Tenocyte response to cyclical strain and transforming growth factor beta is dependent upon age and site of origin

The effect of strain and transforming growth factor beta on equine tendon fibroblasts (tenocytes) was assessed in vitro. Tenocytes were isolated from flexor and extensor tendons of horses from foetal to 10 years of age. These cells were cultured until confluent on collagen-coated silicone dishes. Cyclic biaxial strain of 9+/-1% was applied at 0.5 Hz for 24 hours with or without added TGFbeta1 or 3 (10 ng/ml). Proliferation and synthetic responses were dependent on the tendon of origin. Neither strain nor TGFbeta caused flexor tenocytes to proliferate significantly, while strain alone did proliferate extensor tenocytes. TGFbeta, with or without strain, increased the incorporation of [3H]-proline and the production of types I and III collagen and COMP in both cell types, although the effect on COMP production was more marked in flexor tenocytes, perhaps reflecting the higher levels found in this tendon in vivo. Immature flexor tenocytes synthesised more collagen and COMP than those from mature animals, while age had little effect in extensor tenocytes. Our results suggest that tenocytes become differentiated at an early age and present tendon-specific responses.     

Combining dynamic stretch and tunable stiffness to probe cell mechanobiology in vitro

Cells have the ability to actively sense their mechanical environment and respond to both substrate stiffness and stretch by altering their adhesion, proliferation, locomotion, morphology, and synthetic profile. In order to elucidate the interrelated effects of different mechanical stimuli on cell phenotype in vitro, we have developed a method for culturing mammalian cells in a two-dimensional environment at a wide range of combined levels of substrate stiffness and dynamic stretch. Polyacrylamide gels were covalently bonded to flexible silicone culture plates and coated with monomeric collagen for cell adhesion. Substrate stiffness was adjusted from relatively soft (G′ = 0.3 kPa) to stiff (G′ = 50 kPa) by altering the ratio of acrylamide to bis-acrylamide, and the silicone membranes were stretched over circular loading posts by applying vacuum pressure to impart near-uniform stretch, as confirmed by strain field analysis. As a demonstration of the system, porcine aortic valve interstitial cells (VIC) and human mesenchymal stem cells (hMSC) were plated on soft and stiff substrates either statically cultured or exposed to 10% equibiaxial or pure uniaxial stretch at 1Hz for 6 hours. In all cases, cell attachment and cell viability were high. On soft substrates, VICs cultured statically exhibit a small rounded morphology, significantly smaller than on stiff substrates (p<0.05). Following equibiaxial cyclic stretch, VICs spread to the extent of cells cultured on stiff substrates, but did not reorient in response to uniaxial stretch to the extent of cells stretched on stiff substrates. hMSCs exhibited a less pronounced response than VICs, likely due to a lower stiffness threshold for spreading on static gels. These preliminary data demonstrate that inhibition of spreading due to a lack of matrix stiffness surrounding a cell may be overcome by externally applied stretch suggesting similar mechanotransduction mechanisms for sensing stiffness and stretch.     

Development of fibroblast-seeded collagen gels under planar biaxial mechanical constraints: a biomechanical study

Prior studies indicated that mechanical loading influences cell turnover and matrix remodeling in tissues, suggesting that mechanical stimuli can play an active role in engineering artificial tissues. While most tissue culture studies focus on influence of uniaxial loading or constraints, effects of multi-axial loading or constraints on tissue development are far from clear. In this study, we examined the biaxial mechanical properties of fibroblast-seeded collagen gels cultured under four different mechanical constraints for 6 days: free-floating, equibiaxial stretching (with three different stretch ratios), strip-biaxial stretching, and uniaxial stretching. Passive mechanical behavior of the cell-seeded gels was also examined after decellularization. A continuum-based two-dimensional Fung model was used to quantify the mechanical behavior of the gel. Based on the model, the value of stored strain energy and the ratio of stiffness in the stretching directions were calculated at prescribed strains for each gel, and statistical comparisons were made among the gels cultured under the various mechanical constraints. Results showed that gels cultured under the free-floating and equibiaxial stretching conditions exhibited a nearly isotropic mechanical behavior, while gels cultured under the strip-biaxial and uniaxial stretching conditions developed a significant degree of mechanical anisotropy. In particular, gels cultured under the equibiaxial stretching condition with a greater stretch ratio appeared to be stiffer than those with a smaller stretch ratio. Also, a decellularized gel was stiffer than its non-decellularized counterpart. Finally, the retained mechanical anisotropy in gels cultured under the strip-biaxial stretching and uniaxial stretching conditions after cell removal reflected an irreversible matrix remodeling.     

Applying controlled non-uniform deformation for in vitro studies of cell mechanobiology

Cells within connective tissues routinely experience a wide range of non-uniform mechanical loads that regulate many cell behaviors. In this study, we developed an experimental system to produce complex strain patterns for the study of strain magnitude, anisotropy, and gradient effects on cells in culture. A standard equibiaxial cell stretching system was modified by affixing glass coverslips (5, 10, or 15 mm diameter) to the center of 35 mm diameter flexible-bottomed culture wells. Ring inserts were utilized to limit applied strain to different levels in each individual well at a given vacuum pressure thus enabling parallel experiments at different strain levels. Deformation fields were measured using high-density mapping for up to 6% applied strain. The addition of the rigid inclusion creates strong circumferential and radial strain gradients, with a continuous range of stretch anisotropy ranging from strip biaxial to equibiaxial strain and radial strains up to 24% near the inclusion. Dermal fibroblasts seeded within our 2D system (5 mm inclusions; 2% applied strain for 2 days at 0.2 Hz) demonstrated the characteristic orientation perpendicular to the direction of principal strain. Dermal fibroblasts seeded within fibrin gels (5 mm inclusions; 6% applied strain for 8 days at 0.2 Hz) oriented themselves similarly and compacted their surrounding matrix to an increasing extent with local strain magnitude. This study verifies how inhomogeneous strain fields can be produced in a tunable and simply constructed system and demonstrates the potential utility for studying gradients with a continuous spectrum of strain magnitudes and anisotropies.     

Measurement of intracellular strain on deformable substrates with texture correlation

Mechanical stimuli are important factors that regulate cell proliferation, survival, metabolism and motility in a variety of cell types. The relationship between mechanical deformation of the extracellular matrix and intracellular deformation of cellular sub-regions and organelles has not been fully elucidated, but may provide new insight into the mechanisms involved in transducing mechanical stimuli to biological responses. In this study, a novel fluorescence microscopy and image analysis method was applied to examine the hypothesis that mechanical strains are fully transferred from a planar, deformable substrate to cytoplasmic and intranuclear regions within attached cells. Intracellular strains were measured in cells derived from the anulus fibrosus of the intervertebral disc when attached to an elastic silicone membrane that was subjected to tensile stretch. Measurements indicated cytoplasmic strains were similar to those of the underlying substrate, with a strain transfer ratio (STR) of 0.79. In contrast, nuclear strains were much smaller than those of the substrate, with an STR of 0.17. These findings are consistent with previous studies indicating nuclear stiffness is significantly greater than cytoplasmic stiffness, as measured using other methods. This study provides a novel method for the study of cellular mechanics, including a new technique for measuring intranuclear deformations, with evidence of differential magnitudes and patterns of strain transferred from the substrate to cell cytoplasm and nucleus.     

Combined effects of microtopography and cyclic strain on vascular smooth muscle cell orientation

Cellular alignment studies have shown that cell orientation has a large effect on the expression and behavior of cells. Cyclic strain and substrate microtopography have each been shown to regulate cellular alignment. This study examined the combined effects of these two stimuli on the alignment of bovine vascular smooth muscle cells (VSMCs). Cells were cultured on substrates with microgrooves of varying widths oriented either parallel or perpendicular to the direction of an applied cyclic tensile strain. We found that microgrooves oriented parallel to the direction of the applied strain limited the orientation response of VSMCs to the mechanical stimulus, while grooves perpendicular to the applied strain enhanced cellular alignment. Further, the extent to which parallel grooves limited cell alignment was found to be dependent on the groove width. It was found that for both a small (15microm) and a large (70microm) groove width, cells were better able to reorient in response to the applied strain than for an intermediate groove width (40microm). This study indicates that microtopographical cues modulate the orientation response of VSMCs to cyclic strain. The results suggest that there is a range of microgroove dimensions that is most effective at maintaining the orientation of the cells in the presence of an opposing stimulus induced by cyclic strain.     

Solutions for determining equibiaxial substrate strain for dynamic cell culture

In this work, empirical and analytical solutions of equibiaxial strain on a flexible substrate are derived for a dynamic cell culture system. The empirical formula, which fulfills the mechanistic conditions of the culture system, is based on a regression analysis from finite element analyses for a substrate undergoing large strains (<15%). The analytical (closed-form) solution is derived from the superposition of two elastic responses induced in the equibiaxial strain culture system after applying pressure to a substrate undergoing small strains (microstrains). There is good agreement between the strain predicted from the solutions and from the direct measurement. Using material and geometric properties of the culture system, the solutions developed here are straightforward and can be used to circumvent experimental measurements or finite element analysis to establish substrate pressure–strain relationships.     

An improved texture correlation algorithm to measure substrate-cytoskeletal network strain transfer under large compressive strain

Force-induced deformation of tissues is transduced to the cytoskeletal (CSK) network within cells via focal adhesions. Previous studies have characterized transfer of strains of less than 15% from the substrate to the cell, using mitochondria as surrogate markers for CSK deformation. However, it is unclear if intracellular strains determined from mitochondrial displacement accurately reflect CSK network deformation. Furthermore, previous studies have not characterized substrate-CSK network strain transfer for strain magnitudes exceeding 15%, as can occur in vivo and in cell culture studies. Here, we developed and characterized a texture correlation algorithm to address the image distortion problem caused by large strain. We then used this algorithm to characterize large compressive strain (-40%) transfer from the substrate to the CSK in living cells, using fluorescently tagged actin to perform the tracking and both fluorescently tagged actin and talin to make validation measurements. With this approach, we were able to demonstrate explicitly that substrate strain transfers directly to CSK deformation in living cells undergoing large compressive deformation, and that the strain transfer ratios are independent of cell alignment. The tools and approaches developed here enable improved characterization of cell-matrix interactions under large deformation and in doing so, may reveal new insights into mechanotransduction mechanisms in such circumstances.     

Novel mechanical bioreactor for concomitant fluid shear stress and substrate strain

The two main types of mechanical stimuli used in cellular-level bone mechanotransduction studies are substrate strain and flow-induced shear stress. A subset of studies has investigated which of these stimuli induces the primary mechanotransduction effect on bone cells. The shortcomings of these experiments are twofold. First, in some experiments the magnitude of one loading type is able to be quantitatively measured while the other loading mode is only estimated. Second, the two loading modes are compared using different bioreactors, representing different cellular environments and substrates to which the cells are attached. In addition, none of these studies utilized bioreactors which apply controlled magnitudes of substrate strain and flow-induced shear stress differentially and simultaneously. This study presents the design of a multimodal loading device which can apply substrate stretch and fluid flow simultaneously while allowing for real-time cell imaging. The mechanical performance of the bioreactor is validated in this study by correlating the output levels of flow-induced shear stress and substrate strain with the input levels of displacement and displacement rate. The magnitudes of cross-talk loading (i.e. flow-induced strain, and strain-induced fluid flow) are also characterized and shown to be magnitudes lower than physiological levels of loading estimated to occur in bone in vivo.     

Connexin 43 is localized with actin in tenocytes

Varieties of cell-matrix or cell-cell adhesions are associated with the actin cytoskeleton. However, for gap junctions, which are both channels and adhesions, there has been little evidence for such an association. The purpose of this study was to determine if connexin 43 (Cx43) associates with actin and to determine if this association is altered under mechanical load in tenocytes, a mechanically sensitive cell. Avian tenocytes were subjected to multiple cyclic strain regimens and then fixed and examined immunohistochemically at various times poststrain to determine where Cx43 protein was localized within the cells in relation to actin filaments. Four percent of tenocytes had colocalization of actin filaments and Cx43, which was significantly increased with 5% cyclic strain. To confirm this phenomenon, human tenocytes and COS-7 cells were also subjected to cyclic strain and then fixed at the same times after strain. As with avian tenocytes, Cx43 was colocalized with actin in human tenocytes and COS-7 cells. The colocalization increased significantly after cyclic strain in human tenocytes but not in COS-7 cells. The lack of detectable changes in COS-7 cells may indicate that they are less mechanosensitive than tenocytes perhaps due to the less robust actin cytoskeleton seen in the COS-7 cells when compared to the tenocytes. Furthermore, inhibiting myosin II activity greatly reduced the immunohistochemically-detectable Cx43 on actin filaments. Connexins may associate with actin to stabilize gap junctions at the plasma membrane, ensuring that tenocytes remain coupled during periods of prolonged or intense mechanical loading.     

Strain and strain rate activation of G proteins in human endothelial cells

The endothelium is known to sense and respond to its physical environment, but the underlying mechanisms and early events of endothelial cell mechanotransduction are not well understood. The present study measured G protein activation by mechanical strain in human umbilical vein endothelial cells (HUVEC) directly by photoincorporation of a hydrolysis resistant, radiolabeled GTP analog. Ten percent uniaxial strain at a strain rate of 20% s(-1) over 1min activated a 38kDa Galpha subunit 167+/-17% relative to controls, while 2% cyclic strain failed to significantly activate the protein (117+/-19%). A single cycle of 10% strain at 20% s(-1) strain rate activated the Galpha subunit 152+/-25%, while activation at the same strain but lower strain rate (0.3% s(-1)) was not significantly different from controls (116+/-12%). Western blot analysis identified the 38kDa protein as Galpha(q/11). These results demonstrate the rapid activation of G proteins in HUVEC by cyclic uniaxial strain in a strain- and strain rate-dependent manner.     

IL-1beta decreases the elastic modulus of human tenocytes.

Cellular responses to mechanical stimuli are regulated by interactions with the extracellular matrix, which, in turn, are strongly influenced by the degree of cell stiffness (Young's modulus). It was hypothesized that a more elastic cell could better withstand the rigors of remodeling and mechanical loading. It was further hypothesized that interleukin-1beta (IL-1beta) would modulate intracellular cytoskeleton polymerization and regulate cell stiffness. The purpose of this study was to investigate the utility of IL-1beta to alter the Young's modulus of human tenocytes. Young's modulus is the ratio of the stress to the strain, E = stress/strain = (F/A)/(deltaL/L0), where L0 is the equilibrium length, deltaL is the length change under the applied stress, F is the force applied, and A is the area over which the force is applied. Human tenocytes were incubated with 100 pM recombinant human IL-1beta for 5 days. The Young's modulus was reduced by 27-63%. Actin filaments were disrupted in >75% of IL-1beta-treated cells, resulting in a stellate shape. In contrast, immunostaining of alpha-tubulin showed increased intensity in IL-1beta-treated tenocytes. Human tenocytes in IL-1beta-treated bioartificial tendons were more tolerant to mechanical loading than were untreated counterparts. These results indicate that IL-1beta reduced the Young's modulus of human tenocytes by disrupting the cytoskeleton and/or downregulating the expression of actin and upregulating the expression of tubulins. The reduction in cell modulus may help cells to survive excessive mechanical loading that may occur in damaged or healing tendons.     

Multiscale strain analysis of tissue equivalents using a custom-designed biaxial testing device

Mechanical signals transferred between a cell and its extracellular matrix play an important role in regulating fundamental cell behavior. To further define the complex mechanical interactions between cells and matrix from a multiscale perspective, a biaxial testing device was designed and built. Finite element analysis was used to optimize the cruciform specimen geometry so that stresses within the central region were concentrated and homogenous while minimizing shear and grip effects. This system was used to apply an equibiaxial loading and unloading regimen to fibroblast-seeded tissue equivalents. Digital image correlation and spot tracking were used to calculate three-dimensional strains and associated strain transfer ratios at macro (construct), meso, matrix (collagen fibril), cell (mitochondria), and nuclear levels. At meso and matrix levels, strains in the 1- and 2-direction were statistically similar throughout the loading-unloading cycle. Interestingly, a significant amplification of cellular and nuclear strains was observed in the direction perpendicular to the cell axis. Findings indicate that strain transfer is dependent upon local anisotropies generated by the cell-matrix force balance. Such multiscale approaches to tissue mechanics will assist in advancement of modern biomechanical theories as well as development and optimization of preconditioning regimens for functional engineered tissue constructs.     

Proliferation of human-derived osteoblast-like cells depends on the cycle number and frequency of uniaxial strain

We tested the hypothesis whether the number of applied load cycles and the frequency of uniaxial strain have an effect on proliferation of human bone derived osteoblast-like cells. A new approach was developed in order to differentiate between the effects of frequency and the effects of cycle number and strain duration. Monolayers of subconfluently grown cells were stretched in rectangular silicone dishes with cyclic predominantly uniaxial movement along there longitudinal axes. Strain was applied over 2 days varying the number of applied load cycles (4-3600) at a constant frequency (1Hz) or varying the frequency (0.1-30Hz) at a constant number of applied cycles (1800) or at a constant strain duration (5min). At a constant frequency, proliferative response increases (103%) with the number of applied cycles until a cycle number maximum (1800 cycles) was reached. 3600 cycles reduced cell number (43%) in contrast to the maximum. The variation of the frequency of applied strain tended to result in slight differences with regard to cell proliferation when cycle number was left constant. However, combined with an appropriate number of cycles there was an optimal frequency (1Hz) as stimulus for bone cell proliferation (84%). A higher frequency (30Hz) in combination with a high cycle number (9000) reduced cell number to control level (4%). This study demonstrates a frequency and cycle number dependent proliferative response of human osteoblast-like cells. It could be shown that effects of the frequency should not be considered separately from the effects of the cycle number.     

The functional response of U937 macrophage-like cells is modulated by extracellular matrix proteins and mechanical strain.

Extracellular matrix proteins (ECMs) play a significant role in the transfer of mechanical strain to monocyte-derived macrophages (MDMs) affecting morphological changes in a foreign body reaction. This study investigated how the functional responses of U937 macrophage-like cells differed when subjected to 2 dynamic strain types (nonuniform biaxial or uniform uniaxial strain) while cultured on siloxane membranes coated with either collagen type I or RGD peptide repeats (ProNectin). Biaxial strain caused an increase in intracellular esterase and acid phosphatase (AP) activities, as well as monocyte-specific esterase (MSE) protein levels in cells that were seeded on either uncoated surfaces (shown previously) or collagen, but not ProNectin. Released AP activity, but not released esterase activity, was increased on all surfaces. Biaxial strain increased IL-6, but not IL-8 on all surfaces. When cells were subjected to uniaxial strain, intracellular esterase increased on coated surfaces only, whereas intracellular AP activity was unaffected. Both esterase and AP released activities increased on all surfaces. Uniaxial strain increased the release of IL-6 on all surfaces, but IL-8 on coated surfaces only. This study demonstrated for the first time that ECM proteins could specifically modulate cellular responses to different types of strain. Using this approach with an in vitro cell system may help to unravel the complex function of MDMs in the foreign-body reaction.     

Resident multipotent vascular stem cells exhibit amplitude dependent strain avoidance similar to that of vascular smooth muscle cells

Atherosclerosis is one of the leading causes of mortality worldwide, and presents as a narrowing or occlusion of the arterial lumen. Interventions to re-open the arterial lumen can result in re-occlusion through intimal hyperplasia. Historically only de-differentiated vascular smooth muscle cells were thought to contribute to intimal hyperplasia. However recent significant evidence suggests that resident medial multipotent vascular stem cells (MVSC) may also play a role. We therefore investigated the strain response of MVSC since these resident cells are also subjected to strain within their native environment. Accordingly, we applied uniaxial 1 Hz cyclic uniaxial tensile strain at three amplitudes around a mean strain of 5%, (4–6%, 2–8% and 0–10%) to either rat MVSC or rat VSMC before their strain response was evaluated. While both cell types strain avoid, the strain avoidant response was greater for MVSC after 24 h, while VSMC strain avoid to a greater degree after 72 h. Additionally, both cell types increase strain avoidance as strain amplitude is increased. Moreover, MVSC and VSMC both demonstrate a strain-induced decrease in cell number, an effect more pronounced for MVSC. These experiments demonstrate for the first time the mechano-sensitivity of MVSC that may influence intimal thickening, and emphasizes the importance of strain amplitude in controlling the response of vascular cells in tissue engineering applications.     

Dependence of alignment direction on magnitude of strain in esophageal smooth muscle cells

The response of cells in vitro to mechanical forces has been the subject of much research using devices to exert controlled mechanical stimulation on cultured cells or isolated tissue. In this study, esophageal smooth muscle cells were seeded on flexible polyurethane membranes to form a confluent cell layer. The cells were then subjected to uniform cyclic stretch of varying magnitudes at a frequency of approximately five cycles per minute in a custom made mechatronic bioreactor, providing similar strains experienced in the in vivo mechanical environment of the esophagus. The results show that the orientation response is dependent on the magnitude of cyclic stretch applied. Smooth muscle cells showed parallel alignment to the force direction at low cyclic strains (2%) compared to the hill‐valley morphology of static controls. At higher strains (5% and 10% magnitude), the cells exhibited a consistent alignment perpendicular to the strain. To our knowledge, this is the first time that the alignment direction's dependence on strain magnitude has been demonstrated. MTS analysis indicated that cell metabolism was reduced when mechanical strain was applied, and proliferation was inhibited by mechanical strain. Protein expression indicates a decrease in smooth muscle α‐actin, indicative of changes in cell phenotype, an increase in vimentin, which is associated with increased cell motility, and an increase in desmin, indicating differentiation in stimulated cells. Biotechnol. Bioeng. 2009;102: 1703–1711. © 2008 Wiley Periodicals, Inc.     

Quantitative evaluation of threshold fiber strain that induces reorganization of cytoskeletal actin fiber structure in osteoblastic cells

The cytoskeletal stress fiber structure plays essential roles in various kinds of cellular functions such as shape maintenance, active motility and mechanosensing, and its structure is dynamically reorganized under each functional process. In known reorganization mechanisms of the stress fibers, a change in its mechanical condition has been suggested as one of the key mediators that affect the reorganization process. Some experimental studies have clarified that tension release in the stress fibers induces fiber depolymerization that is considered to be the initial phase of the reorganization process. However, quantitative mechanical values such as strain or stress that induce depolymerization have still not been evaluated. This study is aimed at the quantitative evaluation of the mechanical value that induces stress fiber depolymerization, to gain a basic understanding of the reorganization phenomenon from a mechanical viewpoint. Osteoblastic cells (MC3T3-E1) were cultured on prestretched silicone rubber substrate. Compressive deformation was applied to the cells by uniaxially releasing the prestretched substrate strain and change in the stress fiber structure was observed. The results indicated that the compressive strain magnitude, not in the whole cell body but in the stress fiber itself, is important to induce disassembly of the stress fiber structure. The existence of a threshold strain magnitude for initiating fiber disassembly was also suggested; the threshold strain magnitude was evaluated as approximately -0.20.     

Control of microtubule assembly by extracellular matrix and externally applied strain

A number of studies have suggested that externally applied mechanical forces and alterations in the intrinsic cell-extracellular matrix (ECM) force balance equivalently induce changes in cell phenotype. However, this possibility has never been directly tested. To test this hypothesis, we directly investigated the response of the microtubule (MT) cytoskeleton in smooth muscle cells to both mechanical signals and alterations in the ECM. A tensile force that resulted in a positive 10% step change in substrate strain increased MT mass by 34 +/- 10% over static controls, independent of the cell adhesion ligand and tyrosine phosphorylation. Conversely, a compressive force that resulted in a negative 10% step change in substrate strain decreased MT mass by 40 +/- 6% over static controls. In parallel, increasing the density of the ECM ligand fibronectin from 50 to 1,000 ng/cm(2) in the absence of any applied force increased the amount of polymeric tubulin in the cell from 59 +/- 11% to 81 +/- 13% of the total cellular tubulin. These data are consistent with a model in which MT assembly is, in part, controlled by forces imposed on these structures, and they suggest a novel control point for MT assembly by altering the intrinsic cell-ECM force balance and applying external mechanical forces.     

Design and validation of a corneal bioreactor

Mechanical strain is an important signal that influences the behavior and properties of cells in a wide variety of tissues. Physiologically similar mechanical strain can revert cultured cells to a more normal phenotype. Here, we have demonstrated that 3% equibiaxial (EB) and uniaxial strains confer favorable protein expression in cultured rabbit corneal fibroblasts (RCFs), with approximately 35% and 65% reduction in expression of α‐smooth muscle actin (α‐SMA), respectively. We have designed a novel bioreactor that is capable of imparting up to 7% EB strain and up to 6% EB strain using a cornea‐shaped post. Additional features of the bioreactor include the application of shear stress to cells in culture and the ability to image cells using optical coherence microscopy (OCM) without being removed from the system. Biotechnol. Bioeng. 2012; 109: 3189–3198. © 2012 Wiley Periodicals, Inc.