Modulation of root growth by nutrient-defined fine-tuning of polar auxin transport

Nitrogen is an essential macronutrient and its availability in soil plays a critical role in plant growth, development and impacts agricultural productivity. Plants have evolved different strategies to sense and respond to heterogeneous nitrogen distribution. Modulating root system architecture, including primary root growth and branching, is among the most essential plant adaptions to ensure adequate nitrogen acquisition. However, the immediate molecular pathways coordinating the adjustment of root growth in response to varying nitrogen sources are poorly understood. Here, using a combination of physiological, live in vivo high- and super resolution imaging, we describe a novel adaptation strategy of root growth on available nitrogen source. We show that growth, i.e. tissue-specific cell division and elongation rates are fine-tuned by modulating auxin flux within and between tissues. Changes in auxin redistribution are achieved by nitrogen source dependent post-translational modification of PIN2, a major auxin efflux carrier, at an uncharacterized, evolutionary conserved phosphosite. Further, we generate a computer model based on our results which successfully recapitulate our experimental observations and creates new predictions that could broaden our understanding of root growth mechanisms in the dynamic environment.


44
The ability to sense and adapt to fluctuations in nutrient availability is essential for the survival 45 of all organisms. Every life form on our planet possesses delicate mechanisms for sensing and 46 reacting to the variable nutrient status and adjusts their behavior to maintain growth or cope 47 NO3rapidly enhanced root length and already 6 hours after transfer (HAT), roots were 137 significantly longer compared to these supplied with NH4 + (Fig. S1a). In general, root growth 138 is determined by the elongation of cells, which are constantly produced by the root apical 139 meristem. To study processes that underlie the adaptation of root growth to different forms of 140 N a vertical confocal microscope equipped with a root tracker system was employed. Using 141 this setup, we were able to detect and monitor the earliest root responses with a high cellular 142 resolution 32 . To minimize the interference of physiological conditions for seedling 143 development, a light-dark regime was maintained in course of the root tracking. After the 144 transfer of wild type (Col-0) seedlings to NH4 + containing medium root growth rate (RGR) was 145 enhanced, presumably as a response to stress caused by transfer of seedlings to a fresh plate. 146 Within ~120 min RGR stabilized at an average speed of 1.37 ± 0.025 µmmin -1 . Transition to 147 dark period correlated with a rapid drop of RGR to 0.98 ± 0.029 µmmin -1 , which was 148 maintained during the dark phase and at the light recovered again to 1.27 ± 0.048 µmmin -1 .

149
Seedlings transferred to NO3reacted by an increase of RGR to 1.77 ± 0.042 µmmin -1 and 150 similarly to roots on NH4 + , during the dark period their RGR decelerated and was retrieved to 151 1.81 ± 0.051 µmmin -1 at the light (Fig. 1a, Supplemental video 1). Hence, provision of NO3 -152 caused a rapid enhancement of RGR when compared to NH4 + , but it did not interfere with its 153 circadian rhythmicity 33 . 154 To gain more insight into the mechanistic basis underlying the rapid increase of root 155 length after substitution of NH4 + for NO3 -, we focused on cells in the transition zone (TZ). The 156 TZ is located between the root apical meristem and elongation zone, and cells while passing 6 growth from 2 to 3.76 hours after transfer combined with a tracking of cell membranes pointed 160 at differences in the elongation pattern of epidermal cells in roots supplied with either NH4 + or 161 NO3 -. While, in roots supplemented with NH4 + only a few epidermal cells enter into elongation 162 phase. Provision of NO3increased number of elongating cells in the TZ (Fig. 1b, Supplemental   163 video 2). Next, we analyzed in detail 18 roots 12 HAT on either NH4 + or NO3and measured 164 length of the epidermal cells across the meristematic, transition and the start of the elongation 165 zones. The analyses suggested that on NO3more epidermal cells enter into transition phase, as Supplemental Document 1a-b). Therefore, the growth of epidermal and cortex cells in roots on 184 NH4 + displayed clearly asynchronous behavior (Fig. 1c, d). Additionally, a machine learning 185 approach was applied to regression analysis for assessing the importance of each variable (i.e.  Altogether, these data indicate that roots adopt distinct growth strategies involving fine-204 tuning of cell division and expansion across adjacent tissues to adapt to different forms of N.

205
In roots supplied with NH4 + , the meristematic activity of epidermal cells is attenuated, which 206 results in their earlier transition into the elongation phase when compared to the cortex.

207
Provision of NO3increases the number of epidermal cells in the TZ (Fig. 1b, S1b), which is 208 one of the earliest detectable adaptive responses. Subsequently, within twelve hours, the 209 frequency of cell division in the epidermis increases, which results in shift of balance between 210 cell division and elongation and more synchronized growth of cortex and epidermis.

211
Eventually, a long-term supply of NO3enables enlargement of the root apical meristem 212 compared to roots supplied with NH4 + . 214 The plant hormone auxin is an essential endogenous regulatory cue that determines key 215 aspects of root growth. Interference with auxin biosynthesis 37 , signaling 38 or distribution 39 at 216 the root tip has a significant impact on the meristem maintenance, and transition of 217 meristematic cells into elongation and differentiation phase. Distinct growth patterns observed 218 in roots supplemented with different forms of N prompted us to monitor distribution of auxin 219 at the root tip. Quantification of the LUCIFERASE activity in protein extracts from roots 220 carrying the auxin sensitive DR5::LUCIFERASE reporter revealed that already one hour after 221 transfer to NO3containing medium auxin response increases when compared to roots 222 transferred to NH4 + supplemented medium (Fig. S3a). To closely inspect the auxin distribution 223 in a cell lineage-specific manner a ratiometric degradation based R2D2 auxin reporter was 224 implemented 40 . In accordance with observations based on the DR5::LUCIFERASE reporter, a 225 8 decreased ratio between DII-Venus (green) and mDII-Tomato (red) fluorescent signals 226 indicated increased levels of auxin activity in the central cylinder of roots in response to 227 replacement of NH4 + by NO3 - (Fig. S3b). 228 In addition, we focused on the detailed profiling of the R2D2 reporter in the epidermis 229 and the cortex (Fig. S3c). Interestingly, we detected an overall increase of auxin activity in 230 epidermal cells when compared to cortex cells in roots supplied with NH4 + , whereas no 231 difference between these two cell files were detected in roots on NO3 - (Fig. 2a, b). Furthermore, 232 on NH4 + there was an increase of auxin activity in epidermal cells when compared to cortex 233 cells (starting from ~ 11 th cell from the QC), while in roots supplied with NO3the auxin activity 234 profiles followed similar trends of steady increase in both cortex and epidermal cell files ( Fig.   235 2a, b). Altogether, these analyses indicate that pattern of auxin activity at root meristems might 236 adapt to specific N conditions. In roots supplied with NH4 + , the early steep gradient of auxin 237 signaling in epidermal cells correlates with their early transition into the rapid elongation phase.   The PIN2 auxin efflux carrier is amongst the principal components of PAT mediating 259 basipetal transport of auxin in roots 43,44 . To test whether adjustment of the basipetal auxin flow 260 in response to different sources of nitrogen is dependent on activity of PIN2, we tested eir1-4, 261 a mutant defective in this efflux transporter. In agreement with previous reports 45 , a 262 significantly lower radioactivity in the proximal root zone of the eir1-4 was detected when 263 compared to wild type roots on MS medium (Fig. 3a). Noteworthy, no radioactivity increase 264 in the proximal zone of eir1-4 roots was observed in roots supplied with NO3when compared 265 to NH4 + (Fig. 3a), pointing towards PIN2 function in the flexible adjustment of the basipetal profiles of auxin activity in epidermis and cortex of eir1-4 followed similar trends, 283 characterized by shallow slope along the longitudinal root growth axis (Fig. 3e, Fig. S4b).   Unlike in wild type, no significant increase in root length was detected 1 DAT in either eir1-4 292 or eir1-1 seedlings on NO3when compared to NH4 + supplemented medium (Fig. S5a). Closer 293 inspection of the RGR in real time using vertical confocal -root tracking set up showed that 294 after transfer on NH4 + growth of the eir1-4 roots stabilized at 1.47 + 0.041 µmmin -1 and 1.35 295 + µmmin -1 during light and dark period, respectively. However, no significant increase of RGR 296 after transfer to NO3containing medium could be observed (Fig. 4a). These results strongly 297 support an essential role of PIN2 mediated basipetal auxin transport in rapid adjustment of root 298 growth to form of nitrogen source.

299
To explore whether eir1-4 root growth adapts to different forms of N, elongation 300 patterns of epidermal and cortex cells were analyzed. Measurements of cell lengths along the 301 longitudinal growth axis of eir1-4 roots supplied with NH4 + revealed that unlike in Col-0, 302 epidermal cells undergo gradual, steady elongation growth comparable to that in cortex.

303
Notably, patterns of cortex and epidermal cell growth in eir1-4 appear more synchronous than 304 in wild-type roots on NH4 + (Fig. 4b versus Fig.1c). In eir1-4 roots 12 HAT from NH4 + to NO3 -305 supplemented medium we observed largely synchronized pattern of elongation in both 306 epidermal and cortex cell files, characterized by gradual, steady increase of cell length similar 307 to these observed in Col-0 ( Fig. 4c and Fig. 1d). Consistently with a more synchronous pattern 308 of epidermal and cortex cell growth in both N regimes, no significant differences in frequency 309 of mitotic events between epidermis and cortex were found in eir1-4 roots on medium supplied 310 with either NH4 + or NO3 - (Fig. 4d).

311
Overall, loss of PIN2 activity interfered with enhancement of root growth in response 312 to NO3provision and affected the establishment of tissue specific growth patterns typically 313 adopted by Col-0 roots supplied with different sources of N. Altogether, these results indicate 314 that PIN2 mediated basipetal auxin transport plays an important function in acquiring distinct 315 root growth patterns during adaptation to different N sources. 317 nitrogen source 318 To explore the mechanisms underlying PIN2 function in root growth adaptation to different N 319 sources we examined its expression, abundance at the plasma membrane (PM) and subcellular 320 trafficking in roots supplied with NH4 + or NO3 -. RT-qPCR analyses of 7 DAG roots grown on 321 NH4 + and transferred to media supplemented with either NH4 + or NO3for 1, 6 and 48 hours 322 did not reveal any significant changes in PIN2 transcription in any of the tested conditions ( Fig.   323 S6a). Likewise, expression of neither the PIN2::nlsGFP nor the PIN2::GUS reporter was 324 affected by different N source (Fig. S6b). Interestingly, monitoring of PIN2::PIN2-GFP 325 transgenic seedlings revealed significantly increased abundance of the PM located PIN2-GFP 326 in epidermal and cortex cells of roots supplied with NO3when compared to NH4 + (Fig. 5a). 327 Furthermore, in cortex cells at the transition zone of NO3supplied roots, besides expected 328 localization at the apical PM 39 , enhanced lateralization of PIN2-GFP to the inner and outer 329 PMs could be detected (Fig. 5b, Fig. S6d). Immunolocalisation using PIN2-specific antibodies 330 is fully consistent with the observations of PIN2-GFP and ruled out possible interference with 331 fluorescence of GFP reporter by different N source (Fig. S7a-c). Hence, substitution of NH4 + 332 by NO3seems to affect PIN2 at post-transcriptional rather than at transcriptional level.  faster as compared to roots supplied with NH4 + (Fig. 5e), thus strongly suggesting that delivery 361 of PIN2 towards the PM is differentially regulated by specific forms of N source.

362
Finally, to examine whether the above described different recycling behavior of PIN2 has an 363 impact on the establishment of its apical polar domain, we performed super-resolution imaging 364 employing three-dimensional structured illumination microscopy (3D-SIM). In roots 365 supplemented with either NH4 + or NO3 -, PIN2-GFP accumulated at the apical edge of epidermal 366 cells to the same level. However, in NH4 + supplemented roots, number of the PIN2-GFP 367 positive particles decreased with distance from the cell edge significantly more than in roots 368 supplied with NO3 - (Fig. 5f).

369
In summary, these results suggest that PIN2 subcellular trafficking, and in particular 370 the delivery of PIN2 to the PM is differentially adjusted according to the N source. when compared to NH4 + supplemented medium (Fig. 7a); however, when compared to control 410 seedlings the enhancement of root growth by NO3was less pronounced 4 DAT (Fig. S8b).

411
This suggests that PIN2 S439D -GFP is partially able to mediate distinct root growth responses to 412 different N sources. Roots of eir1-4 expressing PIN2::PIN2 S439A -GFP exhibited delay in 413 adjusting growth to NO3provision and no significant increase in length 6 and 24 HAT to NO3 -414 when compared to NH4 + could be detected ( Fig. 7a; S8b).   Supplementary Fig.10a-d and 10f). To comprehend a necessity for these two essential 463 components in our model, we closely inspected the relation between auxin activity levels and 464 PIN2 fluorescence in our experimental dataset in roots supplemented with NO3or NH4 + . Our 465 analysis revealed that for the same auxin activity two different PIN2 levels were observed in 466 both the cortex and the epidermis that was dependent on distance from the QC -a component 467 missing in Model A (Fig. S10a). This eminently bi-stable feature was important to guarantee 468 the synchrony of cell elongation between adjacent tissues as this feature was compromised in 469 NH4 + grown roots that showed an asynchronous elongation of adjacent cortex and epidermal 470 cells (Fig. S10a). Notably, Model B could successfully capture this relation (Fig. S10c, d). 471 Finally, we coupled auxin activity to cell division and elongation and simulated our root model 472 in both NH4 + and NO3regimes (Fig. 8b). As for previous simulations, the computer model of could predict a precise threshold of auxin levels that was necessary to determine the transition 484 to elongation. This auxin threshold is dynamic as it depends on the actual N source; in 485 particular, higher levels of auxin were required to advance cell elongation on NO3 - (Fig. 8f).       HAT to ammonium or nitrate amended media. "e" denotes epidermis and "c" cortex, 721 respectively. White arrows point to PIN2-GFP protein localization on the lateral membranes.