Spatial constraints control cell proliferation in tissues Sebastian J. Streichana,b,1,2, Christian R. Hoernerb,c,1, Tatjana Schneidtb, Daniela Holzerb, and Lars Hufnagelb,2 aKavli Institute for Theoretical Physics, University of California, Santa Barbara, CA 93106; bCell Biology and Biophysics Unit, European Molecular Biology Laboratory, 69117 Heidelberg, Germany; and cDepartment of Biology, Stanford University, Stanford, CA 94305 Edited by Robert H. Austin, Princeton University, Princeton, NJ, and approved February 28, 2014 (received for review December 11, 2013) Control of cell proliferation is a fundamental aspect of tissue for- We either acutely removed a barrier to release spatial constraints mation in development and regeneration. Cells experience various at the edge of the model epithelium or performed mechanical spatial and mechanical constraints depending on their environ- stretching or compression of the tissue substrate to manipulate mental context in the body, but we do not fully understand if and spatial constraints directly within the model tissue, a method how such constraints influence cell cycle progression and thereby previously applied only to end point assays (21–26). We show that proliferation patterns in tissues. Here, we study the impact of the proliferation rate in tissues is controlled by a mechanosensitive mechanical manipulations on the cell cycle of individual cells within cell cycle checkpoint that monitors the space available to the cell a mammalian model epithelium. By monitoring the response to at the G1–S interface. Using mathematical modeling, we can pre- experimentally applied forces, we find a checkpoint at the G1–S dict the tissue response to changes in spatial constraints and val- boundary that, in response to spatial constraints, controls cell cycle idate the prediction of the model that cell division is required for progression. This checkpoint prevents cells from entering S phase if sustained invasion of a tissue into newly colonized space. the available space remains below a characteristic threshold because of crowding. Stretching the tissue results in fast cell cycle reactiva- Results tion, whereas compression rapidly leads to cell cycle arrest. Our Spatial Constraints Regulate Cell Cycle Progression During Tissue kinetic analysis of this response shows that cells have no memory Expansion. To probe mechanical control of cell cycle progression of past constraints and allows us to formulate a biophysical model in growing tissues, we decided to introduce rapid and temporally that predicts tissue growth in response to changes in spatial con- controlled alterations of spatial constraints in a tissue colonization straints in the environment. This characteristic biomechanical cell assay: We grew an epithelial model tissue consisting of contact- BIOLOGY cycle response likely serves as a fundamental control mechanism inhibited, fully polarized Madin–Darby canine kidney-2 (MDCK-2) DEVELOPMENTAL to maintain tissue integrity and to ensure control of tissue growth cells against a removable barrier (Fig. 1A) (26, 27). To monitor during development and regeneration. cell cycle dynamics, we used a fluorescent ubiquitination-based cell cycle indicator (Fucci; Fig. 1A, Fig. S1,andMovie S1) (28). After cell cycle regulation | mechanical feedback | quantitative biology | barrier removal, the tissue rapidly invaded the available space, and size checkpoint | G1-S transition after a slight delay, cells behind the initial barrier also reactivated their cell cycle by entering S phase (Fig. 1 C–F and Movie S2). This ontrol of cell division during tissue formation is a major was accompanied by a noted increase in the space covered by in- Cregulatory principle of tissue and organ formation, size de- dividual cells, i.e., the cross-sectional cell area (henceforth called termination, tissue regeneration, and tumorigenesis (1, 2). Uni- “cell area”), from the edge of the tissue reaching into the tissue cellular lower eukaryotic organisms, such as the budding yeast (Fig. 1 C′–F′ and Movie S2). Interestingly, cells entered S phase in Saccharomyces cerevisiae, link cell division to cell growth through all regions of increased cell area up to several hundred micro- a size checkpoint during late G1 phase (3–5). This size check- meters behind the initial barrier, whereas cells even further behind point, termed “start” in yeast, is believed to ensure that only cells that have reached a characteristic size enter the cell cycle; it is Significance therefore critical for cell size homeostasis. However, it is not known how cells monitor their own size. The situation is even Spatiotemporal coordination of cell growth underlies tissue less clear in mammalian cells. Although early studies in cultured development and disease. Mechanical feedback between cells cells argued for a size checkpoint similar to that of yeast (6), has been proposed as a regulatory mechanism for growth more recent reports instead proposed the growth rate as a trigger control both in vivo and in cultured cells undergoing contact for cell division (7–11). In this view, reaching a characteristic growth inhibition of proliferation. Evidence beyond theoretical and rate rather than a characteristic size triggers entry into S phase. The correlative observations falls short. In this study, we probe the molecular basis of neither model is well understood. impact of mechanical tissue perturbations on cell cycle pro- The body of animals largely comprises cohesive tissues in gression by monitoring cell cycle dynamics of cells in tissues which cells are not in isolation, but are coupled mechanically subject to acute changes in boundary conditions, as well as through their cell–cell and cell–substrate contacts (12–14). Spa- tissue stretching and compression. Taken together, we con- tial constraints from crowding, i.e., limitations on available space clude that the ability of tissues to support cell cycle progression due to the presence of neighboring cells, impose constraints on adapts to the available space through a memory-free control cell functions, such as cell proliferation. Thus, the regulation of mechanism, which may coordinate proliferation patterns to spatial constraints in tissues possibly represents a tissue-level maintain tissue homeostasis. feedback on the cell cycle regulation of individual cells. Indeed, Author contributions: S.J.S. and L.H. designed research; S.J.S., C.R.H., and T.S. performed experimental and theoretical studies have proposed that physical research; S.J.S. and D.H. contributed new reagents/analytic tools; S.J.S. and C.R.H. ana- parameters, such as cell geometry or local tissue mechanics, lyzed data; and S.J.S., C.R.H., and L.H. wrote the paper. regulate cell division (15–20). However, most of the evidence so The authors declare no conflict of interest. far is based on correlation, and it remains unclear whether me- This article is a PNAS Direct Submission. chanical constraints control cell cycle progression in growing 1S.J.S. and C.R.H. contributed equally to this work. tissues and at what stage of the cell cycle this regulation may act. 2To whom correspondence may be addressed. E-mail: [email protected] or To address this question, we have combined live imaging of [email protected]. the cell cycle state of individual cells over time in a model epi- This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. thelium with experimental perturbation of its spatial constraints. 1073/pnas.1323016111/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1323016111 PNAS Early Edition | 1of6 Downloaded by guest on October 2, 2021 A Cells in S-G2-M phase Cells in G0-G1 Phase B Cell tracks and features Free t t t t time PDMS barrier 0 1 2 3 space Cells at G1-S interface (x0,y0)(x1,y1)(x2,y2)(x3,y3) position s0 s1 s2 s3 state A A A A Area of cross 0 1 2 3 section Generate unconstrained boundary by barrier removal C DEF t = 0 h t = 30 h t = 60 h Time (h) Time Position (µm) Fraction cycling cells C‘ D‘ E‘ F‘ ) 2 m µ Time (h) Time Cell Area ( Position (µm) GIH J 1.2 ) ) 800 0.5 40 2 1 2 Time (h) 700 m 20 m 0.4 µ 200 300 400 500 2 µ 0.8 (h) Time CellArea ( m )(µm2) 600 0.3 0.6 G1G0/G1 phase 500 S−G2−MS-G2-M phase 400 0.2 0.4 S/G2/M 300 0.1 0.2 Relative phase duration G0/G1 200 Cell Area ( Cell Area ( 0 0 200 250 300 350 400 450 500 200 300 400 500 −30 −20 −10 0 10 20 30 Time (h) Cell Area (µm2) Cell Area (µm2) Time (h) Fraction cycling cells Fraction cycling cells Relative phase duration Fig. 1. Cellular area correlates with cell cycle progression during tissue invasion. (A) Fucci cell cycle marker in tissue invasion assay. (Left) Cells in G0–G1 phase constrained by a polydimethylsiloxane (PDMS) barrier. (Right) Shortly before imaging, the boundary is removed, generating free space available to the tissue. (B) Single cell track over time and extraction of features at each time point: position, cell cycle state, and cell area. (C) MDCK Fucci cells in tissue colonization assay. Representative images of the progressing tissue are shown 0 h, 30 h, and 60 h (C–E) after barrier removal. The tissue invades the free space, and cells progress in the cell cycle. Scale bars: 500 μm. (C′–F′) Average cell area at the same time points as in C–E, color-coded as in F′.(D)AsC.(D′)AsC′.(E)AsC.(E′)As C′.(F) Kymograph showing temporal evolution of FCC (color coded). (F′)AsF for cell area. (G) Cell area (blue) and FCC (gray) plotted against time aligned to the first occurrence of 30% cycling cells (indicated by the horizontal black dashed line) in a segment behind the leading edge (t = 0, vertical black dashed line).