© 2025. Published by The Company of Biologists | Journal of Cell Science (2025) 138, jcs264242. doi:10.1242/jcs.264242
RESEARCH ARTICLE
Cell volume regulates terminal differentiation of cultured human
epidermal keratinocytes
Sebastiaan Zijl1,*, Toru Hiratsuka1,2,3, Atefeh Mobasseri1, Mirsana Ebrahimkutty4, Mandy Bö rmel5,
Sergi Garcia-Manyes6,7 and Fiona M. Watt1,4,‡
ABSTRACT comprised of extracellular matrix proteins such as collagen-IV (Watt
To gain insights into the human epidermal stem cell niche, we and Fujiwara, 2011). The basal epidermal layer, which is attached
have previously identified micron-scale topographical substrates that to the basement membrane, contains a patterned distribution of
regulate differentiation of spread keratinocytes. On one substrate (S1), stem cells and cells that have initiated terminal differentiation.
cells interact with circular topographies and differentiation is stimulated; Differentiating cells detach from the basement membrane and move
on the other (S2), cells interact with triangular topographies and through the suprabasal epidermal layers towards the skin surface,
differentiation is inhibited. Cell stiffness on S1 and S2 was similar, from which they are shed (Zijl et al., 2022).
and nuclear localisation of the mechano-sensitive transcriptional Human epidermal keratinocytes can be grown in culture under
regulator YAP1 was decreased on S1 and S2 compared to on flat conditions that support the maintenance of stem cells and the
substrates. However, cells on S2 exhibited reduced cell volume, process of terminal differentiation (Zijl et al., 2022). Cultured
leading us to explore the potential for volume-regulated differentiation. human epidermis can therefore be used to examine, at single-cell
Treatment with polyethylene glycol decreased cell volume and resolution, how individual stem cells make fate decisions based on
inhibited differentiation under a range of conditions. Conversely, external cues from the local microenvironment (Watt, 2016; Louis
deionized water increased cell volume and stimulated differentiation. et al., 2022; Negri and Watt, 2022; Negri et al., 2023).
Bulk RNA sequencing identified several substrate-responsive genes, Previous studies have highlighted the role of cell–substrate
including aquaporins and ion channels. A membrane permeable Ca2+ interactions in controlling the differentiation of human epidermal
chelator and an inhibitor of the water channel aquaporin 3 blocked stem cells (Zijl et al., 2022). When single cells are seeded on
volume-induced differentiation. These studies identify cell volume extracellular matrix (ECM)-coated micro-patterned islands, for
as a mechanism by which keratinocyte–niche interactions regulate example, differentiation is triggered by restricted spreading. This
terminal differentiation. process depends on the ratio of F- to G-actin and activation of serum
response factor (SRF) and its co-factor megakaryocytic acute
KEY WORDS: Stem cell, Niche, Volume control, Cell adhesion, leukaemia (MAL) (Watt et al., 1988; Connelly et al., 2010).
Differentiation, Epidermis Differentiation is also triggered when cells are plated on ECM-
coated soft hydrogels or on hydrogel–nanoparticle composites with
INTRODUCTION high nanoparticle spacing (Trappmann et al., 2012). Mechanistically,
Human skin comprises the epidermis, which is formed of multiple this is mediated via the downregulation of the extracellular signal-
layers of epithelial cells called keratinocytes, and the underlying regulated kinase (ERK)/mitogen-activated protein kinase (MAPK)
connective tissue, the dermis (Rognoni and Watt, 2018). The pathway.
epidermis and dermis are separated by a basement membrane, Although keratinocyte differentiation is typically associated with
reduced cell spreading, we have found that micron-scale substrate
1
topographies can also promote the differentiation of spread cells
Centre for Gene Therapy and Regenerative Medicine, King’s College London, 28th
Floor, Tower Wing, Guy’s Hospital, Great Maze Pond, London SE1 9RT, UK. (Zijl et al., 2019). A substrate comprising circular topographical
2
Department of Molecular Oncology, Graduate School of Medicine, Osaka features of ∼3 µm diameter, 5 µm height and unequal spacing
University, Osaka 541-8567, Japan. 3Department of Oncogenesis and Growth (designated S1) promotes differentiation of spread cells via a
Regulation, Research Center, Osaka International Cancer Institute, Osaka 541-
8567, Japan. 4Directors’ Unit, EMBL, Meyerhofstr. 1, 69117 Heidelberg, Germany. mechanism that is blocked by Rho kinase inhibition or treatment
5
EMBL Electron Microscopy Core Facility, Meyerhofstr. 1, 69117 Heidelberg, with the myosin II inhibitor blebbistatin, but not by SRF inhibition.
Germany. 6Department of Physics, Randall Centre for Cell and Molecular
Biophysics, Centre for the Physical Science of Life and London Centre for
Conversely, a substrate comprising regularly spaced right-angled
Nanotechnology, King’s College London, London SE1 9RT, UK. 7Single Molecule triangles (8 µm sides) of 5 µm height (designated S2) suppresses
Mechanobiology Laboratory, The Francis Crick Institute, London NW1 1AT, UK. differentiation relative to flat surfaces (Zijl et al., 2019). In the
Journal of Cell Science
*Present address: Gurdon Institute, University of Cambridge, Tennis Court Road,
Cambridge CB2 1QN, UK. present study, we have further explored how keratinocytes respond
to each of these substrates, uncovering a potential role for cell
‡
Author for correspondence () volume in the regulation of keratinocyte differentiation.
S.Z., 0000-0002-2856-1331; T.H., 0000-0002-5359-2690; M.E., 0009-0005-
8738-7547; M.B., 0000-0002-4354-891X; S.G.-M, 0000-0001-5140-2606; F.M.W., RESULTS
0000-0001-9151-5154 Keratinocytes on S1 substrates can initiate differentiation
This is an Open Access article distributed under the terms of the Creative Commons Attribution while spread
License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, To enrich for undifferentiated keratinocytes (epidermal stem cells)
distribution and reproduction in any medium provided that the original work is properly attributed.
we seeded single-cell suspensions of keratinocytes, comprising a
Handling Editor: Andrew Ewald mixture of basal and differentiated cells, on collagen-coated S1, S2
Received 24 June 2025; Accepted 29 June 2025 or flat surfaces for 1 h and washed off the non-adherent cells (Jones
1
,RESEARCH ARTICLE Journal of Cell Science (2025) 138, jcs264242. doi:10.1242/jcs.264242
and Watt, 1993; Zijl et al., 2019). Scanning electron microscope representative images of a single field of keratinocytes obtained
(SEM) images of keratinocytes 24 h after plating (Fig. 1A–F) from live-cell imaging experiments. At 6 h, several cells in the
revealed that cells on S1 substrates spread over the individual pillars field were already expressing mCherry and exhibited variable
(Fig. 1C,D). Some pillars at the edges of the substrates were bent, morphologies. Two cells (arrowheads) were mCherry-negative
suggesting that they had been subject to pulling forces by the cells at 6 h and 9 h but expressed mCherry at 12 h and 15 h; in each
(Fig. 1D; Eyckmans et al., 2011), although artefacts caused by case the onset of mCherry expression occurred while the cells were
sample preparation cannot be ruled out entirely. The cytoplasm of spread.
cells on S2 substrates appeared to be draped over the triangular Using LifeAct–EGFP to visualize individual cells, whether
features (Fig. 1E,F). In some cells on S2 substrates, the nucleus or not they had differentiated, we measured the mean mCherry
extended above the surface of the features (Fig. 1F), whereas in fluorescence of 30 cells on S1 that had upregulated mCherry by 20 h
other cells the nucleus appeared to have been accommodated in- (Fig. 2C). The mean mCherry signal from cells increased above
between the features (Fig. 1E), consistent with nuclear distortions baseline from 12 h onwards (Fig. 2D), consistent with earlier studies
observed by light microscopy (Zijl et al., 2019). Heterogeneity in on the kinetics of keratinocyte differentiation (Connelly et al., 2010;
terms of flattened versus protruding nuclei was also evident on flat Zijl et al., 2019; Hiratsuka et al., 2020). These observations establish
and S1 substrates (Fig. 1A–D). that cells plated on S1 can upregulate involucrin expression while
To confirm that cells seeded on S1 initiated differentiation being spread.
without rounding (Connelly et al., 2010), we performed live-cell
imaging of keratinocytes transduced with a pLenti-IVL-mCherry- RNAseq does not reveal unique features of differentiation
LifeAct-EGFP reporter. This reporter expresses LifeAct–EGFP, a on S1
peptide that binds to actin filaments (Riedl et al., 2008; Belin et al., Changes in gene expression associated with commitment to
2014), under the control of the human PGK promoter, and expresses differentiation have previously been described in keratinocytes
mCherry under the control of the involucrin (IVL) promoter undergoing differentiation in single-cell suspension (Mishra et al.,
(Hiratsuka et al., 2020) (Fig. 2A). Flow cytometry confirmed that 2017), including a network of interacting protein phosphatases that
mCherry was expressed by cells of high forward and side scatter are also upregulated during the basal to suprabasal transition of
(enriched for differentiating cells; Jones and Watt, 1993; Connelly keratinocytes in vivo (Reynolds et al., 2021; Negri et al., 2023). To
et al., 2010) and that mCherry-positive cells co-expressed LifeAct– examine whether there was a unique gene expression programme
EGFP (Fig. 2A). We flow sorted LifeActGFP+ cells (Fig. 2A) and associated with differentiation on S1, we seeded keratinocytes on S1
seeded them on the S1 substrates (Fig. 2B). or S2 substrates and extracted RNA at 1 h (initial adhesion), 4 h
Cells expressing the IVL-mCherry-LifeAct-EGFP plasmid were (differentiation commitment; Mishra et al., 2017) and 12 h (IVL
imaged at 1 h intervals starting 4–6 h after seeding. Fig. 2B shows expression on S1; Fig. 2D). RNA was harvested from three
Journal of Cell Science
Fig. 1. SEM of single keratinocytes on flat, S1 and S2 substrates. SEM shows individual well-spread cells adhered to flat surfaces (A,B), S1 substrates
with circular micropillars (C,D) and S2 substrates with triangular pillars (E,F). On the S1 substrate, some peripheral pegs are bent (arrows, D), suggestive of
localized mechanical forces. Nuclear position varies between cells; whereas some cells exhibit nuclei positioned above the underlying substrates (D,F),
others display nuclei positioned within the substrates (C,E). Side views are shown. Scale bars: 10 μm. Data are representative of three independent
experiments.
2
, RESEARCH ARTICLE Journal of Cell Science (2025) 138, jcs264242. doi:10.1242/jcs.264242
Journal of Cell Science
Fig. 2. See next page for legend.
independent experiments and subjected to bulk RNA sequencing substrate). Sequencing and bioinformatic analysis were performed by
(RNAseq) (Fig. 3A). In total 18 samples were submitted for RNAseq Genewiz, Inc. and are deposited in the Gene Expression Omnibus
(three time points with three pooled technical replicates each on each under the accession code GSE303873.
3
RESEARCH ARTICLE
Cell volume regulates terminal differentiation of cultured human
epidermal keratinocytes
Sebastiaan Zijl1,*, Toru Hiratsuka1,2,3, Atefeh Mobasseri1, Mirsana Ebrahimkutty4, Mandy Bö rmel5,
Sergi Garcia-Manyes6,7 and Fiona M. Watt1,4,‡
ABSTRACT comprised of extracellular matrix proteins such as collagen-IV (Watt
To gain insights into the human epidermal stem cell niche, we and Fujiwara, 2011). The basal epidermal layer, which is attached
have previously identified micron-scale topographical substrates that to the basement membrane, contains a patterned distribution of
regulate differentiation of spread keratinocytes. On one substrate (S1), stem cells and cells that have initiated terminal differentiation.
cells interact with circular topographies and differentiation is stimulated; Differentiating cells detach from the basement membrane and move
on the other (S2), cells interact with triangular topographies and through the suprabasal epidermal layers towards the skin surface,
differentiation is inhibited. Cell stiffness on S1 and S2 was similar, from which they are shed (Zijl et al., 2022).
and nuclear localisation of the mechano-sensitive transcriptional Human epidermal keratinocytes can be grown in culture under
regulator YAP1 was decreased on S1 and S2 compared to on flat conditions that support the maintenance of stem cells and the
substrates. However, cells on S2 exhibited reduced cell volume, process of terminal differentiation (Zijl et al., 2022). Cultured
leading us to explore the potential for volume-regulated differentiation. human epidermis can therefore be used to examine, at single-cell
Treatment with polyethylene glycol decreased cell volume and resolution, how individual stem cells make fate decisions based on
inhibited differentiation under a range of conditions. Conversely, external cues from the local microenvironment (Watt, 2016; Louis
deionized water increased cell volume and stimulated differentiation. et al., 2022; Negri and Watt, 2022; Negri et al., 2023).
Bulk RNA sequencing identified several substrate-responsive genes, Previous studies have highlighted the role of cell–substrate
including aquaporins and ion channels. A membrane permeable Ca2+ interactions in controlling the differentiation of human epidermal
chelator and an inhibitor of the water channel aquaporin 3 blocked stem cells (Zijl et al., 2022). When single cells are seeded on
volume-induced differentiation. These studies identify cell volume extracellular matrix (ECM)-coated micro-patterned islands, for
as a mechanism by which keratinocyte–niche interactions regulate example, differentiation is triggered by restricted spreading. This
terminal differentiation. process depends on the ratio of F- to G-actin and activation of serum
response factor (SRF) and its co-factor megakaryocytic acute
KEY WORDS: Stem cell, Niche, Volume control, Cell adhesion, leukaemia (MAL) (Watt et al., 1988; Connelly et al., 2010).
Differentiation, Epidermis Differentiation is also triggered when cells are plated on ECM-
coated soft hydrogels or on hydrogel–nanoparticle composites with
INTRODUCTION high nanoparticle spacing (Trappmann et al., 2012). Mechanistically,
Human skin comprises the epidermis, which is formed of multiple this is mediated via the downregulation of the extracellular signal-
layers of epithelial cells called keratinocytes, and the underlying regulated kinase (ERK)/mitogen-activated protein kinase (MAPK)
connective tissue, the dermis (Rognoni and Watt, 2018). The pathway.
epidermis and dermis are separated by a basement membrane, Although keratinocyte differentiation is typically associated with
reduced cell spreading, we have found that micron-scale substrate
1
topographies can also promote the differentiation of spread cells
Centre for Gene Therapy and Regenerative Medicine, King’s College London, 28th
Floor, Tower Wing, Guy’s Hospital, Great Maze Pond, London SE1 9RT, UK. (Zijl et al., 2019). A substrate comprising circular topographical
2
Department of Molecular Oncology, Graduate School of Medicine, Osaka features of ∼3 µm diameter, 5 µm height and unequal spacing
University, Osaka 541-8567, Japan. 3Department of Oncogenesis and Growth (designated S1) promotes differentiation of spread cells via a
Regulation, Research Center, Osaka International Cancer Institute, Osaka 541-
8567, Japan. 4Directors’ Unit, EMBL, Meyerhofstr. 1, 69117 Heidelberg, Germany. mechanism that is blocked by Rho kinase inhibition or treatment
5
EMBL Electron Microscopy Core Facility, Meyerhofstr. 1, 69117 Heidelberg, with the myosin II inhibitor blebbistatin, but not by SRF inhibition.
Germany. 6Department of Physics, Randall Centre for Cell and Molecular
Biophysics, Centre for the Physical Science of Life and London Centre for
Conversely, a substrate comprising regularly spaced right-angled
Nanotechnology, King’s College London, London SE1 9RT, UK. 7Single Molecule triangles (8 µm sides) of 5 µm height (designated S2) suppresses
Mechanobiology Laboratory, The Francis Crick Institute, London NW1 1AT, UK. differentiation relative to flat surfaces (Zijl et al., 2019). In the
Journal of Cell Science
*Present address: Gurdon Institute, University of Cambridge, Tennis Court Road,
Cambridge CB2 1QN, UK. present study, we have further explored how keratinocytes respond
to each of these substrates, uncovering a potential role for cell
‡
Author for correspondence () volume in the regulation of keratinocyte differentiation.
S.Z., 0000-0002-2856-1331; T.H., 0000-0002-5359-2690; M.E., 0009-0005-
8738-7547; M.B., 0000-0002-4354-891X; S.G.-M, 0000-0001-5140-2606; F.M.W., RESULTS
0000-0001-9151-5154 Keratinocytes on S1 substrates can initiate differentiation
This is an Open Access article distributed under the terms of the Creative Commons Attribution while spread
License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, To enrich for undifferentiated keratinocytes (epidermal stem cells)
distribution and reproduction in any medium provided that the original work is properly attributed.
we seeded single-cell suspensions of keratinocytes, comprising a
Handling Editor: Andrew Ewald mixture of basal and differentiated cells, on collagen-coated S1, S2
Received 24 June 2025; Accepted 29 June 2025 or flat surfaces for 1 h and washed off the non-adherent cells (Jones
1
,RESEARCH ARTICLE Journal of Cell Science (2025) 138, jcs264242. doi:10.1242/jcs.264242
and Watt, 1993; Zijl et al., 2019). Scanning electron microscope representative images of a single field of keratinocytes obtained
(SEM) images of keratinocytes 24 h after plating (Fig. 1A–F) from live-cell imaging experiments. At 6 h, several cells in the
revealed that cells on S1 substrates spread over the individual pillars field were already expressing mCherry and exhibited variable
(Fig. 1C,D). Some pillars at the edges of the substrates were bent, morphologies. Two cells (arrowheads) were mCherry-negative
suggesting that they had been subject to pulling forces by the cells at 6 h and 9 h but expressed mCherry at 12 h and 15 h; in each
(Fig. 1D; Eyckmans et al., 2011), although artefacts caused by case the onset of mCherry expression occurred while the cells were
sample preparation cannot be ruled out entirely. The cytoplasm of spread.
cells on S2 substrates appeared to be draped over the triangular Using LifeAct–EGFP to visualize individual cells, whether
features (Fig. 1E,F). In some cells on S2 substrates, the nucleus or not they had differentiated, we measured the mean mCherry
extended above the surface of the features (Fig. 1F), whereas in fluorescence of 30 cells on S1 that had upregulated mCherry by 20 h
other cells the nucleus appeared to have been accommodated in- (Fig. 2C). The mean mCherry signal from cells increased above
between the features (Fig. 1E), consistent with nuclear distortions baseline from 12 h onwards (Fig. 2D), consistent with earlier studies
observed by light microscopy (Zijl et al., 2019). Heterogeneity in on the kinetics of keratinocyte differentiation (Connelly et al., 2010;
terms of flattened versus protruding nuclei was also evident on flat Zijl et al., 2019; Hiratsuka et al., 2020). These observations establish
and S1 substrates (Fig. 1A–D). that cells plated on S1 can upregulate involucrin expression while
To confirm that cells seeded on S1 initiated differentiation being spread.
without rounding (Connelly et al., 2010), we performed live-cell
imaging of keratinocytes transduced with a pLenti-IVL-mCherry- RNAseq does not reveal unique features of differentiation
LifeAct-EGFP reporter. This reporter expresses LifeAct–EGFP, a on S1
peptide that binds to actin filaments (Riedl et al., 2008; Belin et al., Changes in gene expression associated with commitment to
2014), under the control of the human PGK promoter, and expresses differentiation have previously been described in keratinocytes
mCherry under the control of the involucrin (IVL) promoter undergoing differentiation in single-cell suspension (Mishra et al.,
(Hiratsuka et al., 2020) (Fig. 2A). Flow cytometry confirmed that 2017), including a network of interacting protein phosphatases that
mCherry was expressed by cells of high forward and side scatter are also upregulated during the basal to suprabasal transition of
(enriched for differentiating cells; Jones and Watt, 1993; Connelly keratinocytes in vivo (Reynolds et al., 2021; Negri et al., 2023). To
et al., 2010) and that mCherry-positive cells co-expressed LifeAct– examine whether there was a unique gene expression programme
EGFP (Fig. 2A). We flow sorted LifeActGFP+ cells (Fig. 2A) and associated with differentiation on S1, we seeded keratinocytes on S1
seeded them on the S1 substrates (Fig. 2B). or S2 substrates and extracted RNA at 1 h (initial adhesion), 4 h
Cells expressing the IVL-mCherry-LifeAct-EGFP plasmid were (differentiation commitment; Mishra et al., 2017) and 12 h (IVL
imaged at 1 h intervals starting 4–6 h after seeding. Fig. 2B shows expression on S1; Fig. 2D). RNA was harvested from three
Journal of Cell Science
Fig. 1. SEM of single keratinocytes on flat, S1 and S2 substrates. SEM shows individual well-spread cells adhered to flat surfaces (A,B), S1 substrates
with circular micropillars (C,D) and S2 substrates with triangular pillars (E,F). On the S1 substrate, some peripheral pegs are bent (arrows, D), suggestive of
localized mechanical forces. Nuclear position varies between cells; whereas some cells exhibit nuclei positioned above the underlying substrates (D,F),
others display nuclei positioned within the substrates (C,E). Side views are shown. Scale bars: 10 μm. Data are representative of three independent
experiments.
2
, RESEARCH ARTICLE Journal of Cell Science (2025) 138, jcs264242. doi:10.1242/jcs.264242
Journal of Cell Science
Fig. 2. See next page for legend.
independent experiments and subjected to bulk RNA sequencing substrate). Sequencing and bioinformatic analysis were performed by
(RNAseq) (Fig. 3A). In total 18 samples were submitted for RNAseq Genewiz, Inc. and are deposited in the Gene Expression Omnibus
(three time points with three pooled technical replicates each on each under the accession code GSE303873.
3
