Fig 1.
Schematic representation of the RAM of A. thaliana.
(A) Longitudinal cross-section of the RAM. The different tissues of the RAM that we aimed to describe are indicated with different colors (bright for cells of the PD, and pale for cells of the TD). The distribution of auxin and CK along the longitudinal axis of the RAM is shown at the left. (B) Radial cross-section of the RAM showing the central pro-vascular tissues as the cellular domain that comprises the metaxylem and part of the procambium; and the peripheral pro-vascular tissues that includes the protoxylem, the phloem, part of the procambium and the pericycle. This drawing was made based on a confocal microscopy image of A. thaliana root tip.
Table 1.
Regulatory interactions included in the GHRN models.
Fig 2.
The genetic-hormonal regulatory networks of the RAM of A. thaliana.
Topologies of the (A) GHRN and (B) GHRN1 models. Activating regulatory interactions are represented with directed arrows and inhibitions with blunt arrows. The interactions that represent regulation of protein movement are indicated with dotted lines. Hypothetical interactions are shown with dashed lines; the blue interactions are the hypotheses proposed in this paper.
Fig 3.
Expected and recovered attractors of the cells at the RAM.
(A) Expected attractors. Each activity configuration corresponds to the characteristic genetic expression and hormonal activity profiles of the following cells within the RAM: QC [23,24,27,28,30,36,38,55,63,87,89], Endodermis [28,36,37,39,63,87,89], Peripheral pro-vascular tissues [30,38,64], Central pro-vascular tissues [4,26,29,30,38,42,75,90] and Root Cap [7,31,52–55,75]. The color code is as in Fig 1. Asterisks indicate that the activity of a node can be either 1 or 0. (B) The 11 fixed-point attractors recovered by the GHRN1 model are shown. Ten of the eleven recovered attractors match the expected activity configurations, which is not the case for the central pro-vascular TD3 attractor; for this attractor we indicated in blue the activities of the nodes that disagree with their expected activity.
Fig 4.
Activity of the ARFs that regulate WOX5 in the RAM of A. thaliana.
Inferred expression pattern of ARF10 [87,89] and ARF5 [83,87,89] in the RAM. The expression pattern of JKD [63] and MGP [36] is also shown. Notice the complementary expression patterns of ARF10 and JKD in the RAM, and ARF5 and MGP in the adjacent layer to the stele (delineated in blue).
Fig 5.
Robustness of the GHRN1 Boolean model to perturbations.
A) The histogram shows the frequency of attractor recovery for a population of 1,000 random networks perturbed 100 times. The dashed lines show the 95% quantile and (*) the frequency of recovery for the GHRN1 model with the same number of perturbations. B) For each node it is shown the percentage of conservation of the original attractors in the systematic perturbation of the Boolean function of each node (gray bars), and the number of rows in its truth table (blue line).
Fig 6.
Attractors recovered in the GOF and LOF simulations.
The results of the GOF (A) and LOF (B) simulations of all nodes are shown. The attractors that are not identical to the 11 original attractors are indicated with * (see Methods). The color code is as in Fig 1.
Fig 7.
Nonlinear coupling between the auxin signaling pathway and the gene regulatory network.
The model suggests that the readout of auxin could be mediated by complex interactions between its signaling pathway and the gene regulatory network.
Table 2.
Truth table for the logical rule x(t+1) = y(t) AND z(t).
Table 3.
Truth table of SCR with the rationale behind each output value.
Table 4.
Truth table of ARR1.