Stem cells in plants are crucial for their longevity and usually maintained in meristems, special cellular environments constituting protective niches (Greb and Lohmann, 2016). At key positions in the plant body, we find distinct types of meristems that maintain their activity throughout a plant’s life cycle. Shoot and root apical meristems (SAM, RAM) are located at the tips of shoots and roots, respectively, driving longitudinal growth and the formation of primary tissue anatomy in these organs. Moreover, lateral meristems organized in cylindrical domains at the periphery of shoots and roots execute their thickening. The cambium is the most prominent among these lateral meristems (Fischer et al., 2019). Cambium cells are embedded in two distinct vascular tissues produced in opposite directions by periclinal cell divisions: xylem (wood) and phloem (bast) (Chiang and Greb, 2019; Haas et al., 2022). These tissues carry out fundamental physiological functions: long-distance transport of water and nutrients in case of the xylem and translocation of sugars and a multitude of signaling molecules in the case of the phloem. Based on its tightly controlled bidirectionality of tissue production and resulting bipartite organization, the cambium is a paradigm for bifacial stem cell niches which produce two tissue types in opposite directions and are found across different kingdoms of life (Shi et al., 2017).

Balancing proliferation and differentiation within meristems is essential. In the SAM and the RAM, this balance is maintained via interaction between the pool of stem cells and the organizing center (OC) and the quiescent center (QC), respectively, where the rate of cell division is relatively low. Both domains form a niche within the meristem instructing surrounding stem cells via regulatory feedback loops (Sabatini et al., 2003; Sarkar et al., 2007; Mayer et al., 1998; Daum et al., 2014; Pi et al., 2015). In comparison to apical meristems, functional characterization of cambium domains was performed only very recently. During their transition from stem cells to fully differentiated xylem cells, early xylem cells instruct radial patterning of the cambium, including stem cell activity and, thus, similar to the OC in the SAM, fulfill this role only transiently (Smetana et al., 2019). In addition to influence from the early xylem, phloem-derived DNA-BINDING ONE ZINC FINGER (DOF) transcription factors designated as PHLOEM EARLY DOFs (PEARs) move to cambium stem cells and stimulate their proliferation in a non-cell-autonomous manner (Miyashima et al., 2019). Furthermore, genetically encoded lineage-tracing experiments showed that cell divisions are mostly restricted to individual bifacial stem cells located in the central cambium feeding both xylem and phloem production (Smetana et al., 2019; Shi et al., 2019; Bossinger and Spokevicius, 2018). Altogether, these findings defined functional cambium domains and revealed some of their reciprocal communication.

Another central and well-established mechanism regulating cambium activity in the reference plant Arabidopsis thaliana and beyond (Kucukoglu et al., 2017; Etchells and Turner, 2010; Etchells et al., 2015; Fisher and Turner, 2007) is the action of a receptor-ligand pair formed by the plasma membrane-bound receptor-like kinase PHLOEM INTERCALATED WITH XYLEM (PXY), also known as TDIF RECEPTOR (TDR), and the secreted CLAVATA3/ESR-RELATED 41 (CLE41) and CLE44 peptides. Like the PEAR proteins (Miyashima et al., 2019), CLE41 and CLE44 are expressed in the phloem and thought to diffuse to dividing cells in the cambium area expressing PXY (Etchells and Turner, 2010; Hirakawa et al., 2008). Direct binding of CLE41 to PXY (Hirakawa et al., 2008; Morita et al., 2016; Zhang et al., 2016) promotes the expression of the transcription factor WUSCHEL RELATED HOMEOBOX 4 (WOX4) (Hirakawa et al., 2010), which, in turn, is crucial for maintaining the capacity of cells to proliferate (Kucukoglu et al., 2017; Hirakawa et al., 2010; Suer et al., 2011). At the same time, the PXY/CLE41 module is reported to repress xylem differentiation in a WOX4-independent manner (Hirakawa et al., 2010; Kondo et al., 2014). In this context, PXY stimulates the activity of glycogen synthase kinase 3 proteins (GSK3s), like BRASSINOSTEROID-INSENSITIVE 2 (BIN2) (Kondo et al., 2014). BIN2, in turn, represses the transcriptional regulator BRI1-EMS SUPPRESSOR 1 (BES1), which mediates brassinosteroid (BR) signaling and promotes xylem differentiation (Kondo et al., 2014; Saito et al., 2018). In line with the hypothesis of a dual role in regulating stem cell activity and xylem differentiation, PXY is expressed in the proximal cambium zone containing developing xylem cells and in the central cambium zone containing bifacial cambium stem cells (Shi et al., 2019; Brackmann et al., 2018; Shi et al., 2020).

Distally to the PXY expression domain and oriented towards the phloem, the closest homolog to PXY, the receptor-like kinase MORE LATERAL GROWTH 1 (MOL1), represses cambium activity (Agusti et al., 2011b; Gursanscky et al., 2016). Although their extracellular domains are highly similar, PXY and MOL1 cannot functionally replace each other, indicating that MOL1 activity does not depend on CLE41/44 peptides and that distinct signaling loops act in the proximal and distal cambium domains (Gursanscky et al., 2016). The latter conclusion is supported by the finding that the AUXIN RESPONSE FACTOR5 (ARF5) is expressed in the proximal cambium and promotes the transition from stem cells to xylem cells by directly dampening WOX4 activity (Brackmann et al., 2018; Han et al., 2018). ARF5 activity is enhanced by phosphorylation through the GSK3 BIN2-LIKE 1 (BIL1) which, in contrast to other GSK3s (Kondo et al., 2014), is inhibited by the PXY/CLE41 module (Han et al., 2018).

As the role of multiple communication cascades between different cambium-related tissues is beginning to emerge, it is vital to generate a systemic view on their combined impact on cambium activity and patterning integrated into a dynamic tissue environment. However, although the cambium plays an instructive role for stem cell biology, a dynamic view on its activity is missing due to its inaccessibility for live-cell imaging. Computational modeling, in particular agent-based modeling combining tissue layout with biochemical signaling processes, can overcome these obstacles and help analyzing the interplay between cellular signaling processes, cell growth, and cell differentiation in silico that would otherwise be inaccessible. Here, we present a dynamic, agent-based computational model (Merks and Glazier, 2005) of the cambium integrating the functions of PXY, CLE41, and putative phloem-derived signals into a plant-specific modeling framework. As revealed by informative cambium markers, our model is able to reproduce anatomical features of the cambium in a dynamic manner. It also allows studying the cambium as a flexible system comprised of multiple interacting factors, and the effects of those factors on cell division, differentiation, and tissue patterning.

Growth and development of multicellular organisms are complex nonlinear processes whose dynamics and network properties are not possible to predict only based on information on their individual building blocks and their one-to-one interactions. The rather simple cellular outline along the radial axes of plant organs, growth in only two dimensions, and the recent identification of central functional properties (Smetana et al., 2019; Miyashima et al., 2019; Shi et al., 2019) make radial plant growth an attractive target for a systematic approach to reveal its intriguing dynamics. Here, we developed a computational model representing a minimal framework required for radial plant growth using the VirtualLeaf framework (Merks et al., 2011). In particular, we combined an agent-based model of the tissue layout with an ODE model of the inter-cellular PXY/CLE41 signaling module. By integrating these two modeling and biological scales, we were able to recapitulate not only the complex behaviors that arise as consequence of the cellular interactions (Macal and North, 2005) but also the interplay between cellular layout and intercellular signaling dynamics. Therefore, our model allows analyzing fundamental features of plant organ growth and integrates the PXY/CLE41 module as one central element for cambium patterning and maintenance.

Using positional information mediated by morphogenetic gradients of diffusible chemicals to pattern growing structures is a classical concept in developmental biology that has stirred a long history of fundamental debates (Green and Sharpe, 2015). Initially, we used the PXY/CLE41 module to generate such a gradient instructing cambium cells* to differentiate into xylem cells*, to proliferate, or to differentiate into phloem cells*. Repression of cell division in the distal cambium was achieved by implementing an inhibitory feedback loop of PXY-signaling* on PXY* production. Altogether, this setup was already sufficient to maintain stable radial tissue organization during radial growth and established a maximum of cell division rates in the cambium center as observed by experimental means (Shi et al., 2019). Thus, we conclude that cambium organization and radial patterning of plant growth axes can be maintained by a distinct pattern of radially acting morphogens. Such a role was initially proposed for auxin whose differential distribution, however, seems to be rather a result of tissue patterning than being instructive for radial tissue organization (Bhalerao and Fischer, 2014).

In contrast to expected roles of the PXY pathway in xylem formation based on experiments during primary vascular development (Hirakawa et al., 2008; Hirakawa et al., 2010; Kondo et al., 2014), we observed that the overall amount of proximal tissue production during radial plant growth did not depend on the PXY function. Automated image analysis including object classification revealed that neither the number of cells produced toward the organ center nor the number of vessel elements did change in a pxy mutant background but rather the ratio between parenchyma and fiber cells. Therefore, in contrast to a negative effect of PXY/CLE41 signaling on vessel formation in vascular bundles in leaves (Hirakawa et al., 2008; Kondo et al., 2014), vessel formation during radial plant growth is PXY/CLE41-independent. Instead, fiber formation is positively associated with the PXY/CLE41 module. These observations indicated that xylem formation is unlikely to be instructed by PXY/CLE41 signaling alone and that additional signals are required.

Moreover, the application of markers visualizing cambium organization showed that PXY-deficiency leads to cambium disorganization in some regions of the hypocotyl, whereas in other areas, cambium anatomy is maintained. Since such areas are regularly spaced, this pattern may arise due to factors acting in parallel to PXY/CLE41 and which also carry spatial information. Although ethylene signaling was reported to act in parallel to PXY/CLE41 signaling, spatial specificity does not seem to be a characteristic property of ethylene signaling (Etchells et al., 2012). In contrast, PEAR transcription factors are phloem-derived and stimulate the proliferation of cambium stem cells presumably in a PXY/CLE41-independent manner (Miyashima et al., 2019) and, thus, may act similarly to the PF* factor we introduced in our model. The ERECTA/EPIDERMAL PATTERNING FACTOR-LIKE (ER/EPFL) receptor-ligand pathway acting in concert with the PXY/CLE41 module (Wang et al., 2019; Etchells et al., 2013) represents another candidate for playing such a role. In addition, CLE45 was recently proposed to be expressed in developing sieve elements, the conducting units of the phloem, and repress the establishment of sieve element identity in their immediate environment mediated by the RPK2 receptor protein (Gujas et al., 2020). The PF* factor in our model combines features of these phloem-derived molecules.

In addition to the phloem sending out instructive signals, early xylem cells have been identified to act as an organizing center of cambium patterning (Smetana et al., 2019). Although this finding seems to be at odds with our claim that phloem-derived signals are sufficient for cambium organization, it is important to consider that we ignored, for example, upstream regulation of postulated factors like PXY* or CLE41*, which obviously depends on positional information which could be mediated in plants by auxin or cytokinin signaling (Bishopp et al., 2011). For simplicity, we also ignored organizing effects of signaling longitudinally to cross-sections as it can, for example, be expected for polar auxin transport (Bennett et al., 2016; Ibañes et al., 2009; Fàbregas et al., 2015) in the context of cambium activity or xylem formation. Although being considerably more complex, the establishment of 3D models will be crucial and essential for addressing this aspect.

In this context, it is interesting to note that we deliberately excluded the transition from the initially bisymmetric tissue conformation to a concentric tissue organization as it occurs in hypocotyls and roots (Smetana et al., 2019; Sankar et al., 2014) from our model. Our rationale was that the rather complex change in tissue anatomy from a primary to a secondary conformation in the hypocotyl required more assumptions in our model and would have spoiled the advantages of a relatively simple anatomy for generating a cell-based computational model. Moreover, the differences in primary anatomy of shoots and roots before the onset of radial plant growth (Smetana et al., 2019; Sehr et al., 2010) would have required different cellular outlines for both cases and, thus, would have hampered the generality of our approach.

Interestingly, the front of cambium domains is very stable, that is, almost perfectly circular, in planta but this is not the case for our computational simulations. We believe that instability in the computational models is due to local noise in the cellular pattern leading to differential diffusion of chemicals* with respect to their radial position and to a progressive deviation of domains from a perfect circle. Such a deviation seems to be corrected by an unknown mechanism in planta but such a corrective mechanism is due to the absence of a good indication in planta not implemented in our models. Analyses of wt and pxy lines at different stages (Figure 3—figure supplement 1, Figure 3—figure supplement 2, Figure 3—figure supplement 3, Figure 3—figure supplement 4) revealed ‘gaps’ in the cambium domain already at early stages of pxy development arguing against the possibility that the pxy anatomy is caused by increased front instability. Although a corrective mechanism ensuring front stability in planta is difficult to predict, our model now allows to test respective ideas like directional movement of chemicals or stabilizing communication between cells during cambium activity. In this context, it is interesting that increasing epidermis* ‘stiffness’ increased circularity of the growing organ*, which may be administered by the periderm (Serra et al., 2022), the protective cell layers that we did not consider in our model.

Current research on plant mechanical biology indicates how cell mechanical properties influence cell and tissue morphogenesis. Microtubules, turgor pressure, and cell wall composition are central factors in this regard (Sapala et al., 2018; Altartouri et al., 2019). Due to the geometric constraints in a radially growing plant axis, it becomes challenging to uncouple these factors experimentally and establish the impact of one factor on organ patterning during radial plant growth. By expanding VirtualLeaf to allow for the integration of cell-type-specific wall stiffness, we fundamentally increased the spectrum of potential modeling approaches. In particular, since cell wall stiffness is accessible by the cellular model throughout simulations, it is now possible to simulate and analyze, for example, the dynamics of auxin or brassinosteroid-mediated cell wall loosening (Majda and Robert, 2018; Caesar et al., 2011). In our cambium model, by modulating exclusively cellular ‘stiffness,’ we were able to computationally simplify the ‘physical’ properties and, thereby, develop a hypothesis how inter-tissue forces influence stem cell behavior not only cell autonomously, but also in a non-cell-autonomous manner.

Taken together, we envision that the model presented in this study recapitulates the qualitative and quantitative variation in radial plant growth on multiple levels, found in different mutants or when comparing different dicotyledonous species (Spicer and Groover, 2010). Remarkable features like the establishment of concentric cambium rings often found in the order of Caryophyllales (Carlquist, 2007) or ‘phloem wedges’ found in the Bignonieae genus (Pace et al., 2009) may be recapitulated by adjusting the model’s parameters values or by introducing additional factors. In the future, the model may help to predict targets of environmental stimuli inducing changes of cambium activity like seasonal changes (Bhalerao and Fischer, 2017) or mechanical perturbation (Gerttula et al., 2015), allowing the generation of testable hypotheses. Thus, our dynamic model will be a useful tool for investigating a process not possible to observe in real time and partly develops over exceptionally long periods.

Code files for presented models are deposited at https://github.com/thomasgreb/Lebovka-et-al_cambium-models (copy archived at Lebovka et al., 2023).

1. 1. Agusti J
   2. Herold S
   3. Schwarz M
   4. Sanchez P
   5. Ljung K
   6. Dun EA
   7. Brewer PB
   8. Beveridge CA
   9. Sieberer T
   10. Sehr EM
   11. Greb T

   (2011a) Strigolactone signaling is required for auxin-dependent stimulation of secondary growth in plants

   PNAS 108:20242–20247.

   https://doi.org/10.1073/pnas.1111902108

   • PubMed
   • Google Scholar
2. 1. Agusti J
   2. Lichtenberger R
   3. Schwarz M
   4. Nehlin L
   5. Greb T

   (2011b) Characterization of transcriptome remodeling during cambium formation identifies MOL1 and RUL1 as opposing regulators of secondary growth

   PLOS Genetics 7:e1001312.

   https://doi.org/10.1371/journal.pgen.1001312

   • PubMed
   • Google Scholar
3. 1. Altartouri B
   2. Bidhendi AJ
   3. Tani T
   4. Suzuki J
   5. Conrad C
   6. Chebli Y
   7. Liu N
   8. Karunakaran C
   9. Scarcelli G
   10. Geitmann A

   (2019) Pectin chemistry and cellulose crystallinity govern pavement cell morphogenesis in a multi-step mechanism

   Plant Physiology 181:127–141.

   https://doi.org/10.1104/pp.19.00303

   • PubMed
   • Google Scholar
4. Book

   1. Antonovici C
   2. Peerdeman GY
   3. Wolff HB
   4. Merks RMH

   (2022) Modeling plant tissue development using virtualleaf

   In: Lucas M, editors. In Plant Systems Biology. Springer. pp. 165–198.

   https://doi.org/10.1007/978-1-0716-1816-5_9

   • Google Scholar
5. 1. Arnould O
   2. Capron M
   3. Ramonda M
   4. Laurans F
   5. Alméras T
   6. Pilate G
   7. Clair B

   (2022) Mechanical characterisation of the developing cell wall layers of tension wood fibres by atomic force microscopy

   Peer Community Journal 2:e39.

   https://doi.org/10.24072/pcjournal.138

   • Google Scholar
6. 1. Bennett T
   2. Hines G
   3. van Rongen M
   4. Waldie T
   5. Sawchuk MG
   6. Scarpella E
   7. Ljung K
   8. Leyser O

   (2016) Connective auxin transport in the shoot facilitates communication between shoot apices

   PLOS Biology 14:e1002446.

   https://doi.org/10.1371/journal.pbio.1002446

   • PubMed
   • Google Scholar
7. 1. Bhalerao RP
   2. Fischer U

   (2014) Auxin gradients across wood-instructive or incidental?

   Physiologia Plantarum 151:43–51.

   https://doi.org/10.1111/ppl.12134

   • PubMed
   • Google Scholar
8. 1. Bhalerao RP
   2. Fischer U

   (2017) Environmental and hormonal control of cambial stem cell dynamics

   Journal of Experimental Botany 68:79–87.

   https://doi.org/10.1093/jxb/erw466

   • PubMed
   • Google Scholar
9. 1. Bishopp A
   2. Help H
   3. El-Showk S
   4. Weijers D
   5. Scheres B
   6. Friml J
   7. Benková E
   8. Mähönen AP
   9. Helariutta Y

   (2011) A mutually inhibitory interaction between auxin and cytokinin specifies vascular pattern in roots

   Current Biology 21:917–926.

   https://doi.org/10.1016/j.cub.2011.04.017

   • PubMed
   • Google Scholar
10. 1. Bossinger G
    2. Spokevicius AV

    (2018) Sector analysis reveals patterns of cambium differentiation in poplar stems

    Journal of Experimental Botany 69:4339–4348.

    https://doi.org/10.1093/jxb/ery230

    • PubMed
    • Google Scholar
11. 1. Brackmann K
    2. Qi J
    3. Gebert M
    4. Jouannet V
    5. Schlamp T
    6. Grünwald K
    7. Wallner E-S
    8. Novikova DD
    9. Levitsky VG
    10. Agustí J
    11. Sanchez P
    12. Lohmann JU
    13. Greb T

    (2018) Spatial specificity of auxin responses coordinates wood formation

    Nature Communications 9:875.

    https://doi.org/10.1038/s41467-018-03256-2

    • PubMed
    • Google Scholar
12. 1. Brown CL
    2. Sax K

    (1962) The influence of pressure on the differentiation of secondary tissues

    American Journal of Botany 49:683–691.

    https://doi.org/10.1002/j.1537-2197.1962.tb14997.x

    • Google Scholar
13. Book

    1. Brown CL

    (1964) The influence of external pressure on the differentiation of cells and tissues cultured in vitro

    In: Zimmerman M, editors. The Formation of Wood in Forest Trees. New York: Academic Press. pp. 389–404.

    https://doi.org/10.1016/C2013-0-12178-2

    • Google Scholar
14. 1. Caesar K
    2. Elgass K
    3. Chen Z
    4. Huppenberger P
    5. Witthöft J
    6. Schleifenbaum F
    7. Blatt MR
    8. Oecking C
    9. Harter K

    (2011) A fast brassinolide-regulated response pathway in the plasma membrane of Arabidopsis thaliana

    The Plant Journal 66:528–540.

    https://doi.org/10.1111/j.1365-313X.2011.04510.x

    • PubMed
    • Google Scholar
15. Software

    1. Canty A
    2. Ripley BD

    (2020) Boot: bootstrap R (S-plus) functions

    R Package Version.

    https://cran.r-project.org/web/packages/boot/boot.pdf
16. 1. Carlquist S

    (2007) Successive cambia revisited: ontogeny, histology, diversity, and functional significance

    The Journal of the Torrey Botanical Society 134:301–332.

    https://doi.org/10.3159/1095-5674(2007)134[301:SCROHD]2.0.CO;2

    • Google Scholar
17. 1. Chiang MH
    2. Greb T

    (2019) How to organize bidirectional tissue production?

    Current Opinion in Plant Biology 51:15–21.

    https://doi.org/10.1016/j.pbi.2019.03.003

    • PubMed
    • Google Scholar
18. 1. Daum G
    2. Medzihradszky A
    3. Suzaki T
    4. Lohmann JU

    (2014) A mechanistic framework for noncell autonomous stem cell induction in arabidopsis

    PNAS 111:14619–14624.

    https://doi.org/10.1073/pnas.1406446111

    • PubMed
    • Google Scholar
19. Book

    1. Davison AC
    2. Hinkley DV

    (1997) Bootstrap Methods and Their Applications

    Cambridge: Cambridge University Press.

    https://doi.org/10.1017/CBO9780511802843

    • Google Scholar
20. 1. Etchells JP
    2. Turner SR

    (2010) The PXY-CLE41 receptor ligand pair defines a multifunctional pathway that controls the rate and orientation of vascular cell division

    Development 137:767–774.

    https://doi.org/10.1242/dev.044941

    • Google Scholar
21. 1. Etchells JP
    2. Provost CM
    3. Turner SR

    (2012) Plant vascular cell division is maintained by an interaction between PXY and ethylene signalling

    PLOS Genetics 8:e1002997.

    https://doi.org/10.1371/journal.pgen.1002997

    • PubMed
    • Google Scholar
22. 1. Etchells JP
    2. Provost CM
    3. Mishra L
    4. Turner SR

    (2013) WOX4 and WOX14 act downstream of the PXY receptor kinase to regulate plant vascular proliferation independently of any role in vascular organisation

    Development 140:2224–2234.

    https://doi.org/10.1242/dev.091314

    • PubMed
    • Google Scholar
23. 1. Etchells JP
    2. Mishra LS
    3. Kumar M
    4. Campbell L
    5. Turner SR

    (2015) Wood formation in trees is increased by manipulating PXY-regulated cell division

    Current Biology 25:1050–1055.

    https://doi.org/10.1016/j.cub.2015.02.023

    • PubMed
    • Google Scholar
24. 1. Fàbregas N
    2. Formosa-Jordan P
    3. Confraria A
    4. Siligato R
    5. Alonso JM
    6. Swarup R
    7. Bennett MJ
    8. Mähönen AP
    9. Caño-Delgado AI
    10. Ibañes M

    (2015) Auxin influx carriers control vascular patterning and xylem differentiation in Arabidopsis thaliana

    PLOS Genetics 11:e1005183.

    https://doi.org/10.1371/journal.pgen.1005183

    • PubMed
    • Google Scholar
25. 1. Fischer U
    2. Kucukoglu M
    3. Helariutta Y
    4. Bhalerao RP

    (2019) The dynamics of cambial stem cell activity

    Annual Review of Plant Biology 70:293–319.

    https://doi.org/10.1146/annurev-arplant-050718-100402

    • PubMed
    • Google Scholar
26. 1. Fisher K
    2. Turner S

    (2007) Pxy, a receptor-like kinase essential for maintaining polarity during plant vascular-tissue development

    Current Biology 17:1061–1066.

    https://doi.org/10.1016/j.cub.2007.05.049

    • PubMed
    • Google Scholar
27. 1. Gerttula S
    2. Zinkgraf M
    3. Muday GK
    4. Lewis DR
    5. Ibatullin FM
    6. Brumer H
    7. Hart F
    8. Mansfield SD
    9. Filkov V
    10. Groover A

    (2015) Transcriptional and hormonal regulation of gravitropism of woody stems in Populus

    The Plant Cell 27:2800–2813.

    https://doi.org/10.1105/tpc.15.00531

    • PubMed
    • Google Scholar
28. 1. Greb T
    2. Lohmann JU

    (2016) Plant stem cells

    Current Biology 26:R816–R821.

    https://doi.org/10.1016/j.cub.2016.07.070

    • PubMed
    • Google Scholar
29. 1. Green JBA
    2. Sharpe J

    (2015) Positional information and reaction-diffusion: two big ideas in developmental biology combine

    Development 142:1203–1211.

    https://doi.org/10.1242/dev.114991

    • PubMed
    • Google Scholar
30. 1. Gujas B
    2. Kastanaki E
    3. Sturchler A
    4. Cruz TMD
    5. Ruiz-Sola MA
    6. Dreos R
    7. Eicke S
    8. Truernit E
    9. Rodriguez-Villalon A

    (2020) A reservoir of pluripotent phloem cells safeguards the linear developmental trajectory of protophloem sieve elements

    Current Biology 30:755–766.

    https://doi.org/10.1016/j.cub.2019.12.043

    • PubMed
    • Google Scholar
31. 1. Gursanscky NR
    2. Jouannet V
    3. Grünwald K
    4. Sanchez P
    5. Laaber-Schwarz M
    6. Greb T

    (2016) MOL1 is required for cambium homeostasis in Arabidopsis

    The Plant Journal 86:210–220.

    https://doi.org/10.1111/tpj.13169

    • PubMed
    • Google Scholar
32. 1. Haas AS
    2. Shi D
    3. Greb T

    (2022) Cell fate decisions within the vascular cambium-initiating wood and Bast formation

    Frontiers in Plant Science 13:864422.

    https://doi.org/10.3389/fpls.2022.864422

    • PubMed
    • Google Scholar
33. 1. Han S
    2. Cho H
    3. Noh J
    4. Qi J
    5. Jung HJ
    6. Nam H
    7. Lee S
    8. Hwang D
    9. Greb T
    10. Hwang I

    (2018) BIL1-mediated MP phosphorylation integrates PXY and cytokinin signalling in secondary growth

    Nature Plants 4:605–614.

    https://doi.org/10.1038/s41477-018-0180-3

    • PubMed
    • Google Scholar
34. 1. Hirakawa Y
    2. Shinohara H
    3. Kondo Y
    4. Inoue A
    5. Nakanomyo I
    6. Ogawa M
    7. Sawa S
    8. Ohashi-Ito K
    9. Matsubayashi Y
    10. Fukuda H

    (2008) Non-cell-autonomous control of vascular stem cell fate by a CLE peptide/receptor system

    PNAS 105:15208–15213.

    https://doi.org/10.1073/pnas.0808444105

    • PubMed
    • Google Scholar
35. 1. Hirakawa Y
    2. Kondo Y
    3. Fukuda H

    (2010) TDIF peptide signaling regulates vascular stem cell proliferation via the WOX4 homeobox gene in Arabidopsis

    The Plant Cell 22:2618–2629.

    https://doi.org/10.1105/tpc.110.076083

    • PubMed
    • Google Scholar
36. 1. Ibañes M
    2. Fàbregas N
    3. Chory J
    4. Caño-Delgado AI

    (2009) Brassinosteroid signaling and auxin transport are required to establish the periodic pattern of arabidopsis shoot vascular bundles

    PNAS 106:13630–13635.

    https://doi.org/10.1073/pnas.0906416106

    • PubMed
    • Google Scholar
37. 1. Kondo Y
    2. Ito T
    3. Nakagami H
    4. Hirakawa Y
    5. Saito M
    6. Tamaki T
    7. Shirasu K
    8. Fukuda H

    (2014) Plant GSK3 proteins regulate xylem cell differentiation downstream of TDIF-TDR signalling

    Nature Communications 5:3504.

    https://doi.org/10.1038/ncomms4504

    • PubMed
    • Google Scholar
38. 1. Kucukoglu M
    2. Nilsson J
    3. Zheng B
    4. Chaabouni S
    5. Nilsson O

    (2017) WUSCHEL-RELATED HOMEOBOX4 (WOX4) -like genes regulate cambial cell division activity and secondary growth in Populus trees

    The New Phytologist 215:642–657.

    https://doi.org/10.1111/nph.14631

    • PubMed
    • Google Scholar
39. Software

    1. Lebovka I
    2. Mele BH
    3. Liu X
    4. Zakieva A
    5. Schlamp T
    6. Gursanscky NR
    7. Merks RMH
    8. Großeholz R
    9. Greb T

    (2023) Lebovka-et-al_cambium-models, version swh:1:rev:7b395f343bfd7e5aa33b6883a1b7d7b25455a1c3

    Software Heritage.

    https://archive.softwareheritage.org/swh:1:dir:9088f73bac207da051caf6bb0445ec1ec739d6e2;origin=https://github.com/thomasgreb/Lebovka-et-al_cambium-models;visit=swh:1:snp:01287757944da1bb27adb8dd200f023eee563fb7;anchor=swh:1:rev:7b395f343bfd7e5aa33b6883a1b7d7b25455a1c3
40. 1. Lintilhac PM
    2. Vesecky TB

    (1984) Stress-Induced alignment of division plane in plant tissues grown in vitro

    Nature 307:363–364.

    https://doi.org/10.1038/307363a0

    • Google Scholar
41. Conference

    1. Macal CM
    2. North MJ

    (2005)

    Tutorial on agent-based modeling and simulation

    In Proceedings of the Winter Simulation Conference.

    • Google Scholar
42. 1. Mähönen AP
    2. Higuchi M
    3. Törmäkangas K
    4. Miyawaki K
    5. Pischke MS
    6. Sussman MR
    7. Helariutta Y
    8. Kakimoto T

    (2006) Cytokinins regulate a bidirectional phosphorelay network in Arabidopsis

    Current Biology 16:1116–1122.

    https://doi.org/10.1016/j.cub.2006.04.030

    • PubMed
    • Google Scholar
43. 1. Majda M
    2. Robert S

    (2018) The role of auxin in cell wall expansion

    International Journal of Molecular Sciences 19:951.

    https://doi.org/10.3390/ijms19040951

    • PubMed
    • Google Scholar
44. 1. Mayer KF
    2. Schoof H
    3. Haecker A
    4. Lenhard M
    5. Jürgens G
    6. Laux T

    (1998) Role of WUSCHEL in regulating stem cell fate in the Arabidopsis shoot meristem

    Cell 95:805–815.

    https://doi.org/10.1016/s0092-8674(00)81703-1

    • PubMed
    • Google Scholar
45. 1. Merks RMH
    2. Glazier JA

    (2005) A cell-centered approach to developmental biology

    Physica A 352:113–130.

    https://doi.org/10.1016/j.physa.2004.12.028

    • Google Scholar
46. 1. Merks RMH
    2. Guravage M
    3. Inzé D
    4. Beemster GTS

    (2011) VirtualLeaf: an open-source framework for cell-based modeling of plant tissue growth and development

    Plant Physiology 155:656–666.

    https://doi.org/10.1104/pp.110.167619

    • PubMed
    • Google Scholar
47. 1. Merks RMH
    2. Guravage MA

    (2013) Building simulation models of developing plant organs using virtualleaf

    Methods in Molecular Biology 959:333–352.

    https://doi.org/10.1007/978-1-62703-221-6_23

    • PubMed
    • Google Scholar
48. Software

    1. Merks RMH
    2. Peerdeman GY
    3. Visscher J

    (2021) VirtualLeaf2021

    GitHub.

    https://github.com/rmerks/VirtualLeaf2021
49. 1. Mitsuda N
    2. Iwase A
    3. Yamamoto H
    4. Yoshida M
    5. Seki M
    6. Shinozaki K
    7. Ohme-Takagi M

    (2007) Nac transcription factors, NST1 and NST3, are key regulators of the formation of secondary walls in woody tissues of Arabidopsis

    The Plant Cell 19:270–280.

    https://doi.org/10.1105/tpc.106.047043

    • PubMed
    • Google Scholar
50. 1. Miyashima S
    2. Roszak P
    3. Sevilem I
    4. Toyokura K
    5. Blob B
    6. Heo J-O
    7. Mellor N
    8. Help-Rinta-Rahko H
    9. Otero S
    10. Smet W
    11. Boekschoten M
    12. Hooiveld G
    13. Hashimoto K
    14. Smetana O
    15. Siligato R
    16. Wallner E-S
    17. Mähönen AP
    18. Kondo Y
    19. Melnyk CW
    20. Greb T
    21. Nakajima K
    22. Sozzani R
    23. Bishopp A
    24. De Rybel B
    25. Helariutta Y

    (2019) Mobile pear transcription factors integrate positional cues to prime cambial growth

    Nature 565:490–494.

    https://doi.org/10.1038/s41586-018-0839-y

    • PubMed
    • Google Scholar
51. 1. Morita J
    2. Kato K
    3. Nakane T
    4. Kondo Y
    5. Fukuda H
    6. Nishimasu H
    7. Ishitani R
    8. Nureki O

    (2016) Crystal structure of the plant receptor-like kinase TDR in complex with the TDIF peptide

    Nature Communications 7:12383.

    https://doi.org/10.1038/ncomms12383

    • PubMed
    • Google Scholar
52. Software

    1. Ogle DH
    2. Doll JC
    3. Wheeler AP
    4. Dinno A

    (2023) FSA: simple fisheries stock assessment methods

    R Package Version 0.9.4.

    https://fishr-core-team.github.io/FSA/
53. 1. Pace MR
    2. Lohmann LG
    3. Angyalossy V

    (2009) The rise and evolution of the cambial variant in bignonieae (bignoniaceae)

    Evolution & Development 11:465–479.

    https://doi.org/10.1111/j.1525-142X.2009.00355.x

    • PubMed
    • Google Scholar
54. 1. Pi L
    2. Aichinger E
    3. van der Graaff E
    4. Llavata-Peris CI
    5. Weijers D
    6. Hennig L
    7. Groot E
    8. Laux T

    (2015) Organizer-derived WOX5 signal maintains root columella stem cells through chromatin-mediated repression of CDF4 expression

    Developmental Cell 33:576–588.

    https://doi.org/10.1016/j.devcel.2015.04.024

    • PubMed
    • Google Scholar
55. 1. Qian P
    2. Song W
    3. Zaizen-Iida M
    4. Kume S
    5. Wang G
    6. Zhang Y
    7. Kinoshita-Tsujimura K
    8. Chai J
    9. Kakimoto T

    (2022) A dof-CLE circuit controls phloem organization

    Nature Plants 8:817–827.

    https://doi.org/10.1038/s41477-022-01176-0

    • PubMed
    • Google Scholar
56. Software

    1. Robinson D
    2. Hayes A
    3. Couch S

    (2023) Broom: convert statistical objects into tidy tibbles

    Broom.

    https://broom.tidymodels.org/,https://github.com/tidymodels/broom
57. 1. Sabatini S
    2. Heidstra R
    3. Wildwater M
    4. Scheres B

    (2003) SCARECROW is involved in positioning the stem cell niche in the Arabidopsis root meristem

    Genes & Development 17:354–358.

    https://doi.org/10.1101/gad.252503

    • PubMed
    • Google Scholar
58. 1. Saito M
    2. Kondo Y
    3. Fukuda H

    (2018) BES1 and BZR1 redundantly promote phloem and xylem differentiation

    Plant & Cell Physiology 59:590–600.

    https://doi.org/10.1093/pcp/pcy012

    • PubMed
    • Google Scholar
59. 1. Sankar M
    2. Nieminen K
    3. Ragni L
    4. Xenarios I
    5. Hardtke CS

    (2014) Automated quantitative histology reveals vascular morphodynamics during Arabidopsis hypocotyl secondary growth

    eLife 3:e01567.

    https://doi.org/10.7554/eLife.01567

    • PubMed
    • Google Scholar
60. 1. Sapala A
    2. Runions A
    3. Routier-Kierzkowska A-L
    4. Das Gupta M
    5. Hong L
    6. Hofhuis H
    7. Verger S
    8. Mosca G
    9. Li C-B
    10. Hay A
    11. Hamant O
    12. Roeder AH
    13. Tsiantis M
    14. Prusinkiewicz P
    15. Smith RS

    (2018) Why plants make puzzle cells, and how their shape emerges

    eLife 7:e32794.

    https://doi.org/10.7554/eLife.32794

    • PubMed
    • Google Scholar
61. 1. Sarkar AK
    2. Luijten M
    3. Miyashima S
    4. Lenhard M
    5. Hashimoto T
    6. Nakajima K
    7. Scheres B
    8. Heidstra R
    9. Laux T

    (2007) Conserved factors regulate signalling in Arabidopsis thaliana shoot and root stem cell organizers

    Nature 446:811–814.

    https://doi.org/10.1038/nature05703

    • PubMed
    • Google Scholar
62. 1. Schindelin J
    2. Arganda-Carreras I
    3. Frise E
    4. Kaynig V
    5. Longair M
    6. Pietzsch T
    7. Preibisch S
    8. Rueden C
    9. Saalfeld S
    10. Schmid B
    11. Tinevez J-Y
    12. White DJ
    13. Hartenstein V
    14. Eliceiri K
    15. Tomancak P
    16. Cardona A

    (2012) Fiji: an open-source platform for biological-image analysis

    Nature Methods 9:676–682.

    https://doi.org/10.1038/nmeth.2019

    • PubMed
    • Google Scholar
63. 1. Sehr EM
    2. Agusti J
    3. Lehner R
    4. Farmer EE
    5. Schwarz M
    6. Greb T

    (2010) Analysis of secondary growth in the Arabidopsis shoot reveals a positive role of jasmonate signalling in cambium formation

    The Plant Journal 63:811–822.

    https://doi.org/10.1111/j.1365-313X.2010.04283.x

    • PubMed
    • Google Scholar
64. 1. Serra O
    2. Mähönen AP
    3. Hetherington AJ
    4. Ragni L

    (2022) The making of plant armor: the periderm

    Annual Review of Plant Biology 73:405–432.

    https://doi.org/10.1146/annurev-arplant-102720-031405

    • PubMed
    • Google Scholar
65. 1. Shi D
    2. Tavhelidse T
    3. Thumberger T
    4. Wittbrodt J
    5. Greb T

    (2017) Bifacial stem cell niches in fish and plants

    Current Opinion in Genetics & Development 45:28–33.

    https://doi.org/10.1016/j.gde.2017.02.002

    • PubMed
    • Google Scholar
66. 1. Shi D
    2. Lebovka I
    3. López-Salmerón V
    4. Sanchez P
    5. Greb T

    (2019) Bifacial cambium stem cells generate xylem and phloem during radial plant growth

    Development 146:dev171355.

    https://doi.org/10.1242/dev.171355

    • PubMed
    • Google Scholar
67. Preprint

    1. Shi D
    2. Jouannet V
    3. Agustí J
    4. Kaul V
    5. Levitsky V
    6. Sanchez P
    7. Mironova VV
    8. Greb T

    (2020) Tissue-specific transcriptome profiling of the Arabidopsis thaliana inflorescence stem reveals local cellular signatures

    bioRxiv.

    https://doi.org/10.1101/2020.02.10.941492

    • Google Scholar
68. 1. Smetana O
    2. Mäkilä R
    3. Lyu M
    4. Amiryousefi A
    5. Sánchez Rodríguez F
    6. Wu M-F
    7. Solé-Gil A
    8. Leal Gavarrón M
    9. Siligato R
    10. Miyashima S
    11. Roszak P
    12. Blomster T
    13. Reed JW
    14. Broholm S
    15. Mähönen AP

    (2019) High levels of auxin signalling define the stem-cell organizer of the vascular cambium

    Nature 565:485–489.

    https://doi.org/10.1038/s41586-018-0837-0

    • PubMed
    • Google Scholar
69. 1. Smith RA
    2. Schuetz M
    3. Roach M
    4. Mansfield SD
    5. Ellis B
    6. Samuels L

    (2013) Neighboring parenchyma cells contribute to Arabidopsis xylem lignification, while lignification of interfascicular fibers is cell autonomous

    The Plant Cell 25:3988–3999.

    https://doi.org/10.1105/tpc.113.117176

    • PubMed
    • Google Scholar
70. Conference

    1. Sommer C
    2. Strähle C
    3. Köthe U
    4. Hamprecht FA

    (2011) ilastik: Interactive Learning and Segmentation Toolkit

    In Eighth IEEE International Symposium on Biomedical Imaging. pp. 230–233.

    https://doi.org/10.1109/ISBI.2011.5872394

    • Google Scholar
71. 1. Spicer R
    2. Groover A

    (2010) Evolution of development of vascular cambia and secondary growth

    The New Phytologist 186:577–592.

    https://doi.org/10.1111/j.1469-8137.2010.03236.x

    • PubMed
    • Google Scholar
72. 1. Suer S
    2. Agusti J
    3. Sanchez P
    4. Schwarz M
    5. Greb T

    (2011) WOX4 imparts auxin responsiveness to cambium cells in Arabidopsis

    The Plant Cell 23:3247–3259.

    https://doi.org/10.1105/tpc.111.087874

    • PubMed
    • Google Scholar
73. 1. Taylor NG
    2. Howells RM
    3. Huttly AK
    4. Vickers K
    5. Turner SR

    (2003) Interactions among three distinct cesa proteins essential for cellulose synthesis

    PNAS 100:1450–1455.

    https://doi.org/10.1073/pnas.0337628100

    • PubMed
    • Google Scholar
74. 1. Ursache R
    2. Andersen TG
    3. Marhavý P
    4. Geldner N

    (2018) A protocol for combining fluorescent proteins with histological stains for diverse cell wall components

    The Plant Journal 93:399–412.

    https://doi.org/10.1111/tpj.13784

    • PubMed
    • Google Scholar
75. 1. Wallner E-S
    2. López-Salmerón V
    3. Belevich I
    4. Poschet G
    5. Jung I
    6. Grünwald K
    7. Sevilem I
    8. Jokitalo E
    9. Hell R
    10. Helariutta Y
    11. Agustí J
    12. Lebovka I
    13. Greb T

    (2017) Strigolactone- and karrikin-independent SMXL proteins are central regulators of phloem formation

    Current Biology 27:1241–1247.

    https://doi.org/10.1016/j.cub.2017.03.014

    • PubMed
    • Google Scholar
76. 1. Wallner ES
    2. Tonn N
    3. Shi D
    4. Jouannet V
    5. Greb T

    (2020) Suppressor of MAX2 1-like 5 promotes secondary phloem formation during radial stem growth

    The Plant Journal 102:903–915.

    https://doi.org/10.1111/tpj.14670

    • PubMed
    • Google Scholar
77. 1. Wang N
    2. Bagdassarian KS
    3. Doherty RE
    4. Kroon JT
    5. Connor KA
    6. Wang XY
    7. Wang W
    8. Jermyn IH
    9. Turner SR
    10. Etchells JP

    (2019)

    Organ-specific genetic interactions between paralogues of the PXY and ER receptor kinases enforce radial patterning in arabidopsis vascular tissue

    Development 146:dev.177105.

    • PubMed
    • Google Scholar
78. 1. Wickham H
    2. Averick M
    3. Bryan J
    4. Chang W
    5. McGowan L
    6. François R
    7. Grolemund G
    8. Hayes A
    9. Henry L
    10. Hester J
    11. Kuhn M
    12. Pedersen T
    13. Miller E
    14. Bache S
    15. Müller K
    16. Ooms J
    17. Robinson D
    18. Seidel D
    19. Spinu V
    20. Takahashi K
    21. Vaughan D
    22. Wilke C
    23. Woo K
    24. Yutani H

    (2019) Welcome to the tidyverse

    Journal of Open Source Software 4:1686.

    https://doi.org/10.21105/joss.01686

    • Google Scholar
79. 1. Wolpert L

    (1969) Positional information and the spatial pattern of cellular differentiation

    Journal of Theoretical Biology 25:1–47.

    https://doi.org/10.1016/s0022-5193(69)80016-0

    • PubMed
    • Google Scholar
80. 1. Zhang H
    2. Lin X
    3. Han Z
    4. Qu LJ
    5. Chai J

    (2016) Crystal structure of PXY-TDIF complex reveals a conserved recognition mechanism among CLE peptide-receptor pairs

    Cell Research 26:543–555.

    https://doi.org/10.1038/cr.2016.45

    • PubMed
    • Google Scholar

Decision letter after peer review:

Thank you for submitting your article "Computational modelling of cambium activity provides a regulatory framework for simulating radial plant growth" for consideration by eLife. Your article has been reviewed by 2 experts in the field, and the evaluation has been overseen by a Reviewing Editor and Naama Barkai as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Przemyslaw Prusinkiewicz (Reviewer #2). The Reviewing Editor has drafted this to help you prepare a revised submission.

Major concerns rose by both reviewers relate to presentation, description, and clarification of models. Furthermore, the discussion of key model findings and their relevance should be improved. There are also major concerns on the experimental data quantification, the derived claims and how they relate to the presented models. Note this might also imply a deeper study of the presented models.

Specific comments of the reviewers:

Reviewer #1 (Recommendations for the authors):

Apart of the issues I have already expressed in the Public Review, please find below additional considerations:

– The authors need to work more on the clarity about some aspects of the model. How the growth is implemented in all the models? How diffusion is implemented? How the forces are implemented and the impact of the cell wall thickness? How the final timepoint is determined? I believe equation 5 needs to be revised.

Apart of the issues I have already expressed, please find below additional considerations:

– As I have already previously mentioned, an accurate image quantification with the proper replicates would help understand better the conclusions that can be extracted from the experiments. Otherwise, some statements about the results are not fully supported, and visual inspection of the images can lead to different and contradictory conclusions. For instance:

Figure 2 and Figure S1 look qualitatively different:

– In Figure 2, H4pro:mCherry seems equally expressed in the PXYpro:CFP domain and the SMLX5pro:CFP domain, without a clear maxima.

– In Figure S1, H4pro:mCherry it is much clearer its maxima at the interface of both domains, and is not that clear H4pro:mCherry is really much expression in the PXYpro:CFP domain.

Could the authors clarify these differences, and also how this is related to the statement in line 188? I would strongly recommend to do some kind of quantification, with more replicates.

The expression of PXYpro:CFP in Figure 3A looks like this marker is also reaching xylem, not just cambium. Although this seems contradictory with line 231, in which it is stated the PXYpro:CFP is surrounding the xylem cells. Quantification of the fluorescence and clarification in the text will help understand these apparent contradictions. If there is just a thin PXYpro:CFP band surrounding the xylem, it would be good to understand why this band is narrower sometimes and why is wider (eg Figure 2A)

– To have stronger conclusions about the proliferative state of the cell cycle, apart of doing the proper quantification, it might be useful to see the outline of the cells. Perhaps the combination of such data together with the modelling of the H4 marker could help to proof that cell proliferation is enhanced in that cambium region, and would enable to discard other possible interpretations.

– I would strongly suggest to study in depth the pattern formation capabilities of Model 3, including studying how the main parameters impact in the result of the outcomes (or a simplified version of it, as long as it can recapitulate the main perturbations). Perhaps this model needs to be revised, such that the cambium front instability does not occur. But perhaps the authors are in a misleading region of the parameter space. Also, the dynamics is important of these models, so there might be some observables that would be worth quantifying along time.

Reviewer #2 (Recommendations for the authors):

– The main weakness of this paper is, in my opinion, the presentation of the results. The progression through a sequence of models has some rationale – it presumably depicts the path of refinements through which the authors arrived at the final model – but makes this paper tedious to read and obscures key results. It would be beneficial if the paper focused on a well-organized, crisp, mathematically sound presentation of the final model, including all relevant equations, put carefully in the context of the essential molecular-level information, and with clearly stated discussion of lessons learned and shortcomings of the model. The intermediate steps, which may be of interest to researchers who want to continue this line of study, could be delegated to supplementary materials.

– Summary of key molecular players in lines 76-105 is very dense and requires an explanatory figure illustrating the proposed interactions.

– Although ARF5 is mentioned early on (line 101), the role of auxin is downplayed (c.f. lines 441-443). However, the recent paper by Smetana et al. (ref. 10) attributes a significant role to auxin. This discrepancy should at least be discussed.

– The authors tacitly assume that considering the cross-section of an organ such as the hypocotyl suffices to explain the radial organization of its tissues. Perhaps this is true; nevertheless, it is an assumption, which should be clearly stated as such. In particular, if auxin indeed is involved in the tissue patterning, wouldn't its flow in the longitudinal direction play a role?

– There are previous modeling papers dealing with the organization of tissues in shoots and roots, in particular:

Marta Ibañes et al., Brassinosteroid signaling and auxin transport are required to establish the periodic pattern of Arabidopsis shoot vascular bundles, PNAS 2009

N. Fàbregas et al., Auxin Influx Carriers Control Vascular Patterning and Xylem Differentiation in Arabidopsis thaliana, PLOS Genetics 2015

These papers should be cited.

– The idea of distinguishing real substances, tissues etc. from their models using an asterisk is an interesting one; however, in order to be really helpful it should be applied very consistently throughout. Currently there are inconsistencies: for example, in the legend to the right of Figure 5F symbol DF is not starred, but in the figure caption it is.

– The systems of equations representing the models are not adequately presented. For example, Model 1 deals with xylem, phloem and cambium cells, the latter in two states (Figure 1), yet Equations 2-4 apparently characterize only cambium cells. Further models are described even less rigorously. The list of parameters (Table S1) is obscure, as the equations involving listed parameters are not explicitly given. The readers should not be required to reverse-engineer the code to understand how the models really work.

– The models describe a growing system, yet the equations do not include a term representing a decrease in concentrations of molecules in the cells due to cell growth. Why?

– There are also further questions regarding the models. How and on what basis was the initial tissue template (Figure 1A) specified? What are the initial values of the variables? What are the mathematical conditions for attributing types to cells in the tissue?

– Why it the initial frame ( = the template?) in Model 3B (Movies 5x) different from those used previously?

– The table in Figure 5 G includes the heading "Initial values (after run)": What does this mean? (It sounds like an oxymoron.)

– Hollow phrases such as "minimal framework of intercellular communication loops" (lines 37-38) or "reciprocal and interconnected gradients of regulators along the radial sequence of tissues" (line 483) should be avoided, especially in the absence of a proper description of the mathematical models.

– The section on the physical properties of cambium does not adequately address the problem of radial file formation. The question of factors determining the orientation of cell divisions has been extensively addressed in literature (including numerous models), and references to three old papers do not sufficiently position the issue. The algorithm employed in the models to determine the orientation of the divisions is not explained (presumably, the authors just rely on a default algorithm implemented in Virtual Leaf). The observation that stiffening of xylem cell walls may contribute to the organization of cells into files is of some interest, but begs a biomechanical analysis, which is absent.

– Lines 435-437. Listing Turing (reference 43) as representative of "using positional information mediated by morphogenetic gradients" is inappropriate, given the long history of tension between Turing's reaction-diffusion concept and Wolpert's idea of positional information. The paper by J. Green and J. Sharpe, "Positional information and reaction-diffusion: two big ideas in developmental biology combine", Development 2015, provides an authoritative recent perspective on this tension.

– The discussion seems to be largely disconnected from the content of the paper. In particular, the relation of the text beginning in line 449 to the modeling effort is not clear. Is the phrasing "In contrast to our expectations…" supposed to mean "In contrast to the model"? Then, what is the reason for the discrepancy? Or, were these expectations independent of the model, in which case the question arises, how are they related to the topic of the paper, "Computational modelling of cambium activity…"

– Lines 579-580 state: "All simulations within Model 1, Model 2, and Model 3 […] were repeated at least ten times." Why? If the models were deterministic, they would produce the same results. If they included stochastic terms, the results of different runs would differ, but there is no indication of stochastic terms in the paper.

– Line 592 and following: Which parameters were optimized? The statement "The parameter space contains an interval for each parameter from which the parameter value can be chosen" is unclear: what would be a parameter for which a value cannot be chosen? Also, how many runs were performed to optimize? How was the end of the optimization process decided?

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Computational modelling of cambium activity provides a regulatory framework for simulating radial plant growth" for further consideration by eLife. Your revised article has been evaluated by Naama Barkai (Senior Editor) and a Reviewing Editor.

The manuscript has been improved but there are some remaining issues that need to be addressed, as outlined below:

Most of the remaining issues relate to the either lack of detailed explanations or confusing sentences. I suggest authors go carefully through all comments of Reviewers and try to address those for clarification.

The authors should also provide a quantitative analysis of radial fluorescence profiles as they previously performed in Shi et al. 2019. I believe this analysis will further strengthen authors claims. For ease, there is a number of alternative strategies that Reviewer #1 suggested for authors to consider while doing this analysis.

Reviewer #1 (Recommendations for the authors):

I appreciate the replies of the authors to my comments and the points they addressed, I think the quality of the manuscript has improved. See below some other comments I would like to make of this new revised version of the manuscript.

– I still find some statements about the interpretation of the cambium activity fluorescence reporters might require to be better supported. I understand that a detailed quantification might not be possible, but I would strongly recommend authors to do radial intensity profiles to support their statements, as the authors very nicely did in a previous publication [Shi et al. 2019] (e.g. one could do it with Fiji, with a certain line width and doing some binning along the line to smoothen fluctuations).

For instance, these are interpretations of the authors I did not find convincing:

Lines 284-287: 'PXY promoter reporter activity was observed distally to xylem sectors, whereas the SMXL5 promoter activity was as usual present distally to the PXY activity domain. Interestingly, PXYpro:CFP and SMXL5pro:YFP activity domains were still completely distinct '

I would say PXY promoter activity was observed also within the xylem sectors. I am not sure if PXYpro:CFP and SMXL5pro:YFP would be indeed completely distinct, perhaps they overlap?

It would be good to show the radial quantified profiles in WT as well, to better appreciate the differences with the mutants.

If the authors still find quantification is not the way to go, I would suggest doing zooms of the regions of interest, and showing single and composite channels to facilitate the interpretation.

– In the IRX3pro:CLE41 mutant, the statement about lower number of xylem cells should be better supported; looking at the time course of Figure 3—figure supplement 2, earlier time points in the IRX3pro:CLE41 line might suggest this is not the case. I think this raises the question of having more repeats to support this statement (having said that, more repeats of the pxy mutant would be also desirable). Also, I am wondering whether authors are referring to absolute numbers of xylem cells or fraction of xylem cells, could they clarify? Also, to support the claimed statements, a quantification with ilastik of xylem cells might be realistic to do.

– The authors conclude that the cell wall thickness in the procambial cells is smaller than in its surrounding tissues. The way they show it should be revised. First, it is not clear if they do it with Direct Red 23 (line 418) or Direct Yellow 96 (mentioned in the caption of Figure4—figure supplement 1). Second, it is not clear the mean intensity of Direct Red 23 or Direct Yellow could be a good proxy of cell wall thickness – could the authors justify this? (I am not an expert in this topic, but this should be clear and justified to non-experts as myself). In the case of Direct Yellow 96, the mean intensity might be related to the amount of xyloglucans if I understand it well from Ursache et al. 2018; in the case of Direct Red 23, I understand the fluorescence is related to cellulose content at a given part of the cell wall – but not forcely to thickness, and therefore overall stiffness. Third, the quantification using the Radial Profile function might be very misleading, given there can be other factors affecting the outcome, such as the density of cells at a given radial binning, the cell heterogeneity while being a tissue averaged measure, etc – better to do it at a cellular resolution.

– The diffusion operator as described in the appendix, if I understood it well and I am not wrong, it would not fulfil conservation of mass if it is applied to a tissue made of cells of different sizes. For instance, if you apply your operator to two cells with very two different sizes and make the numbers in terms of the exchange number of molecules (i.e., convert the expression of concentrations to number of molecules and cell volumes), the larger cell will have more flux of molecules than the smaller cell. Given the modelled tissue has cell size heterogeneity that can not be avoided, why didn't the authors use laplacians that could follow a conservation of mass such as in Sukumar and Bolander (2003)? I am wondering whether this violation of mass conservation might affect the presented computational results.

– I appreciate the performed model robustness analysis by the authors. For completeness, I think it would be important to include some additional simulations assessing the effect of diffusion rates (and/or the degradation of the modelled diffusible species), at least for CLE41. I believe some parameter ranges might stabilize the front – my suggestion is not to push to an in depth study on the front instability, which I understood is beyond the scope of this manuscript, but rather to ask for a more complete and robust analysis of essential parameters of the model.

Reviewer #1 (Recommendations for the authors):

Apart of the issues I have already expressed in the Public Review, please find below additional considerations:

– The authors need to work more on the clarity about some aspects of the model. How the growth is implemented in all the models? How diffusion is implemented? How the forces are implemented and the impact of the cell wall thickness? How the final timepoint is determined? I believe equation 5 needs to be revised.

Apart of the issues I have already expressed, please find below additional considerations:

We have provided an expanded description of the simulations in VirtualLeaf in the supporting information including the implementation of growth and diffusion as well as the forces. We have included a detailed description of the changes to the source code of VirtualLeaf for Model 3D and illustrated the impact of the cell wall thickness on the simulations. Simulations were run until central features of the tissue were represented by the model – bidirectional tissue formation, positional cell differentiation and distribution of phloem poles. We thank the reviewer for pointing out the issue with equation 6 (previously equation 5), which has now been revised to list the stimulating chemicals as well as their contribution to DF* production.

– As I have already previously mentioned, an accurate image quantification with the proper replicates would help understand better the conclusions that can be extracted from the experiments. Otherwise, some statements about the results are not fully supported, and visual inspection of the images can lead to different and contradictory conclusions. For instance:

Figure 2 and Figure S1 look qualitatively different:

– In Figure 2, H4pro:mCherry seems equally expressed in the PXYpro:CFP domain and the SMLX5pro:CFP domain, without a clear maxima.

– In Figure S1, H4pro:mCherry it is much clearer its maxima at the interface of both domains, and is not that clear H4pro:mCherry is really much expression in the PXYpro:CFP domain.

Could the authors clarify these differences, and also how this is related to the statement in line 188? I would strongly recommend to do some kind of quantification, with more replicates.

As mentioned below, we agree with these concerns toward the H4 marker used in the initial submission. Because H4 expression is not specifically associated with cell division but with DNA synthesis in general and, thus, with endoreduplication, H4 expression does not report faithfully on cell division. As a response, we removed related figures and now reference our previous study characterizing cell division levels in different cambium domains based on cell linage analyses (Shi et al., 2019). Because this is a far more reliable analysis and convincingly supports our claims, we believe that we thereby addressed this concern. Quantification of images is difficult otherwise due to the stochastic (non‐quantitative) nature of anatomical defects in pxy and IRX3:CLE41 lines.

The expression of PXYpro:CFP in Figure 3A looks like this marker is also reaching xylem, not just cambium. Although this seems contradictory with line 231, in which it is stated the PXYpro:CFP is surrounding the xylem cells. Quantification of the fluorescence and clarification in the text will help understand these apparent contradictions. If there is just a thin PXYpro:CFP band surrounding the xylem, it would be good to understand why this band is narrower sometimes and why is wider (eg Figure 2A)

We agree with the reviewer that this statement was misleading. As reported previously (Shi et al., 2019), PXY promoter activity is not only found in proximal cambium cells but also in the xylem tissue itself. We revised this statement accordingly.

– To have stronger conclusions about the proliferative state of the cell cycle, apart of doing the proper quantification, it might be useful to see the outline of the cells. Perhaps the combination of such data together with the modelling of the H4 marker could help to proof that cell proliferation is enhanced in that cambium region, and would enable to discard other possible interpretations.

See our comments above.

– I would strongly suggest to study in depth the pattern formation capabilities of Model 3, including studying how the main parameters impact in the result of the outcomes (or a simplified version of it, as long as it can recapitulate the main perturbations). Perhaps this model needs to be revised, such that the cambium front instability does not occur. But perhaps the authors are in a misleading region of the parameter space. Also, the dynamics is important of these models, so there might be some observables that would be worth quantifying along time.

We thank the reviewers for suggesting a more throughout parameter analysis. We have now re‐run the parameter search until we obtained 5 distinct parameter sets that recapitulate the central features of cambium activity to ensure that the behavior of the model is robust across the analyzed parameter space. We further implemented experimentally determined ratios for cell wall stabilities in phloem, cambium and xylem cells in an effort to improve the model. This has made the xylem region more compact – and in turn also influenced the shape of the cambium around it. However, this did not solve the issue of the edge instability suggesting that other mechanisms are at play here that are beyond the CEL41‐PXY module and simple mechanic mechanisms like cell type specific cell wall stability values. As we do not know how front stability is achieved in planta and, therefore, do not know how to implement this in the computational model, analyzing front instability is currently beyond the scope of our computational modelling approach but a very interesting aspect nevertheless which is now discussed in the Discussion section.

Reviewer #2 (Recommendations for the authors):

– The main weakness of this paper is, in my opinion, the presentation of the results. The progression through a sequence of models has some rationale – it presumably depicts the path of refinements through which the authors arrived at the final model – but makes this paper tedious to read and obscures key results. It would be beneficial if the paper focused on a well-organized, crisp, mathematically sound presentation of the final model, including all relevant equations, put carefully in the context of the essential molecular-level information, and with clearly stated discussion of lessons learned and shortcomings of the model. The intermediate steps, which may be of interest to researchers who want to continue this line of study, could be delegated to supplementary materials.

(See our comments above) We completely see the point of this suggestion. We discussed the possibility to immediately show Model 3 among the authors but the general opinion was that going through Model 1 and 2 first, will help the readers tremendously to follow why Model 3 is designed as it is. As our aim was to identify the minimal framework around the CLE41/PXY signaling module, we started with the most simplistic model possible and develop Model 3 from there by iterative comparisons of in planta and in silico tissue patterns generating, in our view, a quite attractive storyline. To better explain Model 3, we added all equations and a more detailed description in the main text and the supplemental information.

– Summary of key molecular players in lines 76-105 is very dense and requires an explanatory figure illustrating the proposed interactions.

A large body of literature describes these interactions over many years now, which we also cite. Because we want to avoid the impression that we intend to claim that we represent some of the described factors (e.g. WOX4, BES1, …) by our model, we would abstain from such an explanatory scheme as it, in our opinion, would restrict the readers’ view to look for respective representatives in our model.

– Although ARF5 is mentioned early on (line 101), the role of auxin is downplayed (c.f. lines 441-443). However, the recent paper by Smetana et al. (ref. 10) attributes a significant role to auxin. This discrepancy should at least be discussed.

Thanks for this comment. We discuss the issue of not including the role of auxin and the ‘organizing center’ identified in Smetana et al. in the discussion part.

– The authors tacitly assume that considering the cross-section of an organ such as the hypocotyl suffices to explain the radial organization of its tissues. Perhaps this is true; nevertheless, it is an assumption, which should be clearly stated as such. In particular, if auxin indeed is involved in the tissue patterning, wouldn't its flow in the longitudinal direction play a role?

We agree, that longitudinal signaling is one central aspect of our biological system which we were not able to consider in our simulations, due to the immense effort this would require. This aspect is now mentioned in the discussion.

– There are previous modeling papers dealing with the organization of tissues in shoots and roots, in particular:

Marta Ibañes et al., Brassinosteroid signaling and auxin transport are required to establish the periodic pattern of Arabidopsis shoot vascular bundles, PNAS 2009

N. Fàbregas et al., Auxin Influx Carriers Control Vascular Patterning and Xylem Differentiation in Arabidopsis thaliana, PLOS Genetics 2015

These papers should be cited.

These papers are now cited in the discussion part.

– The idea of distinguishing real substances, tissues etc. from their models using an asterisk is an interesting one; however, in order to be really helpful it should be applied very consistently throughout. Currently there are inconsistencies: for example, in the legend to the right of Figure 5F symbol DF is not starred, but in the figure caption it is.

Thanks for the remark. This has been corrected.

– The systems of equations representing the models are not adequately presented. For example, Model 1 deals with xylem, phloem and cambium cells, the latter in two states (Figure 1), yet Equations 2-4 apparently characterize only cambium cells. Further models are described even less rigorously. The list of parameters (Table S1) is obscure, as the equations involving listed parameters are not explicitly given. The readers should not be required to reverse-engineer the code to understand how the models really work.

Again, thanks for raising this important point. We have now included an entire list of the equations for all model versions and all cell types and more detailed explanations of the models (supplemental methods). Additionally, we have expanded Figure 1 to include a schematic representation of the model structure, where cell‐type specific reactions are colored accordingly. We feel that listing all equations for all cell types in the main text of the manuscript would disrupt the reading flow too much. Also, parameters are now explicitly depicted in Supplementary File 2 (former Table S1). We have also enclosed a table with all conditions for cell differentiation processes for all model structures (Supplementary File 1).

– The models describe a growing system, yet the equations do not include a term representing a decrease in concentrations of molecules in the cells due to cell growth. Why?

The equations for all model species include a degradation term, which means that a steady state between production, diffusion and degradation is reached. As these processes occur on a faster timescale that cell growth, we felt that including a factor describing the dilution by cell growth would provide little improvement on the simulations while notably increasing the computational costs for each simulation.

– There are also further questions regarding the models. How and on what basis was the initial tissue template (Figure 1A) specified? What are the initial values of the variables? What are the mathematical conditions for attributing types to cells in the tissue?

The initial tissue template was designed based on a typical Arabidopsis cross section at the onset of radial growth with some simplifications which we mention in the discussion part. We have now listed the conditions for all cell differentiations in the supporting information as part of a more elaborate description of the models.

– Why it the initial frame ( = the template?) in Model 3B (Movies 5x) different from those used previously?

We have re‐run all simulations with the new version of VirtualLeaf and used in all cases the identical starting frame.

– The table in Figure 5 G includes the heading "Initial values (after run)": What does this mean? (It sounds like an oxymoron.)

We have now expanded the parameter search, added information on the parameter values and fitted model behavior to the supporting information and designed a new Figure 5G. The heading “initial values (after run)” was indeed an oxymoron – we were referring to the cell type proportions at the end of the simulation using the initial parameter set of model 3A. This is now included in the new Figure 5G but named “Model 3A”.

– Hollow phrases such as "minimal framework of intercellular communication loops" (lines 37-38) or "reciprocal and interconnected gradients of regulators along the radial sequence of tissues" (line 483) should be avoided, especially in the absence of a proper description of the mathematical models.

Thanks for this comment. We changed or removed these statements.

– The section on the physical properties of cambium does not adequately address the problem of radial file formation. The question of factors determining the orientation of cell divisions has been extensively addressed in literature (including numerous models), and references to three old papers do not sufficiently position the issue. The algorithm employed in the models to determine the orientation of the divisions is not explained (presumably, the authors just rely on a default algorithm implemented in Virtual Leaf). The observation that stiffening of xylem cell walls may contribute to the organization of cells into files is of some interest, but begs a biomechanical analysis, which is absent.

We agree with the reviewer that a biomechanical analysis would nicely complement our study. In fact, we are doing such an analysis at the moment which, however, will require 2‐3 additional years to be completed and is thus beyond the scope of this work. The point of including a computational approach in this regard in our manuscript is to highlight the potential for our tool to address mechanical aspects of plant growth and to generate related hypotheses which was not possible using the VirtualLeaf platform before. We believe that this will be very useful for readers being interested in this aspect and the stepping stone for further analyses, let it be on the computation and/or the biomechanical side. We included am extended description of our models in the supplemental information which we hope is helpful in this context.

– Lines 435-437. Listing Turing (reference 43) as representative of "using positional information mediated by morphogenetic gradients" is inappropriate, given the long history of tension between Turing's reaction-diffusion concept and Wolpert's idea of positional information. The paper by J. Green and J. Sharpe, "Positional information and reaction-diffusion: two big ideas in developmental biology combine", Development 2015, provides an authoritative recent perspective on this tension.

Again, thanks for this remark. We replaced the original citations with the suggested one and indicated the controversial nature of this concept.

– The discussion seems to be largely disconnected from the content of the paper. In particular, the relation of the text beginning in line 449 to the modeling effort is not clear. Is the phrasing "In contrast to our expectations…" supposed to mean "In contrast to the model"? Then, what is the reason for the discrepancy? Or, were these expectations independent of the model, in which case the question arises, how are they related to the topic of the paper, "Computational modelling of cambium activity…"

We would politely disagree with the statement that the discussion part is largely disconnected from the content of the paper. Still, we agree that the mentioned phrase was ambiguous. In this paragraph we deal with the discrepancy of our results (Figure 3G‐I) with the long‐standing view on the role of PXY in xylem formation and not so much with our modelling efforts. We specified this part accordingly.

– Lines 579-580 state: "All simulations within Model 1, Model 2, and Model 3 […] were repeated at least ten times." Why? If the models were deterministic, they would produce the same results. If they included stochastic terms, the results of different runs would differ, but there is no indication of stochastic terms in the paper.

During tissue simulations in VirtualLeaf, all nodes are moved in a random order, which can affect cell sizes (and with that the timing of cell differentiation) and cellular dimensions (which would influence the direction of cell divisions). To ensure that the model simulations are robust with respect to these fluctuations, we ran repetitions for all simulations in all models.

– Line 592 and following: Which parameters were optimized? The statement "The parameter space contains an interval for each parameter from which the parameter value can be chosen" is unclear: what would be a parameter for which a value cannot be chosen? Also, how many runs were performed to optimize? How was the end of the optimization process decided?

In comparison to the first version, we considerably extended our parameter search. We optimized the parameters, which are most closely related to tissue formation: cell differentiation thresholds, cell division thresholds and maximal cell sizes. All of these parameters, including a brief description of the parameter, are listed in Supplementary File 2. As Model 3A was already able to describe central features of cambium activity, we limited the parameter search space to 1/3‐ and 3‐ times the original parameter value. In total 5,000 runs were performed to explore the parameter space. The final simulation results were then evaluated based on the cell type proportions. We stopped the parameter search, once we had obtained five distinct parameter sets, which can reproduce the central features of the tissue morphology. This procedure is explained in the Materials and methods section.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Reviewer #1 (Recommendations for the authors):

I appreciate the replies of the authors to my comments and the points they addressed, I think the quality of the manuscript has improved. See below some other comments I would like to make of this new revised version of the manuscript.

– I still find some statements about the interpretation of the cambium activity fluorescence reporters might require to be better supported. I understand that a detailed quantification might not be possible, but I would strongly recommend authors to do radial intensity profiles to support their statements, as the authors very nicely did in a previous publication [Shi et al. 2019] (e.g. one could do it with Fiji, with a certain line width and doing some binning along the line to smoothen fluctuations).

For instance, these are interpretations of the authors I did not find convincing:

Lines 284-287: 'PXY promoter reporter activity was observed distally to xylem sectors, whereas the SMXL5 promoter activity was as usual present distally to the PXY activity domain. Interestingly, PXYpro:CFP and SMXL5pro:YFP activity domains were still completely distinct '

I would say PXY promoter activity was observed also within the xylem sectors. I am not sure if PXYpro:CFP and SMXL5pro:YFP would be indeed completely distinct, perhaps they overlap?

Thanks for this comment. We agree that the wording was misleading in this case. As a response, we deleted this sentence as is was also redundant with the sentence before and adapted the text accordingly. The text now reads: “In pxy mutants, the xylem tissue did not have a cylindrical shape, but was instead clustered in radial sectors showing PXYpro:CFP and, at their distal ends, SMXL5pro:YFP activity, whereas regions in between those sectors had little to no xylem and did not show reporter activity (Figure 3E, Figure 2—figure supplement 1, Figure 3—figure supplement 1, Figure 3—figure supplement 2).”

It would be good to show the radial quantified profiles in WT as well, to better appreciate the differences with the mutants.

If the authors still find quantification is not the way to go, I would suggest doing zooms of the regions of interest, and showing single and composite channels to facilitate the interpretation.

Again, thanks for this comment. We tried to perform radial profiling of reporter activities in the different backgrounds but, due to anatomical differences in pxy mutants and in IRX3pro:CLE41 plants, signal noise between samples was too large to generate interpretable results. Instead, we now provide three examples of close‐ups of the cambium domain in wild type and pxy mutants (Figure 3‐figure supplement 3). Careful analysis of these images resulted in a modification of our previous statement mentioned above and now reads: “Interestingly, PXYpro:CFP and SMXL5pro:YFP activity domains were still mostly distinct meaning that PXYpro:CFP activity did not expand further beyond established xylem than in wild type (Figure 3E, Figure 3—figure supplement 3). This discrepancy indicated that, in contrast to our assumption, the CLE41‐PXY signaling module did not restrict PXY promoter activity in the distal cambium. Of note, the sharp border between PXYpro:CFP and SMXL5pro:YFP activity was less pronounced in pxy mutants mostly due to a spread of SMXL5pro:YFP activity towards xylem tissues (Figure 3—figure supplement 3).”

– In the IRX3pro:CLE41 mutant, the statement about lower number of xylem cells should be better supported; looking at the time course of Figure 3—figure supplement 2, earlier time points in the IRX3pro:CLE41 line might suggest this is not the case. I think this raises the question of having more repeats to support this statement (having said that, more repeats of the pxy mutant would be also desirable). Also, I am wondering whether authors are referring to absolute numbers of xylem cells or fraction of xylem cells, could they clarify? Also, to support the claimed statements, a quantification with ilastik of xylem cells might be realistic to do.

Thanks for pointing out our imprecise wording. As one response, we added another comparison between wild type, IRX3 plants and pxy mutants (Figure 3‐figure supplement 2). As another response we exchanged the term ‘xylem cells’ by the more correct term ‘vessel elements’ as those are prominently stained by Direct Red 23 and, thus, are the only xylem‐related cell type which can be non‐ambiguously identified. We did not use this term before because we only speak about ‘xylem cells’ in the context of our model and, for non‐familiar readers, we had the feeling that this may be confusing. As the reviewer will see, the second example confirms a reduced density in vessel elements in IRX3 lines (not in differentiated xylem cell number as previously stated). This is what we now state in the text: “PXYpro:CFP activity was found in irregularly shaped patches containing differentiated xylem vessel elements distributed over the whole cross‐section.”

– The authors conclude that the cell wall thickness in the procambial cells is smaller than in its surrounding tissues. The way they show it should be revised. First, it is not clear if they do it with Direct Red 23 (line 418) or Direct Yellow 96 (mentioned in the caption of Figure4—figure supplement 1). Second, it is not clear the mean intensity of Direct Red 23 or Direct Yellow could be a good proxy of cell wall thickness – could the authors justify this? (I am not an expert in this topic, but this should be clear and justified to non-experts as myself). In the case of Direct Yellow 96, the mean intensity might be related to the amount of xyloglucans if I understand it well from Ursache et al. 2018; in the case of Direct Red 23, I understand the fluorescence is related to cellulose content at a given part of the cell wall – but not forcely to thickness, and therefore overall stiffness. Third, the quantification using the Radial Profile function might be very misleading, given there can be other factors affecting the outcome, such as the density of cells at a given radial binning, the cell heterogeneity while being a tissue averaged measure, etc – better to do it at a cellular resolution.

We fully understand the motivation for this comment. As a response, we first changed ‘Direct Red 23’ to ‘Direct Yellow 96’ in the main text as this was indeed incorrect. As a justification for using Direct Yellow 96 as a proxy for cell wall thickness, we’d like to argue that Direct Red 23 and Direct Yellow 96 behave similarly in stainings of root tips which also includes xylem vessels containing secondary cell walls (Ursache et al., 2018). Moreover, also Direct Red 23 application leads to the same relative staining levels as Direct Yellow 96 when comparing cambium stem cells and surrounding cells, a result which is now included in Figure 4—figure supplement 2. Because Direct Red 23 is indeed considered as a cellulose stain (references in Ursache et al., 2018), we believe that using Direct Yellow 96 staining as a proxy for cell wall thickness (i.e. stiffness) is justified. When quantifying Direct Yellow 96 staining for each cell which we performed for a different study (see included picture), the reviewer will see that differences between cambium stem cells and other cells are quite pronounced in some cases and much less in other cases, making it difficult to choose a specific ratio based on cell‐specific measurements. Our choice of 50% stiffness of cambium stem cells in comparison to surrounding cells is also supported by Atomic Force Microscopy of cell walls being present in the cambium region (Arnould et al., 2022). In this study, tension wood is analyzed with very high spatial resolution also revealing the different layer within the cell walls. Ignoring G‐layers which is specific for tension wood, indentation modulus is ~5 GPa (S1+CML) for cambium cell walls and ~10 GPa for differentiated cells (S2). Taken these points together, we prefer to stick to the original approach and use our previous calculation as an approximation for determining a ration between the stiffness of cambium stem cells and surrounding cells. The main text references now the respective analysis.

– The diffusion operator as described in the appendix, if I understood it well and I am not wrong, it would not fulfil conservation of mass if it is applied to a tissue made of cells of different sizes. For instance, if you apply your operator to two cells with very two different sizes and make the numbers in terms of the exchange number of molecules (i.e., convert the expression of concentrations to number of molecules and cell volumes), the larger cell will have more flux of molecules than the smaller cell. Given the modelled tissue has cell size heterogeneity that can not be avoided, why didn't the authors use laplacians that could follow a conservation of mass such as in Sukumar and Bolander (2003)? I am wondering whether this violation of mass conservation might affect the presented computational results.

Thank you for pointing this out. We have now included a scaling factor in the diffusion between cells that ensures mass conservation for these equations. Here, we decided to work with what had already been programmed in VirtualLeaf and chose a scaling factor that is based on the cells’ area. An explanation of this factor is also included in the supplementary material on simulations in VirtualLeaf. We implemented this correction for all models (Model 1 to Model 4) and we have repeated all simulations, including the parameter search. While small differences between the previous simulations can be observed, the qualitative behavior remains the same.

– I appreciate the performed model robustness analysis by the authors. For completeness, I think it would be important to include some additional simulations assessing the effect of diffusion rates (and/or the degradation of the modelled diffusible species), at least for CLE41. I believe some parameter ranges might stabilize the front – my suggestion is not to push to an in depth study on the front instability, which I understood is beyond the scope of this manuscript, but rather to ask for a more complete and robust analysis of essential parameters of the model.

Thank you for this comment. We have now expanded the parameter search to include the diffusion and degradation rate for CLE41 after implementing the correction to ensure mass conservation. While it seems that some parameter ranges result in more edge instability, it does seem like other mechanisms are at play here that are beyond the CLE41‐PXY signaling module.

Author details

1. Ivan Lebovka

   Centre for Organismal Studies, Heidelberg University, Heidelberg, Germany

   Contribution

   Conceptualization, Resources, Software, Formal analysis, Investigation, Visualization, Methodology, Writing - original draft

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0009-0002-7721-0738

2. Bruno Hay Mele

   Department of Agricultural Sciences, Università degli Studi di Napoli Federico II, Napoli, Italy

   Present address

   Department of Biology, University of Naples, Naples, Italy

   Contribution

   Conceptualization, Software, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-5579-183X

3. Xiaomin Liu

   Centre for Organismal Studies, Heidelberg University, Heidelberg, Germany

   Contribution

   Investigation

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-8592-5752

4. Alexandra Zakieva

   Centre for Organismal Studies, Heidelberg University, Heidelberg, Germany

   Contribution

   Formal analysis, Investigation, Methodology, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-6089-3766

5. Theresa Schlamp

   Centre for Organismal Studies, Heidelberg University, Heidelberg, Germany

   Contribution

   Investigation

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-2986-5134

6. Nial Rau Gursanscky

   Gregor Mendel Institute, Vienna Biocenter, Vienna, Austria

   Contribution

   Supervision

   Competing interests

   No competing interests declared

7. Roeland MH Merks

   1. Mathematical Institute, Leiden University, Leiden, Netherlands
   2. Institute of Biology, Leiden University, Leiden, Netherlands

   Contribution

   Conceptualization, Resources, Supervision, Funding acquisition, Project administration, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-6152-687X

8. Ruth Großeholz

   1. Centre for Organismal Studies, Heidelberg University, Heidelberg, Germany
   2. BioQuant, Heidelberg University, Heidelberg, Germany

   Contribution

   Conceptualization, Data curation, Software, Formal analysis, Funding acquisition, Investigation, Methodology, Writing – review and editing

   For correspondence

   ruth.grosseholz@bioquant.uni-heidelberg.de

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-4604-4538

9. Thomas Greb

   Centre for Organismal Studies, Heidelberg University, Heidelberg, Germany

   Contribution

   Conceptualization, Supervision, Funding acquisition, Visualization, Project administration, Writing – review and editing

   For correspondence

   thomas.greb@cos.uni-heidelberg.de

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-6176-646X

• 1,330

  Page views

• 175

  Downloads

• 2

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.