Constitutive activation of nitrate reductase in tobacco alters flowering time and plant biomass

Abstract

Pyridine alkaloids produced in tobacco can react with nitrosating agents such as nitrite to form tobacco-specific nitrosamines (TSNA), which are among the most notable toxicants present in tobacco smoke. The market type known as burley tobacco is particularly susceptible to TSNA formation because its corresponding cultivars exhibit a nitrogen-use-deficiency phenotype which results in high accumulation of nitrate, which, in turn, is converted to nitrite by leaf surface microbes. We have previously shown that expression of a constitutively activated nitrate reductase (NR) enzyme dramatically decreases leaf nitrate levels in burley tobacco, resulting in substantial TSNA reductions without altering the alkaloid profile. Here, we show that plants expressing a constitutively active NR construct, designated 35S:S523D-NR, display an early-flowering phenotype that is also associated with a substantial reduction in plant biomass. We hypothesized that crossing 35S:S523D-NR tobaccos with burley cultivars that flower later than normal would help mitigate the undesirable early-flowering/reduced-biomass traits while maintaining the desirable low-nitrate/TSNA phenotype. To test this, 35S:S523D-NR plants were crossed with two late-flowering cultivars, NC 775 and NC 645WZ. In both cases, the plant biomass at harvest was restored to levels similar to those in the original cultivar used for transformation while the low-nitrate/TSNA trait was maintained. Interestingly, the mechanism by which yield was restored differed markedly between the two crosses. Biomass restoration in F₁ hybrids using NC 645WZ as a parent was associated with delayed flowering, as originally hypothesized. Unexpectedly, however, crosses with NC 775 displayed enhanced biomass despite maintaining the early-flowering trait of the 35S:S523D-NR parent.

Figure 1

Above-ground biomass of T1 lines under two fertilization levels in 2014. Means ± standard errors of means with the same letter (upper case for the higher fertilization rate, lower case for the lower rate) are not significantly different at a 5% significance level according to REGWQ-based groupings. Sample sizes per line ranged from 29 to 39. TN90e4e5 = wild type; C3-11 = 35S:S523D-NR construct in DH98-325-6#775; GH10-4 = vector control in TN90e4e5; GH3-1, GH8-5, GH5-2, GH8-1, and GH5-5 = 35S:S523D-NR construct in TN90e4e5.

Figure 2

Above-ground biomass of F1 hybrids between the 35S:S523D-NR lines and late-flowering breeding lines and their parents, grown at Clayton in 2017 and 2018. Means ± standard errors of means with the same letter are not significantly different at a 5% significance level according to REGWQ groupings. Sample sizes per line ranged from 26 to 34 in 2017 and 58 to 64 in 2018.

Figure 3

Nitrate (NO3N) content in cured leaves of F1 hybrids between the 35S:S523D-NR lines and late-flowering breeding lines and their parents, grown in 2017 and 2018. Means ± standard errors of means with the same letter are not significantly different at a 5% significance level according to REGWQ groupings. Sample sizes per line ranged from 22 to 33 in 2017 and 58 to 64 in 2018.

Figure 4

Total alkaloid content (% dry weight) in the cured leaves of F1 hybrids between the 35S:S523D-NR lines and late-flowering breeding lines and their parents grown in 2017 and 2018. Means ± standard errors of means with the same letter are not significantly different at a 5% significance level according to REGWQ groupings. Sample sizes per line ranged from 9 to 10 in 2017 and were 10 in 2018.

Figure 5

Total TSNA content in the cured leaves of F1 hybrids between the 35S:S523D-NR lines and late-flowering breeding lines and their parents, grown in 2017 and 2018. Means ± standard errors of means with the same letter are not significantly different at a 5% significance level according to REGWQ groupings. Sample sizes per line ranged from 9 to 10 in 2017 and were 10 in 2018.