Directional transport of auxin is critical for inflorescence and floral development in flowering plants, but the role of auxin influx carriers (AUX1 proteins) has been largely overlooked. Taking advantage of available AUX1 mutants in Setaria viridis and maize, we uncover previously unreported aspects of plant development that are affected by auxin influx, including higher order branches in the inflorescence, stigma branch number, and glume (floral bract) development, and plant fertility.  However, disruption of auxin flux does not affect all parts of the plant, with little obvious effect on inflorescence meristem size, time to flowering, and anther morphology.  In double mutant studies in maize, disruptions of ZmAUX1 also affect vegetative development.  A GFP-tagged construct of Spp1 under its native promoter showed that the SPP1 protein localizes to the plasma membrane of outer tissue layers in both roots and inflorescences, and accumulates specifically in inflorescence branch meristems, consistent with the mutant phenotype and expected auxin maxima.  RNA-seq analysis finds that most gene expression modules are conserved between mutant and wildtype plants, with only a few hundred genes differentially expressed in spp1 inflorescences. Using CRISPR-Cas9 technology, we disrupted SPP1 and the other four AUX1 homologs in S. viridis.  SPP1 has a larger effect on inflorescence development than the others, although all contribute to plant height, tiller formation, leaf, and root development. The AUX1 importers are thus not fully redundant in S. viridis.  Our detailed phenotypic characterization plus a stable GFP-tagged line offer tools for future dissection of the function of auxin influx proteins.

Plant growth, phenotyping, and statistical comparisons

Setaria viridis accessions A10.1 and ME034V were grown in growth chamber and greenhouse conditions, respectively, following Acharya et al. (2017) and Zhu et al. (2018). The original spp1 mutation was isolated from an A10.1 background; ME034V was chosen for CRISPR confirmation of the mutant phenotype because of its high transformation efficiency. Plant height, leaf number, panicle length, and branch number were measured as described in Huang et al. (2017) and Zhu et al. (2018). Fertility was measured as the ratio of spikelets with a fully developed upper floret to total spikelets; bristles were ignored for fertility measurements. Tillers were counted at 37 days after sowing (DAS) and plant height measured at 40 DAS. Stigma and style number were assessed by dissecting one floret from each of ten spikelets per plant, five plants for spp1 and three for A10, for a total of 50 mutant florets and 30 wildtype.  We recorded number of styles and number of stigmas for each floret.  Numbers varied as much within as between plants so the numbers were pooled among all plants to estimate the frequency of each number. Histology and SEM followed Zhu et al. (2018). Inflorescence length, meristem width and height were measured using ImageJ (Schneider et al., 2012) from SEM photos.

For root phenotyping, sterilized seeds were grown either in Murashige and Skoog (MS) medium or germination pouches as described in Huang et al. (2017) and Acharya et al. (2017), respectively.

Auxin rescue experiments followed Marchant (1999) and Yu et al. (2015). 2, 4-D (from Plant Media, in 1mM stock with pure ethanol) and NAA (from Sigma-Aldrich, in 10mM stock with pure ethanol) were added to the medium to a final concentration of 0.1 mM.  MS medium containing 0.1% ethanol was used as a mock control. Seeds were grown on MS medium for three days and then transferred to media containing appropriate concentrations of auxin or mock for three more days. Root hairs were imaged at 4x magnification on a Leica DM750 microscope. Root hair number was counted in the focal plane on the side of the root facing the observer and normalized to root length. Experiments were repeated three times.

In maize, zmaux1 mutant plants were crossed to vt2, bif2 and Bif4 mutants, and F2 segregating populations were grown in the field in Columbia, Missouri in 2017. Plants were genotyped to identify single and double mutants using primers listed in Table S10 and were phenotyped at the eighth week. For the dominant mutant Bif4, both heterozygotes and homozygotes were included for mutant phenotyping analysis. For each mutant and mutant combination we assessed traits of the tassel (length from flag leaf to tassel tip, number of branches, spikelets on main spike, spikelet number per cm) and ear (kernel number, ear row number), and three vegetative traits (height of flag leaf, number of leaves above the lowest elongated internode, and number of tillers).

All pairwise comparisons used Welch’s t-test as implemented in R (R Core Team, 2020).  Single, double and higher order mutants were compared to each other and to wildtype by one-way or two-way Type I or Type II ANOVA as appropriate, followed by Tukey’s HSD test using standard programs in R (R Core Team, 2020).  Comparisons with p>0.05 were considered non-significant.

A ReadMe file has been included with the uploaded data. The files are organized according to the figure numbers in the manuscript.  Each figure number has its own directory, within which are one or more excel files (.xls) presenting the raw measurements.  These are unprocessed although they have been transferred to the excel files in a format for easy conversion to a dataframe compatible with R. Also within each directory are multiple .txt files that present the output of the statistical analyses, whether t-tests, ANOVA, or linear.  Mean and s.d. and p values from the t-test or ANOVA are reported in the appropriate supplemental tables in the manuscript.  Most are also presented as box plots in the figures in the paper.