Perennial and annual lifestyles represent two main life strategies in the plant kingdom, and they are largely defined by flowering behavior. A new study uncovers the genetic basis underlying the diversity of these strategies in the Brassicaceae, in which three floral repressor genes act in an additive fashion to shape continuous variation in flowering behavior, and excitingly, this discovery enables swift perennialization of annual Brassicaceae crops.

While longevity is a long-time focus of biomedical research, it is of much less interest to plant biologists. This is largely because most major food crops (e.g., rice, wheat, and maize) and the primary model plant Arabidopsis thaliana (A. thaliana) are all short-lived annuals. However, in nature long-lived perennials account for > 90% of all plant species;1 our annual-concentrated research therefore greatly limits our knowledge of plants as a whole and overlooks important mechanisms regarding plant longevity. Moreover, understanding the commonly occurred perennial-to-annual transition is of both ecological and agricultural significance, as future climate change has been predicted to result in an increased proportion of annuals1 and perennial grain crops have shown great potential for addressing many sustainability issues in agriculture.2

Perenniality is closely associated with flowering behavior in a large category of perennials, termed as polycarpic perennials, which can flower multiple times across their long lifespans.3 By contrast, monocarpic plants flower only once and die subsequently. Such a difference in parity, however, does not fully explain the diversity of growth habits, which are also characterized by whether or to what extent certain environmental cues are required for floral initiation. The induction of a plant’s flowering by exposure to the prolonged cold of winter is a well-studied process known as vernalization.4 According to this second criterion, monocarpic plants can be further classified into summer annuals, winter annuals, and biennials, whose vernalization requirement is none, facultative (accelerating flowering), and obligate, respectively.5

Following extensive research on control of the floral transition in A. thaliana, a breakthrough that first provided connection between vernalization and parity was achieved by comparing A. thaliana with its close polycarpic perennial relative Arabis alpina.6 Through screening of early flowering mutants lacking vernalization requirement, PERPETUAL FLOWERING 1 (PEP1) was identified to regulate perennial flowering in A. alpina. PEP1 is the orthologue of the A. thaliana floral repressor gene FLOWERING LOCUS C (FLC). Unlike FLC, whose vernalization-induced repressed state is maintained upon return to warmer temperatures, such silencing on PEP1 terminates shortly after the end of vernalization. As such, shoot meristems formed after a vernalization-determined temporal window can be maintained vegetative in A. alpina, whereas in A. thaliana the fate of all auxiliary shoots, regardless of when they are developed, is floral. Although PEP1 is a major determinant of the perennial growth habit, some axillary branches of the pep1 mutant, albeit much less than wild type, can remain vegetative after each flowering round, implying that additional factors are needed to confer perenniality.

To pinpoint these additional factors via gene mapping, Zhai et al.7 had gone to great lengths to screen compatible annual and perennial species from the Brassicaceae. To obtain a boarder view of the whole family, they made use of two pairs of species from two genera — Crucihimalaya and Erysimum. Utilizing the genomes of these four species, which were sequenced and assembled by the group, three intervals were identified to have associations with vernalization requirement. In addition to an expected interval containing a gene homologous to FLC, the other two intervals each contained a homologue of FLC-like genes — FLOWERING LOCUS M (FLM) or MADS AFFECTING FLOWERING (MAF). Intriguingly, the intervals harboring these genes were also detected to be associated with the polycarpic growth habit, supporting functional redundance of these three MADS-box genes in regulating perennial flowering in Brassicaceae plants.

Functional characterization of these three MADS-box genes was carried out by generating knockout mutants of the perennial species and constructing near-isogenic lines carrying the perennial alleles in the annual background. The observed phenotypes were highly diversified, suggesting that, despite targeting many common downstream targets, these MAD-box genes contribute unequally and in a combinatorial manner to perennial flowering behavior, and that the ratios of their contributions also differ among species (Fig. 1). Strikingly, the triple mutant showed an annual-like phenotype, suggesting that these three genes are the only key determinants of perenniality. Subsequent comprehensive expression analysis of the three MAD-box genes revealed that fine-tuning of flowering behavior is also underpinned by intricate modulation of expression dynamics mainly due to various sensitivities to cold environment and distinct stabilities of vernalization-induced silencing (Fig. 1). Collectively, these data outline a rather complex “dual threshold” model responsible for life strategy diversification in the Brassicaceae, in spite of the fact that its core component consists of a mere three genes (Fig. 1).

Three FLC-related MADS-box genes contribute unequally and their combined contribution determines the level of vernalization requirement and the degree of parity. The length of the pink bar represents the degree of contribution, which is concomitantly determined by the gene’s functional strength and current expression level. Another key variable involved is whether silencing of these genes by vernalization is stable or transient. V stands for vernalization.

Building on these mechanistic insights, Zhai et al. actualized the conversion of annual Brassicaceae plants to perennial forms by introducing a single MADS-box gene from perennials. Notably, such an insertion not only can perennialize the compatible annual species in the same genus, but also the annual plant A. thaliana, a process that was previously not accomplished by heterologously expressing PEP1.8 Perhaps more significantly, Zhai et al. also demonstrated that oilseed rape, an important annual Brassicaceae crop, can be perennialized using the same simple genetic modification.

Crop perennilization has succeeded through interspecific hybridization between an annual crop and its perennial relative with subsequent genetic improvements.9 Taking advantage of multiplex genome editing and in-depth functional genomics knowledge of cultivated varieties, de novo domestication of wild perennial species has been proposed as an accelerated route of crop perennilization.10 Unprecedentedly, the genetic basis unraveled by Zhai et al. opens possibilities to perennialize elite annual varieties through minimal genetic manipulation. Facing this prospect, it will be interesting to examine whether a similar genetic mechanism also exists in other plant families and whether or how it governs other perenniality-related physiological traits, particularly the developmental regulation of various perennating organs.