Differences in nicotine metabolism of two Nicotiana attenuata herbivores render them differentially susceptible to a common native predator

Abstract

Background: Nicotiana attenuata is attacked by larvae of both specialist (Manduca sexta) and generalist (Spodoptera exigua) lepidopteran herbivores in its native habitat. Nicotine is one of N. attenuata's important defenses. M. sexta is highly nicotine tolerant; whether cytochrome P450 (CYP)-mediated oxidative detoxification and/or rapid excretion is responsible for its exceptional tolerance remains unknown despite five decades of study. Recently, we demonstrated that M. sexta uses its nicotine-induced CYP6B46 to efflux midgut-nicotine into the hemolymph, facilitating nicotine exhalation that deters predatory wolf spiders (Camptocosa parallela). S. exigua's nicotine metabolism is uninvestigated.

Methodology/principal findings: We compared the ability of these two herbivores to metabolize, tolerate and co-opt ingested nicotine for defense against the wolf spider. In addition, we analyzed the spider's excretion to gain insights into its nicotine metabolism. Contrary to previous reports, we found that M. sexta larvae neither accumulate the common nicotine oxides (cotinine, cotinine N-oxide and nicotine N-oxide) nor excrete them faster than nicotine. In M. sexta larvae, ingestion of nicotine as well as its oxides increases the accumulation of CYP6B46 transcripts. In contrast, S. exigua accumulates nicotine oxides and exhales less (66%) nicotine than does M. sexta. Spiders prefer nicotine-fed S. exigua over M. sexta, a preference reversed by topical or headspace nicotine supplementation, but not ingested or topically-coated nicotine oxides, suggesting that externalized nicotine but not the nicotine detoxification products deter spider predation. The spiders also do not accumulate nicotine oxides.

Conclusions: Nicotine oxidation reduces S. exigua's headspace-nicotine and renders it more susceptible to predation by spiders than M. sexta, which exhales unmetabolized nicotine. These results are consistent with the hypothesis that generalist herbivores incur costs of detoxification, which include the ecological costs of greater predation risks, in addition to the previously demonstrated energetic, physiological and metabolic costs.

Figure 1. Nicotine oxidation does not benefit M. sexta larvae.

Larval (a) mass [(mean± SE) F_{4, 172} = 45.2, P≤0.05, n = 36, 34, 37, 33 and 37 for water, nicotine, NNO, cotinine and CNO, respectively] and (b) mortality (%) after 10 d of feeding on artificial diet containing water (control) or 0.1% (fresh mass) nicotine, NNO, cotinine or CNO (n = 30). (c) Waldbauer assay-based quantification of excreted (%) metabolites by fourth-instar larvae fed artificial diet containing 0.1% (fresh mass) metabolite [(mean± SE) F_{3, 25} = 4.3, P≤0.05, n = 6 for nicotine, NNO and cotinine and 8 for CNO]; (n. a.≡ not applicable). (d) Nicotine, NNO, cotinine or CNO are not degraded in frass over the 24 h period of the Waldbauer assays. Fresh frass was spiked with each metabolite to attain the final concentration of 0.5%; the spiked frass was extracted and analyzed after zero and 24 h of incubation to quantify the recovered metabolite. Every bar represents data from 3 replicates (n = 3). (e) Melanization of cotinine-fed (right) larva. (f) Discharge kinetics of hemolymph-injected (70±2.5 µg≡ 0.001% of larval fresh mass) nicotine, NNO, cotinine or CNO (n = 5). Lower-case letters and asterisks in (a), (c) and (e) indicate significant differences (P≤0.05) by one-way ANOVA; in (b), asterisk indicates significant differences (P≤0.05) in frequencies (and displayed percentages) by Fisher's exact test.

Figure 2. Transcript regulation of M. sexta CYPs in response to ingestion of nicotine and nicotine oxides.

(a) CYP6B46 [(mean± SE) F_{4, 19} = 94.5, P≤0.05, n = 5] (b) CYP4M1 [(mean± SE) F_{4, 20} = 9.5, P≤0.05, n = 5] and (c) CYP4M3 [(mean± SE) F_{4, 20} = 11.1, P≤0.05, n = 5] transcript levels (relative to ubiquitin) in midguts of 48 h old first-instar M. sexta larvae fed artificial diet containing water (control) or 0.1% (fresh mass) nicotine, NNO, cotinine or CNO. Asterisks indicate significant differences (P≤0.05) by one-way ANOVA.

Figure 3. S. exigua oxidizes nicotine, but M. sexta does not.

U(H)PLC/ESI-QTOF-MS-based quantitative analysis of nicotine, NNO, cotinine and CNO in (a) frass (b) hemolymph and (c) headspace of third-instar M. sexta (n = 5) and S. exigua (n = 5) larvae fed artificial diet containing 0.1% (fresh mass) nicotine. Lower-case letters above the S. exigua bars indicate significant differences (P≤0.05) among them, by one-way ANOVA. Asterisks above the M. sexta nicotine bars indicate that they differ significantly (P≤0.05) from the S. exigua nicotine values, as determined by one-way ANOVA. Nicotine oxides were not detected in M. sexta. The detection limit of nicotine was 0.25 ng and 0.5 ng for cotinine, CNO and NNO; efficiency of extraction was >90% for all these compounds.

Figure 4. S. exigua oxidizes nicotine, but M. sexta does not.

U(H)PLC/ESI-QTOF-MS-based quantitative analysis of nicotine, NNO, cotinine and CNO in (a) frass (b) hemolymph and (c) headspace of third-instar M. sexta (n = 5) and S. exigua (n = 5) larvae fed N. attenuata leaves. Lower-case letters above the S. exigua bars indicate significant differences (P≤0.05) among them by one-way ANOVA. Asterisks above the M. sexta nicotine bars indicate that they are significantly different (P≤0.05) from the S. exigua nicotine bars, as determined by one-way ANOVA. Nicotine oxides were not detected in M. sexta. The detection limit of nicotine was 0.25 ng and 0.5 ng for cotinine, CNO and NNO; efficiency of extraction was >90% for all these compounds .

Figure 5. Volatility analysis of nicotine and nicotine oxides.

(a) Schematic of setup used for the collection of evaporated nicotine, NNO, cotinine and CNO placed in the collection vial (1 µg/5 µL methanol). (b) Percentage of evaporated and residual nicotine, NNO, cotinine and CNO (n = 4); one µg of nicotine, NNO, cotinine or CNO (in 5 µL methanol) was incubated for 1 h.

Figure 6. Nicotine is more toxic than the nicotine oxides to S. exigua larvae.

(a) Larval mass of S. exigua [(mean± SE) F_{4, 154} = 6.03, P≤0.05, n = 32, 27, 33, 30 and 33 for water, nicotine, NNO, cotinine and CNO, respectively] and (b) mortality (%) of S. exigua larvae during 10days of feeding artificial diet containing water (control) or 0.1% (fresh mass) nicotine, NNO, cotinine or CNO; each bar represents data from 30 larvae. In (a), lower-case letters above the bars indicate significant differences (P≤0.05) by one-way ANOVA; in (b) asterisk indicates significant differences (P≤0.05) in frequencies (and displayed percentages) by Fisher's exact test.

Figure 7. Nicotine deters C. parallela but nicotine oxides do not.

C. parallela's predation (%) (in 1 h no-choice assay) on second-instar (a) M. sexta and (b) S. exigua larvae fed artificial diet containing water (control) or 0.1% (fresh mass) nicotine, NNO, cotinine or CNO. C. parallela predation (%) (in 1 h no-choice assay) on second-instar artificial diet fed (c) M. sexta and (d) S. exigua larvae coated with water (control) or 0.2% aqueous nicotine, NNO, cotinine or CNO. Each bar represents data from 30 larvae. Asterisks indicate significant differences (P≤0.05) in frequencies (and displayed percentages) by Fisher's exact test.

Figure 8. C. parallela's preference of S. exigua larvae over M. sexta larvae is diminished by topically coating larvae or supplementing their headspace with nicotine.

C. parallela's predation (%) (in 1 h choice assay) on second-instar M. sexta and S. exigua larvae fed (a) artificial diet or artificial diet containing 0.1% (fresh mass) nicotine (b) artificial diet and coated with water (control) or nicotine, and (c) artificial diet containing water (control) or 0.1% (fresh mass) nicotine and having the assay environment nicotine-perfumed using 500 µL of 1 mM nicotine on a cotton swab. Schematics in right panels of (a), (b) and (c) show the effects of various modes of nicotine supplementation to M. sexta and S. exigua larvae on the spider predation behavior. Each bar represents data from 30 larvae. Asterisks indicate significant differences (P≤0.05) in frequencies (and displayed percentages) by Fisher's exact test.