Roots are the anchorage organs of plants, responsible for water and nutrient uptake, exhibiting high plasticity. Root architecture is driven by the interactions of biomolecules, including transcription factors (TFs) and hormones that are crucial players regulating root plasticity. Multiple TF families are involved in root development; some, such as ARFs and LBDs, have been well characterized, whereas others remain less investigated. In this review, we synthesize the current knowledge about the involvement of the large family of homeodomain-leucine zipper (HD-Zip) TFs in root development. This family is divided into four subfamilies (I to IV), mainly according to structural features, such as additional motifs aside from HD-Zip, as well as their size, gene structure, and expression patterns. We explored and analyzed public databases and the scientific literature regarding HD-Zip TFs in Arabidopsis and other species. Most members of the four HD-Zip subfamilies are expressed in specific cell types and several ones from each group have assigned functions in root development. Notably, a high proportion of the studied proteins are part of intricate regulation pathways involved in primary and lateral root growth and development.

Methods used for the analysis of data from the repositories BAR and Root Atlas

Analysis of tissue-specific BAR data
Raw normalized expression levels for the HD-Zip genes and selected marker genes were downloaded from the BAR repository of the University of Toronto (http://bar.utoronto.ca/; Brady et al., 2007; Dinneny et al., 2008; Kilian et al., 2007) (Supplementary Table S1). As markers, we selected genes with known expression in a given tissue or cell type of the root. They were analyzed in the tissue for which they were chosen (image below). Their expression levels were averaged between samples (Supplementary Table S1). Expression was available for tissues in 13 regions. For clearness, we simplified this by collapsing them, i.e. averaging replicates and samples together as follows: regions 1 to 7 of the root tip were considered zone 1; 8 to 10, zone 2; and 11 to 13 zone 3. The inspection of marker distribution means expression in the corresponding tissues, which was dispersed, allowing us to define a conservative expression threshold of 50, for which most of the markers had at least one mean expression value above it. We then applied this threshold to HD-Zip genes mean expression, obtained by an analogous procedure, to classify TFs as expressed, or not, in each tissue or cell type (Supplementary Table S2). Mean expression of marker genes in the corresponding tissue in which they have reported as expressed. Horizontal lines show expression levels of 10, 50, and 100, of which 50 was chosen as a threshold.

Analysis of abiotic stress data from the BAR repository
Raw expression data for Abiotic Stress was downloaded from the BAR repository and averaged between replicates. Responses to different stress factors were the collapsed into Early responses, 15 and 30 min, 1, 3 and 4 hours after treatment; and Late responses, 6, 12 and 24 hours aftertreatment. To recover the expression peak in each time window, the maximum expression level, and not the average, was kept among mean expression values of all considered time points. We then applied the same threshold, defined for tissues (50), which was capable of identifying many HD-Zip I TFs known for their response to several of the abiotic stresses evaluated. Using this threshold, we classified all HD-Zip TFs as expressed, or not, in early or late stress responses (Supplementary Table S3).

Analysis data from the Root Atlas repository
Raw LogNormalized expression data for all HD-Zip genes and 7695 cells were downloaded from the Root Atlas repository (Zhang et al., 2019; http://wanglab.sippe.ac.cn/rootatlas; Supplementary Table S1). Provided cluster assignment for each cell allowed us to analyze cell expression distribution for each HD Zip gene in each of the 24 clusters. In this case, we defined a threshold based on the expression of HZ-Zip III TFs in roots according to the literature. We then defined a gene to be expressed in a cluster if a proportion of 0.03 of the cells in the cluster presented an expression level above 0.5. Based on this criterion, we classified HD-Zip genes as expressed, or not expressed, in each cluster (Supplementary Table S4).