The role of adhesion energy in controlling cell-cell contacts.

Recent advances in microscopy techniques and biophysical measurements have provided novel insight into the molecular, cellular and biophysical basis of cell adhesion. However, comparably little is known about a core element of cell-cell adhesion--the energy of adhesion at the cell-cell contact. In this review, we discuss approaches to understand the nature and regulation of adhesion energy, and propose strategies to determine adhesion energy between cells in vitro and in vivo.

Cell adhesion is a key biological property for maintaining multicellular structures. Regulated adhesion is necessary for several morphogenetic processes, such as cell migration and tissue segregation. During migration, adhesion controls the friction required for efficient translocation between the cell and its substrate [1-4]. Adhesion also plays an important role in regulating the differential cell affinities needed for cell sorting [5] and maintaining tissue integrity during cell segregation [9-12]. In all these processes, the mechanical role of adhesion is to provide the energy (adhesion energy) needed for the optimal attachment of cells to their surroundings [2,13]. The adhesion energy arises from the ionic and hydrogen bonds formed between cells and their adhesion partners. The main proteins mediating cell–substrate interactions are Integrins, while Cadherins typically are responsible for cell–cell adhesion. Both Integrins and Cadherins assemble an adhesion complex on their intracellular tail linking them to the cytoskeleton of the cell [14]. Although more and more components of the adhesion apparatus are being identified, it remains unclear how they function to determine adhesion energy.

The two functions of adhesion energy are to control the adhesive strength and morphology of the cell contact. At this contact, the adhesion energy is antagonized by cell tension, mostly mediated by the contractile actomyosin cell cortex underlying the contact [15]. The combined activities of adhesion energy and cortical tension constitute the interfacial energy at the contact that can expand the contact by a unit of area. Thus, adhesion energy and cortical tension are the key parameters determining how cells interact with their environment [19]. While cortical tension has been extensively studied [7], adhesion energy is still poorly understood.

Here, we want to summarize and discuss the current state of knowledge of adhesion energy, and propose strategies of how to measure and/or calculate it. Although some of the concepts presented here can be applied to both cell–cell and cell–substrate adhesion, we will mainly consider adhesion energy in the context of cell–cell adhesion. First, we will discuss the molecular and cellular basis of adhesion energy. Subsequently, we will describe different strategies and techniques of how to determine and measure adhesion energy in vitro and in vivo.

Molecular and cellular basis of adhesion energy

The main determinants of adhesion energy at the cell contact are the adhesion molecules. For cell–cell adhesion, the force required to separate contacting cells has been shown to linearly depend on the squared total number of E-Cadherin molecules in the cells [25] and the amount of E-cadherin at the plasma membrane [2]. This suggests that the number of classical Cadherins is likely to play a key role in controlling adhesion energy. However, Cadherins are known to undergo dynamic changes in their localization at the plasma membrane through various processes such as Cadherin clustering, endocytosis and recycling [26-28]. Moreover, in epithelial cells of the Drosophila germ band, several distinct plasma membrane pools of E-Cadherin have been identified, each of which bind different populations of Actin [29]. This suggests that the ability of Cadherins to control adhesion energy not only depends on the total amount of Cadherins at the contact, but also on the dynamic spatiotemporal distribution of these molecules and their association with distinct components of the adhesion complex.

Cadherins bind with their extracellular domain to other Cadherins and with their intracellular domain to molecules linking them to the cytoskeleton [30]. Notably, separating Cadherin-mediated cell–cell contacts or directly pulling on Cadherins can lead to the extrusion of plasma membrane tubes or tethers both in vitro [7] and in vivo [32], suggesting that extracellular Cadherin–Cadherin binding can be stronger than the intracellular binding of Cadherins to the cytoskeleton. Consequently, the energy of de-adhesion during contact separation might, at least partly, be determined by the strength of intracellular binding of Cadherins to the cytoskeleton. Therefore, in that case, the energy of de-adhesion would be different from the adhesion energy, which is determined by the binding strength of the adhesion molecules at the contact. It is important to keep this in mind when interpreting results obtained using experimental techniques that probe cell adhesion through de-adhesion measurements (see below).

The intracellular adhesion complex not only modulates adhesion energy by coupling Cadherins to the cytoskeleton, but also through its mechanosensing activity [21,32]. Cadherins at cell–cell contacts show enhanced clustering in response to increased stiffness of the substrate to which the contacting cells bind [34]. Moreover, directly pulling on Cadherin-mediated cell–cell contacts using a micropipette enlarges the cell–cell contact area [35], suggesting that the Cadherin adhesion complex is mechanosensitive. However, the observation that increased cell contractility and consequently tugging on cell–cell contacts not only promotes Cadherin clustering [36] and anchoring to the Actin cytoskeleton [37], but also leads to disassembly of Cadherin-mediated cell–cell contacts [18], indicates that the mechanosensitive response of the Cadherin adhesion complex to contraction varies. While the basis of this variation is still unclear, the association of different Actin networks [29] and isoforms of Myosin motors [41] with the adhesion complex are likely to play an important role therein. To better understand how the Cadherin adhesion complexes react to mechanical force, it will be essential to identify the specific mechanosensitive molecules mediating these effects. α-Catenin plays a crucial role in connecting Cadherins to the Actin cytoskeleton by recruiting Actin and/or Actin binding molecules to the adhesion complex [42,43]. Interestingly, stress-induced conformational changes of α-Catenin have been shown to modulate its ability to recruit Actin [37], suggesting that α-Catenin is a crucial component of the Cadherin adhesion complex for sensing and transducing contact stress. Whether other molecules of the Cadherin adhesion complex are also involved in these processes remains to be established. The development of fluorescent probes that can report the stress applied to specific molecules of the adhesion complex, such as recently reported for Vinculin [44], will be a pivotal tool for constructing a detailed molecular force/stress map of individual adhesion complexes and identify potential candidates involved in mechanotransduction.

Strategies to determine adhesion energy

There are several possible methods to determine adhesion energy. One way is to calculate the energy of adhesion during contact formation on the basis of the binding affinity of the adhesion molecules involved. This, however, requires prior knowledge about the binding affinities of all adhesion molecules present, and the spatial (parallel, serial) arrangements of the bonds formed by these molecules. Because the complete molecular machinery mediating many situations of cell–cell contact are not known, this method is restricted to reduced/artificial systems of cell adhesion where the number and spatial configuration of adhesion molecules at the contact site are predetermined.

Another way to determine adhesion energy is to deduce it from the morphology and interfacial tensions of the contacting cells. Generally, the size of the cell–cell contact results from the minimization of the total surface energy. In order to increase the size of cell–cell contacts, contacting cells must consume energy (adhesion energy) that allows them to deviate from their preferential spherical shape owing to their surface (interfacial) tension [15]. The adhesion energy can thus be obtained on the basis of the geometry of the contacting cells and the interfacial tensions at the cell–cell and cell–extracellular medium interfaces (Box 1) [16,45]. This method of calculation assumes mechanical steady state of the adhering cells and therefore does not capture dynamic changes in adhesion energy during contact formation and maturation.

A key determinant of interfacial tensions is the contractility of the acto-myosin cortex, giving rise to cortex tension [7]. Cell contractility therefore represents an important factor in determining cell–cell contact size [18]. Indeed, the balance between cell contractility and adhesion has been proposed to determine the cell–cell contact size [19]. In a recent study, the relationship between cell–cell contact size, cell contractility and adhesion has been investigated. By measuring the traction exerted by contacting cells on their substrate, the tugging force between these cells was deduced assuming a force balance between tugging and traction forces (Box 2) [35]. The relation between the contact size and tugging force gives the contact stress, which is the property of the cell–cell contact to withstand tensile or compressive forces. Contact stress in turn should correspond to the adhesion energy, and thus knowing about contact stress will provide information about adhesion energy. However, a direct correlation between traction force and tugging force, and between tugging force and contact size has been disputed [40,48-50] and it therefore remains unclear how accurately adhesion energy can be deduced from those experiments.

An alternative approach to obtaining insight into the adhesion energy is to pull apart the cell–cell contact. When the chemical bonds engaged in the contact are separated, adhesion energy is released, producing a force that opposes the separation. In the case of passive elastic materials, the separation force is proportional to the released adhesion energy [51]. For cells, the relationship between the separation or de-adhesion force and the adhesion energy is less clear. The interfacial tension at the cell–cell contact is determined by the combinatorial activities of adhesion energy and cortex tension, which both are likely to undergo dynamic change during contact formation and maturation. In order to understand the relationship between de-adhesion force and adhesion energy, different experimental methods have been developed that allow measuring the force that is needed to mechanically separate contacting cells (Box 2). De-adhesion force measurements can be done both on a molecular and cellular level. On a molecular level, atomic force microscopy (AFM) [52,53] and bioforce probe (BFP) [54] have been employed to measure the de-adhesion forces of single molecules. Experiments with classical Cadherins revealed that E-cadherin has a higher de-adhesion force than N-Cadherin [52,53,55]. Moreover, single molecule studies demonstrated that the de-adhesion force of adhesion molecules strongly depends on the applied separation speed, or loading rate [54,56]. Intermolecular bonds are highly dynamic and change conformation with specific kinetics. Therefore, the pulling speed determines the probability for the paired molecules to be separated in a specific conformation, making it impossible to obtain one single value for the de-adhesion force of two molecules from such experiments. Despite this difficulty, separations performed with similar loading rates can be used to determine relative differences in de-adhesion forces for different molecules [7].

On a cellular level, AFM and the dual pipette assay (DPA) have been used to analyze the function of Cadherins in regulating the de-adhesion force at cell–cell contacts [25]. With AFM, the amount of E-cadherin at the plasma membrane has been shown to linearly scale with corresponding de-adhesion force at cell–cell contacts [2]. Moreover, DPA measurements showed that cell–cell contacts expressing E-Cadherin exhibit a higher de-adhesion force than contacts expressing equal amounts of N-Cadherin [25]. This is consistent with the observation that Cadherins are crucial for determining the adhesion energy at cell–cell contacts [1,2,7], and that E-cadherin molecules exhibit a higher de-adhesion force than N-Cadherin molecules [52,53].

How to deduce the adhesion energy from the measured de-adhesion forces is still a matter of debate. One approach has been to model cells that contact each other via polysaccharides as elastic solids in which cortical tension is uniform (independent of the specific interfaces) [58,59]. However, the observation of interface-specific regulation of cortical tension in cells binding via Cadherins [47] suggests that this approach might not be generally applicable to all cell–cell contacts. Another caveat when using de-adhesion forces to determine adhesion energy is that the molecules involved in contact formation might differ from the molecules involved in contact separation [31], and that thus the energy of adhesion might be different from the energy of de-adhesion (see above). The measurement of separation forces will therefore provide information about the energy of de-adhesion, but not necessarily about the energy of adhesion. Future studies identifying the molecules involved in cell–cell contact formation and separation will be needed to interpret the outcome separation measurements in relation to the adhesion energy.

Methods to measure separation force of cell–cell contacts are currently limited to cultured cells outside of their endogenous environment (ex vivo). However, the separation force critically depends on the specific cell environment such as calcium concentration [25] or substrate attachment [60]. Thus it is impossible to extrapolate the actual forces expected in vivo from the separation force values obtained ex vivo as long as the culture conditions do not precisely correspond to the in vivo situation. Although none of the methods currently available can directly measure cell–cell separation forces in vivo, various micromanipulation techniques allow indirectly determining the adhesive properties of tissues and their constituent cells in their organismal context. By deforming tissues, for example, through micropipette aspiration, several mechanical properties of the tissue, such as surface tension [61], viscosity [62,63], elasticity [63] and compliance [64] can be measured. These measurements can be done in control and experimental cells/tissues where the function of certain adhesion molecules is impaired [5] to obtain insight into the role of these molecules, and consequently of adhesion itself, in controlling tissue mechanics.

Adhesion energy of cell–cell contacts in vivo can also be determined by imaging cell–cell contact dynamics in vivo. While recent advances in image analysis of 2-dimensional time-lapse movies of contacting cells in vivo allow automatic tracking of cell–cell contact dynamics in high spatiotemporal resolution [33,65,66], methods for automatic 3-dimensional tracking are less advanced [67]. Moreover, theoretical models used to determine adhesion energy on the basis of cell–cell contact dynamics either still lack experimental confirmation [58,68,69] or rather provide information about cell–cell interfacial tension which is only partially determined by adhesion energy [17,24,45,70]. Combining the observation of contact size in vivo with direct measurements of mechanical tissue properties might be a good approach to more precisely analyze adhesion energy at cell–cell contacts in vivo.

Conclusion

We would like to argue that in order to understand cell adhesion, insight into the adhesion energy is required, as adhesion energy, together with cortical tension, are the key physical properties determining how cells adhere to each other. There are different ways by which the adhesion energy at cell–cell contacts can be determined: experimentally, it can be obtained on the basis of the cell–cell contact size, the spatiotemporal distribution of adhesion molecules at the contact, and/or the force needed to separate the contact. While these experimental approaches in principle are suitable for determining adhesion energy, questions remain as to the precise spatiotemporal distribution of adhesion molecules at cell contacts and the proper interpretation of cell separation experiments. Currently the most promising approach is therefore to deduce the adhesion energy from the geometry of the contacting cells and their interfacial tensions (Box 1). With the adhesion energy in hand, it is then possible to go back to the corresponding de-adhesion force and adhesion molecule distribution, thereby obtaining insight into the relationship between these variables. In particular, the relationship between adhesion energy and de-adhesion forces will be interesting to explore as the process of adhesion and separation probably involves different molecular bonds with different binding affinities. New methods, such as the recently developed molecular force sensors [44], will be highly useful for gaining mechanistic insight into force transduction at adhesion sites and the generation of adhesion energy. Eventually, a better understanding of adhesion energy will help in unraveling the molecular and cellular mechanisms underlying cell–cell adhesion and its role in determining tissue morphogenesis during development and disease.

References and recommended reading

Papers of particular interest, published within the annual period of review, have been highlighted as:

• of special interest

•• of outstanding interest