Costs and benefits of priming for defense in Arabidopsis Marieke van Hulten*, Maaike Pelser*, L. C. van Loon, Corne �� M. J. Pieterse, and Jurriaan Ton��� Institute of Environmental Biology, Section Phytopathology, Utrecht University, P.O. Box 800.84, 3508 TC, Utrecht, The Netherlands Edited by Frederick M. Ausubel, Harvard Medical School, Boston, MA, and approved February 5, 2006 (received for review November 29, 2005) Induced resistance protects plants against a wide spectrum of diseases however, it can also entail costs due to the allocation of resources or toxicity of defensive products. The cellular defense responses involved in induced resistance are either activated directly or primed for augmented expression upon pathogen attack. Priming for defense may combine the advantages of en- hanced disease protection and low costs. In this study, we have compared the costs and benefits of priming to those of induced direct defense in Arabidopsis. In the absence of pathogen infec- tion, chemical priming by low doses of -aminobutyric acid caused minor reductions in relative growth rate and had no effect on seed production, whereas induction of direct defense by high doses of -aminobutyric acid or benzothiadiazole strongly affected both fitness parameters. These costs were defense-related, because the salicylic acid-insensitive defense mutant npr1-1 remained unaf- fected by these treatments. Furthermore, the constitutive priming mutant edr1-1 displayed only slightly lower levels of fitness than wild-type plants and performed considerably better than the constitutively activated defense mutant cpr1-1. Hence, priming involves less fitness costs than induced direct defense. Upon infection by Pseudomonas syringae or Hyaloperonospora para- sitica, priming conferred levels of disease protection that almost equaled the protection in benzothiadiazole-treated wild-type plants and cpr1 plants. Under these conditions, primed plants displayed significantly higher levels of fitness than noninduced plants and plants expressing chemically or cpr1-induced direct defense. Collectively, our results indicate that the benefits of priming-mediated resistance outweigh the costs in environments in which disease occurs. induced resistance innate immunity plant defense Pthrough lants resist the attacks of harmful microorganisms and insects constitutive and inducible defenses. It is generally believed that inducible defenses have evolved to save energy under enemy-free conditions, but costs still arise upon activation of these defenses under hostile conditions. These costs can result from allocation of limited resources to defensive compounds or toxicity of the defense to the plant���s own metabolism (1). In addition, costs can arise from external factors, when the defensive trait affects a beneficial interaction with another organism in the environment. It is therefore reasonable to assume that plants express their inducible defenses only if the benefits (i.e., protection against the attackers) outweigh the costs of the resistance. Various studies have demonstrated costs related to jasmonic acid (JA)-inducible defense against herbivory. These costs can affect plant growth and reproductive traits (2���4). There are also studies that demonstrated ecological benefits of JA-inducible defense. Agrawal (5) showed that induction of defense in wild radish against insects correlated with enhanced seed production. Additionally, JA-induced defense in wild populations of Nico- tiana attenuata conferred enhanced seed production in popula- tions exposed to herbivory (3). Hence, costs of JA-inducible defenses are outweighed by the benefits of protection when plants are attacked by herbivores. A cost���benefit balance of salicylic acid (SA)-inducible defenses against pathogens has also been supposed. In wheat, Heil et al. (6) demonstrated costs of SA-inducible defenses on growth and seed set. In Arabidopsis, Cipollini (7) showed that exogenously applied SA reduced seed production. Recently, Heidel et al. (8) performed a field exper- iment with two sets of Arabidopsis genotypes: one group that is blocked in SA-inducible defenses and another group that con- stitutively expresses SA-inducible defenses. Both classes of ge- notypes were negatively affected in growth and seed set, sug- gesting that plant fitness reaches an optimum at a certain intermediate level of resistance that balances fitness and defense. Upon appropriate stimulation, plants can increase their level of resistance against future pathogen attack. This phenomenon is known as induced resistance. Based on differences in signaling pathways and spectra of effectiveness, different types of induced resistancehavebeendefined.Theclassicformofinducedresistance is referred to as systemic acquired resistance (SAR) and occurs in distal plant parts upon localized infection by a necrosis-inducing pathogen (9). SAR is controlled by a signaling pathway that depends on endogenous accumulation of SA and the defense regulatory protein NPR1 (10) and is predominantly effective against biotrophic pathogens (11). Selected strains of nonpatho- genic rhizobacteria can also induce systemic resistance, which is referred to as induced systemic resistance (ISR) (12). In Arabidop- sis, ISR triggered by Pseudomonas fluorescens WCS417r functions independently of SA but requires NPR1 and responsiveness to JA and ethylene (13). P. fluorescens WCS417r-mediated ISR has a different spectrum of effectiveness than SAR, and is predominantly effective against pathogens that are sensitive to JA- and ethylene- dependent basal resistance (11). A third type of induced resistance is activated upon application of the chemical -aminobutyric acid (BABA). The signaling pathway controlling BABA-induced resis- tance(BABA-IR)(14)differsfromthatofSARandISR.Although BABA-IR against Pseudomonas syringae depends solely on SA and NPR1 (15), against pathogenic fungi and oomycetes it is controlled by a pathway that involves abscisic acid- and phosphoinositide- dependent signaling (16, 17). BABA-IR is effective against biotro- phic and necrotrophic pathogens, as well as certain types of abiotic stress (14���20). For a long time, it was assumed that protection by induced resistance is based on direct activation of defenses by the resistance- inducing agent. Accumulation of pathogenesis-related proteins is an example that occurs directly upon induction of SAR. However, the suggested contribution of pathogenesis-related proteins to resistance is uncertain and appears insufficient to explain the broad spectrum of protection by SAR (21). Moreover, both rhizobacteria- mediated ISR and BABA-IR are not associated with direct acti- vation of defense-related genes (12, 14). Interestingly, plants ex- pressing SAR, ISR, or BABA-IR exhibit a faster and stronger Conflict of interest statement: No conflicts declared. This paper was submitted directly (Track II) to the PNAS office. Abbreviations: BABA, -aminobutyric acid BABA-IR, BABA-induced resistance BTH, ben- zothiadiazole ISR, induced systemic resistance JA, jasmonic acid RGR, relative growth rate SA salicylic acid SAR, systemic acquired resistance. *M.v.H. and M.P. contributed equally to this work. ���To whom correspondence should be addressed. E-mail: j.ton@bio.uu.nl. �� 2006 by The National Academy of Sciences of the USA 5602���5607 PNAS April 4, 2006 vol. 103 no. 14 www.pnas.org cgi doi 10.1073 pnas.0510213103

activation of specific defense responses after they have been in- fected by a pathogen. This capacity for augmented defense expres- sion is called priming (22). Since the discovery of priming in plant cell suspension cultures by Kauss et al. (23), priming has been demonstrated in different plant species against pathogens, insects, and abiotic stress (22). Hence, priming appears to be a common feature of a plant���s immune system that offers protection against a wide spectrum of environ- mental stresses. In Arabidopsis, Kohler et al. (24) demonstrated that SAR-induced Arabidopsis expressed augmented levels of the de- fense-related PAL gene upon infection by P. syringae. Van Wees et al. (25) and Verhagen et al. (26) demonstrated that treatment of Arabidopsis with ISR-inducing P. fluorescens WCS417r bacteria did not directly activate defense-related genes however, Arabidopsis was primed for enhanced expression of JA- and ethylene-inducible genes upon infection by P. syringae. Treatment with BABA primes Arabidopsis for SA-dependent defenses (15, 20) and enhanced formation of callose-rich papillae that functions independently from SA and NPR1 (16). Recently, we showed that these forms of priming require specific cellular signaling components (17), sug- gesting a regulation mechanism that is exclusively dedicated to priming. Priming accelerates and increases the plant���s ability to activate the defense that is best adapted to cope with a certain stress situation. In this perspective, priming represents an important ecological adaptation to resist environmental stress. As most studies on costs and benefits of induced resistance have focusedonsituationsinwhichthedefenseisactivateddirectlybythe inducing agent, the possibility of studying costs and benefits of priming has so far been overlooked. This lack of data prompted us to determine the costs and benefits of priming in Arabidopsis and compare those to the costs and benefits of induced direct defense. By using BABA as a chemical inducer of priming and a constitutive priming mutant of Arabidopsis, we compared fitness parameters in the absence and presence of pathogen attack. We demonstrate that priming involves considerably fewer costs than induction of direct defense. In addition, we demonstrate that the benefits of priming outweigh the costs when infection by pathogens occurs. Results Effectiveness of Chemically Induced Priming and Direct Defense. To compare the effectiveness of priming and direct defense, 3- and 6-week-old Arabidopsis plants (Col-0) were treated with increasing concentrations of BABA or benzothiadiazole (BTH). Two days later, the 3-week-old plants were mock or challenge inoculated with Hyaloperonospora parasitica, whereas the 6-week-old plants were mock-inoculated or challenge-inoculated with P. syringae. To de- termine the level of priming and or direct defense, leaves were Fig. 1. Chemical induction of priming and direct defense against H. parasitica WACO9 (A���C) and P. syringae pv. tomato DC3000 (D���F). Col-0 plants were soil-drenched with increasing concentrations of BABA or sprayed with BTH and pathogen-inoculated 2 days later. (A) PR-1 gene expression in 3-week-old control plants or BABA- or BTH- treated plants at different time points after inoculation. hpi, hours postinoculation. (B) Callose deposition 2 days after H. parasitica inoculation. (Inset) A representative exam- ple of H. parasitica spores triggering callose deposition in epidermal cells. (Scale bar, 20 m.) n.d., not determined. (C) Induced resis- tance against H. parasitica at 8 days after inoculation. Asterisks indicate statistically different distributions of disease severity classes compared with the water control ( 2 test 0.05). Colonization by the patho- gen was visualized by lactophenol trypan blue staining and light microscopy. (D) PR-1 gene expression in 6-week-old control plants or BABA- or BTH-treated plants at different time points after inoculation. (E) Induced resistance against P. syringae. Shown are means SEM (n 15���20) of the percentage of leaves with symptoms at 3 days after inoculation. Different letters in- dicate statistically significant differences (least significant difference test 0.05). (F) Growth of P. syringae over a 3-day time interval. Shown are means SD (n 5���10). Different letters indicate statistically signif- icant differences (least significant differ- ence test 0.05). All experiments shown were repeated with comparable results. van Hulten et al. PNAS April 4, 2006 vol. 103 no. 14 5603 PLANT BIOLOGY