Twists in the pattern: REM transcription factors determine phyllotaxis in the Arabidopsis inflorescence

Abstract

Patterns that form with mathematical precision are remarkably prevalent in nature. The spirals seen in the scales of a pinecone or the seeds of a sunflower are examples of phyllotaxis, a self-organising process through which lateral organs arise at constant divergence angles around a central axis. These patterns are based on the Fibonacci sequence, where each number is the sum of the preceding two (1, 1, 2, 3, 5, 8, 13, 21, …). Dividing a full circle of 360° by the ratio of two consecutive Fibonacci numbers results in a larger arc of 222.5° and a smaller arc of 137.5°. This ‘golden angle’ of 137.5° is precisely the angle by which leaf or flower primordia are separated from one another. Why this pattern is so prevalent in nature remains unclear; it has been suggested it might improve light capture, maximise organ packing or be due to developmental constraints within the meristem (Reinhardt & Gola, 2022).

At the tissue level, phyllotactic patterns at the shoot apical and inflorescence meristems are shaped by gradients of the plant hormone auxin. Auxin is initially produced at the centre of the meristem and transported to regions at the meristem's periphery where a new leaf or flower primordium will form. The primordium depletes the surrounding area of auxin and thereby creates an ‘inhibitory field’ that defines the spacing between adjacent organs (Galvan-Ampudia et al., 2016). Yet, while auxin's role in establishing phyllotaxis is well understood, additional regulators of this process remain to be identified.

Veronica Gregis and her team study mechanisms regulating reproductive development, with a focus on genetic networks that determine the identity and architecture of reproductive meristems. They investigate how transcription factors drive changes in meristem identity, and how these transcriptional cascades intersect with hormone pathways. They became particularly interested in Arabidopsis REPRODUCTIVE MERISTEM (REM) transcription factors, which are predominantly expressed during inflorescence development. Although identifying the function of these transcription factors has been challenging because of their high redundancy (Mantegazza et al., 2014), Gregis' team discovered that REM34 and REM35 regulate gametophyte development (Caselli et al., 2019) as well as the transition from a vegetative to inflorescence meristem (Manrique et al., 2023).

Further analysis of REM34 and REM35 has been hampered by the lack of mutant alleles for REM35 and by the tight linkage of the two genes on chromosome 4 that effectively prevents the generation of double mutants by crossing. Postdoctoral researcher Francesca Caselli, first author of the highlighted publication, overcame this obstacle using CRISPR/Cas9 technology. Analysis of the newly generated CRISPR mutants showed that the phyllotactic pattern of siliques was altered in rem34 and rem35, with fewer siliques displaying the golden angle of 137.5°. This phenotype was exacerbated in the rem34 rem35 double mutant, with more angles falling into 80°–100° or 170°–190° intervals (Figure 1a). These observations suggest that the two REM genes jointly regulate phyllotaxis.

Caselli and colleagues then used a yeast-2-hybrid interaction screen to identify proteins that act in concert with REM34 and REM35. They identified AUXIN RESPONSE FACTOR 19 (ARF19) as an interactor of REM35, but not of REM34, and the same interaction pattern was observed for the closely related protein ARF7. REM34 and REM35 form both homo- and heterodimers (Caselli et al., 2019), and yeast-3-hybrid assays showed that REM35 can bridge interactions between REM34 and ARF7 as well as ARF7 and ARF19, but not between REM34 and ARF19. ARF7 and ARF19 act redundantly in lateral root development (Okushima et al., 2005) but have no previously described role in inflorescence development. However, Caselli et al. found that phyllotaxis was similarly perturbed in arf7 arf19 and rem34 rem35 double mutants, with even stronger perturbations in the arf7 arf19 rem34 rem35 quadruple mutant. Fluorescent reporter lines showed that all four genes are expressed in the inflorescence meristem, with their expression domains overlapping in the L1 and L2 layers, and in the boundary regions between the meristem and young flower primordia. Overall, these results suggest that REM-ARF protein complexes act in the inflorescence meristem to position new flower primordia.

Aside from a change in phyllotactic pattern, the rem34 rem35 and arf7 arf19 double mutants have an enlarged inflorescence meristem, while overexpression of REM35 reduced the meristem's size. Changes in meristem size are often associated with changes in cell cycle activity. Consistent with this, the S-phase marker Histone H4 showed a widened expression domain in the two double mutants, suggesting an increased number of cells in the S phase.

In the root, ARF7 and ARF19 directly activate the expression of the transcriptional regulators PUCHI and LATERAL BOUNDARY DOMAIN 18 (LBD18) to control lateral root emergence (Kang et al., 2013; Lee et al., 2019). Caselli et al. observed the expression of these two genes also in the inflorescence meristem. The PUCHI and LBD18 promoters harbour putative REM34 binding sites, and ChIP experiments showed that these sites are directly bound by both REM34 and REM35. The REM binding sites are located near auxin response elements bound by ARFs, further suggesting coordinated regulation of these genes by both sets of transcription factors. In agreement with this assumption, the expression patterns of PUCHI and LBD18 were affected in rem34 rem35 and arf7 arf19 mutants: LBD18 expression, which is confined to boundary regions in the wild type, was detected throughout the inflorescence meristem in the double mutants, while PUCHI expression was observed in very few flower primordia in the mutants, but in all primordia in the wild type. Loss-of-function mutants of LBD18 and PUCHI displayed altered phyllotaxis similar to that of rem34 rem35 and arf7 arf19, corroborating the hypothesis that these two genes act downstream of REM-ARF complexes in the control of inflorescence meristem architecture.

This study provides first evidence for a REM-ARF transcription factor complex that regulates inflorescence architecture. While REM34, REM35, ARF7 and ARF19 transcripts are present in the inflorescence meristem, their expression patterns only partially overlap and their function might depend on the presence or absence of their interaction partners. Gregis proposes that different REM-ARF dimers or trimers form simultaneously within distinct meristem domains and fulfil different roles to finetune boundary determination and, consequently, primordia positioning.