Report

A Local Auxin Gradient Regulates Root Cap Self-

Renewal and Size Homeostasis
Graphical Abstract Authors
Carole Dubreuil, Xu Jin,
Andreas Grönlund, Urs Fischer

Correspondence
urs.fischer@slu.se

In Brief
Dubreuil et al. investigated the regulation
of tissue size homeostasis of the apical
root cap in Arabidopsis. They show that
the root cap size is kept constant by
successive cycles of cell division and
separation. Tissue size homeostasis is
coordinated by an auxin gradient with a
maximum in the stem cells and a
minimum in the separating cells.

Highlights
d Root cap size is kept constant through repetitive cell division-
separation cycles

d A local auxin gradient orchestrates cell division and

separation

d Cell division occurs at an auxin maximum—separation at a

minimum

d The gradient is independent of active directional auxin

transport

Dubreuil et al., 2018, Current Biology 28, 2581–2587

August 20, 2018 ª 2018 Elsevier Ltd.
https://doi.org/10.1016/j.cub.2018.05.090
Current Biology
Report
A Local Auxin Gradient Regulates Root Cap
Self-Renewal and Size Homeostasis
Carole Dubreuil,1,3 Xu Jin,1,3 Andreas Grönlund,2 and Urs Fischer1,4,5,*
1Umeå Plant Science Centre, Department of Forest Genetics and Plant Physiology, Swedish University of Agricultural Sciences,
Umeå 901 83, Sweden
2Umeå Plant Science Centre, Department of Plant Physiology, Umeå University, Umeå 907 36, Sweden
3These authors contributed equally
4Present address: KWS Saat SE, Grimsehlstr. 31, 37574 Einbeck, Germany
5Lead Contact
*Correspondence: urs.fischer@slu.se
https://doi.org/10.1016/j.cub.2018.05.090
SUMMARY
Organ size homeostasis, compensatory growth to
replace lost tissue, requires constant measurement
of size and adjustment of growth rates. Morphogen
gradients control organ and tissue sizes by regulating
stem cell activity, cell differentiation, and removal in
animals [1–3]. In plants, control of tissue size is of spe-
cific importance in root caps to protect the growing
root tip from mechanical damage [4]. New root cap
tissue is formed by the columella and lateral root-
cap-epidermal stem cells, whose activity is regulated
through non-dividing niche-like cells, the quiescent
center (QC) [4, 5]. Columella daughter cells in contact
with the QC retain the potency to divide, while
derivatives oriented toward the mature cap undergo
differentiation. The outermost columella layers are
sequentially separated from the root body, involving
remodeling of cell walls [6]. Factors regulating the
balance between cell division, elongation, and sepa-
ration to keep root cap size constant are currently un-
known [4]. Here, we report that stem cell proliferation
induced cell separation at the periphery of the root
cap, resulting in tissue size homeostasis. An auxin
response gradient with a maximum in the QC and a
minimum in the detaching layer was established prior
to the onset of cell separation. In agreement with a
mathematical model, tissue size was positively regu-
lated by the amount of auxin released from the
source. Auxin transporters localized non-polarly to
plasma membranes of the inner cap, partly isolating
separating layers from the auxin source. Together,
these results are in support of an auxin gradient
measuring and regulating tissue size.
RESULTS AND DISCUSSION
Growth Homeostasis by Repetitive Cycles of Stem Cell
Proliferation and Cell Separation
We tracked individual cells of growing Arabidopsis root tips
over a period of several days in order to unravel growth dy-
Current Biology 28, 2581–2587, August 20, 2018 ª 2018 Elsevier Ltd. 2581
namics in root caps (Figures 1A, 1B, and S1A). Mature root
cap apices consisted of one stem cell tier and four layers of
columella and lateral root cap cells, hereafter referred to as
root cap layers, attached to the root body (phase I). At day 5
(phase II), a new columella layer was formed and the outermost
cell layer was removed from the apex of the root cap. During
phase II, the number of root cap layers was kept constant at
four by repetitive cycles of cell division and separation. Cell
growth and elongation compensated for the lost layers during
phase II, and hence, the cap size remained constant (Fig-
ure 1B). To test whether the number of root cap layers is
coupled with stem cell activity, we moved seedlings after
completion of phase I of root cap development to the surface
of medium containing inhibitors or activators of cell division. In
contrast to roots growing into the medium (live-cell tracking),
cell separation of roots grown on the surface was incomplete,
with separated cells of the apical lateral root cap and loosely
attached columella cells of the same layer (Figures S1B–
S1F). Repression of columella stem cell activity by inhibition
of ethylene biosynthesis [7] decreased cell division and sepa-
ration rates to an equal extent, and induction of stem cell divi-
sions in presence of an ethylene precursor increased the rates
of the division-separation cycle (Figures 1C and S1B–S1D).
Similarly, application of the cell division inhibitors hydroxyurea
and aphidicolin co-repressed division and separation rates
equally such that the number of attached root cap layers
was kept constant at four (Figures S1E–S1G). In 55 out of
56 cases, division of stem cells took place prior or simulta-
neously to the separation of the outermost layer (Figure 1D).
Removal of root cap layers, evidenced by fractures between
the very apical lateral root cap cells of the separating layer,
was observed as early as 6 hr after cell division (Figure S1H),
suggesting that cell proliferation triggered separation within
the range of minutes or a few hours.
A Single Quiescent-Center-Derived Factor Is Sufficient
to Coordinate Root Cap Growth
To test whether a single niche-cell-derived factor could provide a
robust signal to orchestrate cell division and separation, we built
a mathematical model assuming that the source of the hypothet-
ical factor are the niche cells, that the factor molecules undergo
random motion, and that the transport between individual
columella and stem cells is quantified by a transmission rate
Figure 1. Stem Cell Division Triggers Cell Separation to Keep Root Cap Size Constant
(A) Live-cell tracking of a root tip, from 3 to 10 days. Phase I, growth and maturation of root cap; phase II, repeated cycles of cell division and separation. Cell
identifiers are as follows: oldest columella layer corresponds to ‘‘z,’’ second oldest to ‘‘y.’’ QC, quiescent center, corresponds to niche cells; SC, columella stem
cells. During phase I, a new columella layer, w, is added to the growing cap by division of the SC. Asterisk represents SC. Green fluorescence, DR5::GFP. Scale
bars represent 20 mm.
(B) Root cap size homeostasis. Average columella length of 5 roots ± SD is shown; positive SD, total root cap length; negative SD, cell length; empty boxes,
separated cells.
(C) 5-day-old seedlings were moved to media containing either an ethylene precursor (1 mM 1-aminocyclopropane-1-carboxylic acid [ACC]) or biosynthesis
inhibitor (10 mM aminoethoxyvinylglycine [AVG]) and grown for an additional 5 days. Number of attached layers, filled bars; separated layers, open bars. Average ±
SD is shown; n = 4; *p < 0.05; **p < 0.01; t test, treatment versus control.
(D) Live tracking with 12 hr interval. Seedlings were 5–7 days old. ‘‘Before,’’ cell division occurred before separation; ‘‘simul.,’’ simultaneous.
(E) Columella length modeled depending on factor release rate from source (blue, low rate; green, high rate). Right side shows concentration gradient after
reaching steady state. AU, arbitrary unit.
See also Figure S1, Video S1, and Data S1.

constant. In this model, elongation and separation of columella
cells depend on the concentration of the niche-cell-derived fac-
tor. Independent of the initial values, the system rapidly reached
a steady state where the concentration of the niche-derived fac-
tor decreased gradually with increasing distance from its source
and where cell separation always succeeded division (Figures 1E
and S1I–S1K; Video S1). Columella length remained constant
within the limits of a fully elongated columella cell. Higher con-
centration of the hypothetical factor in the source resulted in a
longer columella and more rapid cell division-separation cycles

2582 Current Biology 28, 2581–2587, August 20, 2018

Figure 2. An Auxin Gradient Regulates Root Cap Size
(A) Auxin response, DR5::GFP, and fluorescence normalized to quiescent
center (QC). Average fluorescence ± SD is shown; n = 5.
(B and C) DR5::GFP, green; propidium iodide (PI), black. Cell lumina of
separated layer are not PI-stained, indicating that these cells are alive (pink
asterisk). 5-day-old (B) and 8-day-old (C) seedlings are shown. Scale bars
represent 20 mm.
(D and F) Removal of shoot-derived auxin (decapitation). Control is intact
seedlings (D). In decapitated seedlings (F), agar blocks with or without auxin
were placed at the site of decapitation.
(E and G) Removal of local auxin source. Wild-type (WT), Col-0 (E); taa1ckrc1-1;
asa1;asb1 (G).
(F and G) Auxin response, DR5::GFP. Decapitation (F; red line, cut/mock) or
reduced auxin biosynthesis (G; taa1ckrc1-1, red line) result in sharper decay of
auxin response. Presence of exogenous auxin (F, agar blocks containing 1 mM
of the synthetic auxin 1-1-naphthaleneacetic acid [NAA]; G, plates with 100 nM
1-NAA) restores the steepness of the gradient. Averages are shown; n = 4-6;
error bars, SD; *p < 0.05; t test, treatment versus control.
(H) Auxin response (DR5::GFP) in root caps of seedlings grown on exogenous
auxin (100 nM NAA).
(I) Number of columella layers and length on 100 nM 1-NAA.
In (D)–(I) averages ± SD are shown (n = 4–6; *p < 0.05; t test, treatment or
mutant versus control or WT). See also Figure S2.

(Figures S1L and S1P). As compared to changes in source con-

centration, columella length was relatively insensitive to changes
in transmission rate (Figures S1M and S1Q). Similarly, the influ-
ence of threshold concentration, at which separation occurs,
was relatively stronger than the influence of the transmission
rate on the duration of division-separation intervals (Figures
S1N, S1O, S1R, and S1S). Together, a concentration gradient
of a single niche-cell-derived factor, established without the
need of directional active transport, is theoretically sufficient to
provide robust spatiotemporal information for root cap size
homeostasis.

Cell Division Occurs at an Auxin Maximum—Separation

at a Minimum
Motivated by previous observations that auxin regulates cell
proliferation and separation during organ abscission [8–10],
we visualized auxin response (DR5::GFP) in root caps (Figures
1A and 2A–2C). Before the onset of separation, the auxin
response was higher in the outermost layer as compared to
the interior (Figures 2A and 2B). This distribution was gradually
inverted to a response gradient with a maximum in the niche
and a sharp decay in separating cells (Figures 2A–2C).
DR5v2::ndtTOMATO, a supposedly more sensitive auxin re-
porter than DR5::GFP [11], was not co-expressed in all cells
marked with DR5::GFP-derived fluorescence, and we therefore
omitted this reporter from further analysis (Figures S2A and
S2B). In contrast to the basal root cap, where tissue size is
kept constant by programmed cell death [12] rather than
orchestrated cell separation, cells of separated apical layers re-
mained alive (Figure 2C), ruling out that decreased auxin
response in separating cells was a consequence of reduced
viability. Both the auxin gradient and the viability of separated
cells were confirmed by the analysis of DII::VENUS distribution,
a reporter degraded in an auxin-concentration-dependent
manner [13], with a minimum in the QC and a maximum in the
separated layer (Figures S2C and S2D). The observed reporter
dynamics are in line with the idea of auxin as a niche-derived

Current Biology 28, 2581–2587, August 20, 2018 2583


factor permitting cell separation below a certain concentration
threshold.
Root Cap Growth Dynamics Is Regulated by an Auxin
Gradient
To test whether shoot-derived auxin is required for the establish-
ment of a local auxin gradient in the columella, we removed
the shoot prior to the first cell separation event in the root cap.
The auxin response gradient in root caps of decapitated seed-
lings decayed more sharply as compared to intact seedlings,
and both columella length and number of separated layers
decreased, although the number of attached layers remained
constant (Figures 2D, 2F, S2E, and S2F). These defects could
be restored when an agar block containing auxin was placed
on the site of decapitation, indicating that shoot-derived auxin
contributes to the auxin gradient in the root tip and the regulation
of columella size homeostasis. Besides auxin delivered from the
shoot to the root, there is a local tryptophan-dependent source
of auxin biosynthesis in the QC and the columella stem cells
[14, 15]. Mutants defective in this biosynthetic pathway dis-
played a lower auxin response and had lower rates of cell divi-
sion and separation and a reduced columella size (Figures 2E,
2G, and S2G–S2L). By contrast, adding auxin to the growth me-
dium flattened the auxin response distribution, with significantly
higher response in the three outermost layers (Figure 2H), and
increased the number of columella layers and length (Figures
2I and S2M–S2R). Similarly, expression of the bacterial auxin
biosynthesis gene iaaM in the outermost cell layers increased
the number of attached columella layers (Figures S2U–S2W).
2584 Current Biology 28, 2581–2587, August 20, 2018
Figure 3. Auxin Transport Required for
Cell Wall Remodeling, Leading to Cell
Separation
Expression of cell separation markers, 10-day-old
root tips. Green, DR5::GFP; pink, cell separation
makers.
(A–C) Control; LM8, primary antibody against xy-
logalacturonan [12] (A). JIM5, primary antibody
against partially methyl-esterified/unesterified ho-
mogalacturonan [11] (B); JIM13, primary antibody
against arabinogalactan protein [11] (C).
(D–F) 10 mM NPA (auxin transport inhibitor); LM8
(D), JIM5 (E), and JIM13 (F).
Scale bars represent 20 mm; empty arrowhead,
quiescent center; filled arrowhead, outermost
attached layer. See also Figure S3.
These data derived from the manipulation
of the auxin response gradient (Figures
2D–2G) match with the prediction of the
model (Figure 1E), where columella length
and duration of cell-division-separations
cycles strongly depend on the auxin con-
centration in the source. Chemical repres-
sion of cell division rates correlated with
a significantly lower auxin response in
the outermost attached layer (root cap
layer 4) and shorter columella length as
compared to control treatments, whereas
activation of division coincided with a
higher auxin response in root cap layer 5 and longer columella
length (Figures S2S and S2T), indicating that cell division, elon-
gation, and separation rates are co-regulated by auxin.
Auxin Transport Is Required for Columella Cell
Elongation and Maturation
Cell wall remodeling, especially of the pectin component, is a
prerequisite for cell separation in plants. The expression of the
cell wall epitopes JIM13, JIM5, and LM8 [16, 17], marking secre-
tory processes of cell separation, was highly abundant in sepa-
rated apical lateral root cap cells but absent in attached layers
(Figures 3A–3C and S3A–S3C). JIM5, a marker for initiation of
abscission [18], was also present between the separating and
attached columella layers, indicating that separation is initiated
but incomplete in roots grown on the surface of a gel. Notably,
none of the separation markers co-localized to cell walls be-
tween auxin-reporter-expressing cells, suggesting that a local
auxin minimum permits secretion of these epitopes together
with hydrolytic enzymes to induce cell separation. Previous find-
ings that depletion of auxin induces secretion in cell cultures [19]
are in accordance with this hypothesis. Inhibition of auxin trans-
port flattened the auxin response gradient and partially blocked
cell separation, as evidenced by layers of loosely attached cells
and restricted expression domains of the separation markers
(Figures 3D–3F and S3D–S3G). Because the auxin response in
the outer cap of seedlings grown on auxin transport inhibitor
was as low as in control roots, a local auxin minimum alone is
not sufficient to induce cell separation. Decreased columella
cell sizes indicate that reduced auxin response goes along with

Figure 4. Spatiotemporal Regulation of Auxin Transporters Isolates
Separating Layer from Source
(A) 10-day-old WT.
(B) 10-day-old pin3;pin4;pin7 triple mutant.
(C) 10-day-old pin1;pin3;pin4;pin7 quadruple mutant. Asterisks represent
leftover of suspensor. Stem cells aborted, arrowhead. Abortion of early root
development is shown.
(D–L) PIN proteins have stationary expression patterns and are apolarly
localized in the columella.
(D–F) pPIN3::PIN3-GFP; 3 (D), 5 (E), and 7 (F) days old.
(G–I) pPIN4::PIN4-GFP; 3 (G), 5 (H), and 7 (I) days old.
(J–L) pPIN7::PIN7-GFP; 3 (J), 5(K), and 7 (L) days old.
In (A)–(B) and (D)–(L), arrowhead, QC. Scale bars represent 20 mm. See also
Figure S4.
incomplete cell maturation and that such cells are not competent
to be separated.
PIN Activity Is Required for Early Root Cap Development
The columella-expressed auxin efflux carriers of the PIN-
FORMED (PIN) protein family, PIN3, PIN4, and PIN7, facilitate
auxin transport from the quiescent center to the lateral root
cap [20]. Although a pin3;pin4;pin7 triple mutant did not display
any qualitative alterations in root cap organization, removal of
PIN1 function from the triple mutant background aborted cap
development prior to the first separation event (Figures 4A–4C,
S4A, and S4B). In contrast to PIN3, PIN4, and PIN7, PIN1 is
only weakly expressed in the columella [21]. PIN1 expression
in the provasculature presumably provides together with the
columella PINs rootward auxin transport [21]. Hence, PIN-medi-
ated delivery from shoot-derived or recycled auxin to the QC is of
critical importance for early root cap development, whereas the
columella expression domain of the PIN proteins plays a minor
role in root cap development.
Stationary PIN Expression Patterns Partly Isolate
Separating Root Cap Layer from the Auxin Source in the
Niche Cells
After every stem cell division, the newly formed columella
daughter cells acquired PIN3 expression while PIN3 was
switched off in the third outermost layer prior to cell separation
(Figures 4D–4I and S4C–S4J). Similarly, after cell separation,
PIN4 and 7 expression was acquired in the second columella
tier but lost in the second outermost layer (Figures 4D–4I and
S4C–S4J). Hence, relative to the position of the stem cells, the
PIN expression domains remained stationary and loss of PIN
expression in the fourth columella layer partly isolated the sepa-
rating fifth layer from the auxin source in the niche cells. As sug-
gested by a broader PIN expression domain in root caps grown
on exogenous auxin (Figures S4C–S4I), PIN abundance is a
function of auxin concentration, and therefore, loss of PIN
expression in the fourth layer may reinforce the signal for sepa-
ration and contribute to greater robustness of the system. Simi-
larly, synchronization between cell division and separation may
be more robust due to low PIN levels in dividing cells and acqui-
sition of PIN in the newly formed columella cells. During cell divi-
sion, prior to completion of the cell plate, auxin can freely diffuse
between the two daughter cells. However, after division and
repression of PIN at the inner face of the PIN expression domain,
the outer columella is partly disconnected from the auxin source
in the QC, and hence the difference in auxin concentration be-
tween stem cells and the adjacent columella layer might become
bigger and more robustly interpretable (Figure S4K).
All of the columella PINs localized apolarly (Figures 4D–4I), and
therefore, in agreement with the mathematical model, no direc-
tionality of auxin flux mediated by active transporters is to be ex-
pected. In other words, the gradient is shaped by auxin transmis-
sion, decay, dilution due to cell elongation, and loss due to cell
separation rather than by a concentrating mechanism executed
by polar PIN localization. Assuming a single source, this model
for auxin distribution is self-organizing and does not require
any other signal to robustly regulate size homeostasis. By
contrast, polar PIN localization is required for the creation of
auxin maxima instrumental for the initiation of lateral root and
Current Biology 28, 2581–2587, August 20, 2018 2585
leaf primordia [22, 23]. In these cases, auxin is actively pumped
against a concentration gradient and direction of transport
needs to be coordinated over several cells, most likely requiring
additional factors than auxin to establish PIN polarization.
Although auxin gradients within cells could theoretically provide
a cue for PIN polarization in a solely auxin-concentration-depen-
dent manner, it is unlikely that a gradient can be maintained
within single cells due to the rapid movement of this small mole-
cule in the cytosol. Given that the auxin receptors localize to the
nucleus, a relatively small portion of the cell, a cytosolic auxin
gradient would need to be rather steep in order to create a sig-
nificant and interpretable difference in receptor activity between
opposing poles of the nucleus.
Auxin has been implicated in cell separation before. For
example, local auxin applications to leaf petioles indicate that
relatively high auxin concentration on the distal side of the
abscission zone delay, whereas high concentration on the prox-
imal side induces leaf abscission [8, 24]. Besides being a tempo-
ral cue, auxin is also important for the formation of abscission
zones in Arabidopsis siliques and leaf petioles [8–10]. Similarly,
as in columella stem cells, auxin response is high in the dividing
cells that build these abscission zones [8, 10]. The development
of abscission zones results in the formation of a few cell-tier-
thick separation layers, distal to which the fracture occurs.
Although it remains to be tested whether auxin gradients regu-
late the thickness of abscission zones, in a similar fashion as
we observed for tissue size homeostasis in root caps, auxin
response minima co-localize in both valve and leaf separation
with the site of cell separation.
Both hypocotyl and root length can be manipulated by exog-
enous auxin [25, 26], and auxin gradients may be important to
measure the sizes of these organs too. In contrast to root caps,
hypocotyls and root sizes do not oscillate due to sequential
removal and regeneration of cell layers. Size measurement in
these organs could be determined early in development by
cell-autonomous factors. An example of size homeostasis,
more similar to root caps, is the regeneration of intestinal
epithelium. Epithelial stem cell derivatives move from the crypt
to the villus, where they eventually undergo apoptosis and are
removed from the epithelium. Gradients of different morpho-
gens ensure that proliferation and shedding rates are equal,
which prevents overgrowth of the tissue [3]. Interestingly, there
are not only gradients with a maximum in the proliferative cells
(e.g., Wingless and Wnt) but also gradients peaking in the
mature part of the tissue, e.g., bone morphogenetic proteins
(BMPs) [2]. Because BMP represses Wnt action, a simple feed-
back mechanism is in place to co-regulate cell division and
removal rates. Although we did not find evidence for a cue
antagonizing auxin in root caps, and although the mathematical
model supports a one-factor regulation, it is likely that there is
also an information flow from the separating layer to the stem
cells. Unlike on agar plates, mechanical frictions to the root
cap may vary rapidly in time and space due to the heteroge-
neous nature of soil. Premature removal of a root cap layer
most likely has an impact on the shape of the auxin gradient
(steeper decay), but this change might be too small to be
robustly interpreted. Additional outside-to-inside communica-
tion, be it in form of a mechanical or hormonal signal, may
therefore be required to induce compensatory growth of the
2586 Current Biology 28, 2581–2587, August 20, 2018
root cap. Interesting in this respect will be to test whether
ethylene signaling, which regulates auxin biosynthesis in the
root tip and integrates mechanical signals [26, 27], could ac-
count for such a component.
STAR+METHODS
Detailed methods are provided in the online version of this paper
and include the following:
d KEY RESOURCES TABLE
d CONTACT FOR REAGENT AND RESOURCE SHARING
d EXPERIMENTAL MODEL AND SUBJECT DETAILS
d METHOD DETAILS
B Pharmacological treatments
B Live-cell tracking
B Immunohistochemistry
B Model description
d QUANTIFICATION AND STATISTICAL ANALYSIS
SUPPLEMENTAL INFORMATION
Supplemental Information includes four figures, one video, and one data file
and can be found with this article online at https://doi.org/10.1016/j.cub.
2018.05.090.
ACKNOWLEDGMENTS
This work was supported by Bio4Energy, Berzelii Center (Vinnova), and Deut-
sche Forschungsgemeinschaft (DFG) (FI 1668/1-1 to U.F.).
AUTHOR CONTRIBUTIONS
C.D., X.J., and U.F. designed and performed the experiments; A.G. formulated
and analyzed the mathematical model; and all authors interpreted the data and
contributed to the writing of the manuscript.
DECLARATION OF INTERESTS
The authors declare no competing interests.
Received: December 15, 2017
Revised: April 17, 2018
Accepted: May 31, 2018
Published: August 2, 2018
REFERENCES
1. Affolter, M., and Basler, K. (2007). The decapentaplegic morphogen
gradient: from pattern formation to growth regulation. Nat. Rev. Genet.
8, 663–674.
2. Yeung, T.M., Chia, L.A., Kosinski, C.M., and Kuo, C.J. (2011). Regulation
of self-renewal and differentiation by the intestinal stem cell niche. Cell.
Mol. Life Sci. 68, 2513–2523.
3. Vermeulen, L., and Snippert, H.J. (2014). Stem cell dynamics in homeosta-
sis and cancer of the intestine. Nat. Rev. Cancer 14, 468–480.
4. Kumpf, R.P., and Nowack, M.K. (2015). The root cap: a short story of life
and death. J. Exp. Bot. 66, 5651–5662.
5. Fisher, A.P., and Sozzani, R. (2016). Uncovering the networks involved in
stem cell maintenance and asymmetric cell division in the Arabidopsis
root. Curr. Opin. Plant Biol. 29, 38–43.

-Gibouin, M. (2007). Formation and sep-
6. Driouich, A., Durand, C., and Vicre
aration of root border cells. Trends Plant Sci. 12, 14–19.
 7. Ortega-Martı́nez, O., Pernas, M., Carol, R.J., and Dolan, L. (2007).
Ethylene modulates stem cell division in the Arabidopsis thaliana root.
Science 317, 507–510.
8. Jin, X., Zimmermann, J., Polle, A., and Fischer, U. (2015). Auxin is a long-
range signal that acts independently of ethylene signaling on leaf abscis-
sion in Populus. Front. Plant Sci. 6, 634.
9. Sorefan, K., Girin, T., Liljegren, S.J., Ljung, K., Robles, P., Galván-
Ampudia, C.S., Offringa, R., Friml, J., Yanofsky, M.F., and Østergaard,
L. (2009). A regulated auxin minimum is required for seed dispersal in
Arabidopsis. Nature 459, 583–586.
10. van Gelderen, K., van Rongen, M., Liu, A., Otten, A., and Offringa, R.
(2016). An INDEHISCENT-controlled auxin response specifies the separa-
tion layer in early Arabidopsis fruit. Mol. Plant 9, 857–869.
11. Liao, C.Y., Smet, W., Brunoud, G., Yoshida, S., Vernoux, T., and Weijers,
D. (2015). Reporters for sensitive and quantitative measurement of auxin
response. Nat. Methods 12, 207–210, 2, 210.
12. Fendrych, M., Van Hautegem, T., Van Durme, M., Olvera-Carrillo, Y.,
Huysmans, M., Karimi, M., Lippens, S., Gue
rin, C.J., Krebs, M.,
Schumacher, K., and Nowack, M.K. (2014). Programmed cell death
controlled by ANAC033/SOMBRERO determines root cap organ size in
Arabidopsis. Curr. Biol. 24, 931–940.
13. Brunoud, G., Wells, D.M., Oliva, M., Larrieu, A., Mirabet, V., Burrow, A.H.,
Beeckman, T., Kepinski, S., Traas, J., Bennett, M.J., and Vernoux, T.
(2012). A novel sensor to map auxin response and distribution at high spa-
tio-temporal resolution. Nature 482, 103–106.
14. Stepanova, A.N., Hoyt, J.M., Hamilton, A.A., and Alonso, J.M. (2005). A
link between ethylene and auxin uncovered by the characterization of
two root-specific ethylene-insensitive mutants in Arabidopsis. Plant Cell
17, 2230–2242.
15. Petersson, S.V., Johansson, A.I., Kowalczyk, M., Makoveychuk, A., Wang,
J.Y., Moritz, T., Grebe, M., Benfey, P.N., Sandberg, G., and Ljung, K.
(2009). An auxin gradient and maximum in the Arabidopsis root apex
shown by high-resolution cell-specific analysis of IAA distribution and syn-
thesis. Plant Cell 21, 1659–1668.

, M., Santaella, C., Blanchet, S., Gateau, A., and Driouich, A. (2005).
16. Vicre
Root border-like cells of Arabidopsis. Microscopical characterization and
role in the interaction with rhizobacteria. Plant Physiol. 138, 998–1008.
17. Willats, W.G.T., McCartney, L., Steele-King, C.G., Marcus, S.E., Mort, A.,
Huisman, M., van Alebeek, G.J., Schols, H.A., Voragen, A.G.J., Le Goff, A.,
et al. (2004). A xylogalacturonan epitope is specifically associated with
plant cell detachment. Planta 218, 673–681.
18. Roongsattham, P., Morcillo, F., Fooyontphanich, K., Jantasuriyarat, C.,
Tragoonrung, S., Amblard, P., Collin, M., Mouille, G., Verdeil, J.-L., and
Tranbarger, T.J. (2016). Cellular and pectin dynamics during abscission
zone development and ripe fruit abscission of the monocot oil palm.
Front. Plant Sci. 7, 540.
19. Winicur, Z.M., Zhang, G.F., and Staehelin, L.A. (1998). Auxin deprivation
induces synchronous Golgi differentiation in suspension-cultured tobacco
BY-2 cells. Plant Physiol. 117, 501–513.
20. Grieneisen, V.A., Xu, J., Mare
e, A.F.M., Hogeweg, P., and Scheres, B.
(2007). Auxin transport is sufficient to generate a maximum and gradient
guiding root growth. Nature 449, 1008–1013.
21. Omelyanchuk, N.A., Kovrizhnykh, V.V., Oshchepkova, E.A., Pasternak, T.,
Palme, K., and Mironova, V.V. (2016). A detailed expression map of the
PIN1 auxin transporter in Arabidopsis thaliana root. BMC Plant Biol. 16
(Suppl 1 ), 5.
22. Reinhardt, D., Pesce, E.R., Stieger, P., Mandel, T., Baltensperger, K.,
Bennett, M., Traas, J., Friml, J., and Kuhlemeier, C. (2003). Regulation of
phyllotaxis by polar auxin transport. Nature 426, 255–260.
23. Marhavý, P., Vanstraelen, M., De Rybel, B., Zhaojun, D., Bennett, M.J.,
Beeckman, T., and Benková, E. (2013). Auxin reflux between the
endodermis and pericycle promotes lateral root initiation. EMBO J. 32,
149–158.
24. Louie, D.S., and Addicott, F.T. (1970). Applied auxin gradients and abscis-
sion in explants. Plant Physiol. 45, 654–657.
25. Collett, C.E., Harberd, N.P., and Leyser, O. (2000). Hormonal interactions
in the control of Arabidopsis hypocotyl elongation. Plant Physiol. 124,
553–562.
26. Hu, Y., Vandenbussche, F., and Van Der Straeten, D. (2017). Regulation of
seedling growth by ethylene and the ethylene-auxin crosstalk. Planta 245,
467–489.
27. Zhong, S., Zhao, M., Shi, T., Shi, H., An, F., Zhao, Q., and Guo, H. (2009).
EIN3/EIL1 cooperate with PIF1 to prevent photo-oxidation and to promote
greening of Arabidopsis seedlings. Proc. Natl. Acad. Sci. USA 106, 21431–
21436.
28. Zádnı́ková, P., Petrásek, J., Marhavý, P., Raz, V., Vandenbussche, F.,
Ding, Z., Schwarzerová, K., Morita, M.T., Tasaka, M., Hejátko, J., et al.
(2010). Role of PIN-mediated auxin efflux in apical hook development of
Arabidopsis thaliana. Development 137, 607–617.
29. Friml, J., Vieten, A., Sauer, M., Weijers, D., Schwarz, H., Hamann, T.,
Offringa, R., and Jürgens, G. (2003). Efflux-dependent auxin gradients
establish the apical-basal axis of Arabidopsis. Nature 426, 147–153.
30. Vieten, A., Vanneste, S., Wi
sniewska, J., Benková, E., Benjamins, R.,
Beeckman, T., Luschnig, C., and Friml, J. (2005). Functional redundancy
of PIN proteins is accompanied by auxin-dependent cross-regulation of
PIN expression. Development 132, 4521–4531.
31. Blilou, I., Xu, J., Wildwater, M., Willemsen, V., Paponov, I., Friml, J.,
Heidstra, R., Aida, M., Palme, K., and Scheres, B. (2005). The PIN auxin
efflux facilitator network controls growth and patterning in Arabidopsis
roots. Nature 433, 39–44.
32. Zhou, Z.Y., Zhang, C.G., Wu, L., Zhang, C.G., Chai, J., Wang, M., Jha, A.,
Jia, P.F., Cui, S.J., Yang, M., et al. (2011). Functional characterization
of the CKRC1/TAA1 gene and dissection of hormonal actions in the
Arabidopsis root. Plant J. 66, 516–527.
Current Biology 28, 2581–2587, August 20, 2018 2587
STAR+METHODS

KEY RESOURCES TABLE

REAGENT or RESOURCE SOURCE IDENTIFIER

Antibodies
Rat monoclonal anti JIM5 CCRC, University of Georgia, US JIM5
Rat monoclonal anti JIM13 CCRC, University of Georgia, US JIM13
Rat monoclonal anti LM8 CCRC, University of Georgia, US LM8
Goat anti-rat IgG Alexa Fluor 568 nm ThermoFisher Scientific A-11077
Chemicals, Peptides, and Recombinant Proteins
NPA, N-1-naphthylphthalamic acid Duchefa N0926
NAA, 1-naphthaleneacetic acid Sigma-Aldrich N0640
ACC, 1-aminocyclopropane-1-carboxylic acid Sigma-Aldrich A3903
AVG, aminoethoxyvinylglycine Sigma-Aldrich A6685
Hydroxyurea Sigma-Aldrich H8627
Aphidicolin Sigma-Aldrich A0781
Citifluor AF1 Electron Microscopy Sciences Hatfield, PA,USA AF1
Experimental Models: Organisms/Strains
Arabidopsis: Columiba-0 (Col-0, WT) NASC N1093
Arabidopsis: DR5::GFP (DR5rev::GFP), in Col-0 NASC N9361
Arabidopsis: pPIN3::PIN3-GFP, in Col-0 [28] N/A
Arabidopsis: pPIN4::PIN4-GFP, in Col-0 NASC N9576
Arabidopsis: pPIN7::PIN7-GFP, in Col-0 NASC N9577
Arabidopsis: taa1ckrc1-1;DR5::GFP, in Col-0 NASC N66989
Arabidopsis: asa1wei2-1;asb1wei7-1, in Col-0 NASC N16399
Arabidopsis: UAS::iaaM, in Col-0 NASC N9544
Arabidopsis: GAL4-GFP, enhancer trap line, in C24 NASC N9149
Arabidopsis: C24, ecotype NASC N9086
Arabidopsis: DII-Venus, in Col-0 NASC N799173
Arabidopsis: DR5v2 in Col-0 NASC N2105636
Arabidopsis: pin3-5; pin4-3; pin7-1 segregating [28] N/A
for pin1
Software and Algorithms
ImageJ 1.48v NIH https://imagej.nih.gov/ij
Zen 2012 (black) Zeiss N/A

CONTACT FOR REAGENT AND RESOURCE SHARING

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Urs
Fischer (urs.fischer@slu.se).

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Experiments were performed with the Arabidopsis ecotype Columbia-0 (Col-0) or in Col-0 plants carrying DR5::GFP (DR5rev::GFP)
[29]; pPIN3::PIN3-GFP [28]; pPIN4::PIN4-GFP [30]; pPIN7::PIN7-GFP [31]; taa1ckrc1-1 [32] and asa1wei2-1;asb1wei7-1 [14], respectively.
The pin quadruple mutant was isolated from a progeny of pin3-5; pin4-3; pin7-1 segregating for pin1 [28]. Surface sterilized seeds
were kept in darkness at 4
C for 3 days, prior to be cultured in either sealed Petri dishes or, for live-cell tracking, in chambered slides
on 1% Murashige and Skoog (MS) medium (pH 5.8; 1% sucrose; 1% plant agar). Plants were grown under standard greenhouse
conditions with 16 hr light per day and at 23
C.

e1 Current Biology 28, 2581–2587.e1–e3, August 20, 2018

METHOD DETAILS

Pharmacological treatments
For pharmacological treatments, seedlings were transferred to MS plates containing either 10 mM NPA (N-1-naphthylphthalamic
acid, polar auxin transport inhibitor); 100 nM NAA (1-naphthaleneacetic acid, synthetic auxin); 1 mM ACC (1-aminocyclopropane-
1-carboxylic acid, ethylene biosynthesis precursor), 10 mM AVG (aminoethoxyvinylglycine, ethylene biosynthesis inhibitor), 1 mM
hydroxyurea (blocking DNA synthesis) or 20 mg/mL aphidicolin (blocking G-S transition) after completing Phase I of root cap devel-
opment. Roots of seedling treated with NPA, AVG, hydroxyurea and aphidicolin appeared shorter than untreated controls. Roots
containing dead cells in the columella, provasculature or other apical cell types except lateral root cap cells were omitted from further
analysis. In contrast to roots grown into the agar medium, cell separation on the surface of a gel was incomplete, i.e., while the colu-
mella cells retained loosely attached to the root body, apical lateral root cap cells of the same layer completely detached. Such layers
where scored as detached.

Live-cell tracking
To limit effects of observation on root cap development we grew Arabidopsis seedlings on MS medium into sterile chambered mi-
croscope slides (Nunc Lab-Tek), which permitted to keep the observation times short under sterile conditions and to move the cham-
bered slides back into vertical position under greenhouse conditions immediately after examination (Figure S1A). Briefly, 4 mL of MS
medium (pH 5.8; 1% sucrose; 0.8% plant agar) was poured into a sterile chambered microscope slide. A stripe (0.5 cm) of MS me-
dium was removed prior to placing seeds into the groove between the MS medium and cover slide to allow roots growing along the
slide in a vertical growth position. Like this, we manually tracked individual cells expressing DR5::GFP of growing root tips over a
period of several days by confocal microscopy under sterile conditions. Cell sizes and average fluorescence was measured in manu-
ally segmented cells (ImageJ). Relative fluorescence levels were normalized to fluorescence of the quiescent center (QC). Decapi-
tation and placement of agarose blocks were performed in the chambered slides without removing the seedlings. Seedlings were
grown in absence of propidium iodide (PI) since we observed interference of the dye with plant growth. Confocal microscopy was
carried out with a Zeiss LSM510 CLSM system and a C-Apochromat 40x/1.20 N.A. objective. Excitation wavelengths were
488 nm for GFP and PI.

Immunohistochemistry
Whole seedlings were fixed in 4% paraformaldehyde/PBS (pH 7) solution for 30 min under vacuum and washed three times with PBS.
Root tips were dissected, transferred to Superfrost slides (Menzel Gla €ser, Germany), and dried at room temperature. After drying,
root tips were rehydrated in PBS for 10 min and were covered with a coverslip attached with double-sided tape. Blocking was per-
formed with 3% BSA in PBS for 1 hr at room temperature. Primary antibody incubation was performed overnight at 4
C in 3% BSA in
PBS, followed by incubation for 2 hr at 37
C. JIM5 was applied at a dilution of 1:10, JIM13 at a dilution of 1:100, and LM8 at a dilution
of 1:100. ALEXA 568 was used as secondary antibody at 1:100 dilution and was incubated for 2.5 to 3 hr at 37
C. Finally, root tips
were stained with 1 mg/mL DAPI (4’,6-diamidino-2-phenylindole) for 15 min in darkness and mounted in Citifluor AF1. Confocal mi-
croscopy was carried out with either a Leica TCS SP2 AOBS. Excitation wavelengths were 405 nm for DAPI, 488 nm for GFP, 561 nm
for ALEXA 568.

Model description
Change of the qc-derived factor molecules, si , of cell i is due to decay of s and the net difference of s influx and efflux to neighboring
cells. We let the decay rate constant, be denoted D. Reducing the problem to molecular transport in one direction (x-direction), mol-
ecules may leave cell i through cell boundaries with adjacent cells, i + 1 and i – 1. Molecules may likewise enter cell i from either i  1 or
i + 1. The rate of molecules exiting cell i to a neighboring cell (e.g., cell i+1) through the boundary between the two cells is proportional
to the concentration of molecules si /DxiDyiDzi times the surface area DyiDzi of the boundary and the transmission rate from cell i to
cell i+1, Ti,i+1. Multiplying the factors together gives the transmission rate from cell i to cell i+1 as, Ti,i+1si /Dxi. Different transmission
rates between cells can be a result of, e.g., having different levels of transporter factors on different cell surfaces. We now formulate
the change in si over time, due to decay and fluxes between the individual cells, as
   
dsi si si1 si si + 1
=  Ti;i1  Ti1;i  Ti;i + 1  Ti + 1;i  Dsi (Equation 1)
dt Dxi Dxi1 Dxi Dxi + 1
Together with the boundary conditions; 4qc, the influx to the first cell qc, and reflecting boundary at the outermost boundary of the last
cell – we have a complete set of equations that describes the flux and decay of si within the cell system. If we assume the diffusion
rates Ti,j, to be constant and independent of the factor si, we can re-scale time in units of the decay rate constant t = Dt, to see that the
transport properties of the system effectively only depend on the re-scaled transmission rate constant R = T/D. Equation 1 then
reads,
 
dsi si si1 si + 1
= R 2    si : (Equation 2)
dt Dxi Dxi1 Dxi + 1

Current Biology 28, 2581–2587.e1–e3, August 20, 2018 e2

In addition, all cells except qc, are constantly growing. We assume that growth of cell i is proportional to the concentration of s-mol-
ecules. The growth is then described by the following equation
dðDxi Þ si
=K (Equation 3)
dt Dxi
where, as in Equation 2 time is measured in units of the decay rate constant t = Dt. K is the growth rate constant, which is scaled with
the decay rate constant just like the transmission rate constant. Two more components to the model are added. First, we let the stem
cell sc grow from 0.5Dxqc to Dxqc with a growth rate given by Equation 3, and, second cells are separated and removed if they have a
concentration of s below a given threshold qs. Cell separation induces removal of s-molecules from the system and, as we will see,
generates a gradient from the source of s-molecules at qc to the outermost cell. We solve the Equations 2 and 3 numerically to test
which features can be reproduced in the dynamics of root cap growth with the assumption of having uniform and constant transmis-
sion rates between the cells.

QUANTIFICATION AND STATISTICAL ANALYSIS

Sample size (n) and p values (t test) are indicated in the figure legends. Bars represent mean values and error bars are standard de-
viations (SD). Length and fluorescence intensity was measured with ImageJ. For fluorescence intensities cells were manually
segmented and mean fluorescence was measured.

e3 Current Biology 28, 2581–2587.e1–e3, August 20, 2018