502.741 1.819 6.768 4.409 4.144 HSFB2B AT4G36990

14.808 3.881 68.003 n.s. 41.166 HSFA8 AT3G519104.132 2.853 8.599 n.s. 7.041 HSFB2A AT4G11660n.s. n.s. 2.006 n.s. 1.995 HSFA3 AT5G03720

Local

HL HS HL+HS HL(HL&HS)

HS(HL&HS)

n.s. 2.523 2.608 n.s. n.s. TINY-like AT2G258205.433 8.263 9.139 7.018 6.291 CBF3 AT4G25480

21.055 56.156 117.870 n.s. 21.570 CBF1 AT1G12610n.s. n.s. 1.428 n.s. n.s. DREB1A AT1G63030n.s. n.s. 7.198 n.s. n.s. DREB2B AT3G11020n.s. n.s. 1.147 n.s. n.s. AP2-like AT4G31060n.s. 15.165 6.922 7.457 n.s. AP2-like AT1G77640

9.631 26.107 13.659 11.624 16.905 EREBF11 AT1G2837015.286 66.224 36.279 25.277 50.757 TINY-like AT1G33760

n.s. 1.639 1.385 n.s. 1.193 TINY-like AT4G32800n.s. 1.118 0.983 n.s. n.s. Putative AT1G53910n.s. 1.068 n.s. n.s. n.s. AP2-like AT2G20880

1.349 1.685 1.285 1.327 1.542 EREBF7 AT3G203104.678 8.222 3.979 2.357 6.061 ERF2 AT5G472201.058 1.783 0.725 0.473 1.056 EREBF-like AT5G61590n.s. n.s. 1.564 n.s. 1.652 ANT AT4G37750

2.829 2.839 4.054 3.699 4.147 Putative AT1G7808013.979 9.209 51.771 34.538 47.315 DREB2A AT5G054102.562 3.483 2.861 2.404 3.917 ERF1 AT3G232405.027 9.971 5.344 4.201 10.816 ERF1 AT4G17500n.s. 6.467 6.018 n.s. 16.238 EREBF-like AT3G23230

12.450 32.495 39.133 14.810 51.473 Rap2.12 AT4G344103.248 4.271 4.209 4.545 5.575 DRE-bind AT1G221903.162 4.877 4.117 4.545 5.688 ERF4 AT3G15210

10.413 21.199 16.062 17.441 22.633 CBF2 AT4G254705.152 13.169 8.208 9.572 21.264 AP2-like AT1G192101.583 1.944 1.575 1.683 2.477 EREBF8 AT1G53170

26.219 84.043 37.664 40.540 128.180 AP2-like AT1G7493026.947 71.724 28.812 58.944 76.791 CBF1 AT4G25490

n.s. 2.563 n.s. 1.730 2.856 AP2-like AT1G2191016.587 40.489 15.928 29.750 45.759 ERF5 AT5G472301.176 1.348 1.159 1.343 1.503 EREBF3 AT1G50640

15.722 28.236 15.500 22.453 37.621 ERF13 AT2G4484021.668 60.141 26.869 42.218 74.990 ERF6 AT4G1749012.171 21.875 13.934 19.163 26.887 EREBF8 AT5G616003.191 2.592 3.345 3.621 3.437 AIL5 AT5G57390n.s. 2.734 2.909 5.398 5.290 Putative AT4G28140n.s. n.s. n.s. 1.118 n.s. AP2-like AT2G46310n.s. n.s. n.s. 1.377 n.s. AP2-like AT4G23750n.s. n.s. n.s. 8.708 n.s. Putative AT5G61890

3.014 2.794 1.983 2.510 2.546 ERF-like AT5G251905.460 4.633 3.399 2.682 2.879 Rap2.1 AT3G502601.412 1.321 n.s. n.s. n.s. TINY-like AT1G643801.251 1.234 n.s. n.s. n.s. TF-like AT5G075802.771 1.616 n.s. n.s. n.s. Rap2.6 AT1G431601.137 n.s. n.s. n.s. n.s. AP2-like AT1G448302.092 n.s. n.s. n.s. n.s. DREB-like AT2G383406.661 n.s. n.s. n.s. n.s. SHINE2 AT5G253901.678 1.772 1.749 n.s. 1.888 RAP21.273 n.s. 1.044 1.002 1.234 SNZ4.353 n.s. 4.063 n.s. n.s. AP2-like AT5G64750

Local

0 15<

A B

Fig. 4. Overlap between the local transcriptomic responses of two different leaves of the same plant, one subjected to HL and the other to HS simultaneously. (A)Venn diagrams for the overlap between the different local responses of plants subjected to HL, HS, or HL&HS. Yellow arrows represent the number of HL-specific(solid) or HS-specific (dashed) transcripts common between the two different local leaves. (B) Heat maps showing the expression pattern of TFs belonging to theAP2-EREBP, MYB, and HSF families (24) in local tissues of plants subjected to HL, HS, HL+HS, or HL&HS. All transcriptomics experiments were repeated at leastthree times with 40 plants per biological repeat (P < 0.05, negative binomial Wald test followed by Benjamini–Hochberg correction).

13814 | www.pnas.org/cgi/doi/10.1073/pnas.2005077117 Zandalinas et al.

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

11, 2

020

and travel to remote systemic tissues (Fig. 3), the two local leavessubjected to two different stresses can communicate with eachother using stress-specific systemic signals (Fig. 4), affecting geneexpression patterns in each other.

SAA to Light or Heat Stress in Plants Subjected to Abiotic StressCombination Applied to the Same or Two Different Leaves. En-hanced expression of stress-response transcripts in systemic leavescould lead to enhanced plant acclimation (i.e., SAA) (5–7, 15).The differences in transcript expression observed between sys-temic tissues of plants integrating two different systemic signalsoriginating from the same or different leaves (Figs. 1–4 and Ta-ble 1) prompted us to compare their SAA to light or heat stress.For this purpose, we subjected local leaves to a short pretreatmentof HL, HS, HL+HS, or HL&HS; allowed plants to acclimate; andchallenged the systemic leaves with damaging levels of HL or HS(5, 6, 15). Strikingly, although acclimation of systemic tissues tolight (evident by a decrease in ion leakage upon acclimation) orheat (evidenced by the maintenance of high chlorophyll contentupon acclimation) was observed when the different stresses wereapplied to local leaves individually (HL or HS) or simultaneouslyto two different leaves (HL&HS), when the two different stresseswere simultaneously applied to the same leaf (HL+HS), no SAAto light or heat stress was observed (Fig. 5A). This finding dem-onstrates that the outcome of integrating two different systemicsignals generated at the same or different leaves is different notonly in transcript expression (Figs. 1–4 and Table 1), but also inplant acclimation (Fig. 5A).

Regulation of Stomatal Aperture during Stress Combination. Be-cause transcript expression and plant acclimation are associatedwith stomatal responses that could also be coordinated between

different leaves and required for plant acclimation (3, 15, 23, 25),and because stomatal responses to light or heat stress are op-posing (opening during heat stress and closing during intenselight stress) (20), we measured changes in stomatal aperture inlocal and systemic leaves of plants subjected to the two differentstress combinations: HL&HS and HL+HS. As expected, localleaves subjected to light stress displayed stomatal closure, whilelocal leaves subjected to heat stress displayed stomatal opening(Fig. 5B) (15, 22, 23). Interestingly, stomata of local leavessubjected to light stress while the other leaf was subjected to heat[HL(HL&HS)] closed, while stomata of local leaves subjectedto heat stress while the other leaf was subjected to light stress[HS(HL&HS)] closed and then opened (Fig. 5B). A similar closingand opening response was observed in systemic leaves of plantssubjected to the two different stresses applied to two differentleaves (systemic HL&HS). The closing and then opening responseof stomata in systemic (HL&HL) or local leaves of plants subjectedto a combination of HL&HS [HS(HL&HS)] suggest that the sys-temic response to HL (closing) could be faster than the systemicresponse to HS (opening), but the systemic response to HS over-comes the systemic response to HL (Fig. 5B). In contrast, stomatalresponses were dampened, with no significant opening or closingresponses in local and systemic leaves of plants subjected to thetwo different stresses applied simultaneously to the same leaf (localor systemic leaves of HL+HS) (Fig. 5B).

Activation of the ROS Wave during Stress Combination. One of thecentral systemic signaling pathways required for SAA to occur inplants is the ROS wave, an autopropagating cell-to-cell processof ROS production that accompanies the systemic signal (3, 5–8,13–15). Comparing the pattern of ROS wave-associated tran-scripts (6) between local and systemic tissues of plants subjected

Pretreated

0

1

2

3

4

Chl

(μg

mg-

1FW

)

*

* *

Chl AChl B

0

100

200

300

400

EL (%

)

* *

*

HL

HLAcclimation

HS

HL+HS

HL&HSCT

HL HS

HL+HS

HL&HSCT

HSAcclimation

Local Treatments

Local Treatments

Pretreated

A B HSHLLocal SystemicLocal Systemic

a

bc

d

abbc

c

0

1

2

3

4

5

0 2 8 0 2 8Time (min)

bc aba

cbc

a

0

1

2

3

4

5

0 2 8 0 2 8Time (min)

5

4

3

2

1

0

5

4

3

2

1

0

Stom

atal

ape

rture

(μm

)

HL&HSHL+HSLocal Systemic Local (HS) SystemicLocal (HL)

a a a a a a

0

1

2

3

4

5

0 2 8 0 2 8Time (min)

ab

d

e

a

cd

abbc

d cd

0

1

2

3

4

5

0 2 8 0 2 8 0 2 8Time (min)

5

4

3

2

1

0

5

4

3

2

1

0

Stom

atal

ape

rture

(μm

)

Fig. 5. Acclimation of systemic leaves and stomatal aperture responses to stress combination. (A, Top) Schematic representation of how experiments wereconducted. Local leaves were subjected to a short (15 min) pretreatment of HL, HS, HL+HS, or HL&HS, and plants were allowed to acclimate for 45 min.Following acclimation, systemic leaves were challenged with damaging levels of HL or HS, sampled, photographed and subjected to tissue injury assays (ionleakage for HL and chlorophyll content for HS). (A, Middle) Representative systemic leaf images. (A, Bottom) Measurements of systemic leaf injury: increase inion leakage for HL (Left) and decrease in chlorophyll content for HS (Right). All acclimation experiments were repeated at least three times with 10 plants perbiological repeat. Data are mean ± SD. *P < 0.05, two-way ANOVA followed by Tukey’s post hoc test. CT, control; HL or HS, control plants subjected to asystemic HL or HS stress treatment without pretreatment, respectively; Pretreated, plants in which a local leaf was subjected to HL, HS, HL+HS, or HL&HStreatment before the systemic HL or HS treatment; EL, electrolyte leakage. (B) Stomatal aperture in local and systemic leaves of plants subjected to a local HL,HS, HL+HS, or HL&HS treatment. All experiments were repeated at least three times with 500 stomata per plant and 10 plants per biological repeat. Data arepresented as mean ± SD. Different letters denote significance at P < 0.05 (ANOVA followed by a Tukey’s post hoc test).

Zandalinas et al. PNAS | June 16, 2020 | vol. 117 | no. 24 | 13815

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

11, 2

020

to two different stresses applied to the same or two differentleaves revealed that although many ROS wave-associated tran-scripts were up-regulated in the local tissues of plants subjectedto two different stresses applied to the same or different leaves(i.e., local HL of HL&HS, local HS of HL&HS, and localHL+HS), as well as in systemic tissues of plants subjected to twodifferent stresses applied simultaneously to two different leaves(i.e., systemic HL&HS), few ROS wave-associated transcriptswere up-regulated in systemic tissues of plants simultaneouslysubjected to light and heat stress applied to the same leaf (systemicHL+HS) (Fig. 6A and SI Appendix, Fig. S2). This observation wassupported by the abundance of H2O2- and 1O2-response tran-scripts in systemic tissues of HL&HS plants compared with sys-temic tissues of HL+HS plants (Table 1), as well as by in vivomeasurements of ROS accumulation (8) in systemic tissues ofsimilar size and age plants containing an inflorescence stem sub-jected to heat, light, and heat and light combinations applied tothe same or different leaves (Fig. 6B). Although heat and/or lightstresses applied individually or in combination to the same or totwo different local leaves of plants (at the two different develop-mental stages) triggered the ROS wave, compared with all othertreatments, the rate of ROS wave propagation was significantlyfaster when the two different stresses were simultaneously appliedto two different leaves (HL&HS) (Fig. 6B, Table 1, and SI Ap-pendix, Fig. S2).

Accumulation of JA and SA in Local and Systemic Leaves during StressCombination and Altered Systemic ROS Signaling during StressCombination in the Allene Oxide Synthase (aos) Mutant. BecauseSA-regulated transcripts accumulated in local leaves subjected toHL or HS, but not in local leaves of HL+HS plants (Table 1),and because JA was found to play a key role in the protection ofplants subjected to a combination of HL and HS (22), wequantified the levels of SA and JA in local and systemic leaves ofplants subjected to HL, HS, HL+HS, and HL&HS (Fig. 7A). Inagreement with the hormone-response transcript expression re-sults shown in Table 1, JA and SA accumulated in local leaves ofplants subjected to HL or HS. The accumulation of JA wasnevertheless delayed, while the accumulation of SA was com-pletely suppressed in local leaves of HL+HS plants, which dis-play a suppressed ROS wave initiation phenotype (Fig. 6),suggesting that rapid accumulation of these two hormones isneeded to trigger the ROS wave in local leaves and induce SAA(Figs. 5A and 6). In contrast to plants subjected to HL+HS, localleaves of plants subjected to HL&HS accumulated JA at normalrates (Fig. 7A), and this accumulation could be involved in ini-tiating the ROS wave in these plants (Fig. 6).In contrast to JA, the accumulation of SA was completely

blocked in local or systemic leaves in both HL+HS and HL&HS(Fig. 7A), suggesting that SA accumulation is suppressed in sys-temic leaves during stress combination. Interestingly, the sup-pression of local ROS wave initiation during HL+HS did not

0

20

40

60

80

100

120

140

CT HL HS HL+HS HL&HS

LOCAL SYSTEMICHL

(HL&HS)HS

(HL&HS) HL+HS HL+HS HL&HS2 8 2 8 2 8 2 8 2 8

AT1G29670AT4G23870AT2G23290AT3G02840AT2G01300AT4G06880AT2G32210AT4G31210AT1G49750AT5G53160AT4G05330AT1G70290AT4G19530AT5G56750AT1G07280AT4G23300AT5G38510AT5G59750AT4G26260AT1G77640AT4G25810AT5G35735AT1G76650AT3G57930AT4G37260AT3G10190AT3G54810AT4G22690AT1G25400AT1G57990AT1G33610AT1G61470AT4G23810AT5G43620AT3G07790AT5G66070AT4G23180AT1G56520AT1G19180AT4G24160AT3G20600AT5G44350AT1G05575AT4G36500AT5G36925AT2G26530AT3G59080AT1G32928AT5G59550AT3G10930AT3G28340AT2G01180AT5G58430AT5G64660AT1G69900AT2G09225AT2G32030AT5G66650AT3G23450AT1G76680AT3G28910AT1G21910AT2G32200AT2G25735AT5G28630AT1G69890AT4G18197AT2G39360AT4G18205

0.5 10<

A B0 min 15 min 25 min

CT

HL

HS

HL+

HS

HL&

HS

0.0E+00

5.0E+07

1.0E+08

1.5E+08

2.0E+08

2.5E+08

0 10 20 300.0E+00

5.0E+07

1.0E+08

1.5E+08

2.0E+08

2.5E+08

0 10 20 30

Tota

l Rad

iant

Effi

cien

cy (

[p s

-1]

[μW

-1] [

cm²])

Time (min)

CT

HL

HS

HL+HS

HL&HS

Time (min)

0

10

20

Acc

umul

atio

n of

tota

l Rad

iant

Effi

cien

cy (%

)

n.s. n.s. n.s. n.s.

**0 min

15 min

SYSTEMIC

Fig. 6. Systemic accumulation of ROS in mature bolting plants simultaneously subjected to two different stresses. (A) Heat maps showing the expression oftranscripts associated with systemic ROS accumulation (6) in local leaves of plants subjected to light or heat stress while the other leaf is simultaneouslysubjected to the other stress [HL(HL&HS) or HS(HL&HS), respectively] or HL and HS applied to the same leaf (HL+HS), and in systemic leaves of plants subjectedto a local treatment of HL and HS applied to the same (HL+HS) or two different (HL&HS) leaves (SI Appendix, Fig. S2). Transcripts included in A and in SIAppendix, Fig. S2 are significantly different from controls (P < 0.05; negative binomial Wald test followed by Benjamini–Hochberg correction). (B) Repre-sentative images (Left), line graphs showing continuous in vivo ROS measurements (Top Right), and bar graphs showing measurements of ROS at 15 min afterstress application (Bottom Right), at the middle of the inflorescence stem (corresponding to white arrows in images on the left). All experiments were re-peated at least three times with five plants per biological repeat. Data are presented as mean ± SD, **P < 0.01, two-way ANOVA followed by Tukey’s post hoctest. CT, control; n.s., no significant differences with respect to control. (Scale bars, 1 cm.)

13816 | www.pnas.org/cgi/doi/10.1073/pnas.2005077117 Zandalinas et al.

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

11, 2

020

occur in the aos mutant, which is suppressed in JA accumulation(22), supporting a role for JA in suppressing the initiation of theROS wave in local leaves of HL+HS plants (Fig. 6). The findingspresented in Fig. 7 point to important roles for JA and SA insystemic signal integration during stress combination in plants.

Systemic ROS Accumulation and Acclimation of Nonbolting Plantsduring Stress Combination. Because the inability of local leavesof HL+HS plants to rapidly induce the ROS wave (Fig. 6) couldalso result from their metabolic and hormonal state at the repro-ductive stage, we tested whether a similar integration of systemicROS signals (Fig. 6) and plant acclimation (Fig. 5A) occurs inyoung vegetative-stage plants. As shown in Fig. 8A, the systemicROS wave response of young, nonbolting plants to two differentstresses simultaneously applied to the same leaf (HL+HS) wasstronger than that of reproductive-stage bolting plants (HL+HS)(Fig. 6), compared in each developmental stage with the two dif-ferent stresses individually applied to a single leaf (HL or HS). Thisfinding supports the possibility that differences in metabolic state,physiological activity, and/or hormone levels are affecting the in-tegration of systemic signals and the triggering of the ROS wave bylocal leaves subjected to a combination of HL and HS (HL+HS).Nevertheless, even in young vegetative-stage plants, the sys-temic ROS wave response to two different stresses simultaneously

applied to two different leaves (HL&HS) was stronger than thatto two different stresses simultaneously applied to the same leaf(HL+HS) (Figs. 6 and 8A). Similarly, plant acclimation to stresscombination also primarily occurred when the two differentstresses were applied to two different leaves of young vegetative-stage or bolting plants (HL&HS) (Figs. 5A and 8B).

DiscussionTo be effective in inducing plant acclimation (i.e., SAA; Figs. 5Aand 8B), the systemic signal(s) generated at the local leaf(leaves) must trigger several different responses in systemic tis-sues. These include changes in transcript abundance (e.g., pho-tosynthetic- and other light- or heat-specific transcripts; Fig. 9,Table 1, and SI Appendix, Fig. S3) (8–10), metabolite levels (5,18), ROS accumulation (Figs. 6–8) (5, 8), and stomatal responses(Fig. 5B) (15, 22, 23). Moreover, changes in these parametersmust be rapid and coordinated between the different parts of theplant (3, 15, 18, 23). Surprisingly, while different systemic signalsgenerated individually (HL or HS) or simultaneously at two dif-ferent leaves (HL&HS) induced acclimation to light or heat stress,two different stresses applied simultaneously to the same leaf(HL+HS) did not (Figs. 5A and 8B). This finding was supportedby the lower representation of HL-, HS-, and many ROS- andhormone-response transcripts in systemic leaves of plants subjected

0

20

40

60

80

100

120

140

160

CT HL HS HL+HS HL&HS

0 min15 min

Acc

umul

atio

n of

tota

l Rad

iant

Effi

cien

cy (%

)

n.s. n.s.n.s.

* *C

TH

LH

SH

L+H

SH

L&H

S

Tota

l Rad

iant

Effi

cien

cy (

[p s

-1]

[μW

-1] [

cm²])

SYSTEMIC

Time (min) Time (min)

0E+00

1E-06

2E-06

3E-06

4E-06

5E-06

6E-06

7E-06

8E-06

0 10 20 30

CTHLHSHL+HSHL&HS

0E+00

1E-06

2E-06

3E-06

4E-06

5E-06

6E-06

7E-06

8E-06

0 10 20 30

0

10

20

B

0

50

100

150

200

250

300

350

0 2 4 6 8

SA (n

g g-1

FW)

Time (min)

* *

0

10

20

30

40

0 2 4 6 8

JA (n

g g-1

FW)

Time (min)

*

0

10

20

30

40

0 2 4 6 8

JA (n

g g-1

FW)

Time (min)

*

0

10

20

30

40

0 2 4 6 8

JA (n

g g-1

FW)

Time (min)

*

0

10

20

30

40

0 2 4 6 8

JA (n

g g-1

FW)

Time (min)

*

*

HS

HL

0

50

100

150

200

0 2 4 6 8

JA (n

g g-1

FW)

Time (min)

* *

0

50

100

150

200

0 2 4 6 8

JA (n

g g-1

FW)

Time (min)

0

50

100

150

200

0 2 4 6 8

JA (n

g g-1

FW)

Time (min)

*

0

50

100

150

200

0 2 4 6 8

JA (n

g g-1

FW)

Time (min)

*

0

50

100

150

200

250

300

350

0 2 4 6 8

SA (n

g g-1

FW)

Time (min)

**

0

50

100

150

200

250

300

350

0 2 4 6 8

SA (n

g g-1

FW)

Time (min)

*

*

0

50

100

150

200

250

300

350

0 2 4 6 8

SA (n

g g-1

FW)

Time (min)

*

*

0

50

100

150

200

250

300

350

0 2 4 6 8

SA (n

g g-1

FW)

Time (min)

*

*

0

50

100

150

200

250

300

350

0 2 4 6 8

SA (n

g g-1

FW)

Time (min)

0

50

100

150

200

250

300

350

0 2 4 6 8

SA (n

g g-1

FW)

Time (min)

A SYSTEMICLOCAL SYSTEMICLOCAL

0

50

100

150

200

250

300

350

0 2 4 6 8

SA (n

g g-1

FW)

Time (min)

*

*

*

*HL

HS

HLHS

HL+HSHL&HS

0 min 10 min 20 min 30 min

Fig. 7. Quantitative analysis of JA and SA levels in wild-type plants and systemic accumulation of ROS in bolting aos mutants during stress combination. (A)Quantitative analysis of JA and SA in local and systemic wild-type leaves at 0, 2, and 8 min following the application of HL, HS, HL+HS, and HL&HS to localleaves. Hormone analysis was conducted using the same biological material used from transcriptomics analysis. Data are presented as mean ± SD. *P < 0.05,two-way ANOVA followed by Tukey’s post hoc test. (B) Representative images (Left), line graphs showing continuous in vivo ROS measurements (Top Right),and bar graphs showing measurements of ROS at 15 min after stress application (Bottom Right) at the middle of the inflorescence stem (corresponding towhite arrows in images on Left) of aosmutant plants subjected to stress combinations. All experiments were repeated at least three times with five plants perbiological repeat. Data are presented as mean ± SD. *P < 0.05, two-way ANOVA followed by Tukey’s post hoc test. CT, control; n.s., no significant differenceswith respect to control. (Scale bars, 1 cm.)

Zandalinas et al. PNAS | June 16, 2020 | vol. 117 | no. 24 | 13817

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

11, 2

020

to HL and HS simultaneously applied to the same leaf (HL+HS;Fig. 3B and Table 1); the lower representation of transcriptsunique to the stress combination in the systemic leaves of plantssubjected to HL and HS simultaneously applied to the same leaf(HL+HS; Figs. 1–3); and the suppressed stomatal and hormonalresponses of local and systemic leaves of plants subjected to HLand HS simultaneously applied to the same leaf (HL+HS; Figs. 5Band 7), all compared with that of systemic leaves of plants sub-jected to the two different stresses simultaneously applied to twoindividual leaves (HL&HS; Figs. 1–3 and Table 1) (5–8).One plausible mechanistic explanation for this observation is

that the two different stresses applied to the same leaf (HL+HS)suppressed the initiation or reduced the intensity of the ROSwave (Figs. 6–9, Table 1, and SI Appendix, Fig. S2) that is ab-solutely required for SAA to these stresses (5, 6, 15). Anotherplausible explanation could be the opposing effects of the stresscombination applied to the same leaf (HL+HS) on stomatalresponses (Fig. 5B), which could also affect overall plant accli-mation (10, 15, 23, 25). It is also possible that an interplay be-tween JA and SA levels (Table 1 and Fig. 7A) is responsible (18).In support of this explanation is the finding that the suppressionof ROS wave initiation during HL+HS is removed in a mutantdeficient in JA accumulation (aos; Fig. 7B). Therefore, the rateat which JA accumulates at the local leaves of HL+HS plants(compared to HL&HS) and its potential effects on SA accu-mulation (Fig. 7A; also reflected in transcript accumulation inTable 1) could play an important role in how fast and efficientthe single leaf subjected to HL+HS can initiate systemic sig-naling and SAA. Thus, when JA accumulation is suppressed inthe aos mutant, this inhibition is removed, and systemic ROSresponses can occur.Further studies are needed to address the many intriguing

questions related to this new and emerging field of systemic

signal integration during abiotic stress in plants. As revealed bythis work, plants can integrate different systemic signals, and theydo so best when the two different signals originate from twodifferent leaves (Fig. 9). Like more complex multicellular or-ganisms (e.g., animals), plants can therefore integrate differentsystemic signals and improve their chances of survival within thecomplex and rapidly changing environment that occurs in nature.

Materials and MethodsPlant Material and Stress Treatments. A. thaliana Col-0 (cv. Columbia-0) plantswere grown in peat pellets (Jiffy-7; https://www.jiffygroup.com/) at 23 °C underlong-day growth conditions (16-h light/8-h dark; 50 μmol m−2 s−1) for 40 to 55 duntil the inflorescence stem measured 11 to 13 cm. Local leaves were exposed toHL, HS, HL+HS, or HL&HS (SI Appendix, Fig. S1 A and B). HL was applied by sub-jecting a single leaf to a light intensity of 1,700 μmol m−2 s−1 for 2 and 8min usinga ColdVision fiber optic LED light source (Schott; A20980). HS was induced byplacing a heat block 2 cm underneath the treated leaf for 2 and 8 min, increasingthe leaf temperature to 31 to 33 °C (SI Appendix, Fig. S1C) as described previously(23). The temperature of the treated and systemic leaves was continuously mea-sured using an infrared camera (C2; FLIR Systems) (SI Appendix, Fig. S1C).

RNA Sequencing. Treated and untreated local and systemic tissues (SI Appendix,Fig. S1) were sampled at 0, 2, and 8min following stress application; immediatelyfrozen in liquid nitrogen; and used for RNA-seq analysis. Local and systemictissues from 40 different plants were pooled for each biological repeat per timepoint and stress treatment, and the experiment was repeated in three differentbiological replicates. All experiments were conducted at the same time of day (9AM to 12 PM), and all plants used for the experiments were of the same age anddevelopmental stage. Total RNA was isolated and subjected to RNA sequencingand analysis as described previously (6). RNA library construction and sequencingwere performed by the DNA Core Facility at the University of Missouri.

Acclimation Assays, Stomatal Aperture, and In Vivo Measurements of ROSAccumulation. SAA assays were performed as described previously (5, 6)with some modifications. One rosette leaf of each bolting plant was exposed

4

2

0

0 min 10 min 30 min

HL

HS

HL+

HS

HL&

HS

HL

S

HS

S

HL+HS

S

HL

S

HS

Treatment

0

20

40

60

80

100

120

140

160

HL HS HL+HS HL&HS

*0 min10 min

n.s.n.s. n.s.

Acc

umul

atio

n of

tota

l Rad

iant

Effi

cien

cy (%

)

HLHSHL+HSHL&HS

LOCAL SYSTEMIC

Tota

l Rad

iant

Effi

cien

cy (

[p s

-1]

[μW

-1] [

cm²])

0E+00

1E+08

2E+08

3E+08

4E+08

0 10 20 30

HL+HS

HL&HS

0E+00

1E+08

2E+08

3E+08

4E+08

0 10 20 30

HL

HS

0E+00

1E+08

2E+08

3E+08

4E+08

0 10 20 30

HL(HL&HS)

HS(HL&HS)

HL+HS

0E+00

1E+08

2E+08

3E+08

4E+08

0 10 20 30

HL

HS

Time (min) Time (min) Time (min) Time (min)

0

100

200

300

400

EL

(%)

0

1

2

3

Chl

(μg

mg-1

FW)

**

**

*

Pretreated

Chl AChl B

Pretreated

B

A

Fig. 8. Systemic ROS accumulation and plant acclimation of nonbolting wild-type plants during stress combination. (A) Representative images (Left), line graphsshowing continuous in vivo ROSmeasurements at different local and systemic leaves (Top Right; leaves monitored are indicatedwith letters on the left images), andbar graphs showing measurements of ROS at the indicated leaves at 10 min after stress application (Bottom Right). All experiments were repeated at least threetimes with five plants per biological repeat. Data are presented as mean ± SD. *P < 0.05, two-way ANOVA analysis followed by Tukey’s post hoc test. (B) Rep-resentative systemic leaf images (Top) and measurements of systemic leaf injury (increase in ion leakage for HL [Left] and decrease in chlorophyll content for HS[Right]) (Bottom) of nonbolting plants subjected to stress combination. All acclimation experiments were repeated at least three times with 10 plants per biologicalrepeat. Data are presented as mean ± SD. *P < 0.05, two-way ANOVA followed by Tukey’s post hoc test. CT, control; HL or HS, control plants subjected to a systemicHL or HS stress treatments without pretreatment, respectively; Pretreated, plants in which a local leaf was subjected to a HL, HS, HL+HS or HL&HS treatment beforethe systemic HL or HS treatment; EL, electrolyte leakage; n.s., no significant differences with respect to control; S, systemic tissue. (Scale bars, 1 cm.)

13818 | www.pnas.org/cgi/doi/10.1073/pnas.2005077117 Zandalinas et al.

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

11, 2

020

to the different stresses as shown in SI Appendix, Fig. S1 for 15 min. Followingthe local stress treatments, plants were incubated for 45 min under controlledconditions. Following the recovery period, one cauline leaf located 7 to 11 cmabove the rosette leaves in each plant was exposed to 1,700 μmol m−2 s−1 lightusing a fiber optic light source (Schott A20980) for 45min for HL acclimation ordipped into a 42 °C (or 23 °C as a control) water bath for 60 min. The caulineleaves were then photographed and sampled for electrolyte leakage assay (HLacclimation) or for chlorophyll measurement (HS acclimation) immediatelyafter HL or ay 6 d after HS (5, 6). Electrolyte leakage assays were performed asdescribed previously (6). Chlorophyll measurements were performed as de-scribed previously (5). Acclimation assays were repeated at least three timeswith 10 plants per repeat. In vivo measurements of ROS accumulation wereperformed as described previously (8). Stomatal aperture measurements wereperformed as described previously (15, 23).

Quantification of Hormone Levels. Quantification of plant hormones wasconducted using reverse-phase ultra-high-performance liquid chromatog-raphy (UHPLC) coupled with electrospray ionization tandem triple quadru-pole mass spectrometry (ESI/TQ MSMS) using multiple reaction monitoringas described previously (26). In brief, A. thaliana leaves were flash-frozen inliquid nitrogen and ground to powder. Samples (70 mg) were shaken vig-orously in isopropanol:H2O:HCl (2:1:0.002) for 1 h at 4 °C. Following ex-traction, dichloromethane was added, and samples were again shakenvigorously for an additional 30 min at 4 °C. Following centrifugation (3500 ×g at 4 °C), the bottom layer was removed using a glass syringe. Samples werethen dried under nitrogen, redissolved in 0.10 mL of MeOH and 1 mL of 1%

acetic acid, purified over Oasis HLB columns (Waters), washed with 1% aceticacid, and eluted with 80% MeOH and 1% acetic acid. The eluted sampleswere then dried again under nitrogen and redissolved in 25 μL of MeOH and25 μL of 1% acetic acid. A 10-μL volume of each sample was separated on aWaters Acquity UPLC BEH C18 column (2.1 × 150 mm; 1.7 μm) at 60 °C, usinga Waters Acquity UHPLC system. A Waters Xevo TQ MSMS system was usedto identify the different hormones, and absolute quantification wasobtained using stable isotope label internal standards for JA and SA (26).

Statistical Analysis. Differentially expressed transcripts were defined as thosethat had a fold change with an adjusted P < 0.05 (negative binomial Waldtest followed by Benjamini–Hochberg correction). Venn diagram overlapswere subjected to hypergeometric testing using the R package phyper. Ac-climation and stomatal aperture measurement data are presented as mean± SD. Statistical analysis was performed by two-way ANOVA followed byTukey’s post hoc test (*P < 0.05; **P < 0.01). Different letters in stomatalmeasurements denote statistical significance at P < 0.05.

Data Availability.Data supportingthe findingsof thisworkareprovided in themainpaper and SI Appendix. Raw and processed RNA-seq data files were deposited inthe GEO database (https:// www.ncbi.nlm.nih.gov/geo/) (accession no. GSE138196).

ACKNOWLEDGMENTS. This work was supported by funding from the NSF(Award nos. IOS-1353886, MCB-1936590, and IOS-1932639) and the Univer-sity of Missouri.

1. M. Borghi, L. Perez de Souza, T. Yoshida, A. R. Fernie, Flowers and climate change: A

metabolic perspective. New Phytol. 224, 1425–1441 (2019).2. J. Bailey-Serres, J. E. Parker, E. A. Ainsworth, G. E. D. Oldroyd, J. I. Schroeder, Genetic

strategies for improving crop yields. Nature 575, 109–118 (2019).3. H. Kollist et al., Rapid responses to abiotic stress: Priming the landscape for the signal

transduction network. Trends Plant Sci. 24, 25–37 (2019).4. M. Toyota et al., Glutamate triggers long-distance, calcium-based plant defense sig-

naling. Science 361, 1112–1115 (2018).

5. N. Suzuki et al., Temporal-spatial interaction between reactive oxygen species and ab-

scisic acid regulates rapid systemic acclimation in plants. Plant Cell 25, 3553–3569 (2013).6. S. I. Zandalinas, S. Sengupta, D. Burks, R. K. Azad, R. Mittler, Identification and

characterization of a core set of ROS wave-associated transcripts involved in the

systemic acquired acclimation response of Arabidopsis to excess light. Plant J. 98,

126–141 (2019).7. G. Miller et al., The plant NADPH oxidase RBOHD mediates rapid systemic signaling in

response to diverse stimuli. Sci. Signal. 2, ra45 (2009).

HL HS

Stomata close

Stomata open

JA

Acclimation to HL or HS

No Acclimation to HL or HS

HL&HS

ROS wave

Stomata close and then open

Fast ROS wave

HL

Stomata close and then open

ROS wave HS

HL+HS

ROS wave

No change in stomatal

aperture

Slow ROS wave

No change in stomatal

aperture

Stomata close

HS responses

2 8AT1G02390AT1G07985AT1G08430AT1G12130AT1G12610AT1G15010AT1G19210AT1G28370AT1G50745AT1G50750AT1G56300AT1G66100AT1G74930AT1G76650AT1G78760AT1G80440AT1G80840AT2G27080AT2G34600AT2G35930AT2G41640AT2G44840AT3G02840AT3G04640AT3G48520AT3G61190AT4G01130AT4G12900AT4G13395AT4G25470AT4G29780AT5G03545AT5G05430AT5G21960AT5G42380AT5G54490

HL responses

2 8AT1G07985AT1G12610AT1G14250AT1G22680AT1G28370AT1G50745AT1G52040AT1G56300AT1G65390AT1G74930AT1G78760AT1G79360AT1G80840AT2G03250AT2G20870AT2G32210AT2G33020AT2G44840AT3G02840AT3G20180AT3G45140AT4G09820AT4G12900AT4G13395AT4G15200AT4G17490AT4G23680AT4G24140AT4G25470AT4G26415AT4G29780AT5G49615AT5G58280

Photosynthesis

2 8PsaDPsb27PsaOLhca2.1PetHLhcb1.2Lhcb1.1Lhcb1.3PsaKPsaFPsbSPsaGPsbXPsbWLhcb1.4Lhca1ATPase-dLhcb5Lil1.2Lil3.1PsbQPsb28Lil4Lhcb4.1Lhcb3PsaN

HS responses

2 8AT1G02390AT1G07985AT1G08430AT1G12130AT1G12610AT1G15010AT1G19210AT1G28370AT1G50745AT1G50750AT1G56300AT1G66100AT1G74930AT1G76650AT1G78760AT1G80440AT1G80840AT2G27080AT2G34600AT2G35930AT2G41640AT2G44840AT3G02840AT3G04640AT3G48520AT3G61190AT4G01130AT4G12900AT4G13395AT4G25470AT4G29780AT5G03545AT5G05430AT5G21960AT5G42380AT5G54490AT5G62520AT3G29250

HL responses

2 8AT1G07985AT1G12610AT1G14250AT1G22680AT1G28370AT1G50745AT1G52040AT1G56300AT1G65390AT1G74930AT1G78760AT1G79360AT1G80840AT2G03250AT2G20870AT2G32210AT2G33020AT2G44840AT3G02840AT3G20180AT3G45140AT4G09820AT4G12900AT4G13395AT4G15200AT4G17490AT4G23680AT4G24140AT4G25470AT4G26415AT4G29780AT5G49615AT5G58280

Photosynthesis

2 8PsaDPsb27PsaOLhca2.1PetHLhcb1.2Lhcb1.1Lhcb1.3PsaKPsaFPsbSPsaGPsbXPsbWLhcb1.4Lhca1ATPase-dLhcb5Lil1.2Lil3.1PsbQPsb28Lil4Lhcb4.1Lhcb3PsaN

0.5 15<

0.5 15<

1 2

0.5 15<

0.5 15<

1 2

Low representation of acclimation transcripts

High representation of acclimation transcripts

SA

Fig. 9. Model showing that the manner in which plants sense the different stresses that trigger systemic signals during stress combination (i.e., at the same ordifferent leaves) makes a significant difference in how fast and efficient systemic ROS signals and transcriptomic, hormonal, stomatal, and acclimation responsesare triggered. When two different stresses are simultaneously applied to the same leaf (HL+HS), the systemic response is suppressed. This is reflected in theexpression of systemic HL- or HS-response and photosynthetic-associated transcripts (boxed heat maps; SI Appendix, Fig. S3 and Table 1), rates of SA and JAaccumulation (Fig. 7A), accumulation of systemic ROS (dashed orange arrow; Figs. 6 and 8A), and the lack of SAA of systemic leaves (Figs. 5A and 8B). In contrast,when the two different stresses are applied to two different leaves of the same plant, the rate of systemic ROS accumulation is faster (dashed black and redarrows; Figs. 6 and 8A); stomata display a rapid response (Fig. 5B); HL-, HS-, and photosynthetic-associated transcripts accumulate in systemic leaves (boxed heatmaps; SI Appendix, Fig. S3 and Table 1); and SAA is induced (Figs. 5A and 8B). Plants are depicted as being able to integrate different systemic signals: however,themanner in which plants sense the different stresses that trigger these signals makes a significant difference in how fast and efficient they are able to acclimate.

Zandalinas et al. PNAS | June 16, 2020 | vol. 117 | no. 24 | 13819

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

11, 2

020

8. Y. Fichman, G. Miller, R. Mittler, Whole-plant live imaging of reactive oxygen species.Mol. Plant 12, 1203–1210 (2019).

9. C. T. Nguyen, A. Kurenda, S. Stolz, A. Chételat, E. E. Farmer, Identification of cellpopulations necessary for leaf-to-leaf electrical signaling in a wounded plant. Proc.Natl. Acad. Sci. U.S.A. 115, 10178–10183 (2018).

10. A. Christmann, E. Grill, J. Huang, Hydraulic signals in long-distance signaling. Curr.Opin. Plant Biol. 16, 293–300 (2013).

11. M. Szechy�nska-Hebda, J. Kruk, M. Górecka, B. Karpi�nska, S. Karpi�nski, Evidence forlight wavelength-specific photoelectrophysiological signaling and memory of excesslight episodes in Arabidopsis. Plant Cell 22, 2201–2218 (2010).

12. Z. Guo et al., Systemic induction of photosynthesis via illumination of the shoot apexis mediated sequentially by phytochrome b, auxin and hydrogen peroxide in tomato.Plant Physiol. 172, 1259–1272 (2016).

13. W. G. Choi et al., Orchestrating rapid long-distance signaling in plants with Ca2+, ROSand electrical signals. Plant J. 90, 698–707 (2017).

14. S. Gilroy et al., A tidal wave of signals: Calcium and ROS at the forefront of rapidsystemic signaling. Trends Plant Sci. 19, 623–630 (2014).

15. A. R. Devireddy, S. I. Zandalinas, A. Gómez-Cadenas, E. Blumwald, R. Mittler, Co-ordinating the overall stomatal response of plants: Rapid leaf-to-leaf communicationduring light stress. Sci. Signal. 11, eaam9514 (2018).

16. S. Karpinski et al., Systemic signaling and acclimation in response to excess excitationenergy in Arabidopsis. Science 284, 654–657 (1999).

17. A. Kurenda, C. T. Nguyen, A. Chételat, S. Stolz, E. E. Farmer, Insect-damaged Arabi-

dopsis moves like wounded Mimosa pudica. Proc. Natl. Acad. Sci. U.S.A. 116,

26066–26071 (2019).18. F. K. Choudhury, A. R. Devireddy, R. K. Azad, V. Shulaev, R. Mittler, Local and systemic

metabolic responses during light-induced rapid systemic signaling. Plant Physiol. 178,

1461–1472 (2018).19. N. Suzuki, R. M. Rivero, V. Shulaev, E. Blumwald, R. Mittler, Abiotic and biotic stress

combinations. New Phytol. 203, 32–43 (2014).20. R. Mittler, E. Blumwald, Genetic engineering for modern agriculture: Challenges and

perspectives. Annu. Rev. Plant Biol. 61, 443–462 (2010).21. R. Mittler, Abiotic stress, the field environment and stress combination. Trends Plant

Sci. 11, 15–19 (2006).22. D. Balfagón et al., Jasmonic acid is required for plant acclimation to a combination of

high light and heat stress. Plant Physiol. 181, 1668–1682 (2019).23. A. R. Devireddy, J. Arbogast, R. Mittler, Coordinated and rapid whole-plant systemic

stomatal responses. New Phytol. 225, 21–25 (2020).24. S. I. Zandalinas, F. B. Fritschi, R. Mittler, Signal transduction networks during stress

combination. J. Exp. Bot. 71, 1734–1741 (2020).25. J. A. Raven, Speedy small stomata? J. Exp. Bot. 65, 1415–1424 (2014).26. J. Simura et al., Plant hormonomics: Multiple phytohormone profiling by targeted

metabolomics. Plant Physiol. 177, 476–489 (2018).

13820 | www.pnas.org/cgi/doi/10.1073/pnas.2005077117 Zandalinas et al.

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

11, 2

020