Cellular and Molecular Bioengineering, Vol. 7, No. 1, March 2014 (Ó 2013) pp. 26–34 DOI: 10.1007/s12195-013-0307-6 Substrates with Engineered Step Changes in Rigidity Induce Traction Force Polarity and Durotaxis 1 1,2,3 1 1,4 MARK T. BRECKENRIDGE, RAVI A. DESAI, MICHAEL T. YANG, JIANPING FU, 1,5,6 and CHRISTOPHER S. CHEN 1Department of Bioengineering, University of Pennsylvania, Philadelphia, PA 19104, USA; 2Max Planck Institute of Molecular Cell Biology and Genetics, 01307 Dresden, Germany; 3National Institute for Medical Research and University College London, London, UK; 4Department of Mechanical and Biomedical Engineering, University of Michigan, Ann Arbor, MI 48109, USA; 5Department of Biomedical Engineering, Boston University, Boston, MA 02215, USA; and 6The Wyss Institute for Biologically Inspired Engineering, Harvard University, Boston, MA 02115, USA (Received 3 March 2013; accepted 23 September 2013; published online 9 October 2013) Associate Editor Michael R. King oversaw the review of this article. Abstract—Rigidity sensing plays a fundamental role in suggesting that cells could sense substrate rigidity locally to multiple cell functions ranging from migration, to prolifer- induce an asymmetrical intracellular traction force distribu- ation and differentiation (Engler et al., Cell 126:677–689, tion to contribute to durotaxis. 2006;Loet al., Biophys. J. 79:144–152, 2000; Wells, Hepatology 47:1394–1400, 2008; Zoldan et al., Biomaterials Keywords—Durotaxis, Cell migration, Rigidity sensing, 32:9612–9621, 2011). During migration, single cells have Mechanotransduction, Microfabrication. been reported to preferentially move toward more rigid regions of a substrate in a process termed durotaxis. Durotaxis could contribute to cell migration in wound healing and gastrulation, where local gradients in tissue rigidity have been described. Despite the potential impor- INTRODUCTION tance of this phenomenon to physiology and disease, it remains unclear how rigidity guides these behaviors and the Rigidity sensing plays a fundamental role in guiding underlying cellular and molecular mechanisms. To investi- the outcome of multiple dynamic cell behaviors including gate the functional role of subcellular distribution and migration, proliferation, and differentiation.4,8,13,31,37 dynamics of cellular traction forces during durotaxis, we developed a unique microfabrication strategy to generate Single migrating cells have been reported to sense rigidity elastomeric micropost arrays patterned with regions exhib- by preferentially migrating towards the more rigid region iting two different rigidities juxtaposed next to each other. of the substrate in a process termed durotaxis.14,19 After initial cell attachment on the rigidity boundary of the Durotaxis is potentially a prominent guidance cue for cell micropost array, NIH 3T3 fibroblasts were observed to migration as many physiological situations where cell preferentially migrate toward the rigid region of the micro- post array, indicative of durotaxis. Additionally, cells bridg- migration is important, such as wound healing and gas- ing two rigidities across the rigidity boundary on the trulation, are characterized by local changes in environ- micropost array developed stronger traction forces on the mental rigidity.2,36 In addition, pathologic conditions more rigid side of the substrate indistinguishable from forces such as cancer and fibrosis are characterized by local generated by cells exclusively seeded on rigid regions of the increases in tissue rigidity, which may contribute to micropost array. Together, our results highlighted the utility increased trafficking of fibroblasts and immune cells into of step-rigidity micropost arrays to investigate the functional 2,15 role of traction forces in rigidity sensing and durotaxis, the diseased foci. Despite its importance, detailed molecular and cellular understanding of durotaxis remains largely elusive. Studying how matrix mechanics regulates durotaxis requires engineered substrates that present a reproduc- Address correspondence to Jianping Fu, Department of Mechan- ible and quantitatively well-defined substrate rigidity ical and Biomedical Engineering, University of Michigan, Ann Arbor, gradient. Prior studies documenting durotaxis have been MI 48109, USA and Christopher S. Chen, Department of Bioengi- performed by generating a rigidity gradient within syn- neering, University of Pennsylvania, Philadelphia, PA 19104, USA. thetic hydrogels through varying the ratio of hydrogel Electronic mail: [email protected], [email protected] 13 Mark T. Breckenridge and Ravi A. Desai contributed equally to monomer to cross-linker across a hydrogel substrate, this work. using gradients of light to mediate photo-initiated 26 1865-5025/14/0300-0026/0 Ó 2013 Biomedical Engineering Society Durotaxis on Microfabricated Surfaces with Step Rigidities 27 PA crosslinking,28 using gradients of PA pre-polymer increasing the post diameter along one direction of the generated using microfluidic approaches,10,28 or simply array.21 However, changing the post diameter in the applying a tangential strain in the direction away from a PDMS micropost array affects cellular adhesive envi- cell with a microneedle to locally pull PA gels.18,29 These ronment (e.g., the post diameter and density changes can studies have revealed the existence of durotaxis in dif- directly affect the maximum size of individual FAs and ferent types of mechanosensitive adherent cells and FA organization, respectively) that can introduce shown that durotaxis is functionally correlated with additional microenvironmental signals to make substrate rigidity gradient magnitude and is mediated by interpretation of experimental findings difficult. actomyosin-mediated cellular traction forces and cell To assess whether cells sense substrate rigidity attachments to extracellular matrix proteins via focal during durotaxis on a local or cellular length scale, adhesions (FAs).10,18,28,29 A recent study using com- herein we reported a novel microfabrication strategy posite materials containing local rigid adhesive islands that could generate PDMS micropost arrays with dis- grafted onto the surface of a non-adhesive polyacryl- crete step changes in rigidity without using compli- amide hydrogel has suggested that rigidity sensing may cated PECVD or CMP process (see ‘‘Methods’’ be dictated by material compliance across the cell section). In order to keep the tips of the PDMS length.9 Interestingly, a more recent study using high- microposts coplanar, we altered micropost heights by resolution time-lapse traction force microscopy changing the height of underlying base of the sub- (TFM)18 has implicated that traction force fluctuation strate. Cells seeded across post rigidity boundaries and distribution within single mature FAs might be responded by migrating towards more rigid posts, important to regulate durotaxis.18 Together, these indicating the presence of durotaxis on our step-rigidity studies have established the importance of a substrate PDMS micropost array. We then measured traction rigidity gradient in mediating directional cell migration forces of cells bridging step-rigidity boundaries and and suggested that sensing substrate rigidity can be a observed that cells could generate asymmetrical strains local mechanotransductive molecular event occurring and traction forces within each rigidity region, such within individual FAs or a cellular response integrating that subcellular regions could behave analogously to rigidity signals across the whole cell body. However, cells wholly seeded on uniform substrates with that whether or not cells sense gradients of rigidity within rigidity. In particular, traction forces of cells bridging individual FAs or if rigidity is sensed globally between step-rigidity boundaries were greater for subcellular FAs during durotaxis has not been directly investigated. regions on more rigid substrates. These results sug- In addition to PA gels, elastomeric micropost arrays gested that cells could sense substrate rigidity locally, made in polydimethylsiloxane (PDMS) have proven as a and this caused an asymmetrical intracellular traction versatile tool to control substrate mechanics and report force distribution resulting in durotaxis. traction forces with a sub-nanonewton (nN) resolution for single adherent cells.5,20,23,33 Recent studies have further shown the possibility to generate PDMS mi- RESULTS cropost arrays with controlled rigidity profiles while monitoring live-cell traction forces during cell migra- Negative masters for the step-rigidity PDMS micro- tion. The first PDMS micropost array with rigidity post array were fabricated in silicon (Si) wafers using gradients was generated with discrete rigidity bound- standard photolithography and a two-stage deep reac- aries by changing post heights while keeping the post top tive-ion etching (DRIE) protocol (Fig. 1a). The first surface coplanar.6 However, even through durotaxis DRIE step was used to generate negative Si masters with was reported on the step-rigidity PDMS micropost ar- a regular array of microscale cylindrical holes of a uni- ray, this study did not report cellular traction forces form depth across the Si substrate (Supporting Fig. 1a). during durotaxis. Further, the method employed to The depth of microscale holes, which determined the generate the step-rigidity PDMS micropost array post height in soft regions of the step-rigidity PDMS required a complicated microfabrication process micropost array, was precisely