A molecular framework for proximal secondary vein branching in the Arabidopsis thaliana embryo

The establishment of a closed vascular network in foliar organs is achieved through the coordinated specification of newly recruited procambial cells, their proliferation and elongation. An important, yet poorly understood component of this process, is secondary vein branching; a mechanism employed in Arabidopsis thaliana cotyledons to extend vascular tissues throughout the organ’s surface by secondary vein formation. To investigate the underlying molecular mechanism in vein branching, we analyzed at a single-cell level the discontinuous vein network of cotyledon vascular pattern 2 (cvp2) cvp2-like 1 (cvl1). Utilizing live-cell imaging and genetic approaches we uncovered two distinct branching mechanisms during embryogenesis. Similar to wild type, distal veins in cvp2 cvl1 embryos emerged from the bifurcation of cell files contained in the midvein. However, the branching events giving rise to proximal veins are absent in this mutant. Restoration of proximal branching in cvp2 cvl1 cotyledons could be achieved by increasing OCTOPUS dosage as well as by silencing of RECEPTOR LIKE PROTEIN KINASE 2 (RPK2) expression. The RPK2-mediated restriction of proximal branching is auxin and CLE-independent. Our work defines a genetic network conferring plasticity to Arabidopsis embryos to adapt the spatial configuration of vascular tissues to organ growth.

Developing cotyledons were imaged at 7 or 8 days as indicated. Cotyledons and leaves 192 were fixed with 3:1 ethanol: acetic acid, dehydrated in 80% ethanol, and then 100% 193 treated with 10% sodium hydroxide for 1hour at 37˚C and mounted in 50% glycerol.

194
Black and white images were taken and brightness and contrast were adjusted using 195 ImageJ. To visualize GUS staining, we used a staining buffer as previously described  To further corroborate these observations, we monitored the auxin efflux PIN1 protein, 221 whose distribution in torpedo embryos is restricted to procambial and protodermal cells 222 (Fig. 1F). We observed a progressive formation of the midvein concomitant with the veins appears to be different and follow a yet-to-be described mechanism.

242
In cvp2 cvl1 mutants GM cells fail to commit to procambial cell identity

243
To better understand the spatio-temporal arrangement of cotyledon vascular 244 formation during embryogenesis, we decided to exploit the discontinuous cvp2 cvl1   Fig. S3). However, we observed that proximal To gain further insight into the vein network branching defects observed in cvp2 296 cvl1 embryonic cotyledons, we decided first to assess the role of auxin and its PIN1- preventing the extension of procambial cell files into the cotyledon margin area.

360
The partial RPK2-mediated restoration of cvp2 cvl1 vascular phenotype seems 361 to be PIN1-independent To gain further insight into the mechanisms by which RPK2 modulates vein 363 patterning we decided to monitor the transcript profiles of wild-type, cvp2 cvl1 and 364 amiRPK2 cvp2 cvl1 embryos between the early torpedo and bent cotyledon stages.

365
Both mutants showed hundreds of differentially expressed genes (DEGs) when with additional distal branching points (Fig. 4). While these results indicated that polar 519 auxin cues contribute to repress distal branching, they cannot explain the reduced 520 cotyledon vein network complexity observed in cvp2 cvl1 nor its intermittent cotyledon 521 vein pattern. Our observations show that PIN1 polarity appears unaltered in cvp2 cvl1 522 veins (Fig. 3). Additionally, genetic increase of PIN1 dosage does not enhance the 523 branching defects observed in cvp2 cvl1 (Fig. 3), indicating that at least another 524 mechanism independent of PIN1 must be responsible for this phenotype. Previous independent of vascular-specific CLE peptides (Supporting Information Fig. S5), 546 inferring a differential mechanism between vascular formation in the shoot and the root.

547
Our results indicate that the negative RPK2-mediated control of vein patterning is                 . cvp2 cvl1 embryos exhibit aberrant divisions. A) Embryo morphology of WT compared to cvp2 cvl1 embryos using mSP-PI on ovules taken from green siliques and imaged using confocal microscopy. cvp2 cvl1 shows aberrant divisions at all stages of embryogenesis but most abundantly during the globular stage. Yellow arrows point to aberrant divisions, and scale bars are 20μm or 50μm. B) Graphical representation of the percent of aborted embryos occurring in cvp2 cvl1 vs. WT siliques numbered 2 and 3 counted from the apical meristem (containing globular stage embryos). C) Image of a dissected WT silique as compared to cvp2 cvl1 under bright field using a stereomicroscope. Scale bar represents 1inch. Fig. S4. Sensitivity to PIN1-mediated auxin transport is not disturbed in rpk2. A-F) Analysis of cleared leaves of seedlings grown for 4 days (until clear emergence of cotyledons could be detected) transferred to a media supplemented with 10 µM NPA or mock conditions for 5 days. n= 19-35 for each genotype. Scale bars: 500 µm. G-I) Cleared leaves of the indicated genotypes imaged with a stereomicroscope in bright field on a black background. n= 14-21 for each genotype. J-K') Confocal microscopy analysis of PIN1::PIN1-GFP distribution in cells of the root stele of WT and rpk2-2 seedlings. J' and K' represent magnification of the region shown in J and K. Scale bars represent 50µm in J,K and 20µm in J' and K'. L) Root length of seedlings grown as described in A-F) were measured. Note that root length was measured after the treatment with NPA or mock. The root length of seedlings treated with mock were set to 100% and the % shown in the graph represents the % of inhibition of root length by NPA. n=39-56. Analysis of the vein pattern in cleared leaves of 9-day-old seedlings transferred to a medium supplemented with the indicated CLE peptides once the emergence of the cotyledons could be detected. n= 16-46 for each genotype. Scale bars represent 500µm.