Tuning intercellular cohesion with membrane-anchored oligonucleotides

Cohesive interactions between cells play an integral role in development, differentiation, and regeneration. Existing methods for controlling cell-cell cohesion by manipulating protein expression are constrained by biological interdependencies, e.g. coupling of cadherins to actomyosin force-feedback mechanisms. We use oligonucleotides conjugated to PEGylated lipid anchors (ssDNAPEGDPPE) to introduce artificial cell-cell cohesion that is largely decoupled from the internal cytoskeleton. We describe cell-cell doublets with a mechanical model based on isotropic, elastic deformation of spheres to estimate the cohesion at the cell-cell interface. Physical manipulation of cohesion by modulating PEG-lipid to ssDNAPEGDPPE ratio, and conversely treatment with actin-depolymerizing cytochalsin-D, resulted respectively in decreases and increases in doublet contact area. Our data are relevant to the ongoing discussion over mechanisms of tissue surface tension and in agreement with models based on opposing cortical and cohesive forces. PEG-lipid modulation of doublet geometries resulted in a well-defined curve indicating continuity, enabling prescriptive calibration for controlling doublet geometry. Our study demonstrates tuning of basic doublet cohesion, laying the foundation for more complex multicellular cohesion control independent of protein expression.


2/17
Manipulating expression of cohesion-mediating proteins remains the dominant strategy for 26 modulating cohesive interactions between cells [17][18][19] , though a converse strategy is to restrict the 27 spatial positioning of cells with 2D and 3D localization methods such as micropatterning to induce 28 natural cohesion in a predictable manner [20][21][22][23][24] . 29 Cell surface modification with phospholipid-anchored oligonucleotides (ssDNAPEGDPPE) offers 30 a means of inducing physical attachment of cells to substrates or other cells orthogonal to natural 31 adhesion and cohesion biochemistry [25][26][27][28] . Recently, Y. Teramura showed that ssDNAPEGDPPE 32 could be used to investigate the interactions of cohering cell doublets 29 demonstrating dependence of 33 cell-cell interactions on their degree of contact. We have also previously shown that native 34 integrin-mediated adhesion modes take place in parallel with artificially-induced attachment with 35 ssDNAPEGDPPE 28 . 36 The goal of this study is to use artificial ssDNAPEGDPPE-mediated cohesion in a cell doublet 37 model as a tool to manipulate geometric parameters, e.g. their interfacial contact area and extent of 38 membrane deformation, while minimizing the effects of coupling to mechanical, cytoskeletal 39 biochemistry characteristic of other cohesion modes. We show that by tuning the concentration of 40 ssDNAPEGDPPE molecules on the surface of cells before doublet formation, it is possible to achieve 41 a well defined range of doublet geometry distributions characterized by interfacial contact area. 42 1 Materials and Methods 43 44 α-N-hydroxysuccinimidyl-ω-maleimidyl poly(ethylene glycol) (NHS-PEG-Mal, MW 5000), 45 α-Succinimidyl carbonyl-ω-methoxy, polyoxyethylene (PEGDPPE), and 46 1,2-dipalmitoyl-sn-glycerol-3-phosphatidylethanolamine (DPPE) were purchased from NOF

85
Cells i and j interact with a total energy where W Rep is the potential energy contribution due to elastic repulsion and W Adh is the 87 potential energy contribution resulting from adhesion over the surface of contact.

88
To estimate the adhesion energy, we have chosen to model the cell repulsive energy due to elastic 89 deformation by the Hertz-model applied to two compressed elastic spheres described in Landau and 90 Lifschitz's Theory of Elasticity 33,34 shown schematically in Fig. 1 c. Two cells i and j of unequal 91 spherical radii R i and R j are divided by an interfacial contact area A contact of diameter a ij . The 92 two cell centers of curvature are separated by a distance d ij , and the truncated caps due to adhesion 93 are together of length x ij referred to as the indentation depth. The contribution to interfacial energy 94 due to repulsion is whereẼ and ν are the Young's modulus and Poisson's ratio respectively defined for both cells i 96 and j. 97 We assume that the radii of curvature of cells can be represented by spherical radii. We imported 98 the Young's modulus, which we assume for simplicity to be constant in the Hertz model, from  The adhesive contribution to potential energy is where γ is the cohesion represented in units of energy per unit area. The contact area can be 103 rewritten as a function of R i , R j , and d ij such that the potential due to adhesion becomes (4) Thus, given values of R i , R j , and γ, the total interfacial energy between cells i and j can be 105 determined as a function of d ij or x ij . Fig. 1 d shows a plot of potential energies W (magenta),

106
W Rep (red), and W Adh (blue) as a function of interfacial contact area A contact and cohesion γ for a 107 hypothetical pair of fixed volumes V i = V j = 900 µm 3 .

108
The minimum of W can be determined by solving for the root of its derivative with respect to 109 separation distance, the point where the force along the axis extending from cell center to center is 110 zero: and is shown in Fig. 1 To compute the space of possible steady state doublet conformations including contact area as a 116 function of both Young's modulus and cohesion (Fig 2 d), a numerical algorithm was implemented 117 to identify the values of A contact which satisfy the root of dW ddij under an additional constraint of   . This action corrects for irregularities in shape of the interface leading to a sharper distribution 137 of pixel intensities when the image is collapsed into a single horizontal axis. The collapsed intensity 138 profile peak is then integrated and divided by the interface diameter or chord length a ij to obtain 139 the intensity density of the profile in arbitrary intensity units.   potential energy minimum where repulsive and attractive forces are balanced (Fig 1 d).

8/17
We designed and implemented a numerical algorithm to compute the interfacial contact area  doublets prepared from cell groups modified without PEGDPPE (Fig 2 b). 219 We wished to reduce the mechanical integrity of cells, so we treated cells during the 220 ssDNAPEGDPPE modification stage and during doublet formation/imaging with cytochalsin D, a 221 cytoskeletal inhibitor previously shown prevent actin polymerization and reduce cortical 222 integrity 40,41 . Doublets prepared with cytochalsin D (Fig 2 c) exhibited qualitatively largerĀ contact 223 cross-sections in confocal images compared to normal doublets (Fig 2 b). over the average doublet geometry (SEM < 10%) (Fig 2 g).  (Fig 3 b). Forward scatter vs green fluorescence distributions were of highly reproducible 247 shape between conditions besides being shifted in their peak fluorescence according to the with flow cytometry (Fig 3 g) demonstrating prescriptive control over the relative mean 251 concentration of ssDNAPEGDPPE on cell surfaces through the addition of PEGDPPE during 252 modification. 253 We developed a fluorescence spectroscopy-based assay to estimate the surface concentration of 254 FAMSeqAPEGDPPE on membranes of single cells sampled from the same population of cells used 255 for flow cytometry. We used oligonucleotides terminated with a single fluorescein moiety to ensure 256 that the fluorescent emission corresponds to oligonucleotide (SeqA) concentration while leaving 257 PEGDPPE molecules unlabeled (Fig 3 c). We used Benzonase, a nonspecific nuclease capable of 258 digesting ssDNA and dsDNA, to digest away the oligonucleotide-tethered fluorescein molecules, 259 releasing them into the surrounding medium which we collected and analyzed with fluorescence 260 spectroscopy (Fig 3 d). Fluorescent halos marking the contour of the cell membrane in confocal 261 cross-sections could be observed after cells were modified with FAMSeqAPEGDPPE (Fig 3 e), while 262 after Benzonase digestion and removal of the digest supernatant the remaining cells exhibited 263 negligible fluorescence levels when re-imaged (Fig 3 f)  cytometry-based fluorescence intensity, the mean molecular surface concentration correlates inversely 269 with the PEGDPPE:FAMSeqAPEGDPPE molar ratio (Fig 3 h). concentrations computed from fluorescence spectroscopy (Fig 3 j).  (Fig 4 a).

289
These images were then converted to vertically aligned images. Observing the distribution of 290 intensity along the length of these images, intensities were attenuated within 1 µm from the 291 perimeter of the interface, but otherwise apparently uniform throughout the center region ie 292 standard deviations less than 10% in pixel intensity for most cross-sections. The aligned images were 293 converted to collapsed into 1D profiles (Fig 4 b) and integrated to obtain a single mass intensity 294 value for each doublet interface. Across all tested PEGDPPE:FAMSeqAPEGDPPE ratios, we 295 observed a positive correlation between individual doublet mass intensities and their corresponding 296 A contact (Fig 4 c). Importantly, when we analyzed cytochalsin D-treated doublets, we observed a 297 similar positive correlation between mass intensities and their correspondingĀ contact across the 298 range of doublets analyzed, however with an apparently lower modulus than the non-cytochalsin D 299 12/17 treated doublets. 300 We converted integrated intensity profiles to fluorescence intensity-based surface densities by 301 dividing them by the interface diameter a ij while assuming a constant confocal microscope 302 sectioning depth. When the geometrically determined mean cohesion values for different PEGDPPE 303 : ssDNAPEGDPPE ratios are plotted against intensity densities, a nonlinear trend emerges (Fig 4 304  d), and the plot ofĀ contact vs intensity density similarly reveals a nonlinear relationship. The

305
intensity density of the cell-cell interface should correspond to the molecular concentration of 306 fluorescein and thus to the concentration of oligonucleotides at the interface, however because 307 doublet interfaces exhibited intensity densities spanning a range past the upper limit of those taken 308 from images of single cell membrane profiles (Fig 4 e), we were unable to construct a reliable doublet revealed that fluorescently-labeled ssDNA was distributed throughout the interface between 317 the two cells (Fig 1 b)  DNA hybridization is rapid, with the contact-formation phase occurring within seconds. Our system 324 is also likely to involve actomyosin feedback in the form of cortex remodeling which has been shown 325 to occur on timescales of minutes 42,43 in response to cell shape change or external deformation. Our 326 system therefore, while cadherin-free, is nevertheless not completely orthogonal to cytoskeletal 327 dynamics.

328
Our model describes the space of geometric configurations (represented byĀ contact ) in terms of 329 Young's modulus and cohesion, postulating effective independence (Fig 2 d). We hypothesized that 330 modulation of one and/or the other could permit navigation over the configuration space. To explore 331 this principle experimentally, we used a non-hybridizing PEGDPPE during ssDNAPEGDPPE 332 modification to reduce the cohesion of cells and observed that the interfacial contact area decreased 333 compared to doublets without PEGDPPE, and conversely cells treated with cortical actin-inhibiting 334 cytochalasin-D exhibited large interfacial contact areas despite undergoing analogous 335 ssDNAPEGDPPE modification as untreated cells (Fig 2 a, b, c). By preparing 336 complementarily-modified cells with different PEGDPPE : ssDNAPEGDPPE ratios, we were able to 337 systematically shift the distributions of interfacial contact areas determined from microscopy image 338 analyses (Fig 2 e) and converting the measurements of doublet geometry into mean cohesion 339 estimates via equation 6 show that systematic shifts in doublet geometry correspond to analogous 340 shifts in the cohesion (Fig 2 f). For simplicity we assume a constant Young's modulus independent of 341 cortical rearrangement, conceiding that it is likely to be affected in the case of highly deformed 342 doublets. Nevertheless we see that the PEGDPPE relative concentration enables prescriptive 343 adjustment of cell-cell cohesion (Fig 2 g). 344 Our results apply to an ongoing discussion in cytoskeletal biology over what extent spontaneous 345 separation of different cell types within tissues and the interfacial tension of multicellular domains 346 are a result of differential adhesive compatibility analogous to phase separation of polar and 347 nonpolar liquids 14,[44][45][46][47] relative to the cortical stiffness of cells 48,49 . A confounding factor comes 348 from the annular actin reorganization in cadherin-cadherin-cohering cell doublets, which 349 13/17 concentrates at the circumference of the cell-cell interface where cadherins also accumulate 15,50 . The 350 reallocation of actin to the periphery of the contact reduces the compressive resistance that is 351 normally characteristic of surface area minimizing spherical cortical tension. Without direct coupling 352 to the internal cytoskeleton in our case, actin remodeling is likely restricted to that which occurs in 353 response to compressive deformation. The treatment of doublets with cytochalsin-D under normal 354 ssDNAPEGDPPE modification conditions and subsequent increased mean interfacial contact area 355 demonstrates that the doublet geometry is influenced by the elastic mechanics of the cells. In a 356 direct analogy to tissue surface tension, characterized by a minimization of the total exposed surface 357 area multicellular aggregates to total aggregate volume ratio, the surface area to volume ratio of the 358 cell doublet decreases with increasing interfacial contact area indicating a balance of cortical stiffness 359 and cohesion in agreement with current models 51-53 . 360 In summary, we report on the establishment of a model to study cell-cell cohesion with a 361 minimization of the active contribution of cohesion-inducing proteins like cadherins. By tuning the 362 ratio of PEG-only and oligonucleotide-bearing lipid conjugates or by the introduction of 363 stiffness-influencing pharmacological factors it was possible to affect the size of the contact area 364 between cohering cell doublets. We use a mechanical model to describe the differences in contact 365 area under different parameters. The system could be of value to studies which seek to examine the 366 impact of cohesion without a major confounding influence of other cytoskeletal factors.