Mechanical stimuli modulate lateral root organogenesis

Abstract

Unlike mammals, whose development is limited to a short temporal window, plants produce organs de novo throughout their lifetime in order to adapt their architecture to the prevailing environmental conditions. The production of lateral roots represents one example of such postembryonic organogenesis. An endogenous developmental program likely imposes an ordered arrangement on the position of new lateral roots. However, environmental stimuli such as nutrient levels also affect the patterning of lateral root production. In addition, we have found that mechanical forces can act as one of the triggers that entrain lateral root production to the environment of the Arabidopsis (Arabidopsis thaliana) plant. We observed that physical bending of the root recruited new lateral root formation to the convex side of the resultant bend. Transient bending of 20 s was sufficient to elicit this developmental program. Such bending triggered a Ca(2+) transient within the pericycle, and blocking this change in Ca(2+) also blocked recruitment of new lateral root production to the curved region of the root. The initial establishment of the mechanically induced lateral root primordium was independent of an auxin supply from the shoot and was not disrupted by mutants in a suite of auxin transporters and receptor/response elements. These results suggest that Ca(2+) may be acting to translate the mechanical forces inherent in growth to a developmental response in roots.

Figure 1.

Induction of LR formation in curves of roots. A to C, LRs form on the convex side of curves of the primary root (arrowheads) induced by waving (A), gravitropism (B), and encountering barriers to growth (C). D, The site of LR formation can be shifted apically by bend-induced LR formation in gravitropically stimulated roots. The site of LR primordia formation relative to the tip of the root was monitored and compared with vertically grown controls (mean ± sd; n ≥ 25). Note the first LR to form due to bend induction was shifted approximately 200 μm toward the tip relative to plants growing vertically (asterisk, P = 0.0013, t test). These observations suggest that bend induction can recruit pericycle cells to founder pericycle cell fate at a younger age than normally seen in the root. E, Angle threshold for induction of LR to the convex side of a gravitropically induced bend. The position of the LR primordium was scored as being in the region predicted for normal vertical growth based on results shown in D (blue), on the convex side of the bend (red), or in both positions on the same root (green). The proportion of LRs to the convex side of a bend at 10° and above is significantly different from the proportion of left to right sidedness in vertical controls (asterisks, P < 0.05, t test) and in increased total number of LRs compared with vertical controls at 25° (P < 0.05, t test). Results represent means ± se; n = 25 roots from three separate experiments.

Figure 2.

Mechanically induced bends lead to LR formation without gravitropic stimulation. A, A decapped root still forms LRs on the convex side of bends in the main root axis upon bending in response to encountering a barrier to growth. Primary root (green), bend-related LRs produced during agravitropic growth (red), and bend-induced LR formed after the root cap had regenerated and gravitropic growth resumed (yellow) are shown. See Supplemental Figure S6B for non-color-coded version. n = 5; bar = 500 μm. B, Protocol for generating a bend in the main root axis. The phytagel is cut on each side of the root, and by sliding the lower gel sideways the main root axis is bent without reorienting the root tip or the aerial parts of the plant. C and D, LRs are elicited to form by this bending technique, as detectable by either staining for the induction of the early LR primordium marker DR5_pro:GUS (arrow in C) or imaging emerged LRs (arrowhead in D). Representative images of more than 50 plants are shown. Bars = 50 μm (C) and 1 mm (D). E, Quantification of the proportion of plants with emerged LRs in curve versus unbent control. Results represent means ± se; n ≥ 20 in four separate experiments.

Figure 3.

Curve-related LRs can form in positions not seen in unbent controls. A, Unbent wild-type control showing regular alternating LR formation (arrowheads) at approximately 1 mm spacing. This image is representative of more than 100 separate roots. B to D, Mechanical bending of the root leads to multiple LR production sites (asterisks) in patterns not seen in unbent controls, with LRs forming across the vasculature (B), multiple LRs forming throughout the curve (C), or multiple LRs emerging in very close proximity (D). Aberrant LR positioning was seen in 25% of curves. n > 100 roots. Bars = 200 μm.

Figure 4.

Effects of shoot removal on bend-induced LR formation. A, Schematic representation of surgical removal of the hypocotyl to deplete the root of an acropetal source of shoot-derived auxin and replacement of the auxin source with an agar block. B, Removal of the hypocotyl prevented bend-induced LR emergence but still supported bend-induced primordium formation. C, An agar block containing 100 μm 1-NAA was added to the cut surface before or after bending, as indicated, and LR emergence was scored. Note that removal of the hypocotyl blocks LR emergence but not primordium formation in the curve. Replacing the acropetal source of auxin with a 1-NAA-containing agar block restored LR emergence. The proportion of roots showing LR production in response to adding the agar block before or after bending was not significantly different (P > 0.05, t test). Results show means ± se; n ≥ 15 in four separate experiments.

Figure 5.

Effects of mutations in auxin transport/signaling on bend-induced LR formation. A, Plants of each genotype were subjected to bending, and the number of LR primordia plus emerged LRs was scored on the convex side of the bend (gray bars). Data expressed separately as emerged LRs or unemerged primordia are shown in Supplemental Figure S3. Results represent means ± se; n > 15 from more than five separate experiments. Letters represent results significantly different from bent wild-type (WT) controls (P < 0.05, t test). White bars indicate the effect each mutation has on normal LR frequency (calculated as in Supplemental Fig. S3B). Asterisks denote genotypes where bend-induced LRs are at a frequency significantly higher than the values predicted from their normal frequency (P < 0.05, t test). B, aux1-7 shows a wild-type-like frequency of LRs formed at the bend that was induced by growing into a barrier. Col WT, Columbia wild type. C, Roots of the tir1-1 mutant show LR formation in bends induced by growing into a barrier (1) and in curves elicited by the gravitropic response as the root grows past the end of the barrier (2). The results shown are representative of more than 50 separate plants. Bar = 1 mm.

Figure 6.

Effects of transient bending of the root on LR formation. Roots were bent for the indicated time periods and then returned to straight growth, and LR formation was assayed over the following 72 h. Asterisks indicate LR formation at levels significantly higher than the control (P < 0.05, t test). Results represent means ± se; n = 20 in four separate experiments.

Figure 7.

Effects of bending of the root on Ca²⁺ levels. A, Cytoplasmic Ca²⁺ levels were imaged using the YC3.6 Ca²⁺ sensor driven by the CaMV 35S promoter and calculated from the ratio of the sensor's Ca²⁺-dependent FRET signal/cyan fluorescent protein (CFP) signal. Ca²⁺ levels have been pseudocolor coded according to the scale at right. Note the increase in Ca²⁺ in cells under tension. B, Quantitative analysis of Ca²⁺-dependent ratio values from YC3.6 was made in the regions indicated in Supplemental Figure S7. C and D, Roots were pretreated for 5 min with or without 1 mm LaCl₃ and subjected to bending, and Ca²⁺ levels were analyzed as in A and B. Note that bending triggers Ca²⁺ increases in cells under tension. La³⁺ treatment blocks these Ca²⁺ changes elicited by bending the root. Results are representative of 15 separate experiments. E, La³⁺ treatment selectively blocks LR formation in bends of the root. Roots were pretreated for 5 min with or without 1 mm LaCl₃ and subjected to bending for 5 min and then returned to straight vertical growth in La³⁺-free medium. LR formation was scored after 72 h of further growth. Note that La³⁺ treatment selectively blocks LR recruitment to the convex side of the curve in the root but does not block all LR formation. Results show means ± se; n = 15 from four separate experiments.

Figure 8.

Multiple mechanical responses can have a significant effect on root system architecture. A root was grown into a field of horizontal barriers (coverslips) for 7 d according to Massa and Gilroy (2003), and the production of curve-related LRs was tracked. The primary root is color coded green, the initial bend-induced LR red, subsequent bend-induced LRs yellow, and barrier locations blue. Supplemental Figure S6A shows a non-color-coded version of this image. The results shown are representative of five separate experiments.

Figure 9.

Model of a possible role for Ca²⁺ in stretch-induced founder pericycle cell recruitment. A putative stretch-activated, La³⁺-sensitive Ca²⁺ channel responds to membrane tension from root bending and elicits a Ca²⁺-dependent signaling cascade operating in parallel with, and possibly interacting with, an auxin-dependent pathway to founder pericycle cell specification.