# Source paper archive - Source title: Spatial Decoupling of the Endodermal Functions Establishes the Innermost Cortex as the Gravity Sensor in Rice Stems - Author(s): Kashyap, R., Singh, Z., Kumar, A., Gupta, Y., Kumar, A., Singh, S., Jouannic, S., Bergougnoux, V., Prasad, K., Gantet, P., Yadav, S. R. - Source landing URL: https://www.biorxiv.org/content/10.64898/2026.10.31.763306v1 - PDF URL: https://www.biorxiv.org/content/10.64898/2026.10.31.763306v1.full.pdf - PDF SHA-256: 5ec22daa6dc3f0b5d41ec1e33a7e3540749b3f861fdc0fbb8c58dea9bd3fb089 - Converter: pymupdf4llm 0.0.27 - Limits: Automated PDF text extraction; reading order, equations, tables and formatting may be incomplete or incorrect. Images are omitted, and no OCR was performed. Markup and links are neutralized for safe downloading. Consult the original PDF for the authoritative version. --- [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 **Title:** Spatial Decoupling of the Endodermal Functions Establishes the Innermost Cortex as 2 the Gravity Sensor in Rice Stems 3 4 **Authors:** 5 Rohan Kashyap [1], Zeenu Singh [1,5], Akshay Kumar [1,5], Yash Gupta [1], Amit Kumar [1], Shruti Singh [1], 6 Stefan Jouanic [2], Véronique Bergounioux [3], Kalika Prasad [4], Pascal Gantet [2] and Shri Ram 7 Yadav [1,] 8 9 **Author’s Affiliations:** 10 1 Department of Biosciences and Bioengineering, Indian Institute of Technology Roorkee, 11 Roorkee–247667, Uttarakhand, India 12 13 2 UMR DIADE, Université de Montpellier, IRD, CIRAD, 911 Avenue Agropolis, 34394 14 Montpellier cedex 5, France 15 16 3 Czech Advanced Technology and Research Institute (CATRIN), Palacký University 17 Olomouc, Czechia 18 19 4 School of Biology, Indian Institute of Science Education and Research, Pune-411008, 20 Maharashtra, India 21 5 22 These authors contributed equally 23 24 - **Corresponding author** 25 Dr. Shri Ram Yadav 26 Department of Biosciences and Bioengineering 27 Indian Institute of Technology Roorkee 28 Roorkee-247667, Uttarakhand, India 29 Phone: +91-1332-284782 30 [E-mail: shri.yadav@bt.iitr.ac.in](mailto:shri.yadav@bt.iitr.ac.in) 31 32 **Keywords:** Auxin, Endodermis, Gravity response, Rice stem, Statocyte 33 34 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 35 **Abstract** 36 Gravitropism is a remarkable example of coordinated growth, enabling plants to perceive 37 gravity and generate asymmetric cell expansion that restores organ orientation. In _Arabidopsis,_ 38 gravity sensing in shoots occurs within starch-rich endodermis cells. In rice, however, the stem 39 must accommodate gravitropic bending while preserving integrity of tissues surrounding 40 emerging crown roots, raising the possibility that gravity sensing relies on distinct structural 41 and physiological adaptations. Here, using high-resolution imaging together with auxin 42 response analyses, we identify a specialized cell layer at the inner boundary of the rice stem 43 cortex, operationally defined by its position adjacent to the lignified endodermis and associated 44 with gravitropic responses. This layer is characterized by circumferentially elongated cells, 45 starch sedimentation, a confined auxin-response domain, and expression of gravity-sensing 46 markers. Following gravistimulation, starch granules realign and the auxin response domain 47 expands asymmetrically within this layer, coinciding with differential growth during stem 48 bending. Our results further indicate that this response depends on polar auxin transport and 49 conserved components of the gravitropic regulatory network. Together, these findings support 50 a model in which gravity sensing and signaling in rice stems are spatially separated from the 51 lignified endodermal barrier. Such tissue organization may represent a developmental 52 adaptation that allows the stem to reconcile barrier integrity with growth flexibility. 53 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 54 **Introduction** 55 Plants are sessile organisms whose growth and responses to the environment rely on tightly 56 coordinated patterns of cell division and expansion within tissues constrained by rigid cell 57 walls. To support diverse physiological and developmental functions, specialized cell types are 58 organized into distinct tissues and organs. A striking example is the endodermis. In roots, the 59 endodermis forms a selective apoplastic barrier that seals the central vasculature and regulates 60 the movement of water and solutes ( _1_ – _4_ ). In _Arabidopsis_ shoots, the endodermis is commonly 61 referred to as the starch-sheath ( _5_ ). In several species, including _Arabidopsis_, starch-filled 62 amyloplasts sediment within these cells in response to gravity, providing the mechanical cue 63 that initiates gravitropic growth ( _6_ – _8_ ). As a result, barrier formation and gravity sensing have 64 traditionally been viewed as complementary functions of homologous tissues in different 65 organs. Whether this organizational framework is broadly conserved across flowering plants, 66 however, remains an open question. 67 68 In monocot species such as rice ( _Oryza sativa_ L.), the stem serves as a central developmental 69 hub, coordinating the formation of crown roots and tillers, and thereby shaping overall shoot 70 and root systems architectures. The stem base, which is often submerged and is in direct contact 71 with water, must function as a selective interface for water and nutrients uptake. Continuous 72 emergence of post-embryonic crown roots therefore requires maintenance of an uninterrupted 73 apoplastic barrier to prevent uncontrolled entry of water and solute into the vascular system. 74 At the same time, the shoot must perceive and respond to gravity, an essential process for tiller 75 angle determination and lodging resistance ( _9_ – _16_ ). These competing structural and 76 physiological demands raise a fundamental question about tissue organization in species with 77 a lignified stem endodermis as it is the case in rice ( _17_ ). Lignification imparts the hydrophobic 78 properties and mechanical strength required for endodermal barrier function ( _8_, _18_ ), whereas 79 gravitropic responses depend on highly dynamic processes including statolith sedimentation, 80 auxin redistribution, and differential growth ( _19_ – _21_ ). The tissues in the rice stem base display 81 dynamic changes of auxin response ( _19_ – _21_ ), however, it remained unclear whether auxin 82 signaling is spatially organized within this developmental framework. In roots, these functions 83 are spatially separated between the lignified endodermis and gravity-sensing columella cells 84 ( _19_ – _21_ ). In rice seedlings, although amyloplast-sedimentation and presence of Casparian bands 85 have been analyzed ( _22_ – _24_ ), how gravity perception and signaling is organized within the stem 86 in the presence of a lignified endodermis remains poorly understood. 87 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 88 Here, we reveal a previously unrecognized auxin-signaling domain at the stem base of rice that 89 lies immediately adjacent to the lignified endodermis. Based on its topological position 90 overlying the endodermis, we designate this tissue as the innermost cortical cell layer (IC). 91 Using confocal imaging, developmental analysis, physiological assays, data mining from 92 available spatial transcriptomics ( _25_ ) and RNA _in situ_ hybridization, we show that the IC 93 possesses hallmark features of gravity-sensing tissue, including localized auxin responses, 94 statolith accumulation, and enrichment of gravitropism-associated genes. Gravistimulation 95 induces dynamic, asymmetric auxin redistribution within this layer, while pharmacological and 96 genetic analyses demonstrate contribution from polar auxin transport and conserved 97 asymmetric growth regulators ( _26_, _27_ ). Together, our findings support a model in which barrier 98 function and gravitropic signaling are partitioned between adjacent cell layers in the rice stem 99 base, revealing a developmental organization that reconciles the maintenance of a robust 100 endodermal barrier with the cellular flexibility required for gravitropic growth. 101 102 **Results** 103 **Endodermal Lignification Coincides with Auxin Signaling in the IC Layer** 104 In rice and other cereals, post-embryonic development is characterized by the continuous 105 emergence of crown roots, tiller buds, and leaves from the stem (Fig. 1 _A,_ Fig. S1) ( _28_ ). 106 Throughout this process the stem maintains a conserved radial organization in which the central 107 stele is enclosed by a single-layered lignified endodermis that separates it from the surrounding 108 cortex (Fig. 1 _B_ ) ( _23_ ). To identify potential sites of auxin response, we monitored the 109 _DR5rev_ ::erGFP reporter activity ( _29_, _30_ ). Stem sections were co-stained with Basic Fuchsin 110 (BaFu) to visualise the lignified tissues and Fluorescent Brightener 28 (FB28) to reveal overall 111 cell morphology (Fig. 1 _C‒E_ ). Strikingly, we detected strong _DR5_ reporter activity in vascular 112 tissues and at the distal tips of crown root primordia (CRPs) (Fig. 1 _C‒C(ii)_ ). 113 114 In addition, we observed a highly localized _DR5_ signal in the cell layer situated immediately 115 external to the lignified endodermis (Fig. 1 _C(iii)_ ). Since this is the innermost layer of cortex 116 adjacent to the endodermis, we refer to it as the innermost cortical cell layer (IC). Although 117 endodermis-specific markers were not examined, the lignin deposition pattern on the radial cell 118 walls was consistent with Casparian strip formation during early stages of endodermal 119 lignification (Fig. 1 _D_ and _E_ ). As development progressed, these cells became extensively 120 lignified beyond the Casparian strips (Fig. 1 _C_ ( _iii_ )). These anatomical features agree with 121 previous descriptions of endodermal lignification in the mesocotyl and first internode of rice [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 122 seedlings ( _23_ ), supporting the topological assignment of this tissue. Together, these 123 observations identify a previously unrecognized auxin response domain in the IC that emerges 124 concurrently with endodermis lignification, suggesting that auxin responses are spatially 125 organized immediately adjacent to the lignified endodermis of the rice stem. 126 127 **Auxin Signaling and Circumferential Growth Anisotropy Coincide in the IC layer** 128 To determine how this spatial organization is established during development, we analyzed 129 serial sections of 6-day-old stem bases and defined three developmental stages (Fig. 2 _A‒C,_ 130 Fig. S2). During Stage I, the endodermis was not stained by BaFu suggesting that it remained 131 unlignified (Fig. 2 _A_ ). Although CRP had already initiated at this stage, _DR5_ activity was visibly 132 undetectable in the IC cells (Fig 2 _A, Fig._ S2 _D_ ). In Stage II, the endodermis lignification, 133 revealed by BaFu, coincided with the appearance of localized auxin response both at the distal 134 tips of CRPs and within the IC layer (Fig. 2 _B,_ Fig. S2 _D_ ). By Stage III, the lignified stem 135 endodermis formed a continuous structural connection with the endodermis of the emerging 136 crown roots, while _DR5_ activity persisted in the adjacent IC cells (Fig 2 _C,_ Fig. S2 _D_ ). To 137 examine whether auxin perception follows a similar spatial pattern, we analyzed the DII-Venus 138 auxin input reporter ( Fig. S3) ( _31_ ). During Stage I, the DII-Venus fluorescence was relatively 139 uniform throughout the cortex. As development progressed through Stages II and III, however, 140 DII-Venus fluorescence became stronger in the outer cortex, indicating lower auxin 141 accumulation in these cells. This suggests although varying levels of auxin accumulates 142 broadly across the cortex, auxin response is predominantly confined to the IC layer. 143 144 Because cell morphology often reflects tissue specialization, we next quantified it using 145 MorphoGraphX ( _32_ ). We measured _DR5_ signal intensity, along with major axis length and 146 major axis theta (MAθ), which defines the orientation of cell elongation relative to radial axis 147 of the stem (Fig. 2 _D–F,_ Fig. S4). Before endodermal lignification (Stage I), no GFP signal 148 above the background level was detected across the cortex, and MAθ values showed substantial 149 variation. With the onset of lignification in Stage II, both _DR5_ activity and MAθ increased 150 sharply within the IC layer. By Stage III, the MAθ peak became more pronounced, indicating 151 that IC cells had adopted a circumferential growth orientation whereas cells in the outer cortex 152 remained predominantly elongated along the radial axis. Despite this pronounced anisotropy, 153 IC cells underwent only a modest increase in major axis length (46%), compared with 191% 154 increase in the outer cortical cells (Fig. 2 _F_ ). Together, our results show that the emergence of 155 localized auxin response is tightly coupled to the acquisition of a distinctive circumferential [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 156 cell geometry in the IC layer, defining this tissue as a morphologically and functionally 157 specialized domain within the rice stem (Fig. 2 _G_ ). 158 159 **The IC Layer Exhibits a ‘Statocyte-like’ Identity** 160 In _Arabidopsis_ hypocotyls, the unlignified endodermis forms a starch sheath in which starch 161 filled amyloplasts (statoliths) sediment to initiate gravity sensing and ultimately promote auxin 162 redistribution to the adjacent tissues for asymmetric growth ( _6_ – _8_, _33_, _34_ ). In roots, by contrast, 163 the heavily lignified endodermis functions primarily as an apoplastic barrier, whereas gravity 164 sensing is delegated to the columella cells, which also exhibit strong _DR5_ activity ( _3_, _19_ – _21_ ). 165 We therefore hypothesized that rice stems similarly partition barrier and gravity-sensing 166 functions between neighboring tissues. Previous studies have reported an endodermis/starch 167 sheath derived from the ground meristem at the rice stem base ( _35_ ). Consistent with this, high 168 magnification bright-field imaging and Lugol staining revealed dense statolith accumulation in 169 the IC cells, whereas the adjacent endodermis lacked detectable starch granules (Fig. 3 _A_ and 170 _B,_ Fig. S5). 171 172 To further assess the “statocyte-like” identity of the IC layer, we analyzed the high-resolution 173 spatial transcriptomic data from the rice stem base ( _25_ ). Because the reference atlas did not 174 resolve a distinct endodermal cluster, we used the annotated ‘Pericycle-like’, a layer just inner 175 to the endodermis, and ‘Ground tissue’ cell populations as spatial approximations 176 encompassing the stele-cortex boundary, the endodermis, and the surrounding ground tissue. 177 Radial binning of these cells revealed significant enrichment of core gravitropism-associated 178 genes in the innermost bins, which likely correspond to the IC layer (Fig. 3 _C–E, Table S1_ ). 179 These included the gravity-signaling components _OsLAZY6, OsLAZY2,_ and _OsLAZY4_ ( _36_ ), 180 together with auxin transporters ( _OsPIN1a/b_ ), auxin signaling genes ( _OsIAA20_, _OsIAA30_, and 181 _OsARF17_ ), and _OsWOX12_, a regulator of cell expansion during gravitropic bending (27). We 182 validated the expression of selected genes by RNA _in situ_ hybridization. Notably, _OsLAZY2_, 183 _OsLAZY4_, and _OsWOX12_ showed a specific ring-like expression pattern surrounding the 184 vasculature at the stem base (Fig. 3 _F–H’,_ Fig. S6). The expression pattern of these 3 genes 185 overlaps with the expression domain of _OsLAZY1,_ a key regulator of polar auxin transport and 186 shoot gravitropism in rice ( _9_ ). Taken together, the topological position of IC layer adjacent to 187 the lignified endodermis, combined with its localized auxin response, statolith accumulation, 188 circumferential growth anisotropy, and enrichment of gravitropism-related genes, defines this 189 layer as a distinct tissue with statocyte-like characteristics. Although definitive layer-specific [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 190 transcript localization using cell-type-specific reporters will be required to confirm the precise 191 cellular origin of these transcripts, the convergence of anatomical, physiological, and 192 transcriptomic evidence strongly supports the IC as the principal gravity-associated signaling 193 domain in the rice stem. 194 195 **Spatial Separation of Barrier Function and Gravitropic Signaling in the Rice Stem** 196 To test whether the IC and the lignified endodermis perform distinct physiological functions, 197 we next examined their respective roles in barrier formation and gravitropic signaling (Fig. 4). 198 We first assessed the barrier function of the lignified endodermis using the _crown rootless1_ 199 ( _crl1_ ) mutant, which lacks CRPs and therefore retains a continuous endodermis at the stem 200 base (Fig. 4 _A,_ Fig. S7) ( _37_ ). In a propidium iodide uptake assay, the dye was excluded by the 201 lignified boundary, confirming that the endodermis functions as an effective apoplastic barrier 202 independent of crown root emergence (Fig. 4 _B_ and _C,_ Fig. S8). 203 204 We next investigated the role of the IC during gravity response using _DR5rev_ ::erGFP seedlings 205 subjected to a 90° gravistimulation (Fig. 4 _D,_ and S10 _A–F_ ). Gravitropism in the stem involves 206 gravity perception, asymmetric auxin distribution, and differential growth ( _24_ ). Because the IC 207 combines statoliths and a localized auxin response domain, we hypothesized that it contributes 208 to the early stages of this process. Based on this, we first checked the sedimentation of the 209 amyloplasts 1 hour after gravity stimulation (1 HAGS). We observed that the starch granules 210 re-aligned along the direction of the gravity after reorienting seedlings 90° (Fig. S9). At 0 211 HAGS, the _DR5_ activity in the IC closely followed the lignified endodermis boundary (Fig. 212 4 _E–E’’_ ). Within 2 HAGS, well before visible stem bending occurred (Fig. 4 _D_ ), the _DR5_ 213 activity on the lower flank extended apically beyond original alignment with the lignified 214 endodermis (Fig. 4 _F–F’’_ ). As the response progressed, this asymmetric auxin-response domain 215 expanded further in both the apical and lateral directions across multiple cortical cell layers on 216 the lower flank of the stem (Fig. 4 _G–G’’,_ Fig. S10 _G–I’_ ). By 30 HAGS, the _DR5_ activity had 217 further extended upto the base of the coleoptile (the first leaf), a tissue previously implicated 218 in gravity sensing in rice seedlings (Fig. S11 _A_ and _A’_ ) ( _16_, _22_, _24_ ). Notably, a strong 219 independent _DR5_ signal was also evident in the lower flank of the coleoptile by 6 HAGS, 220 suggesting a coordinated gravitropic response in stem and coleoptile (Fig. S11 _B_ ). 221 222 To determine whether this asymmetric auxin response was accompanied by the changes in 223 cellular organization, we quantified cell-shape at the coleoptile-stem junction using [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 224 MorphoGraphX (dotted white square in Fig. S11 _A_ ). We measured both _DR5_ activity and the 225 MAθ at the 30 HAGS. On the upper flank, the _DR5_ peak became broader and less pronounced, 226 whereas on the lower flank the auxin-response domain expanded basally, consistent with its 227 asymmetric distribution (Fig. S11 _C_ ). This spatial asymmetry was mirrored by changes in the 228 cell shape. In the upper flank, MAθ peaked within the IC, while the cells in the remaining 229 cortex averaged approximately 60°. In contrast, cells in the lower flank exhibited an average 230 MAθ of approaching 80°, indicating enhanced circumferential growth anisotropy coincident 231 with the expanding auxin response domain (Fig. S11 _C_ ). Together, these physiological analyses 232 provide functional support for a spatial separation of barrier function and gravitropic signaling 233 within the rice stem. While the lignified endodermis retains its role as a robust apoplastic 234 barrier, the adjacent IC exhibits the dynamic auxin redistribution and growth responses 235 expected of a gravity-responsive tissue, supporting a model in which these complementary 236 functions are partitioned between neighboring cell layers. 237 238 **Polar Auxin Transport and** _**OsWOX12**_ **Coordinate Asymmetric Stem Growth after** 239 **Gravistimulation** 240 To investigate how the IC drives gravitropic responses, we combined pharmacological and 241 genetic approaches. Because polar auxin transport is a central regulator of gravitropism in both 242 shoots and roots ( _20_ ), we first asked whether it is required for the asymmetric auxin response 243 associated with the IC. Seedlings were treated with the auxin transport inhibitor 1-N 244 naphthylphthalamic acid (NPA) and _DR5_ activity was monitored following gravistimulation. 245 Compared with the DMSO control, NPA-treated seedlings showed a marked reduction in 246 gravitropic bending that persisted throughout the 24 h observation period (Fig. 5 _A–B_ ). By 24 247 HAGS, control seedlings had reached an average bending angle of approximately 60°, whereas 248 NPA-treated seedlings bent only about 25° (Fig. 5 _B–C’_ ). Consistent with this impaired 249 response, the asymmetric _DR5_ domain normally expanded along the lower flank in both the 250 apical and lateral directions was largely abolished by NPA treatment (Fig. 5 _D_ and _D’_ ). These 251 results indicate that polar auxin transport is required to establish the asymmetric auxin 252 distribution that emerges from the IC during the early stages of gravitropism. 253 254 We next asked how this localized auxin response is translated into differential growth. Previous 255 studies identified the transcription factors _OsWOX11_ and _OsWOX6/12_ as redundant regulators 256 acting downstream of auxin signaling to promote asymmetric stem growth ( _27_ ). To assess the 257 contribution of this pathway, we examined the gravitropic response of the _wox12_ mutant after [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 258 transferring to dark conditions. Although the mutant seedlings displayed a clear delay in 259 bending relative to wild-type during the first 12 HAGS, this defect gradually diminished, and 260 by 24 HAGS the bending angles of the two genotypes were comparable (Fig. 5 _E–F_ ). Cellular 261 analysis of the lower flank further revealed that cells at the stem-coleoptile junction, where the 262 _DR5_ domain had expanded at 12 HAGS, displayed shorter major axes in the _wox12_ mutant 263 than in the wild type (Fig. 5 _G-H_, Fig. S12). Together, these findings place the IC within a 264 conserved gravitropic regulatory framework. Polar auxin transport is required to establish the 265 asymmetric auxin-response domain associated with the IC, while _OsWOX12_ contributes to the 266 differential cell expansion that follows. These coordinated processes ultimately drive 267 asymmetric stem growth and gravitropic bending. Together, these findings support a model in 268 which IC contributes to gravitropic bending by engaging conserved components of the 269 gravitropic pathway, particularly polar auxin transport, to drive asymmetric stem bending and 270 suggest that this asymmetric stem bending is regulated at least in part by _OsWOX12_ (Fig. 5 _I_ ). 271 272 **Discussion** 273 Gravity acts continuously on all plant cells, yet it is perceived only by specialized cells known 274 as statocytes, which detect the direction of gravity and initiate downstream physiological 275 responses ( _38_ ). In angiosperms, starch-rich statocytes are typically located in the endodermis 276 in shoots and the columella in roots, where sedimentable amyloplasts function as statoliths 277 ( _39_ ). An endodermis can inclusively be defined as a cell layer that has any two of the three 278 criteria; (i) presence of Casparian strips, (ii) topologically the cell layer surrounding the 279 vasculature, and (iii) expression of endodermis-specific genes such as SCR or its recognized 280 ortholog ( _3_ ). In _Arabidopsis_, the non-lignified starch-rich endodermis regulates shoot 281 gravitropism and branch angle ( _3_, _6_, _39_ ). In grasses, however, gravity perception has mainly 282 been associated with the basal region of the coleoptile and leaf sheath ( _16_, _22_, _24_ ). Here, we 283 provide evidence that, in rice stems, gravity-associated signaling is concentrated in a distinct 284 starch sheath layer, which we designate the inner cortex (IC), immediately adjacent to the 285 lignified endodermis. 286 287 This finding revises the classical dual-function model in which the endodermis can act as an 288 apoplastic barrier in root and the gravity-sensing tissue in shoot. Classical anatomical studies 289 have noted that the walls of starch sheath of many stems remain unlignified ( _17_ ). Our results 290 support this distinction in rice stems, where the lignified endodermis appears specialized for 291 barrier formation, the neighboring IC displays structural, transcriptional, and physiological [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 292 characteristics consistent with a gravity-responsive tissue. Thus, barrier function and gravity 293 sensing appear to be partitioned into adjacent cell layers, allowing the stem to maintain both 294 apoplastic integrity and gravitropic responsiveness. A previous study reported that an actin 295 binding Rice Morphology Determinant (RMD) protein regulates light-dependent shoot 296 gravitropism by regulating amyloplast sedimentation in the endodermal cells, however in 297 absence of a complete anatomical tissue context or endodermis-associated marker, its identity 298 as endodermis remained unestablished ( _40_ ). 299 300 Current models propose that gravity perception begins with statolith sedimentation, followed 301 by LAZY protein repolarization, PIN-mediated auxin redistribution, and differential cell 302 expansion ( _24_, _29_, _33_, _34_, _38_ – _41_ ). These processes require dynamic intracellular organization 303 and cellular flexibility, properties that may be incompatible with the rigid, lignified walls of 304 the mature rice endodermis. In contrast, the IC provides a more permissive environment. IC 305 cells sediment amyloplasts, express gravitropism-associated _LAZY_ genes, display localized 306 _DR5_ activity, and show enhanced auxin response after gravistimulation. Although the lack of 307 cell-type-specific markers and mutants prevents direct demonstration of gravity perception at 308 single-cell resolution, the convergence of these anatomical, molecular, and physiological 309 features strongly supports the IC as a statocyte-like tissue associated with gravitropic signaling. 310 311 The developmental sequence also supports this interpretation. CRPs already initiate from the 312 innermost ground tissues of the stem during stage I, before extensive endodermal lignification 313 occurs and IC-specific DR5 activity is established, however at stage II when CRPs progress 314 for morphogenesis, the onset of auxin response coincides in IC and cap cells of the CRPs, 315 raising a possibility that the IC and root cap tissues may share components of a common 316 developmental program. Whether these tissues are developmentally related or evolutionarily 317 homologous remains to be determined. This also raises yet another possibility whether the IC 318 cells, in addition to gravity response, also contributes to root cap formation in the CRPs, similar 319 to the case of lateral root primordia morphogenesis in a few species of Cucurbitaceae and 320 Leguminosae families ( _42_, _43_ ). 321 322 Although gravity-associated signaling appears to originate in the IC, the growth response likely 323 involves coordinated activity across surrounding tissues. Following gravistimulation, the DR5 324 domain expands both apically and laterally before visible bending occurs. This differs from the 325 predominantly lateral auxin redistribution reported in _Arabidopsis_ hypocotyls, where PIN [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 326 proteins redirect auxin laterally toward the epidermis ( _33_, _34_ ). In rice, apical propagation may 327 facilitate communication with younger, more extensible tissues near the stem apex and 328 coleoptile base, coordinating differential growth across the organ. Consistent with this idea, 329 later stages of bending were accompanied by asymmetric cell elongation and differentiation of 330 new lignified tissues on the responding flank. However, because tissue mechanics were not 331 measured directly, the relationship between auxin redistribution and differential growth 332 remains inferential. 333 334 Our findings also extend current models of gravity sensing in grasses. The bases of coleoptile 335 and leaf sheath have been regarded as the principal gravitropic organ in rice seedlings ( _16_, _22_, 336 _24_ ), whereas we now report the IC for gravity response in the rice stem. Previous studies 337 showed that _OsLAZY2_ and _OsLAZY3_ regulate shoot gravitropism and tiller angle through a 338 starch-statolith pathway in leaf sheath base cells, whereas _OsLAZY4_ acts through a parallel 339 pathway. Both converge upstream of _OsLAZY1_ to establish asymmetric auxin distribution and 340 activate _OsWOX6/12_ and _OsWOX11_ ( _16_, _27_, _44_, _45_ ). Rather than replacing this model, our 341 results suggest that the rice stem IC is an additional site of the same gravitropic network. During 342 gravistimulation, auxin response expanded from the stem IC laterally toward the coleoptile 343 base, indicating that these tissues function together as an integrated gravity-sensing system. 344 Our spatial expression pattern analysis reveals that though the gravity-sensing genes, _OsLAZY2_ 345 and _OsLAZY4_ express in the multiple cortical layers, similar to _OsLAZY1_ ( _9_ ), the auxin 346 response and its downstream effector ( _OsWOX12_ ) are activated in a more confined IC layer. 347 348 The ontological origin and molecular basis of IC specification remains unknown. In 349 _Arabidopsis_, the conserved SHORT-ROOT (SHR) pathway specifies the shoot starch sheath 350 endodermis ( _6_, _46_ ), whereas in monocot roots, including maize and _Setaria_, SHR protein 351 movement extends across multiple cortical layers ( _47_ ). It is therefore plausible that related 352 developmental mechanisms specify both the endodermis and IC in rice stems, partitioning 353 endodermal functions between neighboring tissues. Testing this hypothesis will require direct 354 genetic and cell-biological analyses and should clarify the developmental identity of the IC. 355 356 Our findings may also have evolutionary implications. Early monocot lineages are thought to 357 have diversified in aquatic or semi-aquatic environments ( _48_ – _50_ ). In such conditions, increased 358 endodermal lignification may have strengthened the apoplastic barrier while limiting its 359 capacity to support dynamic gravitropic bending. The evolution of a neighboring starch sheath, [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 360 represented here by the IC, could therefore have preserved gravity sensing without 361 compromising barrier function. Although speculative, this scenario provides a framework for 362 understanding how functional specialization of stem tissues may have diverged between 363 monocots and eudicots. Finally, these findings have practical significance for crop architecture. 364 Shoot gravitropism is a major determinant of tiller angle, a trait that strongly influences 365 planting density, canopy architecture, and grain yield in rice ( _9_ – _12_, _16_, _26_, _27_, _51_ ). Identifying 366 the IC as a potential site of gravity-associated signaling expands the current framework of rice 367 shoot gravitropism and provides a foundation for future studies aimed at manipulating cereal 368 architecture through tissue-specific regulation of gravity responses. 369 370 **Materials and Methods** 371 _**Plant material and growth conditions**_ 372 The _DR5rev_ ::erGFP and _wox12_ lines used in the study are in _Oryza sativa_ L. (cv. Taipei309). 373 Mutant for _wox12_ was generated using CRISPR-Cas9, two guide RNAs were designed in the 374 first exon and were cloned in the _p_ RGEB32 vector following Xie et al. 2015 ( _52_ ). The construct 375 was transformed into _Agrobacterium tumefaciens_ LBA4404, and was used for plant 376 transformation as described previously ( _30_, _53_ ). The _crl1_ mutant line is in the cv. Taichung 65 377 background. Seeds were surface-sterilized with 70% ethanol for 90 seconds and 40% (v/v) Rin 378 ALA (commercial bleach) for 20 minutes, followed by repeated washes with sterile water. The 379 seeds were soaked in water for 24 hours and then placed in ½ Murashige and Skoog (MS) 380 media (TS1004, Himedia) containing 1% (w/v) sucrose (Himedia) and 0.05% (w/v) MES 381 (Himedia), solidified with 0.35% (w/v) phytagel (Sigma), and adjusted to pH 5.8. Media was 382 poured into glass jam bottles or magenta boxes. Seedlings were grown at 28°C under a 16 h 383 light:8 h dark photoperiod. For NPA treatment, a 1 mM stock of Naptalam (Sigma) was 384 prepared in DMSO, filtered-sterilized, and added to the autoclaved media before pouring. 385 386 _**Vibratome sectioning and confocal microscopy**_ 387 All tissues were fixed in 4% paraformaldehyde (Sigma) in 1× PBS overnight at 4°C, after 20 388 minutes of vacuum infiltration. The tissues were then washed once with 1× PBS and embedded 389 in 4% agarose (SeaKem) for vibratome sectioning using a Leica VT1000S. Sections of 70-80 390 μm were transferred to the ClearSee solution for 2 days ( _54_ ). The sections were subsequently 391 dual-stained with 0.1% Calcofluor White (Himedia) and 0.2% Basic Fuchsin (Himedia) 392 prepared in ClearSee ( _55_ ), and mounted in ClearSee for confocal microscopy. Imaging was 393 performed using a Leica Stellaris 5 microscope with 63×/1.40 oil–immersion and 20×/0.75 dry [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 394 objectives. Image acquisition parameters were set to 1 Airy unit, a 1024×1024-pixel frame size, 395 line averaging of 2, and a bidirectional scan speed of 400 Hz. Due to the large tissue sizes, all 396 macroscopic images were acquired as tile scans. Fluorescent Brightener 28 (FB28) was excited 397 with a 405 nm laser at 0.87% power (gain: 2.5; emission range: 425–483 nm. Basic Fuchsin 398 was excited with a 561 nm laser at 10.86% power (gain: 23.4; emission range: 600–650 nm). 399 erGFP was excited with a 488 nm laser at 12.31% power (gain: 26; emission range: 503–535 400 [nm). The images were scaled and processed using Fiji software (https://imagej.net/) and Leica](https://imagej.net/) 401 LasX software ( _56_ ). 402 403 _**MorphoGraphX analysis**_ 404 The TIFF files exported from Fiji were used for the MorphoGraphX (MGX) analysis ( _32_ ). 405 Briefly, imported stacks were blurred with a Gaussian kernel and used for edge detection at a 406 threshold of 1. The flat plane of the stack was used to create an overlying mesh, which was 407 segmented and smoothed to a target size of 500 nm. The signal from the FB28 channel was 408 projected onto the mesh from a single z-slice. Cells were manually seeded, followed by 409 watershed segmentation. All the segmentations were manually checked and corrected for any 410 errors. Major axis theta and signal intensities were quantified using the heatmap function after 411 projection of the respective channels. The values were then exported as CSV files for plotting. 412 413 _**Gravity stimulation experiment**_ 414 Seedlings were grown in magenta boxes as described above. Four days post-sterilization, the 415 boxes were rotated 90° and transferred to the dark. Seedlings were imaged every 6 hours, and 416 corresponding tissues were collected for sectioning. The angle of bending was quantified using 417 the angle tool Fiji. 418 419 _**Propidium iodide uptake experiment**_ 420 _crl1_ mutant (cv. Taichung65) seedlings were grown as described above. The stem bases of 6 421 day-old seedlings were collected and immersed in a 15 μg.mL [-1] solution of propidium iodide 422 (PI, Sigma) in water. The stems were vacuum-infiltrated with the PI solution for 20 minutes 423 and incubated for 1 hour at 30°C. Tissues were then washed with water and embedded in 4% 424 agarose. Sections of 200 μm were cut using a Leica VT1000S vibratome and mounted in water. 425 Imaging was performed using a 20x/0.75 DRY objective, followed by a 63x/1.40 oil 426 immersion objective. Image acquisition parameters were set to 1 Airy unit, a 1024x1024 pixel 427 frame size, line averaging 2, and a scan speed of 400 Hz, with bidirectional scanning. PI was [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 428 excited with a 514 nm laser at 2% power (gain: 26.4; emission range: 550–740 nm). Lignin 429 autofluorescence was captured using a 405 nm laser at 15% power (gain: 51.3; emission range: 430 425–469 nm). 431 432 _**Spatial transcriptomics data re-analysis**_ 433 To analyze the spatial expression patterns of the canonical gravitropic regulatory modules, raw 434 transcriptomic data and corresponding cell-type annotations were retrieved from STOmics DB 435 [(http://db.cngb.org/stomics) under the STT0000026 identifier (](http://db.cngb.org/stomics) _25_, _57_ ). The spatial analysis was 436 restricted to the cells annotated as “Pericycle_like” and “Ground_tissue”. To establish a 437 standardized radial axis across varying stem cross–sections, the spatial coordinates for each 438 sample were scaled, and the developmental center of the stem base was defined by calculating 439 a spatial centroid of the pericycle cell population. The Euclidean distance of each cell from this 440 centroid was calculated to determine its absolute radial position. These radial distances were 441 min-max normalized and were partitioned into 10 equally spaced bins. A set of 20 locus IDs ( 442 Table S1) associated with the gravitropic response was used for the analysis. Raw transcript 443 counts (MIDCounts) were normalized to counts per million (CPM). A subsequent 444 log 2 (CPM+1) transformation was applied to stabilize variance across the dataset. For each 445 spatial bin, log 2 (CPM+1) expression values were averaged per sample, followed by the 446 calculation of a global mean expression profile per spatial bin. To generate the heatmap using 447 the pheatmap R package, the global mean expression for each gene was standardized into row 448 wise Z-scores across the 10 radial bins. For individual gene trajectories, the mean expression 449 and standard error of the mean (±SEM) were calculated across all cells within each bin and 450 visualized using the ggplot2 package. To quantify the magnitude of spatial segregation within 451 the inner cortex (Bins 1 and 2) relative to the outer (Bins 3–10), a linear fold-change was 452 calculated as the ratio of inner and outer expression values, which was log2-transformed to 453 generate the final reported _log2_ FC values. To assess the statistical significance of these spatial 454 differences, raw p-values were calculated using Fisher’s exact tests to compare the proportion 455 of expressing versus non–expressing cells between the inner and the outer radial bins. These 456 p-values were corrected for multiple hypothesis testing using the Benjamini-Hochberg False 457 [Discovery Rate (FDR). R version 4.4.1 was used for all the statistical analyses (https://www.R-](https://www.r-project.org/) 458 [project.org/).](https://www.r-project.org/) 459 460 _**RNA in situ hybridization**_ [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 461 Gene-specific fragments of _OsLAZY2, OsLAZY4,_ and _OsWOX12_ were cloned into pBluescript 462 SK+, linearized with HindIII, and transcribed with T3 RNA Polymerase (Roche) for DIG 463 labeled antisense probes. EcoRI linearized clones were used to synthesize the sense probes with 464 T7 RNA Polymerase (Roche). Stem bases of 6-day-old seedlings were fixed in 4% 465 paraformaldehyde (Sigma-Aldrich) in 1X PBS overnight. The tissues were dehydrated using a 466 graded ethanol series (30%, 50%, 70%, and 85%) with 130 mM NaCl (Sigma-Aldrich). This 467 was followed by a final incubation with 95% and 100% ethanol. 10 μm-thick sections were 468 taken on poly-L-lysine-coated slides and were hybridized as described previously. Briefly, the 469 slides were dewaxed with xylene (Himedia), hydrated in an ethanol series, acidified, and 470 treated with 1 μg mL [-1] proteinase K (Sigma-Aldrich). After re-fixing in 4% paraformaldehyde, 471 the slides were hybridized with DIG-labeled probes overnight at 55°C in hybridization 472 chambers, soaked in 50% formaldehyde solution (Rankem). The slides were then washed and 473 treated with 10 μg mL [-1] RNase (Himedia) to degrade unbound probes. 0.5% Boehringer 474 Mannheim reagent was used for blocking, prior to anti-DIG antibody (Sigma-Aldrich) 475 treatment. The signal was detected using NBT-BCIP (Sigma-Aldrich). Dehydrated slides were 476 then mounted in Entellan (Sigma-Aldrich) and were imaged using a Leica Visoria B 477 microscope with 4X and 10X objectives. 478 479 **Acknowledgments** 480 The Department of Biotechnology (DBT), Government of India, the Anusandhan National 481 Research Foundation (ANRF), Government of India, and Indo-French Centre for the 482 Promotion of Advanced Research (IFCPAR/CEFIPRA)/Department of Science and 483 Technology, Government of India, are acknowledged for financial support to the projects of 484 S.R.Y.’s laboratory. Fellowships are provided by the Council of Scientific & Industrial 485 Research (CSIR), India, to R.K., by the Department of Biotechnology (DBT), India, to S.S., 486 and Amit K. and by IIT Roorkee to Y.G., and A.K. We acknowledge the Ashok Soota 487 Molecular Medicine Facility at the Department of Biosciences and Bioengineering, IIT 488 Roorkee, for access to the confocal microscope. [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 489 **References** 490 1. J. A. Raven, D. Edwards, Roots: evolutionary origins and biogeochemical 491 significance. _Journal of Experimental Botany_ **52**, 381–401 (2001). 492 2. S. Naseer, Y. Lee, C. Lapierre, R. Franke, C. Nawrath, N. Geldner, Casparian strip 493 diffusion barrier in Arabidopsis is made of a lignin polymer without suberin. _Proc_ 494 _Natl Acad Sci U S A_ **109**, 10101–10106 (2012). 495 3. N. Geldner, The Endodermis. _Annu. Rev. Plant Biol._ **64**, 531–558 (2013). 496 4. M. Barberon, The endodermis as a checkpoint for nutrients. _New Phytologist_ **213**, 497 1604–1610 (2017). 498 5. A. M. F. Tomescu, The stele – a developmental perspective on the diversity and 499 evolution of primary vascular architecture. _Biol Rev_ **96**, 1263–1283 (2021). 500 6. H. Fukaki, J. Wysocka‐Diller, T. Kato, H. Fujisawa, P. N. Benfey, M. Tasaka, 501 Genetic evidence that the endodermis is essential for shoot gravitropism in 502 _Arabidopsis thaliana_ . _The Plant Journal_ **14**, 425–430 (1998). 503 7. K. Fujihira, T. Kurata, M. K. Watahiki, I. Karahara, K. T. Yamamoto, An 504 Agravitropic Mutant of Arabidopsis, endodermal-amyloplast less 1, That Lacks 505 Amyloplasts in Hypocotyl Endodermal Cell Layer. _Plant and Cell Physiology_ **41**, 506 1193–1199 (2000). 507 8. M. T. Morita, T. Kato, K. Nagafusa, C. Saito, T. Ueda, A. Nakano, M. Tasaka, 508 Involvement of the Vacuoles of the Endodermis in the Early Process of Shoot 509 Gravitropism in Arabidopsis. _Plant Cell_ **14**, 47–56 (2002). 510 9. P. Li, Y. Wang, Q. Qian, Z. Fu, M. Wang, D. Zeng, B. Li, X. Wang, J. Li, LAZY1 511 controls rice shoot gravitropism through regulating polar auxin transport. _Cell Res_ 512 **17**, 402–410 (2007). 513 10. T. Yoshihara, M. Iino, Identification of the Gravitropism-Related Rice Gene 514 LAZY1 and Elucidation of LAZY1-Dependent and -Independent Gravity 515 Signaling Pathways. _Plant and Cell Physiology_ **48**, 678–688 (2007). 516 11. X. Wu, D. Tang, M. Li, K. Wang, Z. Cheng, Loose Plant Architecture1, an 517 INDETERMINATE DOMAIN Protein Involved in Shoot Gravitropism, Regulates 518 Plant Architecture in Rice. _Plant Physiology_ **161**, 317–329 (2012). 519 12. M. Okamura, T. Hirose, Y. Hashida, T. Yamagishi, R. Ohsugi, N. Aoki, Starch 520 reduction in rice stems due to a lack of OsAGPL1 or OsAPL3 decreases grain 521 yield under low irradiance during ripening and modifies plant architecture. 522 _Functional Plant Biology_ **40**, 1137–1146 (2013). 523 13. Z. Li, Y. Liang, Y. Yuan, L. Wang, X. Meng, G. Xiong, J. Zhou, Y. Cai, N. Han, 524 L. Hua, G. Liu, J. Li, Y. Wang, OsBRXL4 Regulates Shoot Gravitropism and Rice 525 Tiller Angle through Affecting LAZY1 Nuclear Localization. _Molecular Plant_ **12**, 526 1143–1156 (2019). [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 527 14. Y. Hu, S. Li, X. Fan, S. Song, X. Zhou, X. Weng, J. Xiao, X. Li, L. Xiong, A. 528 You, Y. Xing, _OsHOX1_ and _OsHOX28_ Redundantly Shape Rice Tiller Angle by 529 Reducing _HSFA2D_ Expression and Auxin Content. _Plant Physiol._ **184**, 1424–1437 530 (2020). 531 15. H. Li, H. Sun, J. Jiang, X. Sun, L. Tan, C. Sun, _TAC4_ controls tiller angle by 532 regulating the endogenous auxin content and distribution in rice. _Plant_ 533 _Biotechnology Journal_ **19**, 64–73 (2021). 534 16. L. Huang, W. Wang, N. Zhang, Y. Cai, Y. Liang, X. Meng, Y. Yuan, J. Li, D. Wu, 535 Y. Wang, _LAZY2_ controls rice tiller angle through regulating starch biosynthesis in 536 gravity‐sensing cells. _New Phytologist_ **231**, 1073–1087 (2021). 537 17. J. L. Seago, Revisiting the occurrence and evidence of endodermis in angiosperm 538 shoots. _Flora_ **273**, 151709 (2020). 539 18. L. J. Gibson, The hierarchical structure and mechanics of plant materials. _J. R. Soc._ 540 _Interface._ **9**, 2749–2766 (2012). 541 19. S. Sabatini, D. Beis, H. Wolkenfelt, J. Murfett, T. Guilfoyle, J. Malamy, P. Benfey, 542 O. Leyser, N. Bechtold, P. Weisbeek, B. Scheres, An Auxin-Dependent Distal 543 Organizer of Pattern and Polarity in the Arabidopsis Root. _Cell_ **99**, 463–472 544 (1999). 545 20. J. Friml, J. Wiśniewska, E. Benková, K. Mendgen, K. Palme, Lateral relocation of 546 auxin efflux regulator PIN3 mediates tropism in Arabidopsis. _Nature_ **415**, 806–809 547 (2002). 548 21. M. Taniguchi, M. Furutani, T. Nishimura, M. Nakamura, T. Fushita, K. Iijima, K. 549 Baba, H. Tanaka, M. Toyota, M. Tasaka, M. T. Morita, The Arabidopsis LAZY1 550 Family Plays a Key Role in Gravity Signaling within Statocytes and in Branch 551 Angle Control of Roots and Shoots. _Plant Cell_ **29**, 1984–1999 (2017). 552 22. K. Abe, H. Takahashi, H. Suge, Localization of Cells Containing Sedimented 553 Amyloplasts in the Shoots of Normal and Lazy Rice Seedlings. _Biol. Sci. Space_ **8**, 554 221–225 (1994). 555 23. H. Watanabe, M. Saigusa, S. Morita, Identification of Casparian Bands in the 556 Mesocotyl and Lower Internodes of Rice ( _Oryza sativa_ L.) Seedlings Using 557 Fluorescence Microscopy. _Plant Production Science_ **9**, 390–394 (2006). 558 24. W. Wang, H. Gao, Y. Liang, J. Li, Y. Wang, Molecular basis underlying rice tiller 559 angle: Current progress and future perspectives. _Molecular Plant_ **15**, 125–137 560 (2022). 561 25. L. Zhong, L. Geng, Y. Xiang, X. Guang, L. Cao, J. Shi, W. Li, J. Wang, W. He, L. 562 Huang, F. Yang, Y.-X. Bai, S. K. Sahu, X. Guo, S. Zhang, G. Zhang, X. Xu, F. Hu, 563 W. Yang, H. Liu, Y. Zhao, J. Lyu, Comparative spatial transcriptomics reveals root 564 dryland adaptation mechanism in rice and HMGB1 as a key regulator. _Molecular_ 565 _Plant_ **18**, 797–819 (2025). [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 566 26. L. Abas, M. Kolb, J. Stadlmann, D. P. Janacek, K. Lukic, C. Schwechheimer, L. A. 567 Sazanov, L. Mach, J. Friml, U. Z. Hammes, Naphthylphthalamic acid associates 568 with and inhibits PIN auxin transporters. _Proc. Natl. Acad. Sci. U.S.A._ **118**, 569 e2020857118 (2021). 570 27. N. Zhang, H. Yu, H. Yu, Y. Cai, L. Huang, C. Xu, G. Xiong, X. Meng, J. Wang, 571 H. Chen, G. Liu, Y. Jing, Y. Yuan, Y. Liang, S. Li, S. M. Smith, J. Li, Y. Wang, A 572 Core Regulatory Pathway Controlling Rice Tiller Angle Mediated by the _LAZY1_ 573 Dependent Asymmetric Distribution of Auxin. _Plant Cell_ **30**, 1461–1475 (2018). 574 28. J.-I. Itoh, K.-I. Nonomura, K. Ikeda, S. Yamaki, Y. Inukai, H. Yamagishi, H. 575 Kitano, Y. Nagato, Rice Plant Development: from Zygote to Spikelet. _Plant and_ 576 _Cell Physiology_ **46**, 23–47 (2005). 577 29. J. Friml, E. Benková, I. Blilou, J. Wisniewska, T. Hamann, K. Ljung, S. Woody, 578 G. Sandberg, B. Scheres, G. Jürgens, K. Palme, AtPIN4 Mediates Sink-Driven 579 Auxin Gradients and Root Patterning in Arabidopsis. _Cell_ **108**, 661–673 (2002). 580 30. T. Garg, Z. Singh, K. Chennakesavulu, K. K. K. Mushahary, A. K. Dwivedi, V. 581 Varapparambathu, H. Singh, R. S. Singh, D. Sircar, D. Chandran, K. Prasad, M. 582 Jain, S. R. Yadav, Species-specific function of conserved regulators in 583 orchestrating rice root architecture. _Development_, dev.200381 (2022). 584 31. G. Brunoud, D. M. Wells, M. Oliva, A. Larrieu, V. Mirabet, A. H. Burrow, T. 585 Beeckman, S. Kepinski, J. Traas, M. J. Bennett, T. Vernoux, A novel sensor to 586 map auxin response and distribution at high spatio-temporal resolution. _Nature_ 587 **482**, 103–106 (2012). 588 32. P. Barbier De Reuille, A.-L. Routier-Kierzkowska, D. Kierzkowski, G. W. Bassel, 589 T. Schüpbach, G. Tauriello, N. Bajpai, S. Strauss, A. Weber, A. Kiss, A. Burian, 590 H. Hofhuis, A. Sapala, M. Lipowczan, M. B. Heimlicher, S. Robinson, E. M. 591 Bayer, K. Basler, P. Koumoutsakos, A. H. Roeder, T. Aegerter-Wilmsen, N. 592 Nakayama, M. Tsiantis, A. Hay, D. Kwiatkowska, I. Xenarios, C. Kuhlemeier, R. 593 S. Smith, MorphoGraphX: A platform for quantifying morphogenesis in 4D. _eLife_ 594 **4**, e05864 (2015). 595 33. H. Rakusová, J. Gallego‐Bartolomé, M. Vanstraelen, H. S. Robert, D. Alabadí, M. 596 A. Blázquez, E. Benková, J. Friml, Polarization of PIN3‐dependent auxin transport 597 for hypocotyl gravitropic response in _Arabidopsis thaliana_ . _The Plant Journal_ **67**, 598 817–826 (2011). 599 34. H. Rakusová, M. Abbas, H. Han, S. Song, H. S. Robert, J. Friml, Termination of 600 Shoot Gravitropic Responses by Auxin Feedback on PIN3 Polarity. _Current_ 601 _Biology_ **26**, 3026–3032 (2016). 602 35. Y. Coudert, Thi, Van Anh Le, P. Gantet, “Rice: A Model Plant to Decipher the 603 Hidden Origin of Adventitious Roots” in _Plant Roots: The Hidden Half_, A. Eshel, 604 T. Beeckman, Eds. (CRC Press, ed. 4, 2013; 605 https://www.taylorfrancis.com/books/9781439846490/chapters/10.1201/b14550606 15), pp. 164–173. [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 607 36. Y. Liang, Y. Yuan, Y. Cai, M. Ma, X. Lin, X. Sun, H. Zhao, J. Zhai, N. Han, J. 608 Zhou, W. Wang, Y. Wang, N. Zhang, HSFA2D – LAZY6–LAZY1 module 609 regulates shoot gravitropism and tiller angle in rice. _New Phytologist_ **247**, 625–636 610 (2025). 611 37. Y. Inukai, T. Sakamoto, M. Ueguchi-Tanaka, Y. Shibata, K. Gomi, I. Umemura, 612 Y. Hasegawa, M. Ashikari, H. Kitano, M. Matsuoka, Crown rootless1, which is 613 essential for crown root formation in rice, is a target of an AUXIN RESPONSE 614 FACTOR in auxin signaling. _Plant Cell_ **17**, 1387–1396 (2005). 615 38. X. Wang, X. Cai, H. Huo, H. Chen, Gravity sensing in plants. _New Phytologist_, 616 nph.71324 (2026). 617 39. N. Kawamoto, M. T. Morita, Gravity sensing and responses in the coordination of 618 the shoot gravitropic setpoint angle. _New Phytologist_ **236**, 1637–1654 (2022). 619 40. Y. Song, G. Li, J. Nowak, X. Zhang, D. Xu, X. Yang, G. Huang, W. Liang, L. 620 Yang, C. Wang, V. Bulone, Z. Nikoloski, J. Hu, S. Persson, D. Zhang, The Rice 621 Actin-Binding Protein RMD Regulates Light-Dependent Shoot Gravitropism. 622 _Plant Physiol._ **181**, 630–644 (2019). 623 41. M. T. Morita, Directional Gravity Sensing in Gravitropism. _Annu. Rev. Plant Biol._ 624 **61**, 705–720 (2010). 625 42. T. T. Xiao, R. Van Velzen, O. Kulikova, C. Franken, T. Bisseling, Lateral root 626 formation involving cell division in both pericycle, cortex and endodermis is a 627 common and ancestral trait in seed plants. _Development_, dev.182592 (2019). 628 43. H. Singh, Z. Singh, R. Kashyap, S. R. Yadav, Lateral root branching: evolutionary 629 innovations and mechanistic divergence in land plants. _New Phytologist_ **238**, 630 1379–1385 (2023). 631 44. Y. Cai, L. Huang, Y. Song, Y. Yuan, S. Xu, X. Wang, Y. Liang, J. Zhou, G. Liu, J. 632 Li, W. Wang, Y. Wang, LAZY3 interacts with LAZY2 to regulate tiller angle by 633 modulating shoot gravity perception in rice. _Plant Biotechnology Journal_ **21**, 634 1217–1228 (2023). 635 45. W. Wang, L. Huang, Y. Song, S. Gui, J. Cao, H. Zhang, M. Du, J. Chen, Z. Wang, 636 J. Zhou, X. Meng, D. Zeng, J. Li, Y. Wang, LAZY4 acts additively with the 637 starch–statolith-dependent gravity-sensing pathway to regulate shoot gravitropism 638 and tiller angle in rice. _Plant Communications_ **5**, 100943 (2024). 639 46. E. K. Yoon, S. Dhar, M.-H. Lee, J. H. Song, S. A. Lee, G. Kim, S. Jang, J. W. 640 Choi, J.-E. Choe, J. H. Kim, M. M. Lee, J. Lim, Conservation and Diversification 641 of the SHR-SCR-SCL23 Regulatory Network in the Development of the 642 Functional Endodermis in Arabidopsis Shoots. _Molecular Plant_ **9**, 1197–1209 643 (2016). 644 47. C. Ortiz-Ramírez, B. Guillotin, X. Xu, R. Rahni, S. Zhang, Z. Yan, P. Coqueiro 645 Dias Araujo, E. Demesa-Arevalo, L. Lee, J. Van Eck, T. R. Gingeras, D. Jackson, 646 K. L. Gallagher, K. D. Birnbaum, Ground tissue circuitry regulates organ 647 complexity in maize and _Setaria_ . _Science_ **374**, 1247–1252 (2021). [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 648 48. M. R. Duvall, G. H. Learn, L. E. Eguiarte, M. T. Clegg, Phylogenetic analysis of 649 rbcL sequences identifies Acorus calamus as the primal extant monocotyledon. 650 _Proc. Natl. Acad. Sci. U.S.A._ **90**, 4641–4644 (1993). 651 49. M. W. Chase, Monocot relationships: an overview. _American J of Botany_ **91**, 652 1645–1655 (2004). 653 50. T. Shi, C. Huneau, Y. Zhang, Y. Li, J. Chen, J. Salse, Q. Wang, The slow-evolving 654 Acorus tatarinowii genome sheds light on ancestral monocot evolution. _Nat. Plants_ 655 **8**, 764–777 (2022). 656 51. G. Tian, L. Gao, Y. Kong, X. Hu, K. Xie, R. Zhang, N. Ling, Q. Shen, S. Guo, 657 Improving rice population productivity by reducing nitrogen rate and increasing 658 plant density. _PLoS ONE_ **12**, e0182310 (2017). 659 52. K. Xie, B. Minkenberg, Y. Yang, Boosting CRISPR/Cas9 multiplex editing 660 capability with the endogenous tRNA-processing system. _Proc. Natl. Acad. Sci._ 661 _U.S.A._ **112**, 3570–3575 (2015). 662 53. S. Toki, N. Hara, K. Ono, H. Onodera, A. Tagiri, S. Oka, H. Tanaka, Early 663 infection of scutellum tissue with _Agrobacterium_ allows high-speed transformation 664 of rice. _The Plant Journal_ **47**, 969–976 (2006). 665 54. D. Kurihara, Y. Mizuta, Y. Sato, T. Higashiyama, ClearSee: a rapid optical 666 clearing reagent for whole-plant fluorescence imaging. _Development_ **142**, 4168– 667 4179 (2015). 668 55. R. Ursache, T. G. Andersen, P. Marhavý, N. Geldner, A protocol for combining 669 fluorescent proteins with histological stains for diverse cell wall components. _The_ 670 _Plant Journal_ **93**, 399–412 (2018). 671 56. J. Schindelin, I. Arganda-Carreras, E. Frise, V. Kaynig, M. Longair, T. Pietzsch, S. 672 Preibisch, C. Rueden, S. Saalfeld, B. Schmid, J.-Y. Tinevez, D. J. White, V. 673 Hartenstein, K. Eliceiri, P. Tomancak, A. Cardona, Fiji: an open-source platform 674 for biological-image analysis. _Nat Methods_ **9**, 676–682 (2012). 675 57. Z. Xu, W. Wang, T. Yang, L. Li, X. Ma, J. Chen, J. Wang, Y. Huang, J. Gould, H. 676 Lu, W. Du, S. K. Sahu, F. Yang, Z. Li, Q. Hu, C. Hua, S. Hu, Y. Liu, J. Cai, L. 677 You, Y. Zhang, Y. Li, W. Zeng, A. Chen, B. Wang, L. Liu, F. Chen, K. Ma, X. 678 Xu, X. Wei, STOmicsDB: a comprehensive database for spatial transcriptomics 679 data sharing, analysis and visualization. _Nucleic Acids Research_ **52**, D1053–D1061 680 (2024). 681 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 682 **Main Figures and Legends** 683 684 685 686 **Fig 1. Endodermal lignification coincides with auxin signaling in the innermost cortical** 687 **cell layer.** _(A),_ A 6-day-old rice seedling with crown roots (CR) originating from the stem, 688 background removed for clarity (Scale bar: 0.2 cm). _(B),_ An illustration showing the general 689 anatomy of the rice stem with stele (St), endodermis (En), cortex (Co), and epidermis (Ep). 690 _(C),_ A section of the stem of a 6-day-old seedling with 3 major auxin signaling domains marked 691 ( _(i)_ to _(iii)_ ; Scale bar: 200 μm). _(i)_, _DR5_ signal in the stelar vasculature of the stem (yellow 692 asterisk; Scale bar: 20 μm). _(ii)_, _DR5_ signal at the tip of the developing crown root primordia 693 (CRP, yellow asterisk; Scale bar: 50 μm). _(iii),_ 3D projection of the lignified En (pink 694 arrowhead) along with the adjacent _DR5_ signal in the innermost cortical cell layer (yellow 695 arrowhead; Scale bar: 50 μm). _(D)_, A Section of a 3-day-old stem base of _DR5rev_ ::erGFP lines 696 with the En and the IC (Scale bar: 50 μm). _(E),_ Dashed yellow square from _D_, an En cell 697 showing the lignin deposition in a ‘casparian strip’ like fashion on the radial walls (dotted lines, 698 Scale bar: 5 μm). FB28: Fluorescent Brightener 28, BaFu: Basic Fuchsin. 699 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 700 701 **Fig 2. Auxin signaling and circumferential growth anisotropy overlap in the IC.** _(A),_ 702 Representative section of a stem at stage I marked with initiating CRP (white-dotted line), 703 unlignified endodermis (En*), and a lack of _DR5rev_ ::erGFP signal in the IC (Scale bar: 100 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 704 μm). _(B)_, Stem section at stage II with _DR5_ signaling at the distal tip of the CRP, lignified En, 705 and the appearance of the auxin response domain in the IC (Scale bar: 200 μm). _(C)_, Stem 706 section at stage III, at which the En of the stem becomes continuous with the En of the CR, and 707 the auxin signaling in the IC persists (Scale bar: 200 μm). _(D–D’),_ Cellular segmentation of the 708 rice stem bases at Stage III showing raw _DR5rev_ ::erGFP signal intensities _(D)_ and the Major 709 Axis Theta _(D’)_ (MAθ, angle in degrees between the Major Axis, MA, of the cell and the 710 reference direction, RD). The RD is measured along the stem radius (En: Endodermis, Ep: 711 Epidermis; IC: Innermost Cortical Cell Layer, Scale bar: 50 μm). _(E)_, Graphical representation 712 of normalized _DR5rev_ ::erGFP signal intensities (Left Y-axis) and MAθ (°, right Y-axis) of 713 each cell plotted against normalized cell coordinates (0 = En and 1 = Ep, n = 6 for each stage, 714 _**** p_ < 0.0001 from Kruskal–Wallis test with Dunn’s multiple comparisons on the MAθ) _._ 715 _(F),_ Graphical representation of major axes of the cells of the IC (normalized cell coordinates 716 between 0.05 and 0.15) and outer cortex (Co, 0.70 to 0.80) (*** _p_ = 0.0003 and **** _p <_ 717 0.0001 from Kruskal-Wallis test with Dunn’s multiple comparisons). _(G)_, An illustration 718 showing the cell shape anisotropy of the cortical cells, and localized auxin response in the IC, 719 adjacent to the lignified En. 720 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 721 722 723 **Fig 3. The Cells of the IC have a ‘statocyte-like’ identity.** _(A–B)_, Sections of _DR5rev_ ::erGFP 724 lines stained with Basic Fuchsin (BaFu) with trans-PMT channel, _(B)_ is the cropped image 725 from the dashed white square of _(A)_, note the occurrence of statoliths (white asterisks) in the 726 cells that display the _DR5_ signal. The pink and yellow arrowheads mark the En and the IC, 727 respectively. _(C)_, The annotated tissue identities (‘Pericycle_like’ and ‘Ground_tissue’) on a 728 spatial map of the rice stem base. The annotated cells were grouped into 10 radial bins, ordered 729 from the innermost to the outermost relative to the centroid. The raw data used for this analysis 730 were obtained from Zhong et al. (2025) (25), CRP: Crown Root Primordia, and St: Stele. _(D),_ 731 Heatmap showing the expression pattern of genes involved in shoot gravitropic response in [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 732 shoots, divided into three processes: gravity sensing, auxin redistribution, and asymmetric 733 growth (first column), displayed across all 10 radial bins. Expression values are shown as row 734 wise z-score calculated from log 2 (CPM+1). The genes were hierarchically clustered by rows. 735 _(E),_ Line plots, mean ± SEM log 2 (CPM+1) expression across radial bins (1 to 10) for 736 representative genes. Log 2 FC and FDR were calculated by comparing bins 1 and 2 against all 737 other bins. _(F–J)_, RNA _in situ_ hybridization using antisense probes of _OsLAZY4 (F)_, _OsLAZY2_ 738 _(G)_, _OsWOX12.1 (H–J)_, note the band-like expression pattern of the genes (black arrowhead) 739 in _F_ and _G_ . _(I),_ Dashed black rectangle from _H. (J),_ Dashed black rectangle from _I_, note the 740 specific expression pattern not only in the IC but also in the bundle sheath (Scale bar: 200 μm 741 for _F_ (inset) _, G_ (inset) _,_ and _H_ ; 100 μm for _F, G,_ and I; and 50 μm for _J_ ). 742 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 743 744 745 **Fig 4.** **Spatial Separation of Stem Barrier Function and Gravitropic Signaling.** _(A)_, 746 Representative seedlings of 6-day-old wild-type (cv. Taichung 65) and _crl1_ mutant, note the 747 absence of crown roots in the mutant (white arrowhead in the WT, Scale bar: 1 cm). _(B)_, 748 Propidium Iodide (PI) uptake assay in the stem of _crl1_ (inset, scale bar: 200 μm), the dotted 749 white square from the inset shows the restricted PI (represented by the white lines in the inset) 750 at the lignified En (pink asterisks, marked with lignin autofluorescence, Scale bar: 50 μm). _(C)_, 751 An illustration showing the continuity of the apoplastic barrier between the crown root (CR) 752 and the stem. “1” and “2” mark the En of the Stem and CR, respectively. The blue line 753 represents the blockage of the apoplastic path. _(D)_, 4-day-old seedlings at 0, 2, and 12 hours 754 after gravistimulation (HAGS, Scale bar: 1 cm). _(E–G’)_, Upper ( _E–G_ ) and lower flanks ( _E’–_ 755 _G’_ ) of _DR5rev_ ::erGFP lines at 0 ( _E_ and _E’_ ), 2 ( _F_ and _F’_ ), and 12 ( _G_ and _G’_ ) HAGS. Note the 756 spatial parity of lignin (pink arrowhead) and _DR5_ signal (yellow arrowhead) in the upper 757 flanks, and its eventual disappearance at 12 HAGS, contrasting with the expanding auxin 758 signaling in the lower flanks (Scale bar: 50 μm). _(D’)_, Illustration of the longitudinal section 759 of a stem base, marking the upper and lower flanks (uf and lf, respectively) used for Fig. _E–_ 760 _G’_ . ( _E’’–G’’_ ), Illustration showing the response of the IC layer HAGS, similar to Fig. _E–G’_ . [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 761 762 763 **Fig 5. Polar Auxin Transport and** _**OsWOX12**_ **Coordinate Asymmetric Stem Growth** 764 **Downstream of the IC Domain.** _(A–A’),_ DMSO and 500 nM NPA-treated seedlings 24 hours 765 after gravity stimulation (HAGS, Scale bar: 1 cm). _(B),_ Graphical representation of the angle 766 of bending (°) in DMSO vs NPA-treated seedlings at continuous HAGS (for DMSO: n=9 for 767 0 HAGS, and n=6 for the rest; for NPA: n=8 for 0, 6, and 12 HAGS; and n=5 for 18 and 24 768 HAGS; ** _p_ = 0.0023 for 12 HAGS, *** _p =_ 0.0008 for 18 HAGS, and _***p_ = 0.0006 for 24 769 HAGS from Student’s t test). _(C–C’),_ Representative longitudinal section of DMSO ( _C_ ) and 770 NPA-treated ( _C’_ ) seedlings 24 HAGS. The dashed-yellow square marks the region for Fig _D_ 771 and _D’_ (Scale bar: 1 mm). _(D–D’),_ Lower flanks of DMSO and NPA-treated seedlings 24 772 HAGS. Note the lack of expansion of the _DR5_ signal in NPA treatment (Scale bar: 100 μm for 773 DMSO and 50 μm for NPA). _(E–E’),_ Wild-type (WT, _E_ ) and _wox12_ ( _E’_ ) seedlings 12 HAGS. 774 Note the reduced bending in the _wox12_ mutant (Scale bar: 1 cm). _(F),_ Graphical representation [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.763306; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.763306) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 775 of the angle of bending (°) in WT and _wox12_ seedlings; n = 19; **** _p_ < 0.0001; and ** _p_ = 776 0.0039 from Student’s t test. _(G–G’)_, Representative longitudinal section of WT ( _G_ ) and _wox12_ 777 ( _G’_ ) seedlings 12 HAGS. The dashed-yellow square marks the region used for segmentation 778 (Scale bar: 1 mm). _(H),_ Graphical representation of major axes (µm) of the cortical cells from 779 the lower flanks of WT and wox12; n > 400 from 3 independent seedlings, **** _p_ < 0.0001 780 from Mann-Whitney U test. _(I)_, An illustration summarizing the spatial decoupling of the 781 endodermal functions within the En and the IC, along with the mechanism of action of the IC. 782 783