# Source paper archive - Source title: The chloroplast-localised transporter RETICULATA controls palisade mesophyll cell division and expansion - Author(s): Locke Bevia, E. A.; Aldorino, E.; Fitzsimons, J. M.; Dillon, J. C.; Whitewoods, C. D. - Source landing URL: https://www.biorxiv.org/content/10.64898/2026.11.03.759069v1 - PDF URL: https://www.biorxiv.org/content/10.64898/2026.11.03.759069v1.full.pdf - PDF SHA-256: fcfbc0027043779c6e647a94b203348cac24fa548c371c4edaac1df9a41bef50 - Converter: pymupdf4llm 0.0.27 - Limits: Automated PDF text extraction; reading order, equations, tables and formatting may be incomplete or incorrect. Images are omitted, and no OCR was performed. Markup and links are neutralized for safe downloading. Consult the original PDF for the authoritative version. --- [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) # The chloroplast-localised transporter RETICULATA controls palisade mesophyll cell division and expansion Edwyn A. Locke Beviá [1,4,†], Emilio Aldorino [2,4,†], James M. Fitzsimons [4], Jean C. Dillon [3,4] and Chris D. Whitewoods [3,4,*] 1 Centre for Organismal Studies (COS), Ruprecht-Karls-Universität Heidelberg, 69120 Heidelberg, Germany. 2 Department of Developmental Genetics, Centre for Plant Molecular Biology (ZMBP), Eberhard-Karls-Universität, Tübingen, D-72076, Germany. 3 Department of Cell and Developmental Biology, John Innes Centre, Norwich Research Park, Norwich, NR4 7UH, UK. 4 Sainsbury Laboratory, University of Cambridge, 47 Bateman Street, Cambridge, CB2 1LR, UK. - Correspondence: chris.whitewoods@jic.ac.uk - These authors contributed equally to this work. ## 1 KEY WORDS 2 Plant development, Air space, Mesophyll, Leaf morphogenesis, Cell identity, Reticulata ## 3 ABSTRACT 4 Plant leaves contain extensive intercellular air spaces, which contribute to photosynthetic ca5 pacity and water use efficiency in different environments. Despite this importance, the devel6 opmental mechanisms that control leaf air space patterning are poorly understood. For over 7 20 years it has been known that plants with disrupted chloroplast biogenesis and metabolism 8 have increased intercellular air spaces, but the underlying molecular mechanism is unknown. 9 Here, we characterise leaf air space morphogenesis in mutants of the chloroplast amino acid 10 transporter _RETICULATA_ ( _RE_ ), which has enlarged air spaces. We show that the _reticulata-6_ 11 ( _re-6_ ) loss of function mutant exhibits increased air space expansion specifically in the pal12 isade, and not the spongy, mesophyll. We also show that _RE_ does not regulate air space 13 patterning via programmed cell death, via a change in stomatal density or a change in cell 14 identity of the mesophyll tissue. Instead, we quantify mesophyll cell morphology and division 15 rates to show that RE acts by promoting early palisade cell division and suppressing later cell 16 expansion. Overall, we show that RE specifically regulated palisade mesophyll morphogene17 sis and air space patterning and hypothesise that chloroplasts may act to regulate air space 18 patterning in the palisade mesophyll in response to light. 1 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) ## 19 INTRODUCTION 20 Leaves are full of air. In terms of volume, intercellular air spaces (IAS) in the leaves of vascular 21 plants range from 3% (Slaton & Smith, 2002) to 73% (Earles _et_ _al._, 2018). The development 22 of IAS was a key innovation during the transition from aquatic to terrestrial habitats because 23 the resistance for CO 2 diffusion is ~10,000-fold lower in air than in water (Earles _et al._, 2018), 24 which meant plants could profit from an increased photosynthetic rate. However, despite the 25 central role of IAS in maximising leaf photosynthetic capacity (Lehmeier _et_ _al._, 2017; Mathers 26 _et_ _al._, 2018), how IAS patterning unfolds during leaf development remains largely unresolved. 27 Many leaves contain two distinct photosynthetic cell layers: the palisade and the spongy 28 mesophyll (Pyke _et_ _al._, 1991; Bar & Ori, 2014). The palisade layer is located beneath the 29 adaxial epidermis and is characterised by having small air spaces and tightly packed cylindrical 30 shaped cells, optimised for efficient light harvesting (Terashima & Inoue, 1984). In contrast, 31 the spongy mesophyll, which is between the palisade mesophyll and the abaxial epidermis, 32 has larger air spaces and irregularly shaped lobed cells that enhance gas exchange and light 33 scattering (DeLucia _et_ _al._, 1996; Smith _et_ _al._, 1997). This coordinated tissue architecture 34 integrates light capture with internal CO 2 diffusion. 35 Throughout the mesophyll, IAS are formed by cell separation at multicellular junctions (Avery 36 Jr, 1933; Jeffree _et_ _al._, 1986; Fox _et_ _al._, 2018). After initial formation, IAS expansion is 37 controlled separately in the palisade and spongy. In the palisade, further cell separation 38 (schizogeny) is regulated in part by growth of the constraining epidermis (Fitzsimons _et_ _al._, 39 2026), whereas in the spongy, air spaces enlarge by preferential expansion of the air space 40 adjacent mesophyll cell wall (expansigeny) (Pyke _et_ _al._, 1991; Jarvis _et_ _al._, 2003; Zhang _et_ _al._, 41 2021). 42 Signals originating from chloroplasts (Li _et_ _al._, 1995; Hricová _et_ _al._, 2006; Pérez-Pérez _et_ 43 _al._, 2013; reviewed in Lundquist _et_ _al._, 2014) and stomata (Masle _et_ _al._, 2005; Dow _et_ _al._, 44 2017; Lundgren _et_ _al._, 2019; reviewed in Baillie & Fleming, 2020) are also known to alter 45 IAS patterning, but the molecular and developmental mechanisms by which they work are 46 unknown. 47 Mutants in chloroplast biogenesis and metabolism have enlarged IAS with irregularly shaped 48 palisade mesophyll cells (Li _et_ _al._, 1995; Chatterjee _et_ _al._, 1996; Keddie _et_ _al._, 1996; Voll _et_ 49 _al._, 2003). The best studied of these is the mutant _reticulata_ ( _re_ ) (Barth & Conklin, 2003; 50 González-Bayon _et_ _al._, 2006; Pérez-Pérez _et_ _al._, 2013; Kuhnert _et_ _al._, 2025), which was identi51 fied by George Rédei in 1964 (Rédei & Hironyo, 1964). This mutant is named for its distinctive 52 leaf colouration of marked dark green veins on a pale green lamina (Fig. 1). _RE_ is a nuclear 53 gene that encodes a basic amino acid transporter localised to the inner plastid envelope, 54 and homologs of this protein are solely found in plastid containing organisms (Kuhnert _et_ 55 _al._, 2025). _RE_ is mainly expressed in the proximal region of young leaves, whereas in older 56 leaves expression is observed in the bundle sheath and surrounding cells, and in the hydath57 odes (Pérez-Pérez _et_ _al._, 2013). _re_ mutants show no detectable differences in chloroplast 2 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) **Fig.** **1.** **The** _**reticulata**_ **mutant** **is** **characterised** **by** **a** **dark** **green** **venation** **on** **a** **pale** **green** **lamina.** (A) First leaf of a 23 days after stratification (DAS) _A._ _thaliana_ wild-type plant. (B) First leaf of a 23 DAS _A._ _thaliana_ _re-6_ mutant plant. Scale bars: 1 mm. 58 number or morphology relative to wild-type (Barth & Conklin, 2003; González-Bayon _et_ _al._, 59 2006), meaning that its effects on leaf structure are likely through changes in metabolites or 60 molecular signals rather than through lack of chloroplast function. 61 Although the mechanism for IAS enlargement in the _re_ mutant is not known, it has been 62 proposed that it is caused by mesophyll cell death due to increased reactive oxygen species 63 (ROS) in mature leaves (Pérez-Pérez _et_ _al._, 2013). In this view, the enhanced air spaces are 64 a pathological outcome of chloroplast dysfunction, and RE (and chloroplasts) may have no 65 physiological role in IAS patterning in wild-type. However, cell death has not been observed 66 in _re_ mutant palisade cells, and ROS has been linked to altered cell expansion during leaf 67 development (Schmidt _et_ _al._, 2016). Therefore, we hypothesise that enlarged IAS in _re_ are 68 not due to cell death but are instead an outcome of either 1– altered palisade division and 69 expansion or 2– a change in adaxial mesophyll cell identity from palisade to spongy. Either of 70 these possibilities would suggest that RE actively regulates IAS patterning. 71 Here we investigate how _RE_ controls air space development. We use confocal microscopy to 72 visualise IAS development in _re-6_ mutant leaves throughout development, quantify cell death 73 using fluorescent dyes and use cell type-specific markers to observe any changes in palisade 74 and spongy mesophyll identity. We show that the _re-6_ mutant’s enlarged air spaces are 75 specific to palisade tissue and form gradually throughout development as a result of reduced 76 palisade cell division, rather than through cell death or a switch in cell identity. This suggests 77 that RE is required for mesophyll morphogenesis and IAS patterning. 3 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) ## 78 RESULTS 79 **The** _**re-6**_ **mutation** **specifically** **disrupts** **IAS** **and** **cell** **morphology** **in** **the** **palisade** **mes-** 80 **ophyll** 81 To compare IAS expansion during leaf development in both _A._ _thaliana_ wild-type and _re-6_ 82 mutant plants, we quantified porosity (percentage of air space) in the first rosette leaf pair 83 across different timepoints, from initial air space formation to leaf maturity. We performed 84 our analysis on the first two true leaves because they are initiated almost simultaneously 85 at opposite positions of the SAM (Van Lijsebettens & Clarke, 1998), and share the reticulate 86 phenotype with older leaves. To quantify porosity we used two different dyes: propidium 87 iodide or Nile Red. For leaves sampled 8 days after stratification (DAS) we used Nile Red in 88 Silicon oil, which enters through stomata and visualises IAS via green fluorescence (Kawase 89 _et_ _al._ (2015); Kawase _et_ _al._ (2016); Fig. 2A,E). Prior to 8 DAS Nile Red does not stain IAS as 90 stomata were not yet open. Instead, we used propidium iodide staining to highlight individual 91 cells (grey), which contrasts with IAS (black) after a process of tissue fixation and clearing 92 (see Materials and Methods and Fig. 2A,E). 93 Since leaf development progresses basipetally (Pyke _et_ _al._, 1991; Donnelly _et_ _al._, 1999; Fox 94 _et_ _al._, 2018), we quantified porosity at the upper half of the leaf blade across the different 95 timepoints (Fig. 2B,C). In the wild-type, and consistent with previous observations (Fitzsimons 96 _et al._, 2026), IAS expansion in the palisade and spongy mesophyll followed distinct trajectories: 97 IAS in the palisade increased slowly before reaching a plateau of ~20% porosity at 14 DAS, 98 whereas the spongy mesophyll expanded more rapidly and plateaued at ~50% (Fig. 2B,C). The 99 steepest phase of IAS expansion occurred between 7-8 DAS in the palisade (3 to 13%) and 100 between 6-8 DAS in the spongy mesophyll (8 to 34%). 101 Although the growth of wild-type and _re-6_ leaves was not significantly different (Fig. 2D), in 102 _re-6_, the rate of IAS expansion in the palisade mesophyll was markedly accelerated (Fig. 2A,B), 103 while the spongy mesophyll followed a wild-type-like pattern (Fig. 2C,E). Palisade porosity in 104 _re-6_ was only slightly elevated at 5 DAS (0.5% for wild-type and 2% for _re-6_ ), and from 6 105 DAS onward became significantly higher than in the wild-type, increasing from 12 to 45% by 106 14 DAS, compared with 2 to 20% in the wild-type. Thus, _re-6_ palisade porosity is 2.3-fold 107 higher than in the wild-type by the time it plateaus, despite both genotypes reaching a plateau 108 at roughly the same time (14 DAS). Cell density and cell shape in the palisade mesophyll of 109 _re-6_ visually differ from wild-type, resembling the lobed morphology of spongy mesophyll 110 cells (Fig. 2A,E). Together, these results suggest that the _re-6_ mutation specifically disrupts 111 palisade mesophyll development, affecting both IAS patterning and cell morphology. 4 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) **Fig.** **2.** **The** _**re-6**_ **mutation** **disrupts** **palisade,** **but** **not** **spongy,** **mesophyll** **development.** (A) Air spaces in the palisade mesophyll of _re-6_ (top row) and wild-type (bottom row) visualised by propidium iodide (5–7 DAS) or Nile Red dye (8–23 DAS). (B,C) Palisade (B) and spongy (C) mesophyll tissue porosity (%) of wild-type and _re-6_ first rosette leaf pair over 23 DAS. Significance letters above violins derive from a one-way ANOVA followed by Tukey’s HSD posthoc test. Yellow horizontal lines on violins indicate mean values. (D) Progression of leaf area over time of _A._ _thaliana_ first rosette leaf pair for wild-type and _re-6_ with a generalised additive model (GAM) fit and 99% confidence interval. (E) Air spaces in the spongy mesophyll of _re-6_ (top row) and wild-type (bottom row) visualised by propidium iodide (6 DAS) or Nile Red dye (8–23 DAS). With propidium iodide, air spaces are black and cells grey; with Nile Red air spaces are green and cell outlines magenta. Each point represents an individual leaf. Scale bars: 50 µm. 5 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 112 **Altered** **adaxial-abaxial** **stomatal** **patterning** **does** **not** **account** **for** **increased** **air** **space** 113 **in** _**re-6**_ 114 Previous work has shown that stomatal density is positively correlated with palisade mesophyll 115 porosity (Dow _et al._, 2017; Lundgren _et al._, 2019). Therefore, to test whether the enlarged IAS 116 in the _re-6_ mutant palisade is a secondary effect driven by stomata, we quantified stomatal 117 density (stomata/mm [2] ) in both the adaxial and abaxial epidermis of the first rosette leaf 118 pair of _A._ _thaliana_ wild-type and _re-6_ plants by using Nile Red staining to visualise stomata 119 (Fig. 3A). We saw no significant difference in stomatal density between wild-type and _re-6_ 120 leaves other than at 10 DAS on the adaxial epidermis and at 8 DAS on the abaxial epidermis 121 (Fig. 3B,C). In both cases, wild-type stomatal density was 16 and 19% lower, respectively, 122 than in _re-6_ . Given the transient nature of these differences during leaf development, these 123 data suggest the specific increase in the palisade mesophyll porosity of _re-6_ is not due to an 124 indirect effect of altered stomatal patterning. **Fig.** **3.** **Stomatal** **density** **in** _**re-6**_ **adaxial** **and** **abaxial** **epidermis** **is** **comparable** **to** **wild-** **type.** (A) Representative adaxial and abaxial epidermis of 23 DAS _re-6_ (top row) and wild-type (bottom row) visualised by Nile Red dye. Cell outlines magenta. (B,C) Adaxial (B) and abaxial (C) stomatal density (stomata/mm [2] ) of wild-type and _re-6_ first rosette leaf pair over 23 DAS. Significance letters above violins derive from a one-way ANOVA followed by Tukey’s HSD posthoc test. Yellow horizontal lines on violins indicate mean values. Each point represents an individual leaf. Scale bars: 100 µm. 125 **Palisade** **cell** **death** **does** **not** **contribute** **to** **IAS** **formation** **in** _**re-6**_ 126 Previous studies have proposed that the increased IAS volume observed in _re_ mutants may 127 result from mesophyll cell death associated with elevated ROS in mature leaves (Barth & 128 Conklin, 2003; Overmyer _et_ _al._, 2008; Pérez-Pérez _et_ _al._, 2013). However, cell death has never 129 been documented during _re_ leaf development. Therefore, to determine if the mechanism for 130 IAS enlargement in _re-6_ is caused by palisade mesophyll cell death, we used Sytox Orange to 131 stain dead cells in rosette leaf one of _A._ _thaliana_ wild-type and _re-6_ plants at 8 and 14 DAS 132 (Fig. 4A). To verify we could reliably detect dead palisade cells with our assay, we performed 133 targeted laser ablations of palisade cells followed by Sytox Orange staining. The ablated 134 palisade cells were successfully stained, confirming the effectiveness of the staining method 6 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 135 for detecting cell death in this tissue (Fig. S1). We detected no significant difference in the 136 proportion of dead palisade cells between wild-type and _re-6_ leaves at either timepoint. In 137 wild-type leaves, dead cells represented 4% of palisade cells at 8 DAS and 0.4% at 14 DAS, 138 whereas in _re-6_, dead cells accounted for 2% of palisade cells at 8 DAS, and no dead palisade 139 cells were detected at 14 DAS (Fig. 4B). Together, these results indicate that the specific 140 increase in palisade mesophyll IAS in _re-6_ is not caused by cell death. Instead, they suggest 141 RE may play a physiological role in IAS patterning in wild-type leaves. 142 **Increased** **air** **space** **in** _**re-6**_ **is** **not** **caused** **by** **a** **switch** **in** **cell** **identity** 143 As palisade mesophyll cell morphology in _re-6_ at the end of leaf expansion resembles the 144 lobed spongy mesophyll cells (Fig. 2A, E), we hypothesised that the enlarged IAS in _re-6_ 145 might result from a switch in mesophyll cell identity from palisade to spongy. To determine if 146 this is the case, we examined palisade and spongy cell identity in wild-type and _re-6_ leaves at 147 14 DAS using the cell type-specific reporters _IQD22::GUS-mCit_ and _CORI3::GUS-mCit_, which 148 mark palisade (Fig. 4C) and spongy (Fig. 4E) mesophyll cells, respectively (Procko _et_ _al._, 2022). 149 _IQD22_ expression in the _re-6_ mutant remained detectable in palisade cells (Fig. 4D), indicating 150 retention of palisade cell identity. Likewise, _CORI3_ expression remained confined to spongy 151 mesophyll cells in _re-6_, whereas palisade cells lacked detectable expression (Fig. 4F) despite 152 their spongy-like morphology. These data suggest that palisade mesophyll cell identity in 153 _re-6_ has not switched to spongy mesophyll. Therefore, the increased palisade mesophyll IAS 154 observed in _re-6_ cannot be explained as a result of altered mesophyll cell identity. 155 **The** _**re-6**_ **mutation** **reduces** **palisade** **cell** **proliferation** **rate** **and** **alters** **cell** **morphology** 156 Since _re-6_ appeared to have altered palisade mesophyll cell density and morphology (Fig. 2A), 157 we quantified these traits throughout development of the first rosette leaf pair of _A._ _thaliana_ 158 wild-type and _re-6_ plants to determine if the enlarged IAS in _re-6_ palisade could be explained 159 by reduced cell density in this tissue and to assess to what extent the _re-6_ mutation affects 160 palisade cell morphology. As above, quantifications were carried out in the upper half of the 161 leaf blade (Fig. 5B-F) using CellposeSAM v4.0.4 cellular segmentation software (Pachitariu _et_ 162 _al._, 2025) on images of palisade mesophyll stained with Nile Red (as in Fig. 2A). An example of 163 the cellular segmentation of the palisade mesophyll for a 23 DAS wild-type leaf can be seen 164 in Fig. 5A. 165 In the _re-6_ mutant, the previously observed accelerated rate of IAS expansion in the palisade 166 mesophyll (Fig. 2A,B) was matched by a proportional and gradual decrease in cell density 167 throughout leaf development relative to the wild-type (Fig. 5B). As with palisade porosity, 168 differences in cell density arose at 6 DAS. From this timepoint onward, _re-6_ consistently 169 exhibited lower cell density, reaching a 2.8-fold reduction compared with the wild-type by 170 23 DAS (Fig. 5B). Given the similar leaf area, these findings suggest that the enlarged IAS 171 phenotype in _re-6_ palisade mesophyll could be explained by a reduced proliferation rate of 172 the palisade cells. To test whether the palisade mesophyll cell division rate was decreased 173 in _re-6_, we performed clonal analysis by using heat shock-inducible CRE-mediated lox site 7 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) **Fig.** **4.** **Increased** **IAS** **in** **the** _**re-6**_ **palisade** **is** **not** **caused** **by** **cell** **death** **or** **a** **switch** **in** **cell** **identity.** (A) Representative images of palisade mesophyll of rosette leaf one stained with Sytox Orange at 8 and 14 DAS (left and right columns, respectively). _re-6_ (top row) and wild-type (bottom row). Chloroplast autofluorescence is visualised in cyan and Sytox Orange in yellow. Scale bars: 20 µm. (B) Proportion of damaged and healthy palisade mesophyll cells of wild-type and _re-6_ rosette leaf one at 8 and 14 DAS. Statistical significance was calculated with Welch’s two-sample t-test (unequal variances): *P<0.05, **P<0.01, ***P<0.001, ns = not significantly different. (C-F) Representative images of transverse sections of leaf one from wild-type _IQD22::GUS-mCit_ (C), _re-6_ _IQD22::GUS-mCit_ (D), wild-type _CORI3::GUS-mCit_ (E) and _re-6_ _CORI3::GUS-mCit_ (F). Calcofluor white is visualised in cyan, mCitrine in yellow and chloroplast autofluorescence in magenta. Scale bars: 100 µm. 8 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 174 recombination to induce YFP expression (Fig. 5G, H) (Kuchen _et_ _al._, 2012; Lee _et_ _al._, 2019; 175 Fitzsimons _et_ _al._, 2026). We induced YFP expression in single cells and counted the number 176 of cells in each clonal sector 3 days later as a proxy for cell division rate. These data show 177 that cell number per sector was significantly different between wild-type (3.94 cells/sector) 178 and _re-6_ (2.59 cells/sector; Fig. 5I), the latter having on average 1.35 less cells per sector 179 compared to the wild-type. This is equivalent to 0.63 divisions per day in wild type and 0.42 180 divisions per day in _re-6_, indicating that the proliferation rate of palisade cells in _re-6_ is 181 around 30% lower than that of the wild-type in the 4 to 7 DAS period of leaf development. 182 Notably, reduction of palisade cell division rate in _re-6_ occurs right before the steepest phase 183 of IAS expansion (7-8 DAS), when IAS go from 17 to 35% (Fig. 2B). Thus, it seems likely 184 that the enlarged IAS phenotype in _re-6_ palisade mesophyll is caused, at least in part, by a 185 reduction in the cellular division rate early on in leaf development. 186 Throughout leaf development, _re-6_ palisade cells appear to be bigger than those of the wild187 type, although this difference only becomes significant from 14 DAS onward (Fig. 5C). By 23 188 DAS, _re-6_ palisade cells are on average 1.8 times larger than those of the wild-type. Given 189 that from 6 DAS onward palisade cell density was lower in _re-6_ (Fig. 5B), but clear differences 190 in cell size emerge only from 14 DAS onward (Fig. 5C), it is possible that the enlarged cell 191 size phenotype in the _re-6_ palisade arises as an indirect effect to compensate for the reduced 192 palisade cell division rate early in leaf development or that RE has an effect on cell division 193 early on and an effect on cell expansion later. 194 To further characterise how cell morphology changes over time, we quantified three shape 195 descriptors for palisade cells of _re-6_ and wild-type in the paradermal plane: aspect ratio, 196 circularity and solidity. Aspect ratio is defined as the ratio of the major to the minor axis 197 of each cell, such that higher values indicate more elongated cells. Circularity is calculated 198 as (4π·area)/perimeter [2] ; a value of 1 indicates a perfect circle, whereas values approaching 199 0 indicate increasingly elongated or irregular shapes. Solidity is defined as the ratio of the 200 cell area to its convex area, the latter being the area of the smallest convex polygon that 201 fully encloses the cell. A solidity of 1 indicates that a cell has no indentations, while values 202 <1 reflect the presence of lobes, concavities, or irregular boundaries. Wild-type palisade cells 203 start off as small, densely packed polyhedral cells, and gradually acquire their characteristic 204 columnar shape, which looks roughly like a circle in a transverse cut (Fig. 2A). The palisade cells 205 in _re-6_ initially also follow this developmental trajectory, however, after 14 DAS their aspect 206 ratio values increase, becoming notably more elongated than those of the wild-type (Fig. 5D). 207 In terms of the circularity, _re-6_ palisade cells were more circular than wild-type cells up until 208 10 DAS, probably because they are less tightly packed in _re-6_ . But after 12 DAS, instead 209 of plateauing like the wild-type cells, the circularity values of _re-6_ palisade cells dropped 210 significantly (Fig. 5E), showing that the palisade of _re-6_ undergoes drastic cell morphological 211 changes. Likewise, this trend was mirrored by the solidity values (Fig. 5F), demonstrating the 212 emergence of deep lobes and concavities in _re-6_ palisade cells. Although visually the palisade 213 cells of _re-6_ seem to acquire a spongy-like morphology at 23 DAS (Fig. 2A,E), the values 214 for the three shape descriptors were significantly different between the palisade mesophyll 9 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) **Fig.** **5.** **Palisade** **mesophyll** **cells** **of** _**re-6**_ **divide** **more** **slowly** **and** **are** **larger** **than** **wild-** **type.** (A) Snapshot of the palisade mesophyll of a 23 DAS wild-type leaf segmented with CellposeSAM v4.0.4. (B–F) Palisade cell (B) density (cells/mm [2] ), (C) area (µm [2] ), (D) aspect ratio, (E) circularity and (F) solidity of wild-type and _re-6_ first rosette leaf pair over 23 DAS. Each data point represents the mean value for a leaf. (G, H) Wild-type (G) and _re-6_ (H) palisade mesophyll clonal sectors expressing YFP in a 7 DAS leaf 3 days after heat-shock induction. Chlorophyll autofluorescence is visualised in grey, YFP in yellow. (I) Quantification of cells in clonal sectors expressing YFP in palisade mesophyll of 7 DAS leaves induced at 4 DAS. Data represent individual sectors. Statistical significance was calculated with Welch’s two-sample t-test (unequal variances): *P<0.05, **P<0.01, ***P<0.001, ns = not significantly different. Yellow horizontal lines on violins indicate mean values. Scale bars: 50 µm. 10 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 215 cells of _re-6_ and the spongy mesophyll cells of wild-type at 23 DAS ( _P_ =0.00065 for aspect 216 ratio, _P_ =0.00007 for circularity and _P_ =0.00002 for solidity), consistent with the hypothesis 217 that palisade mesophyll identity in _re-6_ has not switched to spongy mesophyll. Together, 218 the quantification of these shape descriptors throughout palisade development show that _re-_ 219 _6_ palisade cells undergo progressive morphological changes that complement the observed 220 reduction in cell density and division rate, and the increase in IAS. Thus, indicating that 221 the _re-6_ mutation not only decreases palisade cell proliferation but also disrupts palisade 222 differentiation, leading to enlarged, irregularly shaped cells in the mutant. 223 To continue with the quantitative characterisation of the spongy mesophyll of _re-6_, we per224 formed the same analysis on the first rosette leaf pair of _A. thaliana_ wild-type and _re-6_ plants 225 at 23 DAS (Fig. S2), as described above for the palisade. Consistent with the qualitative ob226 servations shown in Fig. 2E and the wild-type-like patterning of spongy mesophyll porosity 227 throughout leaf development (Fig. 2C), spongy mesophyll cell density at 23 DAS remained wild228 type-like (Fig. S2A). Likewise, spongy mesophyll cell area (Fig. S2B), circularity (Fig. S2D), and 229 solidity (Fig. S2E) remained unchanged in _re-6_ . In contrast, cell aspect ratio was slightly but 230 significantly higher in _re-6_ than in the wild-type ( _P_ =0.0494; Fig. S2C). Together, these find231 ings indicate that, unlike the palisade mesophyll, the spongy mesophyll is largely unaffected 232 by the _re-6_ mutation, exhibiting only a subtle change in cell shape at 23 DAS. ## 233 DISCUSSION 234 Here we show that RE is required for normal palisade mesophyll morphogenesis: in _re-6_, re235 duced proliferation in the palisade is followed by increased cell area and the development 236 of irregularly-shaped, lobed cells, while spongy mesophyll morphogenesis occurs as normal. 237 These changes occur without altered stomatal density, with no detectable increase in pal238 isade cell death, and without a spatial shift in molecular markers of cell identity. Together, 239 our findings demonstrate that _RE_ is required for palisade mesophyll morphogenesis and IAS 240 patterning. 241 **RE** **is** **specifically** **required** **for** **normal** **palisade** **mesophyll** **morphogenesis** 242 The observation that loss of RE function affects palisade, but not spongy, mesophyll mor243 phogenesis suggests that these cell types differ in their dependence on RE-mediated plastid 244 function. This selectivity occurs even though _RE_ expression is not restricted to the palisade 245 mesophyll (Pérez-Pérez _et al._, 2013), raising the question of how a broadly expressed gene can 246 specifically affect morphogenesis of one cell type. There is precedent for molecular mecha247 nisms that regulate air space patterning differentially between the spongy and palisade: our 248 recent work showed that global alterations in brassinosteroid signalling affect IAS expansion 249 in the palisade, but not the spongy mesophyll (Fitzsimons _et_ _al._, 2026). Similarly, changes in 250 stomatal density specifically alter palisade mesophyll cell density (Masle _et_ _al._, 2005; Lee _et_ 251 _al._, 2015; Dow _et_ _al._, 2017; Lehmeier _et_ _al._, 2017; Lundgren _et_ _al._, 2019). 11 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 252 In the _re-6_ mutant, palisade cells acquire a more spongy-like morphology but retain _IQD22_ 253 expression and lack detectable _CORI3_ expression. These expression patterns and the fact 254 that _re-6_ palisade cell shape is still significantly different to wild-type spongy mesophyll cell 255 morphology suggest that there is no wholesale conversion of palisade cell identity and high256 light that palisade cell identity and morphology can at least be partially uncoupled. Overall, 257 these data suggest that RE may act downstream of palisade cell identity to control palisade 258 morphogenesis, although its mode of action remains unknown. 259 _**RE**_ **regulates** **air** **space** **patterning** **via** **palisade** **proliferation,** **independently** **of** **cell** 260 **death** **and** **stomatal** **density** 261 The data presented here show that RE controls early palisade proliferation: in _re-6_, cell density 262 and tissue porosity diverge from wild-type around 6 DAS, and clonal analysis demonstrates 263 reduced cell proliferation around this time. As increased cell size becomes significant at 264 around 14 DAS, it is possible that RE primarily affects early palisade cell division, with later 265 effects on cell morphology arising as indirect results of compensation for reduced cell number, 266 as overall leaf size is maintained (Ferjani _et_ _al._, 2007). However, it is also possible that RE 267 affects both directly. 268 Previous studies in _A. thaliana_ and wheat have shown that the degree of mesophyll porosity is 269 positively correlated with the presence of open functional stomata (Dow _et al._, 2017; Lundgren 270 _et_ _al._, 2019). In _re-6_ the density of functional stomata in the adaxial and abaxial epidermis is 271 comparable to the wild-type, but IAS is enlarged. In addition, IAS enlargement begins before 272 stomata are open. Therefore it is unlikely that RE affects IAS in the palisade mesophyll 273 indirectly via a change in stomatal density. 274 Other studies have proposed that increased IAS volume in the _re_ mutant is caused by mes275 ophyll cell death due to increased reactive oxygen species (ROS) in mature leaves (Barth & 276 Conklin, 2003; Overmyer _et_ _al._, 2008; Pérez-Pérez _et_ _al._, 2013). However, none of these stud277 ies documented any palisade cell death during _re-6_ leaf development. Here, we show that, 278 during early leaf development where air spaces diverge from wild-type, cell death is not ac279 countable for the enlarged air spaces in the palisade of _re-6_ . Instead, our findings suggest 280 that enlarged air spaces in _re-6_ are due to the reduced proliferation rate of the palisade cells 281 early on in leaf development. 282 **Linking** **plastid** **amino** **acid** **homeostasis** **to** **palisade** **development** 283 The recent identification of RE as a chloroplast amino acid transporter provides a biochem284 ical starting point to understand its developmental function. RE transports arginine, lysine, 285 ornithine and citrulline across the chloroplast inner envelope, and mutants have disrupted 286 amino acid homeostasis (Kuhnert _et_ _al._, 2025). 287 One possible connection to cell proliferation is through polyamine metabolism. In _A._ _thaliana_, 288 polyamines such as putrescine and spermidine are synthesised from arginine through the argi289 nine decarboxylase pathway (Hanfrey _et_ _al._, 2001). Polyamines have been linked to cell pro290 liferation in tobacco leaves (Paschalidis & Roubelakis-Angelakis, 2005), in mesophyll-derived 12 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 291 cell cultures of several species (Tiburcio _et_ _al._, 1986; Majewska-Sawka _et_ _al._, 1997) and in _A._ 292 _thaliana_ roots (Ko _et_ _al._, 2026). Therefore, disruption of arginine homeostasis in _re-6_ could 293 alter polyamine levels and reduce early palisade cell proliferation. Quantifying polyamine 294 abundance and function in a cell type-specific manner during early leaf development would 295 determine whether this pathway contributes to the _re_ phenotype and if a limitation in the 296 supply of polyamines would affect the development of the palisade but not the spongy mes297 ophyll. 298 The cellular site of RE action is also currently unresolved. High levels of expression in the 299 bundle sheath have led to the hypothesis that the bundle sheath supplies either essential 300 metabolites (Rosar _et al._, 2012) or a differentiation signal for mesophyll development (Knappe 301 _et_ _al._, 2003; Pérez-Pérez _et_ _al._, 2013; reviewed in Lundquist _et_ _al._, 2014). Alternatively, RE 302 may act within palisade cells themselves to control development cell-autonomously. Tissue303 and clonal sector-specific rescue would help distinguish these possibilities, while single-cell 304 metabolomics and transcriptomics of _re_ may identify specific plastid metabolic defects that 305 are associated with altered leaf structure and show how RE affects development differently 306 in the palisade and spongy mesophyll. 307 In addition, since the reticulate phenotype of _re_ can be rescued when grown in a short-day 308 photoperiod (Overmyer _et al._, 2008; Pérez-Pérez _et al._, 2013) and under very low light (Barth & 309 Conklin, 2003; Overmyer _et al._, 2008), it is possible that under these conditions the metabolic 310 demand of the developing leaf is reduced, allowing the system to compensate and thereby 311 suppress the reticulate phenotype. Thus, it is plausible that RE function links the regulation 312 of palisade development in response to light levels and the metabolic state of the cell. ## 313 CONCLUSIONS 314 Overall, our findings demonstrate that RE regulates palisade mesophyll morphogenesis and 315 IAS patterning by promoting early palisade cell division and regulating later cell expansion 316 and morphology. Understanding how RE-dependent plastid amino acid homeostasis regulates 317 these processes will help reveal how chloroplast metabolism shapes leaf development. More 318 broadly, future work in this area may reveal how chloroplasts integrate environmental signals 319 to shape plastic leaf development. ## 320 MATERIALS AND METHODS 321 **Plant** **material** **and** **growth** **conditions** 322 The Col-0 wild-type and _re-6_ mutant plants were obtained from the Nottingham Arabidopsis 323 Stock Centre (NASC). To generate transgenic plants for sector analysis, wild-type plants were 324 transformed with the 71891 plasmid by floral dip and lines selected based on resistance to 325 BASTA and the ability to produce fluorescent sectors after heat shock. A sector line was then 13 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 326 crossed to _re-6_ to generate sectors in the _re-6_ mutant background. 327 Prior to sowing, seeds were sterilised by incubating first for 10 min with 70% ethanol and 328 0.05% (w/v) sodium dodecyl sulfate (SDS) and later for 5 min with 100% ethanol. Plants 329 were grown on plates containing MS media [0.441% Murashige and Skoog including vitamins, 330 0.05% (w/v) MES, 0.8% Difco agar, pH to 5.7]. Sterilised seeds were stratified in the dark 331 at 4°C for 2 days before growing in a controlled environment chamber at 20°C, 50% relative 332 humidity and 16 h of illumination (150 µmol·m [−2] - s [−1] ). 333 **Construct** **generation** 334 The _IQD22::GUS-mCit_ and _CORI3::GUS-mCit_ lines were obtained from Carl Procko (Procko _et_ 335 _al._, 2022) and crossed with the _re-6_ mutant to generate, respectively, palisade and spongy 336 mesophyll-specific marker lines in the _re-6_ background. Heterozygous plants were imaged 337 for _IQD22::GUS-mCit_ / _re-6_, and homozygous plants were imaged for _CORI3::GUS-mCit_ / _re-6_ . 338 The heatshock sector construct 71891 was generated by golden gate cloning (see Table S1 for 339 cloning scheme, and supplemental file 71891.gb for construct sequence). Construct design was 340 based on the brainbow 1.0 construct where induction generates mutually exclusive sectors of 341 either YFP or RFP (Livet _et_ _al._ (2007), Fig. S3). 342 **Staining** **methods** **and** **confocal** **imaging** 343 Nile Red dye (Sigma-Aldrich) was dissolved in 1.0 cs Silicon oil (Aston Chemicals) by shaking 344 overnight to a final concentration of 10 µM. Leaves were stained by soaking in Nile Red 345 solution for 10 min before mounting on a slide under a coverslip and imaging on a Leica SP8 346 confocal microscope using a 20× dry objective. Nile Red bound to cell membranes was excited 347 with a 561 nm wavelength laser and visualised between 565 and 640 nm, and Nile Red free 348 in solution was excited with a 488 nm wavelength laser and visualised between 490 and 560 349 nm. 350 Staining with propidium iodide (PI) was performed by incubating _A._ _thaliana_ seedlings with 351 shaking for 2 h in three parts 100% ethanol and one part glacial acetic acid, then we washed 352 the seedlings in 10x PBS, followed by staining with 250 ng/mL of PI for 40 min and then 353 leaving the seedlings in ClearSee (Kurihara _et_ _al._, 2015) overnight. PI was excited with a 354 561 nm wavelength laser and visualised between 590 and 650 nm. To quantify leaf porosity 355 (percentage air space in a region of tissue) unbound Nile Red fluorescence or areas with no 356 PI fluorescence were quantified by automatic thresholding in ImageJ as described previously 357 (Kawase _et_ _al._, 2015, 2016), and porosity was defined as the percentage area of a 2D region 358 that contained air space. To accurately identify palisade and spongy mesophyll tissue through359 out development before defining morphological characteristics emerged, palisade and spongy 360 mesophyll tissue were defined by position: palisade as the most adaxial layer of subepidermal 361 mesophyll cells and spongy as the most abaxial layer of subepidermal mesophyll cells. 362 In leaf tissue, dead cells were visualised by staining with Sytox Orange, as described previously 363 by Truernit & Haseloff (2008). Sytox Orange (Molecular Probes, Eugene, USA) was dissolved in 14 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 364 DMSO and diluted to a concentration of 250 nM in water. Leaves were stained by soaking in 365 Sytox Orange for 5 min before mounting on a slide under a coverslip and imaging on a Leica 366 SP8 confocal microscope using a 20× dry objective. Sytox Orange was excited with a 568 nm 367 wavelength laser and visualised between 580 and 610 nm. 368 To confirm that Sytox Orange could stain dead cells in the mesophyll, laser ablations were 369 performed on the Leica SP8 using the ROI function, focusing a laser of 568 nm wavelength 370 for 30 seconds on a single point. This caused the cell to rupture and the dye to penetrate and 371 fluoresce. Ablation experiments with Sytox Orange were conducted at 8 DAS and 14 DAS. 372 To visualise _IQD22::mCit_ and _CORI3::mCit_ expression in mesophyll cells, leaves were embedded 373 in 3% agarose and sectioned using a vibratome (Leica VT 1200S) to a thickness of 150 µm. 374 Sections were stained for 10 minutes in a 1% calcofluor white solution then rinsed in water 375 and visualised using a Leica SP8 confocal microscope. Calcofluor white was excited with a 405 376 nm wavelength laser and visualised between 450 and 480 nm. mCitrine was excited with a 514 377 nm wavelength laser and visualised between 524 and 575 nm. Chlorophyll autofluorescence 378 was observed by exciting the sample with a 405 nm wavelength laser and visualised between 379 680 and 740 nm. 380 **Leaf** **area** **and** **cell** **counts** 381 Whole leaves were imaged using a Keyence VHX8000 microscope, and leaf area quantified 382 using ImageJ. Cell density measurements were calculated by using the CellposeSAM v4.0.4 383 cellular segmentation software (Pachitariu _et_ _al._, 2025) to generate a ROI from confocal 384 images of stained leaves like the ones from Fig. 2A,E. The ROI was then imported to ImageJ 385 to obtain a cells/mm [2] value for each leaf. Quantification of cell area and shape descriptors 386 were performed in ImageJ using the measurements command with the ROI generated on 387 CellposeSAM v4.0.4. Stomata density measurements were calculated by manually counting 388 stomata number in a 500×500 µm region of epidermis tissue from a leaf stained with Nile Red. 389 Cell death proportions in the palisade mesophyll were calculated by counting Sytox Orange 390 stained and non-stained cells in a 500×500 µm region of the palisade mesophyll. 391 **Sector** **analysis** **and** **heat** **shock** **induction** 392 Expression of CRE recombinase was transiently induced in 4-day old plate-grown seedlings 393 transformed with the 71891 plasmid by heat shocking them for 60-90 seconds by floating 394 plates in a 39°C water bath. Plants were then returned to the growth room and left to 395 grow. Heat shocks of this length cause CRE expression in very few cells and generate clonal 396 sectors. In individual cells that were induced, mutually exclusive clonal sectors of YFP/RFP 397 were detected using a Leica SP8 confocal microscope and cell numbers per clone were manually 398 counted 3 days post-induction. 399 **Statistical** **analyses** **and** **figure** **design** 400 Statistical analyses, graph plotting and figure design were performed in R. 15 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) ## 401 ACKNOWLEDGEMENTS 402 We thank the facilities and technical staff for general support. ## 403 AUTHOR CONTRIBUTIONS 404 Conceptualisation: E.A.L.B., E.A., C.D.W.; Funding acquisition: C.D.W.; Investigation: E.A.L.B., 405 E.A., J.M.F., J.C.D., C.D.W.; Methodology: E.A.L.B., E.A., C.D.W.; Project administration: C.D.W.; 406 Resources: E.A., J.M.F., J.C.D.; Supervision: E.A.L.B., E.A., C.D.W.; Visualisation: E.A.L.B., E.A., 407 C.D.W.; Writing – original draft: E.A.L.B.; Writing – review & editing: E.A.L.B., E.A., J.C.D., C.D.W. ## 408 FUNDING 409 This work was funded by a Career Development Fellowship awarded to C.D.W. by the Gatsby 410 Charitable Foundation (award G110004). ## 411 REFERENCES 412 **Avery** **Jr** **GS** . **1933** . Structure and development of the tobacco leaf. _American_ _Journal_ _of_ 413 _Botany_ : 565–592. 414 **Baillie** **AL,** **Fleming** **AJ** . **2020** . The developmental relationship between stomata and meso415 phyll airspace. _New_ _Phytologist_ **225** : 1120–1126. 416 **Bar** **M,** **Ori** **N** . **2014** . Leaf development and morphogenesis. _Development_ **141** : 4219–4230. 417 **Barth** **C,** **Conklin** **PL** . **2003** . The lower cell density of leaf parenchyma in the arabidopsis 418 thaliana mutant lcd1-1 is associated with increased sensitivity to ozone and virulent pseu419 domonas syringae. _The_ _Plant_ _Journal_ **35** : 206–218. 420 **Chatterjee** **M,** **Sparvoli** **S,** **Edmunds** **C,** **Garosi** **P,** **Findlay** **K,** **Martin** **C** . **1996** . DAG, a gene 421 required for chloroplast differentiation and palisade development in antirrhinum majus. _The_ 422 _EMBO_ _journal_ **15** : 4194–4207. 423 **DeLucia E, Nelson K, Vogelmann T, Smith W** . **1996** . Contribution of intercellular reflectance 424 to photosynthesis in shade leaves. _Plant,_ _Cell_ _&_ _Environment_ **19** : 159–170. 425 **Donnelly** **PM,** **Bonetta** **D,** **Tsukaya** **H,** **Dengler** **RE,** **Dengler** **NG** . **1999** . Cell cycling and cell 426 enlargement in developing leaves of arabidopsis. _Developmental_ _biology_ **215** : 407–419. 16 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 427 **Dow** **GJ,** **Berry** **JA,** **Bergmann** **DC** . **2017** . Disruption of stomatal lineage signaling or tran428 scriptional regulators has differential effects on mesophyll development, but maintains coor429 dination of gas exchange. _New_ _Phytologist_ **216** : 69–75. 430 **Earles** **JM,** **Théroux-Rancourt** **G,** **Roddy** **AB,** **Gilbert** **ME,** **McElrone** **AJ,** **Brodersen** **CR** . 431 **2018** . Beyond porosity: 3D leaf intercellular airspace traits that impact mesophyll conduc432 tance. _Plant_ _Physiology_ **178** : 148–162. 433 **Ferjani** **A,** **Horiguchi** **G,** **Yano** **S,** **Tsukaya** **H** . **2007** . Analysis of leaf development in fugu 434 mutants of arabidopsis reveals three compensation modes that modulate cell expansion in 435 determinate organs. _Plant_ _Physiology_ **144** : 988–999. 436 **Fitzsimons** **JM,** **Rock** **AB,** **De** **Falbe** **RL,** **Fox** **S,** **Whitewoods** **CD** . **2026** . Brassinosteroid 437 controls leaf air space patterning non-cell autonomously by promoting epidermal growth. 438 _Development_ **153** : dev205110. 439 **Fox** **S,** **Southam** **P,** **Pantin** **F,** **Kennaway** **R,** **Robinson** **S,** **Castorina** **G,** **Sánchez-Corrales** 440 **YE,** **Sablowski** **R,** **Chan** **J,** **Grieneisen** **V,** _**et**_ _**al.**_ **2018** . Spatiotemporal coordination of cell 441 division and growth during organ morphogenesis. _PLoS_ _biology_ **16** : e2005952. 442 **González-Bayon** **R,** **Kinsman** **EA,** **Quesada** **V,** **Vera** **A,** **Robles** **P,** **Ponce** **MR,** **Pyke** **KA,** 443 **Micol** **JL** . **2006** . Mutations in the RETICULATA gene dramatically alter internal architecture 444 but have little effect on overall organ shape in arabidopsis leaves. _Journal_ _of_ _experimental_ 445 _botany_ **57** : 3019–3031. 446 **Hanfrey** **C,** **Sommer** **S,** **Mayer** **MJ,** **Burtin** **D,** **Michael** **AJ** . **2001** . Arabidopsis polyamine 447 biosynthesis: Absence of ornithine decarboxylase and the mechanism of arginine decarboxy448 lase activity. _The_ _Plant_ _Journal_ **27** : 551–560. 449 **Hricová A, Quesada V, Micol JL** . **2006** . The SCABRA3 nuclear gene encodes the plastid RpoTp 450 RNA polymerase, which is required for chloroplast biogenesis and mesophyll cell proliferation 451 in arabidopsis. _Plant_ _physiology_ **141** : 942–956. 452 **Jarvis** **M,** **Briggs** **S,** **Knox** **J** . **2003** . Intercellular adhesion and cell separation in plants. _Plant,_ 453 _Cell_ _&_ _Environment_ **26** : 977–989. 454 **Jeffree** **C,** **Dale** **J,** **Fry** **S** . **1986** . The genesis of intercellular spaces in developing leaves of 455 phaseolus vulgaris l. _Protoplasma_ **132** : 90–98. 456 **Kawase** **T,** **Sugano** **SS,** **Shimada** **T,** **Hara-Nishimura** **I** . **2015** . A direction-selective local457 thresholding method, DSLT, in combination with a dye-based method for automated three 17 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 458 dimensional segmentation of cells and airspaces in developing leaves. _The_ _Plant_ _Journal_ **81** : 459 357–366. 460 **Kawase** **T,** **Sugano** **SS,** **Shimada** **T,** **Hara-Nishimura** **I** . **2016** . Differential and simultaneous 461 visualization of cells and airspaces in plant leaves. _Bio-protocol_ **6** : e1826–e1826. 462 **Keddie** **JS,** **Carroll** **B,** **Jones** **J,** **Gruissem** **W** . **1996** . The DCL gene of tomato is required 463 for chloroplast development and palisade cell morphogenesis in leaves. _The_ _EMBO_ _journal_ **15** : 464 4208–4217. 465 **Knappe** **S,** **Löttgert** **T,** **Schneider** **A,** **Voll** **L,** **Flügge** **U-I,** **Fischer** **K** . **2003** . Characterization 466 of two functional phosphoenolpyruvate/phosphate translocator (PPT) genes in arabidopsis– 467 AtPPT1 may be involved in the provision of signals for correct mesophyll development. _The_ 468 _Plant_ _Journal_ **36** : 411–420. 469 **Ko** **D,** **Ruonala** **R,** **Faille** **A,** **Hellmann** **E,** **Help** **H,** **Liu** **H,** **Nielsen** **R,** **Haakonsson** **A,** **De** 470 **Diego** **N,** **Paatero** **A,** _**et**_ _**al.**_ **2026** . Recruitment of bifunctional regulator thermospermine to 471 methylated ribosomes directs xylem fate. _Science_ **391** : 694–699. 472 **Kuchen** **EE,** **Fox** **S,** **Barbier** **de** **Reuille** **P,** **Kennaway** **R,** **Bensmihen** **S,** **Avondo** **J,** **Calder** 473 **GM,** **Southam** **P,** **Robinson** **S,** **Bangham** **A,** _**et**_ _**al.**_ **2012** . Generation of leaf shape through 474 early patterns of growth and tissue polarity. _Science_ **335** : 1092–1096. 475 **Kuhnert** **F,** **Westhoff** **P,** **Valencia** **V,** **Krüger** **S,** **Vogel** **K,** **Lundquist** **PK,** **Rosar** **C,** **Goss** **T,** 476 **Weber** **AP** . **2025** . RETICULATA1 is a plastid-localized basic amino acid transporter. _Nature_ 477 _Plants_ **11** : 1890–1902. 478 **Kurihara** **D,** **Mizuta** **Y,** **Sato** **Y,** **Higashiyama** **T** . **2015** . ClearSee: A rapid optical clearing 479 reagent for whole-plant fluorescence imaging. _Development_ **142** : 4168–4179. 480 **Lee** **KJ,** **Bushell** **C,** **Koide** **Y,** **Fozard** **JA,** **Piao** **C,** **Yu** **M,** **Newman** **J,** **Whitewoods** **C,** **Avondo** 481 **J,** **Kennaway** **R,** _**et**_ _**al.**_ **2019** . Shaping of a three-dimensional carnivorous trap through 482 modulation of a planar growth mechanism. _PLoS_ _Biology_ **17** : e3000427. 483 **Lee** **JS,** **Hnilova** **M,** **Maes** **M,** **Lin** **Y-CL,** **Putarjunan** **A,** **Han** **S-K,** **Avila** **J,** **Torii** **KU** . **2015** . 484 Competitive binding of antagonistic peptides fine-tunes stomatal patterning. _Nature_ **522** : 485 439–443. 486 **Lehmeier** **C,** **Pajor** **R,** **Lundgren** **MR,** **Mathers** **A,** **Sloan** **J,** **Bauch** **M,** **Mitchell** **A,** **Bellasio** 487 **C,** **Green** **A,** **Bouyer** **D,** _**et**_ _**al.**_ **2017** . Cell density and airspace patterning in the leaf can be 488 manipulated to increase leaf photosynthetic capacity. _The_ _Plant_ _Journal_ **92** : 981–994. 18 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 489 **Li H, Culligan K, Dixon RA, Chory J** . **1995** . CUE1: A mesophyll cell-specific positive regulator 490 of light-controlled gene expression in arabidopsis. _The_ _Plant_ _Cell_ **7** : 1599–1610. 491 **Livet** **J,** **Weissman** **TA,** **Kang** **H,** **Draft** **RW,** **Lu** **J,** **Bennis** **RA,** **Sanes** **JR,** **Lichtman** **JW** . 492 **2007** . Transgenic strategies for combinatorial expression of fluorescent proteins in the ner493 vous system. _Nature_ **450** : 56–62. 494 **Lundgren** **MR,** **Mathers** **A,** **Baillie** **AL,** **Dunn** **J,** **Wilson** **MJ,** **Hunt** **L,** **Pajor** **R,** **Fradera-** 495 **Soler** **M,** **Rolfe** **S, Osborne** **CP,** _**et**_ _**al.**_ **2019** . Mesophyll porosity is modulated by the presence 496 of functional stomata. _Nature_ _Communications_ **10** : 2825. 497 **Lundquist** **PK,** **Rosar** **C,** **Bräutigam** **A,** **Weber** **AP** . **2014** . Plastid signals and the bundle 498 sheath: Mesophyll development in reticulate mutants. _Molecular_ _Plant_ **7** : 14–29. 499 **Majewska-Sawka** **A,** **Niklas** **A,** **Jażdżewska** **E** . **1997** . The effect of polyamines on the devel500 opment of sugar beet protoplasts. _Biologia_ _plantarum_ **39** : 561–567. 501 **Masle** **J,** **Gilmore** **SR,** **Farquhar** **GD** . **2005** . The ERECTA gene regulates plant transpiration 502 efficiency in arabidopsis. _Nature_ **436** : 866–870. 503 **Mathers** **AW,** **Hepworth** **C,** **Baillie** **AL,** **Sloan** **J,** **Jones** **H,** **Lundgren** **M,** **Fleming** **AJ,** 504 **Mooney** **SJ,** **Sturrock** **CJ** . **2018** . Investigating the microstructure of plant leaves in 3D with 505 lab-based x-ray computed tomography. _Plant_ _methods_ **14** : 99. 506 **Overmyer** **K,** **Kollist** **H,** **Tuominen** **H,** **Betz** **C,** **Langebartels** **C,** **Wingsle** **G,** **Kangasjärvi** 507 **S,** **Brader** **G,** **Mullineaux** **P,** **Kangasjärvi** **J** . **2008** . Complex phenotypic profiles leading to 508 ozone sensitivity in arabidopsis thaliana mutants. _Plant,_ _cell_ _&_ _environment_ **31** : 1237–1249. 509 **Pachitariu** **M,** **Rariden** **M,** **Stringer** **C** . **2025** . Cellpose-sam: Superhuman generalization for 510 cellular segmentation. bioRxiv. 511 **Paschalidis** **KA,** **Roubelakis-Angelakis** **KA** . **2005** . Sites and regulation of polyamine 512 catabolism in the tobacco plant. Correlations with cell division/expansion, cell cycle 513 progression, and vascular development. _Plant_ _Physiology_ **138** : 2174–2184. 514 **Pérez-Pérez** **JM,** **Esteve-Bruna** **D,** **González-Bayón** **R,** **Kangasjärvi** **S,** **Caldana** **C,** **Hannah** 515 **MA,** **Willmitzer** **L,** **Ponce** **MR,** **Micol** **JL** . **2013** . Functional redundancy and divergence within 516 the arabidopsis RETICULATA-RELATED gene family. _Plant_ _physiology_ **162** : 589–603. 517 **Procko** **C,** **Lee** **T,** **Borsuk** **A,** **Bargmann** **BO,** **Dabi** **T,** **Nery** **JR,** **Estelle** **M,** **Baird** **L,** **O’Connor** 518 **C,** **Brodersen** **C,** _**et**_ _**al.**_ **2022** . Leaf cell-specific and single-cell transcriptional profiling reveals 19 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 519 a role for the palisade layer in UV light protection. _The_ _Plant_ _Cell_ **34** : 3261–3279. 520 **Pyke** **K,** **Marrison** **J,** **Leech** **A** . **1991** . Temporal and spatial development of the cells of the 521 expanding first leaf of arabidopsis thaliana (l.) heynh. _Journal_ _of_ _Experimental_ _Botany_ **42** : 522 1407–1416. 523 **Rédei** **GP,** **Hironyo** **Y** . **1964** . Linkage studies. _Arabidopsis_ _Information_ _Service_ **1** : 9–10. 524 **Rosar** **C,** **Kanonenberg** **K,** **Nanda** **AM,** **Mielewczik** **M,** **Bräutigam** **A,** **Novák** **O,** **Strnad** **M,** 525 **Walter** **A,** **Weber** **AP** . **2012** . The leaf reticulate mutant dov1 is impaired in the first step of 526 purine metabolism. _Molecular_ _plant_ **5** : 1227–1241. 527 **Schmidt** **R,** **Kunkowska** **AB,** **Schippers** **JH** . **2016** . Role of reactive oxygen species during 528 cell expansion in leaves. _Plant_ _Physiology_ **172** : 2098–2106. 529 **Slaton** **MR,** **Smith** **WK** . **2002** . Mesophyll architecture and cell exposure to intercellular 530 air space in alpine, desert, and forest species. _International_ _Journal_ _of_ _Plant_ _Sciences_ **163** : 531 937–948. 532 **Smith** **WK,** **Vogelmann** **TC,** **DeLucia** **EH,** **Bell** **DT,** **Shepherd** **KA** . **1997** . Leaf form and 533 photosynthesis. _BioScience_ **47** : 785–793. 534 **Terashima** **I,** **Inoue** **Y** . **1984** . Comparative photosynthetic properties of palisade tissue 535 chloroplasts and spongy tissue chloroplasts of camellia japonica l.: Functional adjustment of 536 the photosynthetic apparatus to light environment within a leaf. _Plant_ _and_ _Cell_ _Physiology_ 537 **25** : 555–563. 538 **Tiburcio** **AF,** **Kaur-Sawhney** **R,** **Galston** **AW** . **1986** . Polyamine metabolism and osmotic 539 stress: II. Improvement of oat protoplasts by an inhibitor of arginine decarboxylase. _Plant_ 540 _Physiology_ **82** : 375–378. 541 **Truernit** **E,** **Haseloff** **J** . **2008** . A simple way to identify non-viable cells within living plant 542 tissue using confocal microscopy. _Plant_ _methods_ **4** : 15. 543 **Van** **Lijsebettens** **M,** **Clarke** **J** . **1998** . Leaf development in arabidopsis. _Plant_ _Physiology_ _and_ 544 _Biochemistry_ **36** : 47–60. 545 **Voll** **L,** **Häusler** **RE,** **Hecker** **R,** **Weber** **A,** **Weissenböck** **G,** **Fiene** **G,** **Waffenschmidt** **S,** 546 **Flügge** **U-I** . **2003** . The phenotype of the arabidopsis cue1 mutant is not simply caused by a 547 general restriction of the shikimate pathway. _The_ _Plant_ _Journal_ **36** : 301–317. 20 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.03.759069; this version posted November 4, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.03.759069) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 548 **Zhang** **L,** **McEvoy** **D,** **Le** **Y,** **Ambrose** **C** . **2021** . Live imaging of microtubule organization, cell 549 expansion, and intercellular space formation in arabidopsis leaf spongy mesophyll cells. _The_ 550 _Plant_ _Cell_ **33** : 623–641. 21