If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
It is often desirable to evaluate the ability of cells to move in an unrestricted manner in multiple directions without chemical gradients. By combining the standard radial migration assay with injection-molded gaskets and a rigid fixture, we have developed a highly reliable and sensitive method for observing and measuring radial cell migration. This method is adapted for use on high-throughput automated imaging systems. The use of injection-molded gaskets enables low-cost replacement of cell-wetted components. Moreover, the design enables secondary placement of attractants and co-cultures. This device and its enhanced throughput permit the use of therapeutic screening to evaluate phenotypic responses, for example, cancer cell migration response due to drugs or chemical signals. This approach is orthogonal to other 2D cell migration applications, such as scratch wound assays, although here we offer a noninvasive, enhanced-throughput device, which currently is not commercially available but is easily constructed. The proposed device is a systematic, reliable, rapid application to monitor phenotypic responses to chemotherapeutic screens, genetic alterations (e.g., RNAi and CRISPR), supplemental regimens, and other approaches, offering a reliable methodology to survey unbiased and noninvasive cell migration.
Many biological questions are concerned with the rate(s) of cellular migration, for example, tissue development and healing, cancer metastasis, immune cell trafficking, and regeneration, among others. Understanding the precise mechanisms by which secreted factors contribute to cellular microenvironments and regulate downstream migratory responses is a critical aspect of cellular biology.
However, adoption of this approach has been relatively slow because placement of the gasket for high-throughput studies is tedious. Therefore, we set out to adapt the radial migration assay for enhanced-throughput applications by reducing the complexity and time necessary to set up and perform the assay and interpret the data.
One of the most popular 2D assays of cell motion is the scratch wound or wound closure assay.
One of the disadvantages to this technique, however, is that creating a linear scratch (wound) in a cell monolayer potentially initiates the uncontrolled release of inflammatory cytokines, chemokines, and wound repair signals, which undoubtedly could influence cellular responses and trigger cellular communication programs.
Indeed, these responses are critical when studying wound repair mechanisms modeling physically damaged tissue. Conversely, in cancer research, examining intrinsic cell migratory capacity free of influence from wound healing or damage responses (i.e., damage-associated molecular patterns, alarmins, etc.; see Bianchi
) may be the desired outcome. While we did not experimentally determine this claim in our models, several reviews of in vitro motility and migration modeling have described this as a potential negative outcome.
In fact, recent adaptations to the wound closure assay have been implemented by the use of polydimethylsiloxane inserts (e.g., Ibidi culture inserts) that are placed in the cell culture well and removed once cells reach confluency, and wound or gap closure rates can be overserved with minimal disruption to the cell monolayer. Despite this advancement, the procedure remains laborious, time-consuming, costly, and offers minimal throughput for larger studies. Like the wound closure assay, the radial migration assay permits the evaluation of 2D multidirectional, undirected cell migration and enables communication between bulk cells and leading-edge migratory cells. Our radial migration device offers a higher-throughput approach and much lower consumable costs, all with a simple user interface to gather important 2D cell motility data.
Independent of migration assays, robust imaging techniques have been developed for tracking the migration of cells from precise locations.
These techniques acquire an initial image that represents the size distribution of the starting population and tracks the shape and growth in size of the cells as they migrate radially from a central point, thus permitting the measurement of either bulk cellular migration or single-cell dynamics. These data help probe the relationship between phenotypic behavior and cellular pathways. Coupling high-throughput migration assays with automated imaging provides an attractive avenue to evaluate intrinsic cell mobility and, further, evaluate the potential use of targeted therapies to slow or inhibit cell migration reproducibly. A highly reproducible assay requires fewer replicates, as fluctuations are low, saving time and resources. This is particularly useful in the design of high-value chemotherapeutics that aim to inhibit cancer metastasis or understand the relationship between genetic mediators of cancer cell migration or invasion, with pertinent examples described in the literature.
Here, we describe our newly designed radial migration device that enables the rapid and precise application of the radial migration assay for basic and translational biology, without inducing a damage or inflammatory response. Our enhanced-throughput radial migration tool (Fig. 1A) is a simple fixture based on sound engineering design principles. Herein we describe the method and device design and then provide modeling of the forces and characterization of the approach. We demonstrate the use of the integrated device to study the migration of cancer cells as a proof of principle. Collectively, this new tool permits enhanced-throughput adaptation of a previously tedious and time-consuming assay. When paired with high-content imaging (e.g., BioTek BioSpa, BioTek, Winooski, VT, and Cytation 5 incubator/imaging platform or IncuCyte Live-Cell Analysis platforms), this new assay provides valuable information regarding unabated cell migration and offers a novel method to evaluate migratory responses.
Materials and Methods
We designed the device in SolidWorks (Dassault Systemes, Vélizy-Villacoublay, France) 3D design software. The files provided in the Supplementary Materials (S1–S6) were exported and uploaded to the digital fabricator Protolabs (Maple Plain, MN), who then milled the parts out of 6061 aluminum with a tolerance of ±0.005 inches. Three parts were fabricated by Protolabs’ machining division, including the gasket fixture plate, which holds the gaskets in place (S1), the base for aligning the top plate with a well plate (S2), and a lock to hold the well plate in place (S3). The parts were then tapped, the edges broken, and then bead blasted for finishing. The final part, a lid designed to prevent contamination of the sample, was fabricated by Protolabs’ 3D printing division out of WaterShed XC 11122 material (S4). The aluminum parts were anodized by Alpha Metal Finishing (Dexter, MI) in a hard-coat gray anodization. The remainder of the device components were purchased from McMaster-Carr (Elmhurst, IL). The components are provided in tabular form in the Supplementary Materials for those who wish to use the device (S6). The purchased components include two clamps (McMaster, 6385K110), two springs (McMaster, 9434K820), two 1/8-inch-diameter alignment pins (McMaster, 97395A454), one set screw (McMaster, 91658A154), two 4-inch-long ¼′′-20 threaded rods (McMaster, 90322A652), one ½-inch-long ¼′′-20 screw for locking the well plate in place (McMaster, 91251A527), 005 size silicone gaskets (McMaster, 1173N005), and square cross section 005 size silicone gaskets (McMaster, 1182N005).
Breast cancer cell lines MDA-MB-231 and HCC1937 were acquired from the American Type Culture Collection (Manassas, VA) and maintained in Gibco RPMI-1640 supplemented with 10% fetal bovine serum (FBS; cat. 35-001-CV, Corning, Corning, NY), 5 μg/mL gentamycin (cat. 30-234-CI, Corning), 2 mM l-glutamine (cat. 25-005-CI, Corning), and 1× antibiotic-antimycotic (cat. 30-004-CI, Corning). Cells were cultured at 37 °C and 5% CO2. MDA-IBC-3 cells were acquired from Dr. Wendy Woodward (MD Anderson) and maintained in Gibco DMEM-F12, 10% FBS, 5 μg/mL gentamycin, 1 μg/mL insulin (cat. I9278, Sigma-Aldrich, St. Louis, MO), 1 μg/mL hydrocortisone (cat. 354203, Corning), and 1× antibiotic-antimycotic (cat. 15240096, Gibco, Carlsbad, CA). Cells were cultured at 37 °C and 10% CO2. Cells were subcultured at 1:4 ratios and only utilized in experimental assays for three to five passages following revival from frozen early passage stocks.
Radial Device Preparation and High-Content Microscopy
Each cell model was seeded at 5.0 × 104 per well of the radial migration device in a 24-well tissue culture-treated dish (cat. 1156F00, Denville Scientific, Metuche, NJ). Additional 24-well plates examined for this study included Costar 24-well tissue culture-treated plates (cat. 3524, Corning) and Falcon 24-well tissue culture-treated plates (cat. 353047, Corning). Seeded cell lines were maintained at 5% CO2 or 10% CO2 (cell line dependent) and 37 °C overnight (~16 h) to permit adherence to a 24-well culture dish. The radial device was removed and cells were washed 3× with sterile phosphate-buffered saline (PBS; Gibco) to remove nonadherent cells and/or debris and maintained in 500 μL of complete media. Brightfield images at low magnification (i.e., 1.25× objective) were taken on a BioTek BioSpa and Cytation 5 automated incubator high-throughput imaging platform (BioTek) every 6 h for 2 days. The BioSpa incubator system and Cytation 5 imager were maintained at 37 °C and 5% CO2. Data were analyzed on BioTek Gen5 v3.08, and statistical analyses were performed in GraphPad Prism v8.3.0 (GraphPad, La Jolla, CA).
Device Principle and Experimental Approach
The approach we present here (Fig. 1A,B) is simply designed to alleviate barriers in the use and interpretation of radial migration assays. The method uses an aluminum fixture made of two parts—a base and a gasket alignment plate. The gasket alignment plate, which harbors an extruded shaft, aligns using steel pins and applies pressure to the gasket within each well of a 24-well plate. Commercially available gaskets are suitable for this application and are inserted into recesses in the bottom of the locating shafts. Alternatively, custom gaskets can be created for more advanced studies. Clamps are located on both ends of the device and apply pressure between the upper and lower plates. Alignment detents are used for three-point kinematic locating of the well plate on the base plate.
To use the device, we placed a 24-well cell culture plate against the detents (Fig. 1C). Then we clamped the gasket alignment plate against the well plate. For sterility, we autoclaved the entire system, including the gaskets, and then cleaned with ethanol or plasma. The central opening of each well was filled with 100 μL of cell culture media and placed in a vacuum bell jar with applied vacuum (10–100 mTorr) for ~2–3 min to remove air bubbles at the bottom of each well, after assembly. The device was loaded with cells into the existing media in each well. While cell concentration should be empirically optimized, we find that ~50,000 cells per well is a sufficient amount to form a confluent monolayer of cells in each well (see Fig. 3F). We covered the wells with a lid for sterility and humidity control after loading cells, and the device was placed in the incubator to allow cells to become adherent. We released the clamps, and the gasket alignment plate was removed after the cells had adhered. Springs will lift the gaskets and gasket alignment plate off the well plate in a uniform way to minimize peeling and other damage to the radial cell monolayer. We removed nonadherent cells and debris with three to five successive PBS washes. At this time, readers could supplement cells with media components or drug treatments, or the cell monolayer may be deposited with Matrigel to observe 3D matrix invasion. Following cell treatment, the user would then initiate imaging to observe migratory and invasive responses as a function of time.
Modeling Gasket Sealing Forces
To ensure proper and uniform sealing of the gaskets against each well, we combined a finite element analysis (FEA) model with accepted practices used in designing face gaskets. First, considering the model for a single gasket, this was of the axial static face type with a target of 20% compression. For a round gasket with a cross-sectional diameter of 1.78 mm and a circumference of 13.6 mm, the force per linear millimeter applied to the seal can be tabulated from a standard table. For a Shore hardness of 50A, the expected range is between 0.79 and 2.45 N/mm. We chose the value of 2.45 N/mm for a Shore hardness of 50A because the compressive modulus of silicone is higher than that of other elastomeric materials. This results in a force of 33.4 N per gasket. Because we are sealing 24 wells, the total force for the plate is approximately 802.8 N (7.5 MPa). Therefore, the applied load from each of the two clamps must be 401.4 N, and the selected ¼′′-20 bolts are rated for a clamping force of 5870 N. The fixtures were computer numerical control machined and lightly bead blasted, resulting in a 0.8 µm finish. These estimated forces should seal against the surface roughness of 0.8 µm specified on the gasket alignment plate. Finally, the gland inset that houses the gaskets uses a 5% interference fit between the gland and the gasket. Given the thickness of the gasket at 1.78 mm and a compression of 0.38 mm, the displacement represents an additional 3%. This would require a nominal amount of additional pressure and lead toward a balance of deflection and compression.
To understand how the force transferred from the gasket alignment plate varies between the gaskets and the well plate, we used an FEA model to simulate a range of loads and fixture thicknesses. The FEA model was performed using SolidWorks’ built-in FEA system. The applied analysis was determined for a static force applied by the two clamps. The solver was a direct sparse matrix solver, h-adaptive. A nonpenetrating contact constraint was applied between the gaskets, and the top and bottom fixtures, as well as between both fixtures. We specified aluminum 6061 as the material for the fixtures and a 50A Shore durometer hardness variant of silicone rubber for the gaskets to match the gaskets purchased for experimentation.
We applied different loads to the model to observe the impact on the contact pressure between the gaskets, fixtures, and their deformation. In addition, the model was used to understand how the effective thickness of the fixture can be optimized to reduce the bending curvature in the device due to the end loads, which would prevent sealing in the center of the device. The loads tested were 100–1000 N for 4, 5, and 6 mm supports with ribbing (Fig. 2). Our data show the contact pressure across the gasket array, with the smallest contact pressure occurring in the center of the device (left panel, contact pressure map of the gaskets) and the highest on the edges, with a minimum pressure of 0.06 MPa (4 mm fixture, Fig. 2A, middle panel) and a maximum pressure of 18.6 MPa (6 mm fixture, Fig. 2A, right panel). The pressures near both the edges and the center are higher, with an increase in effective fixture thickness adjusted using a trapezoidal rib across the top of the fixture. Additionally, Figure 2B displays these trends, but with respect to the resultant deformation of the gaskets and the fixture plate. The left panel shows exaggerated bending due to the clamps on each end, and the middle panel shows why the rib thickness must be optimized, as a 4 mm thickness separates the gasket from the well (16 µm at 1000 N) in the center, while a 5 or 6 mm thick rib enables clamping in the center (–8 µm and –32 µm, respectively). The edge deformations are not significantly different near the loads (all approach –188 µm). Finally, the strain on the gaskets is mapped in Figure 2C (left panel), and the center and edge strains for each load and thickness are plotted in the middle and right panels, respectively.
Device Prototype Analysis
Based on the results from FEA modeling (Fig. 2), we fabricated a prototype of the device. The final platform is shown in Figure 3A. Deformation of gaskets clamped against a 24-well plate is shown in Figure 3B. To validate the sealing and performance of the device, we seeded MDA-MB-231 breast cancer cells into the device as described in the “Device Implementation and Experimental Approach” section of this paper. A radial cell monolayer postseeding can be observed in Figure 3C. Upon deposition of cell media into each well of the device, we observed several instances of bubble formation near the gasket–cell culture plate interface resulting in improper radial cell monolayer formation (Fig. 3D). To avoid the formation of bubbles at this interface, we find that application of vacuum (whole device placed inside desiccator bell jar; see “Device Implementation and Experimental Approach” section for further detail) greatly reduces bubble formation, significantly enhancing the reproducibility of the assay.
Next, we sought to determine the reproducibility of the size of the radial cell monolayer across the device. First, gaskets with round cross sections were compared against gaskets with square cross sections. The resulting diameters of the cell spot were measured by brightfield microscopy (Fig. 3E). Variability is not significantly different between the two cross section shapes, and thus we chose the more readily available round cross section gaskets for all remaining experiments. The average diameter of radial cell monolayers was 2.66 mm, with a standard deviation of 0.11 mm (n = 24) (Fig. 3E). The maximum diameter was 2.81 mm, and the minimum diameter was 2.5 mm. Given that the specified gasket diameter is 2.56 mm, this is consistent as the inner diameter of the rounded gasket will be slightly larger than the nominal gasket diameter (Fig. 3E, blue). The gasket alignment plate is designed with a hard stop to ensure that 20% of the gasket height is compressed. The coefficient of variation percentage for well-to-well variability was 4.3%.
Proof-of-Principle Experimental Approach
To display the utility of our device and assay, we aimed to determine the migratory capacity of HCC-1937 human breast cancer cells in the presence or absence of growth serum. Using BioTek’s Gen5 analysis software, we utilized built-in algorithms that provided robust identification and labeling of radial cell monolayers (Fig. 4A). We then quantified radial cell areas postexpansion, measured every 6 h. As expected, we observe increases in radial migration when HCC-1937 cells are provided growth serum when compared with serum free conditions (Fig. 4B). Kinetic analysis in the Gen5 platform permits measurements at a rate of micrometers per minute as a function of radial expansion area measurements over time. Using this methodology, we can quantify radial expansion velocities (μm/min) for each cell monolayer, averaged over the entire experimental analysis. Radial expansion velocities are a function of the area measurements and provide a more concise display of the data (Fig. 4C). Collectively, these data show the capacity for this assay to demonstrate significant changes in migratory rates, depending on the cellular environment. Indeed, these results could be confounded by the influence of proliferation, and the user should be made aware of this particular caveat. Regardless, this assay may be applied to observe phenotypic changes in response to drug regimens or to monitor the role a genetic component may play in regulating migration, among various other scenarios.
We have developed and described a new device that improves an important but currently underutilized migration assay: the radial migration assay. Prior challenges to this method include the limited tools available, resulting in laborious manual setup and highly variable results. This requires an extensive number of replicates to produce significant results. In contrast, we offer a higher-throughput version of the radial migration assay. Due to its scalability, precision design, and compatibility with automation, it produces less variable data, at a higher rate, on a user-friendly platform.
The primary limitation of the system we present here is the nonuniformity of the clamping pressure. Future design iterations of the device will address this by modeling alterations to the clamping setup or perhaps modality. Consistently obtaining proper gasket contact pressure is currently challenging, although implementation of a thread stop may aid in applying the correct clamping force on a more consistent basis. Additionally, while application of vacuum effectively removes bubbles formed by the deposition of cell culture media, we do occasionally encounter bubbles forming at the gasket–culture plate interface. This technical error must be controlled by the user by utilizing several technical replicates per experiment to allow for radial formation error (i.e., bubble formation or poor cell adherence). Lastly, significant care must be taken to utilize clean instruments (e.g., forceps) when placing or handling the gaskets, such that dirt or oils are not deposited on the gasket face. Regardless of these prototyping issues, we show that the approach is robust, as evidenced by our results, and technical error can be minimized by careful practices of the user.
Migration and motility are crucial phenotypic characteristics that most cells intrinsically harbor. Cellular motility is an imperative aspect of tissue generation, organ formation, wound healing, and inflammatory responses, among many other physiological functions that are essential for life.
While many fields of biological research have significant interest in cellular migration, it is cemented as a hallmark of cancer and considered one of the key cellular functions that enable metastatic spread, which ultimately causes mortality. In essentially all tumor types, metastasis is the single most important identifier of patient mortality. Brain tumors are an exception to this rule, but cancer cell motility plays a major role in brain metastases and in the progression of brain tumors as well. Despite recent advances in the development of targeted cancer therapeutics, patients who are diagnosed at a late stage (i.e., stage III or later), with locally advanced or distant metastatic tumors, have an overall poor prognosis and low survival rates. This has highlighted the need for therapeutic regimens that are aimed at diminishing or managing metastasis via inhibition of cellular migration and invasion. In this context, our device would be useful to survey migratory and invasive phenotypes in response to metastasis-targeted therapies in real time. An important caveat of this approach, like wound closure assays, is that this assay determines bulk cellular migration with contributions from proliferating cells and important cell–cell communication programs. Moreover, in the context of tumor modeling, bulk cell communication is an imperative aspect of cellular migration. The user should acknowledge this important pitfall and mitigate it with the use of additional assays to better understand migration at the single-cell level in the absence of proliferative responses and bulk cell influence.
To circumvent or control for the contributions of proliferation, cell tracking fluorescent dyes that are not passed to daughter cells could be utilized to determine the population of proliferating cells versus migrating cells.
Our device and assay are useful in examining the role of molecular mediators of cellular migration and invasion postgenetic modification of the target, for example, RNAi, overexpression, and conditional expression. Additionally, approaches to examine the efficacy of a therapeutic targeted toward inhibiting cellular migration could be applied. Indeed, our device is currently limited to 24-well formats, which is not consistent with large, high-throughput screening studies using 96- or 384-well applications. Despite this, our device is a significant improvement in throughput capacity compared with prior approaches, which are often implemented in a one-by-one time-consuming and labor-intensive fashion. In fact, when multiple devices are used, in tandem with an automated high-content imaging system (e.g., BioTek Cytation and BioSpa system), significant amounts of phenotypic screening data can be acquired. While our design offers significant improvement in reproducibility and labor over current radial migration assays, we aim to develop this platform into 48- and 96-well applications to offer truly high-throughput screening approaches.
An important point that differentiates our method from other 2D migration assays is that the radial migration assay does not mechanically damage cell monolayers. This potentially mitigates responses from wounded or dying cells that are poorly controlled and offers information about the intrinsic, nondirectional motility of a cell. Indeed, nondirectional, inherently acquired cell motility by tumor cells is the precursor to metastatic invasion. Furthermore, this system may be deployed to understand how controlled environmental cues or signals alter cellular motility. For instance, our system is designed to be utilized in a standard 24-well culture plate and could be paired with co-culture systems to understand how other immune cells of the tumor microenvironment, for example, tumor-associated macrophages, impact cancer cell motility. Additionally, collagen or extracellular matrix deposition on cells following removal of the device would allow evaluation of cellular invasion.
In summary, we advance this device in the hope that it can benefit a variety of researchers broadly interested in cellular migration and motility. Our newly manufactured device coupled to a simple user interface permits the acquisition of migration data that were previously laborious and time-consuming to acquire. The ease of use of the device and its straightforward implementation, along with the simplicity of data acquisition and analysis, permit a variety of users with diverse backgrounds to utilize our device and assay. While high-throughput imaging permits the acquisition of more robust kinetic data, users without this capability can still simply obtain initial and endpoint images to determine migratory capacity.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for this project is supported by the Breast Cancer Research Foundation (S.D.M., A.C.L., C.R.O.); Rogel Cancer Center core grant NIH-P30-CA046592-29, P30CA046592 (S.D.M., A.C.L., C.R.O.), T32CA009676 (C.R.O.), and UL1TR002240 (T.M.W.); and the METAvivor Foundation (S.D.M.).