SNF1-related kinases allow plants to tolerate herbivory by allocating carbon to roots

Abstract

Herbivore attack elicits costly defenses that are known to decrease plant fitness by using resources that are normally slated for growth and reproduction. Additionally, plants have evolved mechanisms for tolerating attack, which are not understood on a molecular level. Using 11C-photosynthate labeling as well as sugar and enzyme measurements, we found rapid changes in sink-source relations in the annual Nicotiana attenuata after simulated herbivore attacks, which increased the allocation of sugars to roots. This herbivore-induced response is regulated by the beta-subunit of an SnRK1 (SNF1-related kinase) protein kinase, GAL83, transcripts of which are rapidly down-regulated in source leaves after herbivore attack and, when silenced, increase assimilate transport to roots. This C diversion response is activated by herbivore-specific elicitors and is independent of jasmonate signaling, which regulates most of the plant's defense responses. Herbivore attack during early stages of development increases root reserves, which, in turn, delays senescence and prolongs flowering. That attacked GAL83-silenced plants use their enhanced root reserves to prolong reproduction demonstrates that SnRK1 alters resource allocation so that plants better tolerate herbivory. This tolerance mechanism complements the likely defensive value of diverting resources to a less vulnerable location within the plant.

Fig. 1.

Experimental setup. (A) Numbers denote the mature (source) leaves used for either ¹¹CO₂ pulse feeding (+3) or elicitation (+2, +4, and +5); immature (sink) leaves are labeled with negative numbers. The sequence indicates the leaf age; the larger the number, the older the leaf. (B) Scheme of detection areas. The load leaf was separately measured to control ¹¹CO₂ pulses. (C) Scheme showing the positions of the shoot and root detectors as well as the lead and tungsten shielding (collimation) needed to separate the field of view of the different detectors.

Fig. 2.

C allocation in N. attenuata. (A) Relative change (mean ± SE, n = 3–6) of the root-partitioned C fraction of asGAL83, WT, and asLOX plants after 5 h in response to different types of induction (C, control; W, wounding; R, R elicitation; FAC, application of FACs; WS, wounding of sink leaves; SR, aboveground sink removal) as measured by ¹¹CO₂ application. Asterisks indicate significant difference from WT C (for each comparison with WT C, Mann–Whitney U test, P < 0.05). (B) Fraction (mean ± SE, n WT = 45, n asGAL83 = 27) of assimilates partitioned to roots of unelicited plants (Mann–Whitney U test, U < 462.5, P = 0.0134).

Fig. 3.

Enzyme activities of alkaline invertase, soluble acid (vacuolar) invertase and cell wall invertase, and soluble sugar contents (sucrose, glucose, and fructose) of source leaves (A), sink leaves (B), and roots (C) 5 h after W and R treatment (see Fig. 1). (A) Mann–Whitney U test, sucrose, U < 25, P = 0.009; fructose, U < 24, P = 0.0163; n = 5. (B) Mann–Whitney U test, U < 20, P < 0.01; n = 5. (C) Mann–Whitney U test, wounding, U < 9, P < 0.05; R elicitation, U < 9, P < 0.05; n = 4. SuSy activities (data not shown) were not changed after any treatment. ∗, P < 0.05; ∗∗, P < 0.01.

Fig. 4.

Northern blot of GAL83 and SNF1 (α-subunit) mRNA expression in N. attenuata sink and source leaves (see Fig. 1A) 30 min after 4 h of repeated wounding and application of M. sexta R. Total RNA was pooled from 27 replicates for each treatment. GAL83 transcripts were down-regulated locally in treated source leaves and slightly down-regulated systemically after treatment of sink leaves. The transcript of the α-subunit, SNF1, was not regulated in leaves. rRNA was used as a loading control.

Fig. 5.

Seed capsules per g of biomass and percentage of unripe capsules of the total capsules (A) and root (dry) masses (B) of WT and asGAL83 plants 54 days after elicitation. (A) Seed capsule number per g of biomass, mean ± SE (percentage of unripe capsules of total capsules is shown by gray bars). Asterisks over capsules/biomass bars (control and wounding) indicate significant differences between lines (Mann–Whitney U test, P < 0.01). Asterisks over capsules/biomass bars (R elicitation and herbivory) indicate significant differences between treatment and control plants (Mann–Whitney U test, P < 0.01). Asterisks over unripe capsules bars indicate differences compared with control (Mann–Whitney U test, P < 0.01). (B) Final root mass, mean ± SE. The asGAL83 controls do have a significantly larger root mass than WT controls (unpaired t test, DF = 26, T = 2.071, P < 0.05). All elicited asGAL83 root masses are significantly smaller than those of asGAL83 control plants (ANOVA, F_3,46 = 15.525, P < 0.0001, post hoc P < 0.001). ∗, P < 0.05; ∗∗, P < 0.01; ∗∗∗, P < 0.001.

Fig. 6.

Flower production of WT and asGAL83 plants after elicitation (mean ± SE). (A) Control and wounding treatment. (B) R elicitation and herbivory. (A and B) Fully opened flowers were measured. Watering of plants was gradually reduced over a 10-day period (see Supporting Text). Asterisks indicate that asGAL83 plants produced significantly more flowers than did WT (unpaired t test, t >3, P < 0.05). (C) Late flowers of WT controls (left) are smaller than those of WT H-treated plants (right) on day 49 after elicitation; the same difference was observed for asGAL83 flowers.