Mechanochemistry in Materials
CHAPTER 1: Mechanochemistry: Inspiration from Biology
Published:24 Oct 2017
Mechanochemistry refers to the study of the evolution of the formation and disruption of chemical bonds upon application of an external force. In this chapter, the roles of mechanical forces in different biological systems are highlighted along with mechanisms and mechanotransduction pathways showing how complex biological systems can provide inspiration for materials design. Examples of how mechano-based systems have been mimicked by other scientists are also discussed including self-healing systems.
1.1 Introduction and Historical Perspective
Mechanochemistry from the contraction of µηχανικóς mēkhanikos (mechanic) and χηµíα khēmia (chemistry) is the study of the evolution of the formation and disruption of chemical bonds upon application of an external force. The terminology “mechanochemistry” was first coined by Ostwald as “the coupling of mechanical energy and chemical energy”.1 Unlike electromagnetic force or gravity, mechanical force implies contact. Indeed, for a force to be applied on a given object or organism, it entails a connection or at least transmission of forces through a medium. Mechanical energy, often in the form of applied force, is a lesser-known way to initiate chemical reactions than conventional stimuli (e.g. heat, light and electricity). At first glance, it is especially interesting and somewhat intriguing to think of chemistry triggered by forces, as it is not something that is typically taught in your General Chemistry 101 class. However, bond ruptures upon mechanical action surround us from random chain scission as one tears through packaging material or simply as one presses on a surface and senses the mechanical deformation. In prehistory, our ancestors took advantage of mechanochemical phenomena for survival, e.g. drilling wood for fire. Theophrastus of Eresus (ca. 315 BC), Aristotle’s student, first recorded that grinding cinnabar in a copper mortar using a brass pestle could reduce it to mercury.2 In the 19th century, Faraday and Lea applied sliding and grinding to study the chemical reactions of solid substances.3
In many other ways though, for the better part of its existence, humankind has tried to use materials that would minimize their alteration upon application of force. For instance, Damascus steel, which dates back to 900 AD, was a metal of choice for weaponry as it remarkably exhibited superplasticity along with incredible levels of hardness. Interestingly, it is also one of the early examples (though unwittingly) of nanotechnology and mechanochemistry as carbon nanotubes as well as cementite nanowires were proven to form during the forging and annealing of Indian wootz steel. It is indeed believed that the combination of heat and mechanical action along with impurities in the ore (including carburizing wood and leave additives) was responsible for the catalyzing the process responsible for the formation of these nanostructures and the characteristic wavy patterns.4 The quest for materials that would best resist mechanical constraints has been a technological driving force throughout the Anthropocene and the field of polymer science is no exception to this pursuit. For example, worldwide many groups have been looking at nanocomposites as a source of reinforcement of lightweight materials, with many of these solutions having made it to commercial products in the fields of transportation, construction or even appliances.5
From Staudinger to Melville,6,7 mechanical force was studied early on and the random scission often seen as a foe engendering undesirable chain scission and weakening or worse causing the rupture of polymeric materials. However, in the early 2000s, a new idea emerged in the realm of polymer science.8 Namely, it became apparent that one could utilize mechanical force to one’s advantage rather than combating it. Sensing, repair or self-stiffening were some of the many putative functionalities proposed in the context of mechanochemistry (which have been achieved today and are described in detail in the following chapters). Interestingly and regardless of whether Nature was a source of inspiration for some of the seminal work, it is worth noting that these very functions are ubiquitous in biological systems. Like in many other areas of the sciences, materials researchers recognized early on that one could learn a tremendous amount of information from studying biological materials and understanding how millennia of evolutions have shaped the functionalities of biological systems. Consequently, the present chapter focuses on the mechanochemical strategies developed by Nature to achieve some of the aforementioned functions. It aims to familiarize materials scientists (who may not always have the opportunity to investigate them) with some of the biomechanics and biophysics reports that cover these intricate (and sometimes not entirely understood) processes. This chapter was written from a polymer scientist’s perspective and will therefore remain partial and sometimes simplistic. This “editorial line” is substantiated by two main reasons. First, the audience of this book is presumed to be mostly polymer researchers and it would be illusory to condense the complexity of biological mechanotransduction pathways and make it fully accessible in a few pages. Second, the point of this chapter is rather to inspire materials scientists, and provide them with a general overview of the strategies utilized by living organisms to sense and adapt to mechanical constraints. Consequently, a broad stroke approach seems more adequate as it would appear nonsensical and unachievable to reproduce exactly what Nature does to transduce mechanical forces. Instead, one can draw overarching guiding principles upon which to base his/her reflection and ultimately the design of biomimetic mechanotransducers. Interestingly, there are already several reports of materials which are (wittingly or not) akin to biological systems and, whenever suitable, we will draw a parallel between artificial and biological systems.
1.2 Biomimetism and Rationale for Emulating Mechanotransduction Pathways
Before delving deeper into the intricacies of mechanotransduction, it is worth doing two things: (i) remembering some of the key principles of how to approach biomimetism and (ii) explaining the reason why mechanotransduction is critical for life.
1.2.1 Principles of Biomimetism and Strategies to Implement It
In 1994, a team of researchers led by David Tirrell laid the foundation for the study of materials of biological origin, specifically hierarchically organized structures.9 Such hierarchy is also present in mechanotransduction schemes and it stands to reason that some of these general ideas developed in their report are readily applicable to mechanotransduction. Tirrell and coworkers highlighted some of the commonalities in materials of biological origin, which are worth recalling. Particularly, we will highlight the properties as they pertain to mechanotransduction.
Evolutionary engineering has advanced by means of an iterative process whereby the structure was refined slowly at each generation. As will become apparent, transduction pathways often find a great deal of commonalities, probably originating from shared ancestral strategies. Many of the elementary units (evidently nuclear bases and amino acids but not only, e.g. microfibrils) are recurrent such that function often comes from specific assembly rather than the building blocks themselves. This means for instance that orientation control plays a critical role in determining the assembly responsiveness. The latter is usually adapted and gradual, and varies according to the task performed (e.g. slow adaptation vs. imminent danger). Shape plays a pivotal role in dictating and modulating the response and shape complexity ensures the right response to the right cue. Since the (mechanotransduction) tasks often prove repetitive, resistance, durability and resiliency are essential. These properties are often predicated upon subjacent principles of reversibility or better yet of mendability. To achieve these paramount functions, Nature often capitalizes on the utilization of non-covalent forces as well as out-of-equilibrium dynamics.10 While the former are readily achievable and have been realized in the field of mechanochemistry (see Chapter 5), the latter is harder to implement in synthetic systems as dissipative structures are still a research curiosity. It is interesting to note that many of the biological systems involved in mechanotransduction in cells are either in kinetically trapped states (e.g. folded proteins, cell membranes) or simply in a dissipative state (e.g. the formation/rearrangement of the cytoskeleton to accommodate deformations).
Also of interest is the realization that the mechanical coupling usually occurs between objects of disparate sizes. Interestingly, the interactions maintaining these objects together are often weak and their chemical and thermal stabilities moderate. In that sense, artificial systems are likely to display superior performance and will therefore be more adapted to some of the harsh requirements of materials used in various applications (e.g. transportation or armament). By combining this enhanced resistance of man-made assemblies with the hierarchical notions of biological materials, one can anticipate the creation of tantalizing responsive structures. When considering the latter, one must also keep in mind the necessity to direct the self-assembly processes and to have them happen at a fast rate. While biological materials assemble at a slow speed (often due to concurrent interactions), any sort of technologically relevant artificial assembly would require high-throughput procedures to warrant economical viability. While Nature can afford long times for assembly to sustain life, it is indeed desirable to operate at greater celerity as one aims towards the widespread implementation of smart materials (i.e. materials capable of adapting their environment). Indeed, such materials typically require long fabrication time and are consequently rather costly. Of concern also when designing next-generation mechanoresponsive polymers is the introduction of environmental cost in the equation of the design of said materials. Much like Nature chose to use limited raw materials, it would make sense in the long run to limit oneself to non-deleterious starting materials and/or materials that can be easily recycled/reused or degraded.
Whenever possible, the best approach to ensure the success of biomimetic strategies takes advantage of the synergistic combination of diverse expertise (e.g. biophysicists, to synthetic chemists, material scientists, computational chemists and bioengineers). It is only through the combination of these know-hows that we will succeed in developing not only passively smart materials (i.e. that will be pre-programmed to respond by means of set transitions) but also actively smart materials (i.e. whereby a feedback loop serves to modulate the response as proposed in Chapter 7 of this monograph).
1.2.2 Introduction of the Importance of Mechanotransduction Pathways for Living Organisms
The ability to sense force is of paramount importance for the survival and the favorable evolution of complex organisms comprised of many cells. The latter themselves are constantly experiencing a slew of mechanical actions: flow, elongation, pressure and pressure waves. The ability to change as a response to the nature of their surrounding environment is capital in the development and the subsistence of living organisms. This feedback is necessary for a variety of key biological functions such as proliferation, differentiation, motility or cell death.11 It also plays a role in organ growth, bone adaptability and homeostasis (viz. the maintenance of parameters, e.g. pH, sugar, within a normal range for an organism). By analogy, for smart materials, one could anticipate systems that adapt to mechanical load or vibration, detect and report failure as well as initiate needed mending cycles. The implementation of feedback loops and graduated responses is a necessary condition for the future development of better smart materials. In living organisms, deficient feedback loops in mechanotransduction are responsible for numerous diseases such as cardiomyopathies, cancer metastasis or muscle degeneration. At the cellular level, mechanical deformation triggers diverse signaling pathways or induces changes in ionic concentrations, which in turn modulates the response appropriately (adjustment of the stiffness of the cell, its shape or even cross-talk with the extracellular matrix).
When one is thinking about the notion of mechanotransduction, the first things that probably come to mind are the senses of touch or hearing. Sensory cells are responsible for both. These cellular structures are hyper-specialized and have evolved to transduce mechanical inputs into given signals (e.g. flow of Ca2+ ions through transmembrane proteins to elicit an action potential in neurons). Sensory cells were logically amongst the first ones to be studied, as they offered a convenient model for mechanosensing. However, mechanotransduction goes far beyond the mere sensory pathways. It is important in the proper functioning of the vascular system and, in particular, in the role of the heart and its development.12 It shapes the bones and their strength as a result of muscle contractions and external forces (e.g. gravity or hits). Think of the ability of a martial arts expert to strengthen his/her shins or forearms to endure greater impact and administer harsher blows, or even patients encouraged to walk to limit osteoporosis and encourage bone growth. This self-strengthening mechanism is highly desirable and has been the source of inspiration for exquisite studies in the field of synthetic mechanochemistry. For instance, Black Ramirez et al. have demonstrated the possibility to induce mechanochemical strengthening as a response to shear forces in a poly(butadiene) system containing gem-dibromocyclopropane along the chain and infused with the ditetrabutylammonium salt of sebacic acid.13 Upon ring-opening of the three-membered propane cycle, one bromo-group thus formed can react with the dicarboxylate, thereby promoting the sought-after cross-linking reaction and subsequent reinforcement. Embryonal cell fate is also directed by external forces applied to the cell, which helps in the coordination of tissue growth. Also, said forces have been shown to influence stem-cell differentiation and have been proposed as a way to replace diseased tissues. Respiration cycles and the mechanical feedback loop associated with them contribute to the lung homeostasis.
Interestingly, in the body, all these circuits work in parallel, meaning that the overall response is the fruit of multiple mechanotransduction events. Wang et al. introduced the concept of gating in polymers as a means to control stress-response. The incorporation of cyclobutane served as a “gate” to control the activation of gem-dichlorodicyclopropanes. The single-chain force spectroscopy measurement indeed demonstrated the power of that gating mechanism for the design of next-generation materials. While this is rather limited in scope as compared to much more sophisticated biological systems, it is certainly a step in the right direction. Therefore, by combining mechanophores and/or specific architectures, one can readily envision creating systems that will get closer to emulating the intricate mechanotransduction pathways.
Mechanotransduction approaches can be broadly classified into those involved in sensory mechanisms (i.e. involving ion channels) and those mediated by more complex signaling pathways (i.e. entailing the coordination of a cascade of biochemical processes leading to the desired response). The former rely on the concerted rearrangement of vast supramolecular ensembles, such as lipid bilayers and pore-forming structures, while the latter are more often than not the consequence of protein unfolding and molecular conformational changes. The former are typically fast, while the latter often have latency times in excess of minutes. Interestingly, and as stated previously, both strategies involve non-covalent interactions. By reducing the number of bonds to break and form, Nature optimizes its energetic savings and limits the penalty associated with what would be a complex return to the pre-load state (equilibrium or metastable). Orr and coworkers have proposed the hypothesis that specialized cells employ mechanisms for mechanotransduction similar to the rest of the cells but obtain greater sensitivity by magnifying the strains experienced by the primary transducers.14
1.2.3 Adaptivity in Bones
Biological structures are not mere static systems. Instead, they are actively responding to the stimuli from their milieu,15 constantly adjusting their architecture to adapt their properties to their surroundings. There are, for instance, repair mechanisms that are used to maintain the integrity of the system and improve its performance or simply to repair damage caused by external factors. These very principles constitute the basis of mechanobiology, which has the mantra “form follows function” as it assumes that an organism changes its structure to fit its role. Perhaps a good example of this mechanobiological approach is observed in bone development. In 1962, Chalmers and Ray carried out a classic experiment to investigate the influence, if any, of mechanical loading on the structure adopted by a bone during osteogenesis.16 They used a cartilage model of a femur and transplanted it into the spleen – an area without mechanical loading. The cartilage developed into a femur, but did not have any of the refinements that are associated with a functioning femur (e.g. its characteristic shape), thereby illustrating the necessity of mechanical force in osteogenesis to fit function.
The bone is able to fine-tune its properties in response to changes in the mechanical environments thanks to the permanent remodeling of its constitutive material. This adaptation occurs on two different occasions: (i) during growth, or (ii) as a response to new mechanical requirements. Both can, however, be explained by the same remodeling process. In cancellous bone (i.e. the spongy portion of osseous tissue as opposed to cortical bone, its compact counterpart), a cavity is formed when osteoclasts (large bone cells that absorb bone tissue) resorb a “bone packet”.15,17,18 This cavity is then filled by osteoblasts to form a new bone matrix (Figure 1.1). Gradually, the matrix absorbs minerals, rapidly at first then slowly, increasing its stiffness. This bone remodeling process occurs throughout an animal’s lifetime in response to mechanical strains exerted on the bone cells. Fratzl mentions that an adult skeleton contains between 1 and 2 million remodeling sites.15
In addition to the effect of mechanical loading during bone growth, a fascinating fact about bone tissue is the control that said mechanical forces have in the bone remodeling process. This role has been known as early as the late 19th century, with studies of cancellous bone structures showing that bone remodeling was a mechanically controlled process rather than a random replacement of old bone matrix. The Wolff–Roux law states that “bone is deposited wherever mechanically needed and resorbed whenever there is no mechanical use”, in a “use it or lose it” system.15 Frost further proposed that there is a feedback loop that regulates bone mass and architecture, functioning like a “mechanostat”, where an increase in mechanical loading leads to bone deposition and vice versa.19 This level of adaptability is evidently the Holy Grail of materials science where the system could readily stiffen or soften at will. In bones, increase in mechanical loading can be caused by a new intensive exercise activity (e.g. squash and tennis) while a loading decrease can be caused by inactivity. Frost estimates that increases in local strains of between 0.15 and 0.3% will lead to an increase in bone mass while resorption can be caused by a 0.01–0.03% decrease in local strains. As one considers the possibility to have switchable mechanomorphing materials, it is indeed interesting to note this asymmetry between growth and resorption, or as the case would be in man-made materials between stiffening and softening. Evidently, biological systems present the advantage of having a “constant” supply of building blocks brought on site through the vasculature, which is harder to implement due to the difficulty to implement such a system. It is, however, worth noting the remarkable work of Sottos, Moore, White and coworkers, on the fabrication of vascular structures for self-repair, whereby, for instance, dicyclopentadiene is polymerized via ring-opening metathesis polymerization to heal cracks.20 Evidently, these vascular systems are based on damage and not simply mechanical force and also cannot currently help in the implementation of reversible softening and hardening through a desirable feedback loop. Nonetheless, these approaches could ultimately be used to circulate the necessary components to reversibly stiffen or soften as a response to mechanical force. Unlike bone growth, which requires a complex process, one can readily imagine a system whereby the kinetics of adaptation could be dictated by the kinetics of a given reaction (Figure 1.2).
In seeking to understand mechanical force impact on bone development, a number of experiments are often performed, probing factors such as frequency, magnitude and direction of the forces. An in vivo experiment is the avian ulna model (Figure 1.4a), where the ulna is only exposed to external loading. Results have shown that dynamic mechanical stimulus produces a bone forming response but a static stimulus does not. An interesting observation from this experiment is that bone loss can also be prevented by high frequencies (30 Hz) even when accompanied by sub-physiological stimuli. Interestingly, this cycling concept has not really been explored, except in materials science. Indeed, to the best of our knowledge, nothing has been done to reproduce what bones often experience, namely repeated impact at the limit of elasticity. By playing with frequencies and amplitude, one can anticipate some conditions under which mechanical activation would be enhanced. Similar observations could be made whereby deformations in the a priori elastic regime could upon cycling result in activation.
While proper load cycling has not been attempted, Craig and coworkers have demonstrated how soft spiropyrane-bearing polydimethylsiloxane systems could be used to detect damage by opening to the merocyanine form and reversibly closed to afford the “original” material (Figure 1.3).21 Likewise, Imato et al. have used the reversibility in diarylbibenzofuranone-based mechanophores to achieve cycling in thermoplastic polyurethane.22 Both of these examples take advantage of the reversible nature of the mechanophore linkages and repeated deformation is applied after a recovery step, whereby the broken link is regenerated.
Interestingly, Larsen et al. have used repeated activation to demonstrate the concept of flex-activated mechanophores, whereby multiple compressions after folding were used to achieve maximum activation. Here, again, this concept is different from that of oscillatory stress. Instead, the material is repeatedly damaged and plastically deformed. Also of interest though is the work of Beiermann et al.,23 who looked into the alignment of chains (as a result of tensile stretch rate) as a factor impacting greatly the activation of mechanophores. They found that increasing stretch rates led to a greater accumulation of stress (ergo, a more pronounced response). One could anticipate that repeated application of stress in one direction may lead to a rearrangement of chains in one dimension and a more likely activation.
While most in vivo experiments like those mentioned above seek to understand the end results of the mechanical forces on bond cells, in vitro studies aim to relate the bone cells’ biological response to the Wolff–Roux law and investigate the mechanotransduction pathway for such processes. They help elucidate how mechanical stimulus is sensed by the bone cells and transduced into cellular signals, which are then relayed to osteoclasts and osteoblasts. In bone cells, small channels called canaliculi are responsible for the supply of nutrients to the osteocytes. Loading of bones will cause interstitial liquid to be squeezed through the canaliculi, leading to stimulation of the osteocytes by both the fluid shear stress and electromagnetically due to the ionic composition of the tissue fluid. Studying these mechanotransduction processes through has also shed more light on the conditions that facilitate osteoporosis. For example, in response to pulsating fluid flow, osteocytes have been observed to produce higher levels of PGE2, a signaling molecule, which promotes anabolic actions leading to bone formation. PGE2 response to strain has also been shown to be enhanced by estrogen. This observation has been used to explain why there is a higher rate of osteoporosis in women during menopause. Response to assemblies to fluid motion is indeed a very interesting concept that has been investigated by Holme and coworkers26 and then further revisited by the Zumbuehl group (Figure 1.5).27 One idea is to create shear-sensitive liposome assemblies that will deliver their therapeutic payload as a function of shear stress in a liquid environment. For instance, one may consider such delivery systems for the treatment of vasodilation or clot lysis.
Another interesting feature of bone cells is the unique healing process, which occurs in response to fracture. The bone system is exceptional when compared to other tissues in that it is able to completely restore the original bone form and functionality instead of leaving scar tissue. This regeneration occurs in steps similar to the skeletal development process. On the extreme side, living organisms such as salamanders can actually regrow their limbs completely. While complete regrowth is not possible (yet?!) in humans, the bone-healing process still manages to replace/repair damage. In addition to biological influences, mechanical stimuli play an important role in the healing process. In a primary fracture, the fragments are first immobilized and the remodeling occurs through the creation of new bone cells to connect the two fragments. While the healing process is considered ideal, it is a very delicate process. The fragments are required to be free of any mechanical loading and the bone is very weak until complete healing. However, since all the treatment procedures require some degree of motion, primary fracture healing is very rare. Secondary fracture healing, on the other hand, is less delicate. During bone fracture, the blood supply is disrupted, leading to the cell death. The inflammation that occurs on the affected area is instrumental in cleaning off the dead materials, restoration of the blood vessels and to congregate mesenchymal stem cells (MSCs). As will be discussed later in this chapter, mechanical forces are instrumental in the differentiation of these MSCs into cartilage, fibro-cartilage and bone cells. A hard bone shell forms around the fracture site, which is filled with cartilage. The soft tissue mineralizes and is finally replaced by bone, before the shell is removed by resorption. While the biological machinery involving multiple step recovery (e.g. inflammation, clean up, resorption) is hard to replicate with materials, it is still interesting to point out some of the principles behind healing as one considers ways to emulate Nature. For instance, the principle of using a scaffold for rapid repair followed by a resorption of said scaffold is an interesting concept that has not been exploited. It is, however, worth mentioning the work of White et al., whereby a gelation step facilitates the recovery of large damage volumes, which in spite of the lack of resorption uses the principle of scaffolding (Figure 1.6).28
Studies of bone healing in Nature influence the design of self-healing materials. Remarkably, the effect of mechanical loading in bone cells can be used as an inspiration to develop materials whose function determines their structure or those capable of self-healing in response to damage by mechanical forces.29 The latter has seen a number of materials developed, with potential use of self-healing materials in electronics, coatings and transportation industries. These systems are usually designed using three approaches: capsule-based, vascular (vide supra) and intrinsic methods. In the capsule-based method, a healing agent is stored in capsules. Upon material fracture, the capsules rupture, releasing the healing agent. The vascular approach employs a matrix of hollow tubes throughout the material, storing the healing agent (Figure 1.7). Upon fracture, the healing material is released. The intrinsic self-healing polymeric systems often contain materials with a latent functionality that can trigger the healing process in the event of damage.
One of the earliest examples of a capsule-based self-healing system was demonstrated by White et al. in 2001 with an epoxy matrix.20 The matrix was loaded with urea-formaldehyde shells encapsulating dicyclopentadiene (DCPD) as the healing agent. A small amount of a Grubbs catalyst was dispersed in the matrix and, upon cracking of the matrix, the microcapsules were observed to rupture, releasing the healing agent. Upon contact with the Grubbs catalyst, the DCPD polymerized, rebonding the cracked matrix. Since then, many other groups have exploited the principle of capsules for damage reporting.32–35
1.3.1 Via Protein Unfolding
For decades, studies in various biological systems have shown how mechanical forces on the small scale lead to responses from simple microscale levels in cells to large-scale ones that can lead to even larger consequences like changes in transcription and protein levels. The mechanical stimuli responsible for these changes includes hydrodynamic forces, shears and stresses. Early work by Klibanov and Ishimori showed that often simple enzyme conformational changes brought upon by mechanical forces can lead to changes in enzyme activity.36 Klibanov et al. covalently attached chymotrypsin and trypsin to nylon fibers and observed that when stretched, the enzymes lost their ATPase activity. Later, Ishinori also observed a decrease in activity on glucose oxidase attached to polyvinyl chloride membranes. From these observations, they hypothesized that enzyme deformation due to the stretching were responsible for the loss in activity. The influence of mechanical force on enzyme is actually no stranger to the field of polymer science as this was the topic of the thesis of Tirrell in 1977.37
This reliance on mechanical perturbations by various enzymes has also been observed in proteins that possess cryptic sites or sites that are hidden when in the protein’s natural state. The presence of mechanical force leads to unfolding of the proteins exposing the hidden reactive sites that are responsible for initiating signaling pathways. One such example is the cytoskeletal protein, Talin, active in the cell adhesion process.36 Talin is responsible for linking integrins to the cell’s cytoskeleton by binding to vinculin, another cytoskeletal protein. In its resting state, the binding locations are inaccessible to the vinculin but when stretched by tensional forces, the sites become accessible leading to the vinculin binding and force transfer to the extracellular matrix (ECM). Similarly, the von Willebrand Factor (vWF), a protein instrumental in the blood clotting process also uses the cryptic site approach. In the case of vascular damage, the hydrodynamic forces from the blood lead to a conformational change from the vWF’s coiled state to an unfolded state, which exposes the binding sites for platelet receptors, which play an important role in the blood clotting process. Fibronectin (FN), a protein instrumental in extracellular matrix (ECM) binding to the cells, also exhibits these cryptic sites as shown in Figure 1.8.38 In FN, the spatial positioning of its loops (RGD and PHSRN) controls the cell binding. Small forces can unfold FN, increasing the distance between the two loops, and exposing its binding sites in its type III domains. In this stretched state, α5β3 integrin binding is favored while the unstretched state allows for αvβ3 binding. In a simple way, Nature uses changes in protein conformation to control access to binding sites, thereby resulting in the ability to regulate complex biochemical process using low energy. These cryptic-site-based systems show how small mechanical forces can elicit large-scale responses in the body.
Work done by Discher et al. in their investigation of red blood cell (RBC) protein, β-spectrin, also showed how mechanical forces can influence specific binding of a protein to cysteine residues.39 In RBCs, blood flow induces stresses that can lead to deformation. The membrane of RBC is reinforced by a cytoskeleton constituted in part of spectrin chains connected to F-actin through its cysteine residues (Figure 1.9).40 F-actin is an essential protein instrumental in the deformation process. Spectrin’s α and β chains contain 20 and 15 cysteines, respectively. In the experiment, RBCs were reversibly lysed and filled with a reactive cysteine-modified fluorophore before being reformed and sealed (“ghost RBCs”). They were then exposed to a range of conditions at varying temperatures, while some were sheared using a microfluidic device. They were then relysed and the results showed that the sheared RBCs had about 66% more fluorescent dye compared to the static ones, suggesting that the folding hides specific domains.
The biomimetic inspiration from these simple cryptic site systems has been adopted in the rising field of soft-mechanochemistry, which is carving out a new approach in polymer and materials science design. The idea of mechanochemistry, the process of transforming a mechanical force into a (bio)chemical response has been around for a long time. However, most systems often require harsh conditions in the form of “high-energy forces and high energy input” and are mostly irreversible. Therefore, the possibility of using small mechanical forces to elicit large changes has the ability to significantly increase the application of polymer science in both biological and non-biological systems. Impressive work in soft-mechanochemistry has been done by Jierry and coworkers at the University of Strasbourg, whereby they have demonstrated polymer-based systems that utilize cryptic sites to reduce the energy demand.36 Using a surface-functionalized elastic layer of polydimethylsilane (PDMS), they grew polyethylene glycol (PEG) brushes on the surface before attaching biotin, a ligand for the protein streptavidin. In the unstretched state, the PEG brushes completely covered the biotin on the surface. When a streptavidin solution was added to the surface, no binding was observed, but, when stretched, the streptavidin bound to the biotin, with a linear increase in binding with stretching observed. This binding was reversible, with the strong lateral pressure from the PEG brushes breaking the strong non-covalent bond between the streptavidin and the biotin. When the biotin was replaced with RGD, a cell adhesion protein, the setup showed a reversible binding to osteoprogenitor cells.
These cryptic site-based systems have also been successfully used to design mechanoresponsive catalytic systems by the same group. In one such system, they developed a catalytically active 3-D network with partial reversibility by using a maleimide-modified version of the enzyme B-galactosidase (Figure 1.10a).41 The network consisted of polyelectrolyte multi-layer films (PEMs) of thiopyridyl-modified poly-l-lysine (PLL-S-TP) and hyaluronic acid (HA) deposited on PDMS. After initial carbodiimide cross-linking in the PEMs, the modified enzyme was then introduced, and diffused into the layers before permanently embedding into the layers through the thiol-maleimide click reactions with deprotected thiopyridyl moieties as shown in Figure 1.10b. Stretching the PEM films resulted in a catalytic activity decrease of up to 40% (with fluorescein di(β-d-galactopyranoside) as the substrate). They hypothesized that the stretch-induced enzyme conformational change led to the decrease in activity.
Perhaps the earliest influence of mechanical forces in human growth is evident in the heart, the first functional organ to develop in an embryo. Cardiogenesis relies on the stresses exerted on the heart for growth and cell differentiation. In trying to understand cardiac tissue development, Discher et al. postulated a model centered on cell-matrix interaction in which the proliferation of fibroblasts is limited by the stiffness of their environment, which in turn correlates with their collagen matrix density.12 Collagen production is mainly done by fibroblasts; therefore, collagen mRNA is expected to be proportional to the fibroblast composition. The fibroblast composition is determined by the stiffness of the environment, itself controlled by the ECM. The overall point of the model is that turnover (i.e. expression and lysis) of key structural proteins is mechanoregulated. Cells transduce stresses/strains or any forces that are exerted on them into a change in protein level, which is reflected at the transcription level (hence the link to RNA). A model which supports the idea that forces affect protein synthesis/degradation and transcription means that a “use it or lose it” mechanism likely controls key circuits.
The approach based on Discher’s model seems to suit the observations that have been made during growth from an embryonic heart to a full adult. The density and stiffness of the extracellular matrix (ECM) regulates the function and structure of the neonatal cardiomyocytes (cardiac muscle cells). In a seemingly self-reinforcing model for heart development (Figure 1.11), the stresses exerted through neighboring cells is propagated through the ECM and cell–cell adhesions onto cardiac fibroblasts (connective tissue producing collagen). In response to the strain, the fibroblasts divide and produce more ECM, which in turn cause an increase in production of microfibril proteins. The resulting reorganization of the microfibril produced exerts a strain on the fibroblasts, restarting the cycle. This somewhat simplified model of fibroblast and cardiomycyte balance during cardiogenesis can provide ideas on the design of “self-aware” systems, which use the presence of mechanical forces around them to control their growth or perform their functions.
Research in muscular dystrophies and other conditions linked to skeletal muscle cells has also helped elucidate the importance of mechanotransduction and the complexity of the pathways that allow for signal transmission from the extracellular matrix to the nucleus to elicit responses at a transcription level.11 The transmission of extracellular forces into signals that reach the nucleus is initiated in the extracellular matrix (Figure 1.12), which as mentioned earlier consist of tissue-specific proteins such as collagen, laminin and fibronectin. The ECM is linked to the cytoskeleton (actin) through adhesion complexes that are present at the cell surface. In skeletal muscle, this adhesion complex consists of dystrophin and a dystrophin-associated protein complex (DAPC). Forces transmitted through the cytoskeletal network are then transferred to the nucleus through coupling to nesprins and other proteins on the outer nuclear membrane. Nesprins interact with inner nuclear membrane proteins (such as SUN 1 and SUN 2), which in turn interact with lamins and other nuclear envelope proteins such as emerin. Lamins complete the force transmission pathway by forming stable structures which are able to bind to DNA. A number of studies targeting different components of this mechanotransduction pathway have shown that changes in any of these sites could result in altered cellular functions. These changes can be brought about by factors such as mutations, and changes in cellular environments or structure. From a materials science perspective, it is interesting to see the hierarchy and the level of modularity found in living organisms. Quite manifestly, it is thus far nearly impossible to emulate this level of intricacy. Nevertheless, as we aim to fabricate ever more complicated systems, it is quite intriguing to notice that parallel pathways involving multiple components are involved in graduated responses.
The genetic disorder Duchenne muscular dystrophy is an example of the effect of changes in one component of the pathway. Mutations in the gene responsible for encoding dystrophin production lead to disruption in the dystrophin-associated protein complex’s ability to transmit force signals between cytoskeleton and the ECM, eventually leading to progressive muscle degeneration and other consequences in cell function and viability. For example, work by Loufrani et al. shows that muscle-fibers with a dystrophin deficiency lead to decreased vascular density in cardiac muscle.42 Defects and mutations in the nuclear envelope proteins such as nesprins and emerin have also been shown to result in muscular dystrophies. In general, disorders that are caused by disruptions in the mechanotransduction pathways are often caused by changes in three components that directly affect the pathways: the extracellular environment, cell structure and organization, and defects in the sensing and signaling (Figure 1.13).
As Figure 1.13 shows, the number of possible factors that can lead to defective mechanotransduction pathways cannot be overstated, and therefore the study of the effect of mechanotransduction in physiology is clearly important. Understanding the role of mechanotransduction in systems can lead to an increase in the role polymer scientists can play in the design of possible remedies.
1.3.2 Via Ion Channel Opening
Mechanosensory hair cells convert mechanical energy from sound waves into electrochemical signals that can be further transferred into the nervous system. Mechanically gated ion channels in the mechanotransduction pathway are located at the tip of the stereocilia (apical modifications of the cells), giving the hair cells a specialized sensitivity that is not observed in other cell types in response to mechanical forces (Figure 1.14).43,44 The stereocilia are connected to the mechanotransduction channels by tip links, extracellular filaments which are able to transmit information on the differential deflection of the stereociliary tips to the transduction channels. The tip links’ ability to sense direction of the hair stereociliary tips is understood to be instrumental in the state of the ion channels. Deflection towards taller stereocilia activates the opening of the ion channels and, consequently, a cation influx while deflection towards the shorter stereocilia leads to gate closure.45 However, these characteristics are only a small part of the mammalian hair cell mechanotransduction puzzle. The transduction pathway is complex and consists of many components. While several of these components are known, there is not enough information on their roles in the mechanotransduction pathway. Recent models that have postulated that the complexity of the process is necessary for frequency discrimination of the mechanotransduction pathways have been identified.
Knowing more about hair cells and the transduction pathway will help with understanding and coming up with solutions and remedies for malfunctions in the mammalian cochlea. Of greater interest, however, is the potential of systems that can be developed from this knowledge. Developing systems that have the discriminative mechanosensitivity possessed by hair cells will open up exciting applications in both sensing systems and noise-cancelling systems. One such system was recently reported by Asadnia et al., who developed novel “miniature all-polymer flow sensors that closely mimic the intricate morphology of the mechanosensory ciliary bundles in biological hair cells” (Figure 1.15).47 Using the flexible polydimethylsilane (PDMS), they fabricated rows of pillars to mimic hair bundles in fish. Mimicking the geometry of the hair bundles imparted flow direction sensing on the bundles. The PDMS pillars were then connected to piezoelectric polvinylidine fluoride (PVDF), which acted as tip links. A hyaluronic acid-based hydrogel was then used to cover the pillars to enhance their sensitivity. This simple, all-polymer system showed sensitivity to water flow with a very low threshold detection of 8 µm s−1.
In addition to our audial ability, the sense of touch is another fascinating example of human senses for which mechanotransduction plays a starring role. While conventional touch is essential for survival, allowing one to observe and respond to mechanical perturbations in their surroundings,48 discriminative touch allows us to accomplish uniquely human operations, from playing musical instruments to tweeting about successes in the lab. A number of mechanosensitive neurons innervate the skin,49 with each of the sensations often observed by the skin being mediated by different receptors. For example, light touch is mediated by Aβ afferents with a lower mechanical threshold. Painful touch, on the other hand, is mediated by nociceptors that have a higher mechanical threshold.
While there is still a large gap in knowledge about the mechanism of the transduction pathways, it is agreed that mechanical stimuli play an active role in directly activating these transduction channels. This mechanical gating can be activated in a number of ways, such as the use of forces in the membrane bilayer, or through the extracellular membrane (Figure 1.16).
There are four main classes of mechanoreceptors and their classifications are determined by factors such as morphology, sensitivity to dynamic and static events, mechanical threshold, firing pattern and the mechanical stimulus to which they respond.50,51 These classifications are slow adapting-type I (SA I), slow adapting-type II (SA II), rapidly adapting-type I (RA I) and rapidly adapting-type II (RA II). SA I receptors are often represented by Merkel cells, which are critical for detection of texture. SA II afferents consist of Ruffin endings and perceive the stretch of the skin between fingernails to determine an object’s shape and for motion detection. FA I receptors consist of Meissner’s corpuscles, which are responsible for low-frequency vibrations, slip and motion detection. FA II receptors, represented by Pacinian corpuscles, show high-frequency sensitivity. These afferents can also be classified based on their electrophysiological properties such as action-potential propagation into Aβ, Aδ and C-fibers. Just like the studies of human hair cells, the understanding of the mechanotransduction pathways in touch sensitivity still lags behind, on a molecular and cellular level. However, recent discoveries such as the identification of molecular markers for several types of touch receptors have given scientists the ability to probe individual sensory neurons. Interestingly, to the best of our knowledge, this approach relying on transmembrane gating proteins has not been pursued as an avenue for mechanochemistry in polymers (Figure 1.17).
Piezo proteins, which are a group of mechanically activated ion channels, are a class of recently discovered proteins that have provided more answers to the transduction pathways in mammalian touch.48 They exist as tetramers, with each tetramer consisting of over 30 transmembrane proteins. They can be activated by a variety of stimuli such as touch, suction and shear stress, which points to the fact that they are activated by membrane deformation. Isolated Piezo1 proteins have been observed to form ion-conducting pores in lipid bilayers. More so, it has been observed that disease-causing mutations slow these gating channels. Piezo2 proteins, which are part of this protein family, are known to be responsible for mechanotransduction in most receptors. High concentration of the Piezo2 protein transcripts in dorsal root ganglion (DRG) neurons suggests that the gene is responsible for encoding transduction channels in touch receptors. Loss of Piezo2 proteins has been shown to lead to complete loss of gentle-touch responsiveness. The discovery of these Piezo proteins further helps to illuminate the role of mechanical forces and their transfer in the mammalian touch process. Further studies on these proteins will help us to better understand the mechanotransduction pathway in mammalian touch, and provide inspiration in the design of other systems in mechanochemistry.
Just like hair-cell sensitivity, the ability to mimic touch sensitivity has a vast range of applications, from prosthetics to sensors. Consequently, a number of systems that seek to mimic touch sensitivity have been developed. Although these examples may not fall in the conventional category of polymer mechanochemistry, we felt it was necessary to bring them to light as they can provide an important inspiration for the field and its potential direction. While most of them do not have the ion channels in their transduction process, they seek to follow the same principles of using mechanical perturbations on their surfaces to detect touch. In analyzing the biomimetic systems that have been developed in artificial tactile sensing, Lucarotti et al. mainly divided these methods into two approaches, namely synthetic skin and transduction mechanisms, and bio-artificial mechanisms.51 Bio-artificial approaches are also divided into bio-hybrid systems (synthetic sensors within tissue-engineered skin) and fully biological sensors. These bio-artificial systems often find applications in dermal substitution and repair. Fully biological tactile sensors often consist of an outer layer made from tissue-engineered hydrogel (consisting of cell cultures such as ECM materials) and gelatin containing artificial skin.51 Using the inspiration from skin cells, Cheneler et al. designed a bio-hybrid tactile sensor that is often used to monitor responses of mammalian cell types to normal and tangential loads.52 In their design (Figure 1.18), they etched microchannels into a silicon substrate to act as a supply route for nutrients to cultured cells that were placed in a well above the substrate and separated by a nanoporous polycarbonate membrane. Conductivity sensors were attached on either side of the well with electrodes and a constant current was kept flowing through the electrodes. Fluctuations in conductance due to extracellular ions caused by applied loads were then detected and analyzed to study the response depending on the load applied.
Synthetic tactile transduction methods often employ capacitative, optoelectric, inductive, piezoelectric and piezoresistive techniques in their design (Figure 1.19). Piezoresistive systems transduce force variations into changes in resistance, which can be detected by simple measuring systems.53 Piezoelectric systems take advantage of the electrical charges produced when some materials are under a mechanical force in the presence of an electrical dipole moment. Capacitance-based systems monitor changes in the area (A) and distance between electrodes (d) induced by the forces applied to the system. The changes in these dimensions are then used to determine the capacitance (C = εA/d) based on the material’s dielectric constant (ε).
Recently, Jung et al. designed a piezoresistive tactile sensor capable of discriminately measuring multidirectional forces.54 The inspiration for the design of the sensor was based on the organization of mechanoreceptors in the skin. The dermal layer has multiple mechanoreceptors, which allows it to respond to different types of stimulations. The signals are sent to the central nervous system without complex processing of the transduced signals because of the use of different nerve fibers. In Jung’s design, the sensor system consists of multiple sensing elements embedded to measure the magnitude and direction of applied forces (Figure 1.20). Each of the sensing elements consisted of two composite films made from PDMS mixed with carbon nanotubes (CNT/PDMS). A conductive adhesive is used to connect the CNT/PDMS films to conductive polymer blocks, which act as electrodes. The force exerted on the contact area reduces the area between two CNT/PDMS films, which in turn causes a change in the resistance of the interlocking films. The sensitivity of the sensing elements was increased by fabricating microdome structures on the film sides facing each other. The changes in the resistance of each element were then used as a measure of the direction and magnitudes of the forces applied. The sensor was shown to successfully measure the shear and normal pressures as low as 5.28 kPa and 128 Pa, respectively.54 In addition to this high sensitivity, the minimal interference was observed between the sensing elements. Results from this open up the possibility of such accurate tactile responsiveness in applications such as slip sensitivity in prosthetic limbs.
The skin’s sensory ability has also been increasingly mimicked in the development of the so-called electronic skin (e-skin), which seeks to develop systems which even exceed the resolution and rapid response of the skin itself.53 The Javey group has been one of the groups that have been producing interesting work in the development of pressure sensitive e-skins.55,56 They have even developed a user-interactive e-skin that provides instant visual feedback through mapping the applied pressure onto a built-in organic light emitting diode (OLED) matrix.57 Using transistor arrays, they are able to transduce and amplify the pressure signal into electronic signal. In their design (Figure 1.21), they laminated an OLED matrix between a transistor matrix and a pressure sensitive rubber (PSR). Increases in the pressure applied on the rubber lead to a decrease in resistance by pushing closer embedded nanoparticles that are in the PSR. The decrease in resistance will increase the current flowing through the OLED, thereby increasing the intensity of the output light. This “system-on-plastic” design demonstrates how easy it has become to get instantaneous feedback on artificial tactile systems, something which was not possible only a few decades ago. A significant amount of design in the old systems went into the process of analyzing the data from the mechanoresponse in order to deduce the type of stimuli.
A significant number of these synthetic skin and transduction systems such as those mentioned above have found a variety of applications, from prosthetics, computer touch screens and pressure mapping. In a self-completing cycle, some of these wearable artificial tactile sensors are now being used to monitor the skin – using its own technique to monitor it.58 While power consumption has been a traditional deterrent in the use of some of these systems, the emergence of better design techniques has led to more energy-efficient systems. In fact, some systems are now able to self-power, with flexible solar cells59 and piezoelectric nanogenerators60 recently gaining prominence. Wang et al. have a good review that discusses some of the most recent and exciting developments in the development of e-skin.53
In this chapter, the role of mechanical forces in a variety of biological systems was highlighted. The conversion of these mechanical forces into chemical bonds is also important. A number of mechanisms and mechanotransduction pathways in a variety of systems have also been highlighted, providing inspiration on how the complex biological systems can provide ideas to design systems that serve as remedies for when these systems are malfunctioning or for other systems altogether. Some of the ways in which these mechano-based systems have been mimicked by other scientists have also been highlighted.
The importance of mechanical forces was shown in the bone development, from growth to repair, and the effect of mechanical forces in space exploration was also highlighted. We also got a glimpse of some applications that are mainly based on the bones such as self-healing systems. Work by Discher et al. showed the importance of mechanical forces in exposing binding sites in folded proteins. This work has given the idea of cryptic sites, which were exploited by Jierry and coworkers in the developing field of soft mechanochemistry, where the use of small mechanical forces is exposing these cryptic sites and leading to chemical changes with larger energy requirements. The effects of failures in mechanotransduction pathways in living organisms was also observed, with a long list of diseases associated with defects in mechanotransduction, from cancer, central nervous system disorders and immune system disorders. Such cases provide challenges to polymer chemists to use this knowledge of various mechanotransduction pathways in order to develop materials that can perform the same functions to act as remedies or outright replacements. Lastly, the effect of mechanical forces on the sensitivity in both hair and skin was also discussed. The mechanoresponsiveness and sensitivity have now been mimicked to develop systems with cochlear and tactile sensing abilities that can serve not only as replacements in the case of permanent injuries and genetic defects, but in other applications such as pressure mapping, energy production and even humanoid skin. Such work can serve as a reminder that Nature is always an endless source of inspiration.