Immunological "Memory" in the Induced Accumulation of Nicotine in Wild Tobacco

Abstract

Nicotine, an inducible defense in Nicotiana sylvestris, is produced in the roots and transported throughout the plant after leaf damage. Root nicotine production is activated by an endogenously produced signal, jasmonic acid (JA), which is synthesized in response to leaf damage and either directly through transport or indirectly via another signal, increases JA pools in the roots. The addition of JA or its methyl ester (MJ) to the roots of hydroponically grown plants stimulates nicotine production in the same way that leaf damage does. In this paper we use MJ to stimulate the induced defense without damaging plants in order to examine the hypothesis that these plants have an immunological memory of prior inductions. Such a memory can be measured as a change in either the speed or the amount of induced nicotine production or accumulation that results from prior inductions. Memory is an important component of the induced defenses of animals and is predicted to evolve in systems where an initial attack is a reliable predictor of future attacks. We induced 200 plants 1, 2, or 3 times during an 18—d period of rosette growth, allowed 6 d between inductions for the relaxation of the response between inductions and quantified changes in biomass, whole—plant nicotine pools and de novo rates of nicotine production from 15N—labeled nitrate acquired at the time of induction. Induced changes in nicotine pools and the rates of nicotine production varied across the three induction periods: the highest rates and induced pools occurred during the second induction period, but the highest concentrations were found during the first induction period. The rates of de novo nicotine production from 15N—labeled nitrate were dramatically increased by MJ stimulations during all three induction periods. We found no evidence for alterations in the rates of nicotine production in plants not currently exposed to MJ; 6 d after an induction, plants had relaxed their rates of nicotine production to levels that were not significantly different from those found in plants without a prior induction history. Moreover, the maximum pools of 15N— labeled nicotine were equivalent among all induced plants regardless of prior induction history. In summary, one or two prior inductions did not markedly affect nicotine production from recently acquired nitrate as measured in either induced or uninduced plants. However, the ecologically more relevant measure of the speed of induction is the change in whole—plant nicotine pools, and plants with one and two prior inductions increased their whole—plant nicotine pool significantly faster during the third induction period than did plants not previously induced. Plants with two prior inductions attained significant increases in their nicotine pools 2 d earlier than did plants with one or no prior inductions. Moreover, the effect was additive: plants with two prior inductions were faster than plants with one. The effects of prior inductions on the speed of the induced response did not translate into increases in the magnitude of the response at the end of the 18—d experiment. The lack of change in the magnitude of the induced response may reflect the methods used in our experiment more than the plant's capabilities. The whole—plant induced response is limited by the amount of signal transported to the roots, and since we used the same amount of MJ in all stimulations, our experiment would not detect memory in the production of the signal in the damaged leaf or its transport to the roots. These results demonstrate that plants, like animals, alter their induced defense in response to their prior experiences.