Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2022 Dec 12;14(50):55307–55319. doi: 10.1021/acsami.2c07971

Multiplexed Detection of Molecular Interactions with DNA Origami Engineered Cells in 3D Collagen Matrices

Melika Shahhosseini †, Peter E Beshay †, Ehsan Akbari ‡, Niksa Roki †,∥, Christopher R Lucas †,∥, Alex Avendano §, Jonathan W Song †,∥,*, Carlos E Castro †,‡,*
PMCID: PMC9785045  PMID: 36509424

Abstract

graphic file with name am2c07971_0005.jpg

The interactions of cells with signaling molecules present in their local microenvironment maintain cell proliferation, differentiation, and spatial organization and mediate progression of diseases such as metabolic disorders and cancer. Real-time monitoring of the interactions between cells and their extracellular ligands in a three-dimensional (3D) microenvironment can inform detection and understanding of cell processes and the development of effective therapeutic agents. DNA origami technology allows for the design and fabrication of biocompatible and 3D functional nanodevices via molecular self-assembly for various applications including molecular sensing. Here, we report a robust method to monitor live cell interactions with molecules in their surrounding environment in a 3D tissue model using a microfluidic device. We used a DNA origami cell sensing platform (CSP) to detect two specific nucleic acid sequences on the membrane of B cells and dendritic cells. We further demonstrated real-time detection of biomolecules with the DNA sensing platform on the surface of dendritic cells in a 3D microfluidic tissue model. Our results establish the integration of live cells with membranes engineered with DNA nanodevices into microfluidic chips as a highly capable biosensor approach to investigate subcellular interactions in physiologically relevant 3D environments under controlled biomolecular transport.

Keywords: DNA origami, DNA nanotechnology, subcellular interactions, cell membrane engineering, cell microenvironment

1. Introduction

Living cells interact with a myriad of extracellular signaling molecules, such as nucleic acids, growth factors, and cytokines. For instance, antigen-presenting dendritic cells (DC) are activated by cytokines in the local environment, leading to their secretion of proinflammatory cytokines, activation of T-cells, and regulation of immune response and homeostasis.1 DCs also respond to extracellular microRNAs (miRNAs) and tumor-secreted DNA, to modulate immune responses in cancer and other diseases.2−4 Yet, our understanding of interactions between cells and extracellular ligands, especially in native three-dimensional (3D) tissue environments, remains incomplete. Many cellular responses to external cues occur rapidly and are contingent on the spatial context of the signals (e.g., molecular gradients) within the local milieu. Therefore, improvements in cell-based biosensor technologies with dynamic subcellular readouts and improved spatiotemporal resolution are needed to further our understanding of cell interactions with extracellular signaling molecules in their microenvironment. Thus, the goal of this work is to develop a robust approach to monitor cell interactions with the surrounding environment in real time in a physiologically relevant 3D microenvironment, which would enhance studies focused on understanding cell biology and developing effective therapeutics.

Many of the conventional molecular biochemistry assays to characterize cell interactions, such as flow cytometry, enzyme-linked immunosorbent assay (ELISA), immunostaining, and polymerase chain reaction (PCR), focus on measuring the presence of specific molecular markers that are indicative of cell response (e.g., membrane receptors, mRNA, or activated signaling molecules). However, they require staining, washing, and manipulation before imaging and only provide end-point results. These methods are further constrained by requiring the measurement of many cells as opposed to single-cell analysis. They also do not provide a direct indication of the extracellular ligand that leads to cell activation (only measure cell response), thereby failing to monitor interactions at the cell surface in real time. More recent studies have engineered cell surfaces with synthetic probes,5,6 labeled proteins,7 or fluorescent DNA constructs.8 These advances have allowed for improved capabilities for measurements at the single-cell level,9 but they still are only capable of understanding cell interactions with a single target molecule and cannot monitor the interaction of cell membranes with multiple biomolecules.

Cell interactions are highly affected by their microenvironments, including the physical and biochemical properties of the surrounding extracellular matrix (ECM).10,11 Living systems are complex and difficult to control and interrogate efficiently at cellular length scales, making it difficult to study detailed mechanisms efficiently in animal models. 3D tissue models are useful surrogates for animal models that recapitulate key aspects of living tissues while enabling control over the environment with simpler measurement readouts; hence, these are useful systems to study underlying mechanisms of cell response in physiologically relevant microenvironments. Although prior work has demonstrated that engineered cells with sensing capabilities can be monitored in live animal models,5,8 they have not yet been widely deployed in reconstituted 3D tissue models, which provide simpler measurement readouts and control over the microenvironment composition. The incorporation of cells engineered with molecular detection capabilities into 3D tissue models would enable novel insights into the spatiotemporal effects of soluble signals and intercellular communication, including mechanisms that mediate disease progression.

Structural DNA nanotechnology12 has emerged as a versatile approach to make biocompatible nanodevices with precise structures that can be functionalized with a large range of molecules, making them attractive for biological applications, including engineering cell membranes.13 The molecular self-assembly process known as DNA origami14 allows for programming complex nanoscale geometry,15,16 tunable mechanical and dynamic properties,17−19 and the incorporation of one or many molecules with nanometer precision.20−25 DNA origami nanodevices have been recently used in applications including drug delivery,26−29 ion and molecular transport,30,31 and imaging,32 as well as molecular sensing, manipulation, and measurement.21,33,34 In addition, 3D DNA nanodevices were successfully incorporated into the cell membranes to control adhesion between two living cells20 and facilitate cell–cell communication,35 and other advanced functions like membrane sculpting36 or cargo transport37 have been demonstrated on synthetic membranes. While these nanodevices have been used in a variety of biological assays including cell culture,29,38 cell spheroids,39 and animal models,27 they have not been implemented into 3D ECM model systems, which are ideal for probing biological mechanisms in native tissue environments.

Here, we establish a method to sense multiple biomolecules on the membrane of living cells in a 3D tissue model. We designed a DNA origami cell sensing platform (CSP) capable of detecting the presence of two specific molecules. We focused on detecting nucleic acid sequences on the surfaces of both CH12-LX B cells (suspension) and MutuDC 1949 dendritic cells (adherent) with fluorescence-based reporting both in cell culture and in 3D collagen matrices. Using microfluidics to control the ECM structure formation and the localized transport of target molecules, we show how multifunctional DNA origami devices can be used to probe the temporal interactions of cells and their local environment with subcellular resolution in a tissue model system.

2. Results and Discussion

2.1. Stable DNA Origami Device Detects Multiple Targets

We designed the CSP structure using DNA design software caDNAno.40 The structure consists of 40 double-stranded DNA (dsDNA) helices organized into three layers with gaps in the middle layer (Figure 1A), inspired by a prior nanorod design that exhibited efficient folding and robust stability in cell culture media.29 The full caDNAno design and corresponding oligonucleotide sequences are provided in Figure S1 and Table S1. The CSP design allows for the selective incorporation of up to 30 single-stranded DNA (ssDNA) overhangs on the membrane-facing side (bottom overhangs) and up to 12 overhangs on the membrane-opposing side (top overhangs, Figure 1A,D). To ensure proper molecular self-assembly and optimize the ion concentration and isothermal annealing temperature,41 CSP was subjected to thermal annealing by rapid heating to 65 °C followed by slow cooling to 4 °C over 2.5 days in different MgCl2 concentrations (10–24 mM). Gel electrophoresis confirmed the folding of CSP in a wide range of 14–24 mM MgCl2 (Figure S2A). We chose 18 mM for subsequent folding. Self-assembled nanostructures were then folded at various isothermal annealing temperatures in 18 mM MgCl2 over 4 h and subjected to agarose gel electrophoresis. Gel electrophoresis revealed the successful folding of CSP over the annealing temperature range of 42–59 °C (Figure S2B). We chose 52 °C for subsequent folding with isothermal annealing. To ensure that the structural integrity is preserved in cell culture media, we tested the stability of CSP in Roswell Park Memorial Institute (RPMI) 1640 cell culture medium or 1× phosphate-buffered saline (PBS) supplemented with 2 or 8% fetal bovine serum (FBS) and 1 mM MgCl2 after 4 h incubation in 37 °C via agarose gel electrophoresis. Figure 1B shows that CSP nanostructures remained intact under each cell culture medium condition compared to the control structures in storage buffer, 1× FOB solution (5 mM Tris, 5 mM NaCl, 1 mM ethylenediaminetetraacetic acid (EDTA)) supplemented with 10 mM MgCl2. The slight difference in the migration speeds is likely due to the different buffer conditions. Leading bands were excised and visualized using transmission electron microscopy (TEM) to confirm that CSP preserved its structural integrity in cell culture media (Figures 1C and S3).

Figure 1.

Figure 1

Open in a new tab

Design and characterization of CSP. (A) Isometric schematic illustrating the structure design and locations of the top (red) and bottom (blue) overhangs. (B) Gel analysis confirms CSP structural stability after 4 h in cell culture conditions (37 °C) compared to CSP in storage buffer (left to right, 1 kb DNA ladder, the 7249 M13mp18 scaffold starting material, CSP in storage buffer, CSP in RPMI supplemented with 1 mM MgCl2, CSP in RPMI supplemented with 1 mM MgCl2 and 2% FBS, CSP in RPMI supplemented with 1 mM MgCl2 and 8% FBS, CSP in 1× PBS, CSP in 1× PBS supplemented with 1 mM MgCl2 and 2% FBS, CSP in 1× PBS supplemented with 1 mM MgCl2 and 8% FBS, 1 kb DNA ladder). (C) TEM images confirm that the structure is preserved after incubation under cell culture conditions for 4 h in storage buffer (top) and RPMI supplemented with 2% FBS and 1 mM MgCl2 (bottom). Scale bars: 100 nm. (D) Two-dimensional (2D) top-view and side-view schematics of CSP with DNA detection modules on C3 (red) and C4 (green) locations. Before the introduction of DNA targets, the quenching oligo is bound to the respective top overhangs, placing the quencher in the proximity of the fluorophore, leading to a low fluorescence signal. After the addition of DNA targets, the quenching oligo is displaced with the target oligo, resulting in a high fluorescence signal (shown in (E), dashed lines). (E) Ensemble fluorescence measurement showing the fluorescence signal of CSP folded with quenching oligo (wQ, dark solid lines), folded without the quenching oligo (woQ, light solid lines), and folded with quenching oligo followed by the addition of the target oligo (wQwT, dashed lines).

The CSP enables incorporation of up to 12 top overhangs (Figure 1D) which could potentially allow for 12 distinct detection sites for various biomolecular targets. Here, to enable the detection of two distinct DNA target strands, two overhangs with unique sequences were incorporated at locations 3C and 4C onto the CSP. Two quencher-labeled oligonucleotides (QOs) complementary to each of the top overhangs were designed to provide a 5-base pair toehold for the binding of target oligos (Figure 1D). Moreover, one Cy3-labeled oligonucleotide and one Cy5-labeled oligonucleotide were incorporated into the CSP platform to create the two-channel internally labeled CSP. The Cy3 and Cy5 oligonucleotides were incorporated so that the Cy3 molecule is proximal to the quencher at the 4C overhang location, and Cy5 is proximal to the quencher at the 3C overhang location (Figure 1D). Thus, Cy3 and Cy5 are quenched in the initial configuration and hence emit a low fluorescence signal (solid dark green (wQ Cy3) and solid dark red (wQ Cy5) in Figure 1E). The addition of the target DNA strands causes the QO strands to be removed via toehold-mediated strand displacement42 (Figure 1D). After target A (DNA target corresponding to the Cy3 channel) displaces the QO, the Cy3 molecule emits a higher fluorescence signal as is shown with the dashed green spectra (wQwT Cy3) in Figure 1E. The Cy3-fluorescence signal of the CSP folded without the quenching oligo was also shown as a control to display the maximum fluorescence signal (solid light green spectra (woQ Cy3) in Figure 1E). The design principles are the same for the Cy5-labeled oligo and the 3C overhang. The Cy5 fluorescence signal is low before the addition of target B (DNA target corresponding to the Cy5 channel), as shown with solid dark red spectra (wQ Cy5) in Figure 1E. Incubation of target B with the CSP causes strand displacement and an increase in the Cy5 fluorescence signal (dashed red spectra (wQwT Cy5) in Figure 1E). The solid light red spectra (woQ Cy5) in Figure 1E show the Cy5 signal from the CSP folded without QO, as the control shows that essentially all quencher strands are displaced. As shown in Figure 1E, the absolute value of the maximum intensity of Cy3 and Cy5 fluorophores (solid light green and red spectra, respectively) are different and Cy5 intensity is consistently higher than Cy3. Previous studies have shown that Cy3 and Cy5 fluorophore molecules have different brightness intensities and the Cy5 brightness index is higher than the Cy3 brightness index. Moreover, the fluorescence efficiency of Cy3 and Cy5 is strongly sequence-dependent,43,44 which can justify the differences we see in intensity values. However, the change in relative fluorescence intensity before and after the addition of the target sequence is significant and we use the relative value for the cell detection experiments. The quantitative comparison of maximum fluorescence intensity between different conditions with independent replicates is presented in Figure S4.

2.2. CSP Detects DNA Targets on the Cell Membrane

To detect DNA targets on the membrane of living cells, the CSP was incorporated onto the extracellular side of cell membranes using cholesterol-conjugated oligonucleotides based on a previously published method20 (Figure 2A). The functionalized cells with CSP were immobilized in a glass chamber and either imaged directly using epifluorescence microscopy or imaged after introducing 1 μM target nucleic acid strands into solution and incubating for 15 min at 37 °C. Representative fluorescence images in Figure 2B show: (I) cells functionalized with non-QO-labeled CSP, (II) cells functionalized with QO-labeled CSP, (III) cells with QO-labeled CSP after incubation with target A, (IV) cells with QO-labeled CSP after incubation with target B, (V) cells with QO-labeled CSP after incubation with both DNA targets, and (VI) cells functionalized with non-QO-labeled CSP after incubation with both DNA targets.

Figure 2.

Figure 2

Open in a new tab

Detection of ssDNA targets on CH12-LX suspension cells. (A) Schematics of the detection of ssDNA targets on the cell membrane. CSP is incorporated into the cell membrane. The addition of DNA targets results in the displacement of QO and a high fluorescence signal. (B) Fluorescence and differential interference contrast (DIC) images representing controls and steps taken to detect target oligos on the cell membrane. (I) CH12-LX cells functionalized with non-QO-labeled CSP (control). (II) CH12-LX cells functionalized with QO-labeled CSP. (III) The sample in (II) after the addition of target A. (IV) The sample in (II) after the addition of target B. (V) The sample in (II) after the addition of both targets. (VI) The sample in (I) after the addition of both targets (control). Scale bars, 10 μm. (C) Boxplot shows the median and interquartile range of cells’ mean fluorescence intensity normalized relative to the average of the mean fluorescence intensity for conditions I and II. The signal is attributed to the CSP bound to the surface of 100–300 single cells in three independent experiments for each condition (Red: Cy5 signal, green: Cy3 signal). (D) Detection of target A on the membrane of CH12-LX cells in real time. DIC and fluorescence images show the increase of Cy3 signal during 10 min. (E) The fluorescence intensity attributed to the CSP bound to the surface of the bottom cell was quantified for each time point around the membrane, and data from time = 1, 7, and 10 min are shown (scale bars, 10 μm).

The fluorescence intensity attributed to CSP on the surface of cells was measured using a custom MATLAB code and parameterized in terms of the mean fluorescence intensity around the perimeter of individual cells.20 The mean fluorescence intensity for individual cells was normalized with respect to the overall average of the mean fluorescence intensity under conditions I (i.e., cells functionalized with non-QO-labeled CSP) and II (i.e., cells functionalized with QO-labeled CSP) as the maximum and minimum, respectively. Based on the results in Figure 2C the cells’ mean fluorescence intensity from the CSP increased robustly after the addition of the respective DNA targets A and B compared to condition II (cells functionalized with QO-labeled CSP). The median of means (MoM) fluorescence intensity of the cells incubated with target A (condition III, Cy3 plot in green) is close to the MoM of cells functionalized with non-QO-labeled CSP (condition I, Cy3 plot in green). Similarly, MoM from the cells functionalized with QO-labeled CSP and incubated with target B (condition IV, Cy5 plot in red) is as high as MoM in cells functionalized with non-QO-labeled CSP (condition I, Cy5 plot in red). Importantly, comparing cells with QO-labeled CSP after incubation with targets A or B (conditions III and IV) to the minimal signal (i.e., cells functionalized with QO-labeled CSP, condition II) shows that the addition of each DNA target only affects the corresponding fluorescence signal and does not affect the fluorescence intensity in the other channel. In addition, exposing cells with QO-labeled CSP to both targets (condition V) leads to robust fluorescence increases in both fluorescence channels, illustrating both specific and multiplexed sensing of molecular targets.

To demonstrate CSP as a tool to monitor cell–biomolecule interactions in real time, we demonstrated the detection of target A on the cell membrane over a time span of 10 min. Figure 2D shows brightfield and fluorescence images of three single cells, which were initially functionalized with QO-labeled CSP. At time t = 0 min, target A was added to the imaging dish at a final concentration of 1 μM, and images were taken every minute to monitor the temporal increase in CSP fluorescence intensity due to detection of target A. We quantified the fluorescence intensity around the membrane of the bottom cell at each time point, and the data from 1, 7, and 10 min after the incubation are shown (Figure 2E). The results show up to a 4-fold increase of the Cy3 intensity at some angles, which highlights the ability to resolve interactions with subcellular resolution around the cell surface. Interestingly, we observed a peak in intensity at 7 min. This nonuniform distribution of the detection signal may be due to inhomogeneous distribution of targets or binding events at the cell surface due to spatial localization of CSP nanodevices to cholesterol- and sphingolipid-rich regions of the plasma membrane.45 However, the 10 min time point shows a more even distribution of CSP, suggesting that the inhomogeneity at earlier time points is due to the spatiotemporal variation of target binding around the cell possibly due to inhomogeneous distribution of targets in solution. DNA can interact with the membrane itself,46,47 which might cause some inhomogeneity in the distribution of DNA targets around the cell. These results highlight that it would be beneficial to be able to consistently track the presence of the CSP separately from the detection signal directly on the same cell. To demonstrate this capability, we also folded CSP with internal Cy5, without the quencher oligo and the overhang, and Cy3 along with its quencher oligo and overhangs. In this modification, Cy5 acts as an internal control signal, and we can use the Cy3 channel to detect the corresponding DNA target. We ran a concentration titration experiment using this platform, and the results are shown in Figure S10. We observed that the Cy5 signal stays constant across all samples and the Cy3 channel responds to the DNA target in a concentration-dependent manner, suggesting that this normalization approach could be useful for quantifying concentrations of targets.

2.3. CSP Detects Cell Membrane Interactions in the 3D ECM Model

Interactions between cells and their microenvironment are highly affected by ECM biophysical and biochemical properties. The design and fabrication of integrated microfluidic devices with localized 3D ECM compartments advance the study of complicated living systems,46,47 hence aiding the understanding of detailed biological mechanisms. These devices enable controlling the biophysical properties of the microenvironment such as collagen density while enabling measurements with subcellular spatial resolution in real time. Here, we used a tissue model to establish the successful incorporation of functionalized cells into the collagen matrix to investigate the cell–target molecule interaction in a 3D microenvironment that is more representative of native biological conditions.

To determine how target biomolecules interact with functionalized cells in the 3D collagen matrix, we developed a microfluidic platform that features fluid flowing through two adjacent channels, which we call side channels, on either side of a central channel that contains a 3D collagen extracellular matrix (ECM). We refer to this microfluidic system as an ECM probing chip (EPC). The EPC has two side channels (50 μm in height) that are spaced 1 mm apart, each with its own individual inlet and outlet. The individual inlets and outlets in the side channels (Figure 3A—purple channels) allow for the control of flow through the side channels and across the middle ECM channel, which contains a collagen matrix that can be seeded with live cells. Along the device, there are six apertures (100 μm in width and 50 μm apart) that allow for components introduced into a side channel (e.g., target strands) to flow through the 3D collagen matrix (Figure 3A—green channel). In these experiments, dendritic cells were functionalized and mixed with collagen I and were seeded into the middle ECM channel of the EPC and incubated for 30 min at 37 °C to ensure collagen polymerization prior to running detection experiments.

Figure 3.

Figure 3

Open in a new tab

Detection of ssDNA targets on DCs in a 3D collagen matrix using the EPC. (A) Schematic images of the microfluidic platform used for the detection. (Left) Top view of the channel featuring the localized region of a mixture of cells and the collagen gel (green) and the channels that will be used to apply the flow of targets (purple). Each side channel has independent input and outlet ports, allowing the control of flow in both channels. (Middle) Close-up view of the boxed area in the left, showing apertures that allow the connection of the target channels with the cell channel. (Right) Close-up view of the boxed area in the middle, showing cells functionalized with CSP in collagen gel seeded into the middle channel of the EPC. (B) Confocal images representing cells functionalized with CSP in collagen gel seeded into the middle channel of the EPC (scale bar: 50 μm). (Left) Brightfield image showing cells seeded into the middle channel and one aperture connecting the middle channel to the top channel. Using confocal microscopy, the formation of collagen fibers (green), successful binding of CSP to the cells (red and yellow channels) and incorporation of cells into the collagen matrix are shown. (C) Fluorescence and DIC images representing the detection of target A on the cell membrane in the collagen matrix. (I) DC before the addition of target A. (II) DC after the flow of target A into the collagen channel. The boxplot represents the mean fluorescence intensity of ∼50 cells (for each condition) in the collagen matrix, before and after the flow of the target in five different EPCs. The individual cells are examples of DCs functionalized with QO after the addition of target A (II) (scale bars, 10 μm). (D) Detection of target A on the surface of DC in the collagen matrix in real time. DIC and fluorescence images show the increase of the Cy3 signal over 10 min (left). The average of the Cy3 signal on the membrane of seven cells seeded into three different EPCs over 10 min. The fluorescence signal is normalized based on the average signal of all of the cells at times 0 and 10 min, and the rate of photobleaching (scale bars, 10 μm).

First, we showed with fluorescence and confocal microscopy the successful formation of collagen fibers and the stability of functionalized cells incorporated into collagen matrices. Figure 3B shows (left to right) brightfield, Cy3, Cy5, and reflectance images of DC incorporated into the middle channel, located close to an aperture. Yellow and red images represent Cy3 and Cy5 signals, respectively, demonstrating the stability of CSP on the membrane of the DCs at 2 h after the preparation of the samples, and the reflectance image shows the collagen matrix structure. The cell periphery also shows up in the reflectance image, which may be due to collagen binding or accumulation at the surface of the cell.48

To establish the functionality of CSP on the membrane of DC, we repeated the two-channel detection experiment with DCs. Figure S5 shows the successful multiplexed detection of targets A and B on the membrane of DCs deposited in an imaging dish. Comparing woQwoT samples in both CH12.LX and DCs indicates that CH12.LX cells display a higher level of CSP incorporation, which is likely due to the differences in the molecular composition of the cell membranes;20,49 however, there is still sufficient CSP on the surface to carry out target detection. We tested the detection of target A on the DCs seeded within the ECM in the EPC. DCs functionalized with QO-labeled CSP were seeded into the middle channel of five EPC devices, and DIC and fluorescence images were taken from ∼50 individual cells (condition I in Figure 3C). Lastly, 1 μM target A was introduced into the inlet channel (target channel) and 1× PBS was added to the outlet channel. By applying a droplet of the DNA target on top of both reservoirs of the inlet channel, a pressure difference was applied across the collagen channel, causing the flow of the DNA target through the collagen. Near 50 individual cells were again imaged after 15 min (condition II in Figure 3C) and the mean fluorescence intensity was measured. The boxplot in Figure 3C shows that MoM for DC fluorescence intensity increased by approximately 6 times after the addition of the target DNA.

We further demonstrated the capability of our platform to study the interaction of biomolecules with the cell membrane in situ in real time. The EPC was seeded with DCs functionalized with QO-labeled CSP and was mounted on the imaging stage. Target DNA and 1× PBS were introduced into the inlet and outlet channels, respectively. Then, two droplets of the DNA target were applied on top of the inlet reservoirs to help drive the flow through the middle channel, and individual cells were imaged every minute over 10 min. Figure 3D shows DIC and fluorescence images of two CSP-functionalized DCs imaged over time. The intensity of the Cy3 signal increased significantly over 10 min after introduction of the target DNA. The mean fluorescence intensity of seven individual cells seeded in three EPCs at each time point was measured, and the plot in Figure 3D shows its average relative to the average signal at time zero over the time span of the experiment. We normalized individual cells’ mean Cy3 signal with respect to the average Cy3 signal at time 0. To account for photobleaching, we also normalized the mean fluorescence intensity of each cell at each time point to the average cell fluorescence intensity for the case where no target was added to cells with non-QO-labeled CSP (Figure S6). The Cy3 reporter signal increases steadily, more than doubling over 10 min. This experiment was repeated, and the cells were imaged for 50 min with 5 min intervals (to prevent significant photobleaching), and the results are shown in Figure S7. These experiments revealed that the signal increased mainly over the first 10 min of the experiment and saturated by approximately 20 min.

To test whether CSP could be used for longer-term experiments, we tested the stability of membrane anchoring on both DCs in the ECM and CH12.LX cells in suspension (Figure S9). The level of CSP on the surface remained similar for up to 4 h for the DCs in ECM and decreased by less than 10% in the CH12.LX cells in suspension. For the CH12.LX cells, we also compared the stability to cholesterol-labeled dsDNA strands, where we observed a 55% decrease in the presence of DNA strands on the cell surface at 4 h, and the majority of signals appeared to come from internalized strands that remained near the cell surface (Figure S9A). Hence, the CSP provides a distinct advantage for cell surface applications at a time scale of several hours. It is worth noting that photobleaching would limit the frequency of imaging, especially for longer time points.

To assess the potential effects of the CSP on cell function, especially at these longer time points, a live/dead assay was performed on DCs seeded into the collagen model. These results showed that DC viability was consistently high (∼90%) after 4 h of incubation in the EPC device and drops to 70% (Figure S8) over 24 h for both CSP-labeled and unlabeled DCs. These results indicate functionalizing cells with CSP does not have an impact on their viability for up to 24 h. Other potential effects of the CSP on cell function could include occluding binding of surface receptors or cell stimulation, especially for immune cells.50 To assess these potential effects, we evaluated surface markers of DCs labeled with CSP via flow cytometry. These results revealed no difference in the binding of an antibody to the DC marker CD11c in cells labeled with CSP versus unlabeled cells for up to 6 h (Figure S12), suggesting that this level of CSP labeling does not impede access to the CD11c surface receptors (although a decrease in antibody binding to CD11c was observed at 24 h). We also quantified the level of CD69, CD40, CD80, and CD86 (all surface markers of immune cell activation) on DCs labeled with CSP and compared them to cells treated with PBS or lipopolysaccharide (LPS) as negative and positive controls, respectively (Figure S12). In summary, these results revealed little to no increase in these activation markers for up to 2 h and significant increases at 6 h and longer time points. The cellular activation caused by CSP was comparable to LPS, which could be due to residual endotoxin remaining from the bacterial scaffold production.50 Collectively, these results suggest that CSP does not significantly impact cell viability, surface receptor binding, and expression of surface activation markers over a time scale of 2 h. However, cellular activation is important to consider for future studies relating molecular detection to cellular function. Moreover, more careful production (e.g., endotoxin removal50,48) or modification methods (e.g., polymer coating51) could provide routes to mitigate cellular activation caused by CSP.

3. Conclusions

The ability to monitor the interactions of cells with the instructive cues originating from the local molecular and extracellular environment is essential for understanding the determinants of cell fate and function,52 for example, regulating cellular spatiotemporal organization,53,54 processes such as neurotransmission,55 wound healing, and inflammation,56 and progression of diseases including metabolic disorders, autoimmune diseases, and cancer.57−60 Here, we established a method to functionalize living cells with DNA origami sensors to spatiotemporally monitor interactions of cells with biomolecules in the local ECM microenvironment using our microfluidic 3D tissue model device. We demonstrated the multiplexed sensing on the surface of CH12-LX B cells and demonstrated the capability of our method to monitor subcellular interactions of MutuDC 1949 cells in the 3D ECM. The main driving force in delivering the target molecules to CSP structures on the cell membrane is target diffusion, which acts differently in free solution versus the extracellular matrix. This difference is due to the combination of many factors such as cell–matrix interactions, molecule–matrix interactions, transport hindrances, and external stimuli.61 Studies of dynamic nanoparticle diffusion in a 3D cell culture environment show that this phenomenon is hampered by the interaction with the collagen fibers.62 Moreover, mechanical forces within the 3D tissue can redistribute molecules,61 and other factors such as interactions between the cell and the matrix might influence target binding through physical obstruction or influencing cell morphology. So, it is crucial to investigate cell–target interactions in both 2D and 3D environments.

In this study, our results established the integration of DNA origami nanodevices, cell membrane engineering, and microfluidic tissue model systems as a powerful approach to probe biological interactions in physiologically relevant 3D environments with high spatial (i.e., micron scale) and temporal (i.e., minute scale) resolution. The time resolution could easily be improved to the second scale to study faster interactions; however, this could require additional strategies to inhibit photobleaching.63,64 Moreover, our approach allows real-time detection, but only of the first encounter event. We believed that single detection devices can still be highly useful, since we can detect the molecules and potentially measure cells’ downstream signal and activation in relevant applications. However, using DNA nanotechnology, it is feasible to incorporate a resettable design into the CSP to repeatedly probe target detection. Alternatively, weaker affinity binding (e.g., shorter base-pairing segments) could be used to allow for transient binding to molecular beacons with tunable dwell times.65

Here, we focused on nucleic acid targets as a proof of concept to enable the detection of relevant targets such as miRNA and circulating tumor DNA (ctDNA) that are present in extracellular circulation and ECM and can play an important role in activation of immune cells.66,67 This approach could be expanded to study cell response to a variety of molecules. For example, the incorporation of aptamers could enable the detection of target growth factors or cytokines68 and the measurement of local environmental factors such as pH.69 Other DNA constructs have been demonstrated as tools to measure forces19,21,70,71 or ion concentrations.72 The programmability and stability of the platform in a wide range of solution conditions allow for the incorporation and functionality of many DNA aptamers.8 Thus, it is possible to look at the underlying mechanisms of cooperative action between multiple types of biomolecules or biomolecules and other local factors (pH, forces, ions).

In this study, we multiplexed two organic fluorophores to label our DNA nanostructures. Common fluorescence imaging systems could multiplex 3–4 channels. Further, multiplexing could be achieved using novel fluorescence imaging techniques such as metafluorophores,73 frequency multiplexed DNA-PAINT,74 or fluorescence nanoparticles with controllable spectral properties75 that could expand the distinct number of channels and minimize cross-talk and photobleaching. For example, luminescent quantum dots (QDs) have been proven as a promising alternative to traditional organic dyes in various fluorescence-based applications such as multicolor live cell imaging over a week,76 and QDs have been used for Förster resonance energy transfer (FRET)-based studies of biomolecular interactions.77

We successfully established a reliable method to study subcellular interactions in a 3D tissue model using a microfluidic device. Microfluidic devices are enabling novel approaches for probing important biological questions in cellular microenvironments with precise control over biophysical and biochemical parameters.78 Moreover, cutting-edge microfluidic models allow for recapitulating native cell-ECM structures in microfluidics that orchestrate physiological processes such as angiogenesis, vessel branching, and tissue morphogenesis. Prior work has utilized microfluidic devices to actuate DNA origami nanodevices inside cell-sized microfluidic compartments.79 Here, we establish the integration of DNA nanodevices, cell membrane engineering, and microfluidic tissue model systems to enable a unique approach to study subcellular biological phenomena in a 3D ECM. Combining the programmability of DNA origami with versatile designs of microfluidic chips allows for the investigation of the subcellular interactions under controlled biomolecular transport and fluid mechanical stimuli with unprecedented spatial and temporal resolution.

4. Experimental Section

4.1. Design and Fabrication of the CSP

The CSP was designed using software caDNAno40 and fabricated using protocols developed by Castro et al.15 CSP is a 65 nm × 30 nm × 6 nm platform, with 42 potential overhangs (Figure 1). Staple sequences were specified in caDNAno and ordered from a commercial vendor (Integrated DNA Technologies, Coralville, IA). The CSP was folded using p7249 from the M13mp18 genome with a total of 186 ssDNA staples. To fold the CSP with two DNA detection modules on locations C3 and C4, staple strands with an end directly adjacent to the overhang locations (green oligos in Figure S1 and Table S1) were replaced with one DNA oligo labeled with a Cy5 and one DNA oligo labeled with a Cy3 fluorophore (Table S2). Moreover, CSP was folded with 30 overhangs on the membrane-facing side (bottom overhangs, purple oligos in Figure S1 and Table S1) and 2 overhangs on the opposite side (top overhangs, Table S2) on locations C3 and C4. Briefly, a purified scaffold at 20 nM concentration was combined with a 5-fold molar excess of staples (each staple at 100 nM) and a 10-fold molar excess of QO (if needed, in 200 nM each) in folding buffer (1× FOB: 5 mM Tris, 5 mM NaCl, 1 mM EDTA, supplemented with 18 mM MgCl2).

To find the optimal salt concentration needed for structural folding, initially, folding reactions with different salt concentrations (10, 12, 14, 16, 18, 20, 22, 24 mM MgCl2) were prepared. Then, the self-assembly reaction was done by rapidly heating the reactions to 65 °C followed by slow cooling to 4 °C over 2.5 days in a thermal cycler (Bio-Rad, Hercules, CA). The folding reaction was brought to 65 °C and stepped from 65 to 24 °C in 1 °C steps at varying time increments. From 65 to 62 °C, the temperature was held for 1 h; from 61 to 59 °C, the temperature was held for 2 h; from 58 to 46 °C, the temperature was held for 3 h; from 45 to 40 °C, the temperature was held for 1 h; from 39 to 24 °C, the temperature was held for 30 min. After being held at 24 °C for 1 h, the temperature was lowered to 4 °C. Folding reaction products were subjected to agarose gel electrophoresis15 in a 2% agarose (Life Technologies) gel (0.5× Tris–borate–EDTA (TBE)) in the presence of 11 mM MgCl2 and 1 μM ethidium bromide (EtBr). The highest intensity on the gel image was observed in the 18 mM MgCl2 band, and after TEM confirmation, it was chosen as the optimized concentration for folding the CSP (Figure S1).

Next, using folding reactions at 18 mM MgCl2, the optimal isothermal annealing temperature was found.80 The folding mixtures were subjected to thermal annealing by rapid heating to 65 °C followed by isothermal annealing for 4 h at different temperatures (60, 58.4, 56.0, 52.5, 47.8, 44.4, 41.7, 40) and rapid cooling to 4 °C. Folding reaction products were subjected to agarose gel electrophoresis, and the results revealed the successful folding of CSP over a wide range of 42–59 °C (Figure S1); 52 °C was chosen for subsequent folding with isothermal annealing, since it has the highest intensity on the gel image. Therefore, to fold CSP, the folding mixtures at 18 mM MgCl2 were subjected to thermal annealing by rapid heating to 65 °C followed by 4 h at 52 °C and rapid cooling to 4 °C.

4.2. DNA Nanostructure Purification

Folded DNA nanostructures were purified by mixing folding reaction products with equivolume of 15% poly(ethylene glycol) (PEG) 8000 (Sigma-Aldrich, St. Louis, MO) supplemented with 500 mM NaCl and centrifugation for 30 min at 16 000 rcf to remove excess staple strands (protocol modified from ref (81)). After removing the supernatant, purified structures were resuspended in 1× FOB supplemented with 10 mM MgCl2 (storage buffer). To fully remove the excess staple strands, the PEG purification procedure was repeated two times.

4.3. Transmission Electron Microscopy (TEM)

TEM grids were prepared as described in Castro et al.15 Briefly, 4 μL of a ∼1 nM purified DNA nanostructure was deposited onto a copper TEM grid coated with carbon and formvar (Electron Microscopy Sciences, Hartfield, PA) and incubated for 4 min at room temperature. The sample was removed by gently touching filter paper to the edge of the grid, and 10 μL of 2% uranyl formate negative stain was applied to the grid and immediately removed with filter paper as a washing step. Then, 20 μL of 2% uranyl formate was immediately added, incubated for 40 s, and finally removed with filter paper. The grid was allowed to dry for at least 30 min before being visualized on a Tecnai G2 BioTWIN transmission electron microscope (FEI, Hillsboro, OR) in The Ohio State University Campus Microscopy & Imaging Facility (CMIF).

4.4. Functionality of CSP in Detection of DNA Targets

To evaluate CSP strand displacement, structures (folded with QO) at 5 nM concentration were incubated with 1 μM final concentration of each DNA target (A or B) at 37 °C for 15 min followed by fluorescence measurement on a fluorometer (FluoroMax-4, Horiba, Japan). The samples were excited with 510 and 610 nm lasers for targets A and B, respectively, and the emission was measured from 530 to 700 nm for target A and 630 to 700 nm for target B. The fluorescence intensity of CSP folded with or without QO in the same concentration (5 nM) was also tested on the fluorometer to demonstrate the maximum and minimum fluorescence signals.

4.5. Stability of the CSP in Cell Culture Conditions

To confirm CSP structural integrity under cell culture conditions, folded structures were purified using the PEG centrifugal purification process, resuspended in different cell culture media, and incubated for 4 or 24 h at 37 °C. Purified structures were incubated in CH12-LX clear cell culture media (RPMI 1640 without l-glutamine (Corning)) supplemented with 1 mM MgCl2, RPMI supplemented with 2% heat-inactivated fetal bovine serum (FBS, Atlas Biologicals) and 1 mM MgCl2, RPMI supplemented with 8% FBS and 1 mM MgCl2, 1× phosphate-buffered saline (PBS) without MgCl2/CaCl2 (Corning, Cat# 21-031-CV), 1× PBS supplemented with 2% FBS and 1 mM MgCl2, and 1× PBS supplemented with 8% FBS and 1 mM MgCl2 following PEG purification. Agarose gel electrophoresis assay in a 2% agarose gel in the presence of 11 mM MgCl2 and 1 μM ethidium bromide followed by visualization of the excised gel bands on TEM was used to confirm the stability of the structures in the cell culture media (Figures 1 and S3).

4.6. Cell Culture

CH12-LX is a murine B-cell lymphoma obtained from Dr. Gail Bishop.82 To culture the CH12-LX B cells, RPMI 1640 without l-glutamine (Corning) was supplemented with 10% heat-inactivated FBS (Atlas Biologicals), 1% penicillin–streptomycin-–glutamine (100×, Thermo Fisher), 1% sodium pyruvate (100 mM, Life Technologies), 1% N-(2-hydroxyethyl)piperazine-N′-ethanesulfonic acid (HEPES) (1 M, Life Technologies), 1% minimum Eagle’s essential medium (MEM) non-essential amino acids (100×, Life Technologies). For experiments, CH12-LX cells were washed once with 1× PBS, followed by one wash with the experimental medium (clear RPMI 1640 without l-glutamine (Corning) supplemented with 2% heat-inactivated FBS and 1 mM MgCl2). Finally, CH12-LX cells were resuspended in the experimental medium at 4000 cells/μL for incubation with cholesterol-conjugated oligos.

The wild-type MutuDC1940 dendritic cell (DC) line is derived from mouse spleen tissues and is GFP positive due to the GFP reporter in the CD11c:SV40LgT transgene. This cell line was obtained from abm (Richmond BC, Canada) and was maintained in supplemented IMDM in a 37 °C incubator at 5% CO2. To make the complete growth medium, IMDM (1×) + GlutaMAX (Gibco Ref: 31980-030) was supplemented with 10% heat-inactivated fetal bovine serum (FBS, Atlas Biologicals), 1% of 7.5% sodium bicarbonate solution (Life Technologies), 50 μM β-mercaptoethanol, 1% HEPES (1 M, Life Technologies), and 1% penicillin–streptomycin–glutamine (100×, Thermo Fisher). For passaging purposes, the cells were washed with 1× PBS without MgCl2/CaCl2, followed by their detachment from the culture flask using a 1:1 ratio of 1× PBS and 0.25% trypsin–EDTA (1×) (Corning). The cells were incubated with a PBS–trypsin mixture for 3–4 min at room temperature. Trypsin was then neutralized using the culture media. For functionalization experiments, the cells were washed once with 1× PBS followed by one wash with the experimental medium (clear RPMI 1640 without l-glutamine (Corning) supplemented with 2% heat-inactivated fetal bovine serum (FBS, Atlas Biologicals) and 1 mM MgCl2). Finally, DCs were resuspended in the experimental medium at 4000 cells/μL for incubation with cholesterol-conjugated oligos.

4.7. Cell Membrane Functionalization

The cells were functionalized with CSP using the protocol explained in Akbari et al.20 Briefly, cells at a density of 4000 cells/μL were resuspended in the experimental medium (clear RPMI 1640 without l-glutamine (Corning) supplemented with 2% heat-inactivated FBS (Atlas Biologicals) and 1 mM MgCl2) and incubated with 10 μM cholesterol-conjugated oligonucleotide (Table S3) for 5 min at 37 °C. The cells were then washed once in the experimental media to remove the excess cholesterol-conjugated oligonucleotide. The cells were then incubated with the 60-base bridge oligo at 1 μM for 5 min at 37 °C, followed by the addition of the 20-base pair fortifier oligos at 1 μM, incubation for 5 min at 37 °C, and a wash with experimental media. Finally, the cells were incubated with the nanostructures at 5 nM, for 5 min at 37 °C. The cells were washed a final time to remove excess nanostructures using the experimental media and were resuspended in either 1× PBS or the experimental media for the experiment.

4.8. DNA Target Detection on the Membrane of Suspension Cells

CSP folded with QO and without QO were incorporated into the membrane of two separate subpopulations of cells (CH12-LX or DCs). The subpopulation with QO then was divided into four smaller subpopulations, transferred to an eight-well imaging dish, and incubated with either Cy3-target at 1 μM concentration, Cy5-target at 1 μM concentration, or both DNA targets (A and B) at 1 μM final concentration or 2 μL of ddH2O (to maintain the same buffer composition) in 37 °C for 15 min. Cells functionalized with CSP devices without QO were divided into two subpopulations and incubated with either 1 μM final concentration of each DNA target A or B or 2 μL of ddH2O on the imaging dish in the same condition. All six conditions were imaged using DIC, Cy3, and Cy5 imaging settings (see the Supporting Information). For real-time experiments, the Cy3-target was added at 1 μM final concentration to the subpopulation of cells functionalized with QO on the microscope stage after capturing the first image at time = 0. The DIC and Cy3-fluorescence images were captured every minute. The fluorescence signal on the membrane of each cell was measured and analyzed using a home-built MATLAB code described below.

4.9. DNA Target Detection on the Membrane of DCs Seeded into the EPC

DCs were functionalized with CSP using the same protocol from ref (20). While preparing the cells, a collagen mixture was prepared according to the vendor’s instructions. Briefly, the pH of high-concentration rat tail collagen I stored in acidic solution (Corning Life Sciences) was neutralized to a pH of 7.4 using NaOH in 10× PBS without MgCl2/CaCl2 so that the final concentration of PBS was 1×. The mixture was prepared in a 4 °C ice bath and incubated at the same temperature for 10 min. Finally, functionalized cells were resuspended in 1× PBS without MgCl2/CaCl2 and mixed with collagen so that the concentrations of collagen and the cells were 2 mg/mL and 7000 cells/μL, respectively. The final mixture was seeded into the middle channel of the EPC and incubated at 37 °C for 30 min to allow for formation of the collagen matrix before imaging. Each EPC was mounted onto the imaging stage and approximately 20 cells were imaged using DIC and Cy3-fluorescence settings. DNA target A at 1 μM concentration and 1× PBS were added to the EPC inlet and outlet channels, respectively, and the EPC was incubated for 15 min at room temperature. Then, 20 cells were imaged using the same settings to study DNA displacement on their membrane. For real-time experiments, the Cy3-target A at 1 μM concentration and 1× PBS were added to the EPC inlet and outlet channels on the microscope stage after capturing the first image at time = 0. The DIC and Cy3-fluorescence images were captured either every minute or every 5 min. The fluorescence signal on the membrane of each cell was measured and analyzed using a home-built MATLAB code.

4.10. Cell Periphery Fluorescence Analysis

A home-built MATLAB code previously developed by Akbari et al.20 was used to analyze the fluorescence signal on the membrane of the cells. The code measures the intensity of each pixel around the circumference of each cell and reports the mean fluorescence signal along with the angular distribution of the fluorescence signal. Between 100 and 300 cells were analyzed for each suspension experiment, and nearly 50 cells were analyzed for collagen-seeded cell experiments.

Acknowledgments

Electron and confocal microscopies were done at the Campus Microscopy & Imaging Facility (CMIF), The Ohio State University. The clean room microfabrication procedure was performed at Nanotech West Laboratory, The Ohio State University.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsami.2c07971.

  • Additional experimental details; materials; and results including flow cytometry gating schemes, live/dead assays, and long-term detection results (PDF)

Author Contributions

J.W.S., C.E.C., and M.S. conceived this study. M.S. executed the majority of the DNA structure design, fabrication, experiments, and analysis. P.E.B. supported imaging and analysis and prepared microfluidic tissue model systems. E.A. supported experiment design and conception of the research and plan. N.R. and C.R.L. executed flow cytometry and helped prepare materials for DNA origami device fabrication. A.A. designed the microfluidic chip. The manuscript was written through contributions of all authors.

This work was supported by the National Institute of Health (grant R01HL141941 awarded to J.W.S. and C.E.C.). M.S., P.E.B., and N.R. are recipients of the Pelotonia Fellowship from The Ohio State University Comprehensive Cancer Center. This work was also supported in part by the Mary Wieczynski Furnivall Cancer Research Fund.

The authors declare no competing financial interest.

Supplementary Material

am2c07971_si_001.pdf (2.7MB, pdf)

References

  1. Blanco P.; Palucka A. K.; Pascual V.; Banchereau J. Dendritic Cells and Cytokines in Human Inflammatory and Autoimmune Diseases. Cytokine Growth Factor Rev. 2008, 19, 41–52. 10.1016/j.cytogfr.2007.10.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Liang H.; Kidder K.; Liu Y. Extracellular MicroRNAs Initiate Immunostimulation via Activating Toll-like Receptor Signaling Pathways. ExRNA 2019, 1, 9 10.1186/s41544-019-0009-x. [DOI] [Google Scholar]
  3. Ranganathan P.; Ngankeu A.; Zitzer N. C.; Leoncini P.; Yu X.; Casadei L.; Challagundla K.; Reichenbach D. K.; Garman S.; Ruppert A. S.; Volinia S.; Hofstetter J.; Efebera Y. A.; Devine S. M.; Blazar B. R.; Fabbri M.; Garzon R. Serum MiR-29a Is Upregulated in Acute Graft-versus-Host Disease and Activates Dendritic Cells through TLR Binding. J. Immunol. 2017, 198, 2500–2512. 10.4049/jimmunol.1601778. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Kang T. H.; Mao C.-P.; Kim Y. S.; Kim T. W.; Yang A.; Lam B.; Tseng S.-H.; Farmer E.; Park Y.-M.; Hung C.-F. TLR9 Acts as a Sensor for Tumor-Released DNA to Modulate Anti-Tumor Immunity after Chemotherapy. J. ImmunoTher. Cancer 2019, 7, 260 10.1186/s40425-019-0738-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Laughlin S. T.; Baskin J. M.; Amacher S. L.; Bertozzi C. R. In Vivo Imaging of Membrane-Associated Glycans in Developing Zebrafish. Science 2008, 320, 664–667. 10.1126/science.1155106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Amaral A. J. R.; Pasparakis G. Cell Membrane Engineering with Synthetic Materials: Applications in Cell Spheroids, Cellular Glues and Microtissue Formation. Acta Biomater. 2019, 90, 21–36. 10.1016/J.ACTBIO.2019.04.013. [DOI] [PubMed] [Google Scholar]
  7. Giepmans B. N. G.; Adams S. R.; Ellisman M. H.; Tsien R. Y. The Fluorescent Toolbox for Assessing Protein Location and Function. Science 2006, 312, 217–224. 10.1126/science.1124618. [DOI] [PubMed] [Google Scholar]
  8. Zhao W.; Schafer S.; Choi J.; Yamanaka Y. J.; Lombardi M. L.; Bose S.; Carlson A. L.; Phillips J. A.; Teo W.; Droujinine I. A.; Cui C. H.; Jain R. K.; Lammerding J.; Love J. C.; Lin C. P.; Sarkar D.; Karnik R.; Karp J. M. Cell-Surface Sensors for Real-Time Probing of Cellular Environments. Nat. Nanotechnol. 2011, 6, 524–531. 10.1038/nnano.2011.101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Lun X. K.; Bodenmiller B. Profiling Cell Signaling Networks at Single-Cell Resolution. Mol. Cell. Proteomics 2020, 19, 744. 10.1074/MCP.R119.001790. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Muncie J. M.; Weaver V. M. The Physical and Biochemical Properties of the Extracellular Matrix Regulate Cell Fate. Curr. Top. Dev. Biol. 2018, 130, 1. 10.1016/BS.CTDB.2018.02.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Beshay P. E.; Cortes-Medina M. G.; Menyhert M. M.; Song J. W. The Biophysics of Cancer: Emerging Insights from Micro- and Nanoscale Tools. Adv. NanoBiomed Res. 2022, 2, 2100056 10.1002/anbr.202100056. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Seeman N. C. Nucleic Acid Junctions and Lattices. J. Theor. Biol. 1982, 99, 237–247. 10.1016/0022-5193(82)90002-9. [DOI] [PubMed] [Google Scholar]
  13. Huo S.; Li H.; Boersma A. J.; Herrmann A. DNA Nanotechnology Enters Cell Membranes. Adv. Sci. 2019, 6, 1900043 10.1002/advs.201900043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Rothemund P. W. K. Folding DNA to Create Nanoscale Shapes and Patterns. Nature 2006, 440, 297–302. 10.1038/nature04586. [DOI] [PubMed] [Google Scholar]
  15. Castro C. E.; Kilchherr F.; Kim D.-N.; Shiao E. L.; Wauer T.; Wortmann P.; Bathe M.; Dietz H. A Primer to Scaffolded DNA Origami. Nat. Methods 2011, 8, 221–229. 10.1038/nmeth.1570. [DOI] [PubMed] [Google Scholar]
  16. Douglas S. M.; Dietz H.; Liedl T.; Högberg B.; Graf F.; Shih W. M. Self-Assembly of DNA into Nanoscale Three-Dimensional Shapes. Nature 2009, 459, 414–418. 10.1038/nature08016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Marras A. E.; Zhou L.; Su H.-J.; Castro C. E. Programmable Motion of DNA Origami Mechanisms. Proc. Natl. Acad. Sci. U.S.A. 2015, 112, 713–718. 10.1073/pnas.1408869112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Zhou L.; Marras A. E.; Su H. J.; Castro C. E. DNA Origami Compliant Nanostructures with Tunable Mechanical Properties. ACS Nano 2014, 8, 27–34. 10.1021/nn405408g. [DOI] [PubMed] [Google Scholar]
  19. Hudoba M. W.; Luo Y.; Zacharias A.; Poirier M. G.; Castro C. E. Dynamic DNA Origami Device for Measuring Compressive Depletion Forces. ACS Nano 2017, 11, 6566–6573. 10.1021/acsnano.6b07097. [DOI] [PubMed] [Google Scholar]
  20. Akbari E.; Mollica M. Y.; Lucas C. R.; Bushman S. M.; Patton R. A.; Shahhosseini M.; Song J. W.; Castro C. E. Engineering Cell Surface Function with DNA Origami. Adv. Mater. 2017, 29, 1703632 10.1002/adma.201703632. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Le J. V.; Luo Y.; Darcy M. A.; Lucas C. R.; Goodwin M. F.; Poirier M. G.; Castro C. E. Probing Nucleosome Stability with a DNA Origami Nanocaliper. ACS Nano 2016, 10, 7073–7084. 10.1021/acsnano.6b03218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Ding B.; Deng Z.; Yan H.; Cabrini S.; Zuckermann R. N.; Bokor J. Gold Nanoparticle Self-Similar Chain Structure Organized by DNA Origami. J. Am. Chem. Soc. 2010, 132, 3248–3249. 10.1021/ja9101198. [DOI] [PubMed] [Google Scholar]
  23. Chhabra R.; Sharma J.; Ke Y.; Liu Y.; Rinker S.; Lindsay S.; Yan H. Spatially Addressable Multiprotein Nanoarrays Templated by Aptamer-Tagged DNA Nanoarchitectures. J. Am. Chem. Soc. 2007, 129, 10304–10305. 10.1021/ja072410u. [DOI] [PubMed] [Google Scholar]
  24. Maune H. T.; Han S.; Barish R. D.; Bockrath M.; Goddard W. A. III; Rothemund P. W. K.; Winfree E. Self-Assembly of Carbon Nanotubes into Two-Dimensional Geometries Using DNA Origami Templates. Nat. Nanotechnol. 2010, 5, 61–66. 10.1038/nnano.2009.311. [DOI] [PubMed] [Google Scholar]
  25. Bui H.; Onodera C.; Kidwell C.; Tan Y.; Graugnard E.; Kuang W.; Lee J.; Knowlton W. B.; Yurke B.; Hughes W. L. Programmable Periodicity of Quantum Dot Arrays with DNA Origami Nanotubes. Nano Lett. 2010, 10, 3367–3372. 10.1021/nl101079u. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Douglas S. M.; Bachelet I.; Church G. M. A Logic-Gated Nanorobot for Targeted Transport of Molecular Payloads. Science 2012, 335, 831–834. 10.1126/science.1214081. [DOI] [PubMed] [Google Scholar]
  27. Zhang Q.; Jiang Q.; Li N.; Dai L.; Liu Q.; Song L.; Wang J.; Li Y.; Tian J.; Ding B.; Du Y. DNA Origami as an In Vivo Drug Delivery Vehicle for Cancer Therapy. ACS Nano 2014, 8, 6633–6643. 10.1021/nn502058j. [DOI] [PubMed] [Google Scholar]
  28. Zhao Y.-X.; Shaw A.; Zeng X.; Benson E.; Nyström A. M.; Högberg B. DNA Origami Delivery System for Cancer Therapy with Tunable Release Properties. ACS Nano 2012, 6, 8684–8691. 10.1021/nn3022662. [DOI] [PubMed] [Google Scholar]
  29. Halley P. D.; Lucas C. R.; McWilliams E. M.; Webber M. J.; Patton R. A.; Kural C.; Lucas D. M.; Byrd J. C.; Castro C. E. Daunorubicin-Loaded DNA Origami Nanostructures Circumvent Drug-Resistance Mechanisms in a Leukemia Model. Small 2016, 12, 308–320. 10.1002/smll.201502118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Langecker M.; Arnaut V.; Martin T. G.; List J.; Renner S.; Mayer M.; Dietz H.; Simmel F. C. Synthetic Lipid Membrane Channels Formed by Designed DNA Nanostructures. Science 2012, 338, 932–936. 10.1126/science.1225624. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Burns J. R.; Seifert A.; Fertig N.; Howorka S. A Biomimetic DNA-Based Channel for the Ligand-Controlled Transport of Charged Molecular Cargo across a Biological Membrane. Nat. Nanotechnol. 2016, 11, 152–156. 10.1038/nnano.2015.279. [DOI] [PubMed] [Google Scholar]
  32. Jungmann R.; Avendaño M. S.; Woehrstein J. B.; Dai M.; Shih W. M.; Yin P. Multiplexed 3D Cellular Super-Resolution Imaging with DNA-PAINT and Exchange-PAINT. Nat. Methods 2014, 11, 313–318. 10.1038/nmeth.2835. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Wang D.; Fu Y.; Yan J.; Zhao B.; Dai B.; Chao J.; Liu H.; He D.; Zhang Y.; Fan C.; Song S. Molecular Logic Gates on DNA Origami Nanostructures for MicroRNA Diagnostics. Anal. Chem. 2014, 86, 1932–1936. 10.1021/ac403661z. [DOI] [PubMed] [Google Scholar]
  34. Nickels P. C.; Wünsch B.; Holzmeister P.; Bae W.; Kneer L. M.; Grohmann D.; Tinnefeld P.; Liedl T. Molecular Force Spectroscopy with a DNA Origami–Based Nanoscopic Force Clamp. Science 2016, 354, 305–307. 10.1126/SCIENCE.AAH5974. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Li J.; Xun K.; Pei K.; Liu X.; Peng X.; Du Y.; Qiu L.; Tan W. Cell-Membrane-Anchored DNA Nanoplatform for Programming Cellular Interactions. J. Am. Chem. Soc. 2019, 141, 18013–18020. 10.1021/jacs.9b04725. [DOI] [PubMed] [Google Scholar]
  36. Franquelim H. G.; Khmelinskaia A.; Sobczak J.-P.; Dietz H.; Schwille P. Membrane Sculpting by Curved DNA Origami Scaffolds. Nat. Commun. 2018, 9, 811 10.1038/s41467-018-03198-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Rubio-Sánchez R.; Barker S. E.; Walczak M.; Cicuta P.; Di Michele L. A Modular, Dynamic, DNA-Based Platform for Regulating Cargo Distribution and Transport between Lipid Domains. Nano Lett. 2021, 21, 2800–2808. 10.1021/acs.nanolett.0c04867. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Schüller V. J.; Heidegger S.; Sandholzer N.; Nickels P. C.; Suhartha N. A.; Endres S.; Bourquin C.; Liedl T. Cellular Immunostimulation by CpG-Sequence-Coated DNA Origami Structures. ACS Nano 2011, 5, 9696–9702. 10.1021/nn203161y. [DOI] [PubMed] [Google Scholar]
  39. Wang Y.; Benson E.; Fördős F.; Lolaico M.; Baars I.; Fang T.; Teixeira A. I.; Högberg B. DNA Origami Penetration in Cell Spheroid Tissue Models Is Enhanced by Wireframe Design. Adv. Mater. 2021, 33, 2008457 10.1002/adma.202008457. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Douglas S. M.; Marblestone A. H.; Teerapittayanon S.; Vazquez A.; Church G. M.; Shih W. M. Rapid Prototyping of 3D DNA-Origami Shapes with CaDNAno. Nucleic Acids Res. 2009, 37, 5001–5006. 10.1093/nar/gkp436. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Sobczak J.-P. J.; Martin T. G.; Gerling T.; Dietz H. Rapid Folding of DNA into Nanoscale Shapes at Constant Temperature. Science 2012, 338, 1458–1461. 10.1126/science.1229919. [DOI] [PubMed] [Google Scholar]
  42. Yurke B.; Turberfield A. J.; Mills A. P.; Simmel F. C.; Neumann J. L. A DNA-Fuelled Molecular Machine Made of DNA. Nature 2000, 406, 605–608. 10.1038/35020524. [DOI] [PubMed] [Google Scholar]
  43. Kretschy N.; Sack M.; Somoza M. M. Sequence-Dependent Fluorescence of Cy3- and Cy5-Labeled Double-Stranded DNA. Bioconjugate Chem. 2016, 27, 840–848. 10.1021/acs.bioconjchem.6b00053. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Harvey B. J.; Perez C.; Levitus M. DNA Sequence-Dependent Enhancement of Cy3 Fluorescence. Photochem. Photobiol. Sci. 2009, 8, 1105–1110. 10.1039/b905078g. [DOI] [PubMed] [Google Scholar]
  45. Kraft M. L. Sphingolipid Organization in the Plasma Membrane and the Mechanisms That Influence It. Front. cell Dev. Biol. 2016, 4, 154 10.3389/fcell.2016.00154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Avendano A.; Cortes-Medina M.; Song J. W. Application of 3-D Microfluidic Models for Studying Mass Transport Properties of the Tumor Interstitial Matrix. Front. Bioeng. Biotechnol. 2019, 7, 6 10.3389/fbioe.2019.00006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Huang C. P.; Lu J.; Seon H.; Lee A. P.; Flanagan L. A.; Kim H.-Y.; Putnam A. J.; Jeon N. L. Engineering Microscale Cellular Niches for Three-Dimensional Multicellular Co-Cultures. Lab Chip 2009, 9, 1740–1748. 10.1039/b818401a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Noteborn W. E. M.; Abendstein L.; Sharp T. H. One-Pot Synthesis of Defined-Length SsDNA for Multiscaffold DNA Origami. Bioconjugate Chem. 2021, 32, 94–98. 10.1021/acs.bioconjchem.0c00644. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Rivel T.; Ramseyer C.; Yesylevskyy S. The Asymmetry of Plasma Membranes and Their Cholesterol Content Influence the Uptake of Cisplatin. Sci. Rep. 2019, 9, 5627 10.1038/s41598-019-41903-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Lucas C. R.; Halley P. D.; Chowdury A. A.; Harrington B. K.; Beaver L.; Lapalombella R.; Johnson A. J.; Hertlein E. K.; Phelps M. A.; Byrd J. C.; Castro C. E. DNA Origami Nanostructures Elicit Dose-Dependent Immunogenicity and Are Nontoxic up to High Doses In Vivo. Small 2022, 18, 2108063 10.1002/smll.202108063. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Ponnuswamy N.; Bastings M. M. C.; Nathwani B.; Ryu J. H.; Chou L. Y. T.; Vinther M.; Li W. A.; Anastassacos F. M.; Mooney D. J.; Shih W. M. Oligolysine-Based Coating Protects DNA Nanostructures from Low-Salt Denaturation and Nuclease Degradation. Nat. Commun. 2017, 8, 15654 10.1038/ncomms15654. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Ferreira L.; Karp J. M.; Nobre L.; Langer R. New Opportunities: The Use of Nanotechnologies to Manipulate and Track Stem Cells. Cell Stem Cell 2008, 3, 136–146. 10.1016/j.stem.2008.07.020. [DOI] [PubMed] [Google Scholar]
  53. Birnbaum M. E.; Mendoza J. L.; Sethi D. K.; Dong S.; Glanville J.; Dobbins J.; Özkan E.; Davis M. M.; Wucherpfennig K. W.; Garcia K. C. Deconstructing the Peptide-MHC Specificity of T Cell Recognition. Cell 2014, 157, 1073–1087. 10.1016/j.cell.2014.03.047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Suzuki T. Regulation of Intestinal Epithelial Permeability by Tight Junctions. Cell. Mol. Life Sci. 2013, 70, 631–659. 10.1007/s00018-012-1070-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. The Principles of Nerve Cell Communication. Alcohol Health Res. World 1997, 21 ( (2), ), 107–108. [PMC free article] [PubMed]
  56. Eming S. A.; Krieg T.; Davidson J. M. Inflammation in Wound Repair: Molecular and Cellular Mechanisms. J. Invest. Dermatol. 2007, 127, 514–525. 10.1038/SJ.JID.5700701. [DOI] [PubMed] [Google Scholar]
  57. Mayor R.; Carmona-Fontaine C. Keeping in Touch with Contact Inhibition of Locomotion. Trends Cell Biol. 2010, 20, 319–328. 10.1016/j.tcb.2010.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Burdick M. M.; McCarty O. J. T.; Jadhav S.; Konstantopoulos K. Cell-Cell Interactions in Inflammation and Cancer Metastasis. IEEE Eng. Med. Biol. Mag. 2001, 20, 86–91. 10.1109/51.932731. [DOI] [PubMed] [Google Scholar]
  59. Sakaguchi S.; Yamaguchi T.; Nomura T.; Ono M. Regulatory T Cells and Immune Tolerance. Cell 2008, 133, 775–787. 10.1016/j.cell.2008.05.009. [DOI] [PubMed] [Google Scholar]
  60. Lu L.; Walker W. A. Pathologic and Physiologic Interactions of Bacteria with the Gastrointestinal Epithelium. Am. J. Clin. Nutr. 2001, 73, 1124S–1130S. 10.1093/ajcn/73.6.1124S. [DOI] [PubMed] [Google Scholar]
  61. Griffith L. G.; Swartz M. A. Capturing Complex 3D Tissue Physiology in Vitro. Nat. Rev. Mol. Cell Biol. 2006, 7, 211–224. 10.1038/nrm1858. [DOI] [PubMed] [Google Scholar]
  62. Belli V.; Guarnieri D.; Biondi M.; della Sala F.; Netti P. A. Dynamics of Nanoparticle Diffusion and Uptake in Three-Dimensional Cell Cultures. Colloids Surf., B 2017, 149, 7–15. 10.1016/j.colsurfb.2016.09.046. [DOI] [PubMed] [Google Scholar]
  63. Hoebe R. A.; Van Oven C. H.; Gadella T. W. J.; Dhonukshe P. B.; Van Noorden C. J. F.; Manders E. M. M. Controlled Light-Exposure Microscopy Reduces Photobleaching and Phototoxicity in Fluorescence Live-Cell Imaging. Nat. Biotechnol. 2007, 25, 249–253. 10.1038/NBT1278. [DOI] [PubMed] [Google Scholar]
  64. Bogdanov A. M.; Kudryavtseva E. I.; Lukyanov K. A. Anti-Fading Media for Live Cell GFP Imaging. PLoS One 2012, 7, e53004 10.1371/JOURNAL.PONE.0053004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Wang K.; Tang Z.; Yang C. J.; Kim Y.; Fang X.; Li W.; Wu Y.; Medley C. D.; Cao Z.; Li J.; Colon P.; Lin H.; Tan W. Molecular Engineering of DNA: Molecular Beacons. Angew. Chem., Int. Ed. 2009, 48, 856–870. 10.1002/ANIE.200800370. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Raisch J.; Darfeuille-Michaud A.; Nguyen H. T. T. Role of MicroRNAs in the Immune System, Inflammation and Cancer. World J. Gastroenterol. 2013, 19, 2985. 10.3748/WJG.V19.I20.2985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Leone K.; Poggiana C.; Zamarchi R. The Interplay between Circulating Tumor Cells and the Immune System: From Immune Escape to Cancer Immunotherapy. Diagnostics 2018, 8, 59 10.3390/diagnostics8030059. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Kimoto M.; Yamashige R.; Matsunaga K.; Yokoyama S.; Hirao I. Generation of High-Affinity DNA Aptamers Using an Expanded Genetic Alphabet. Nat. Biotechnol. 2013, 31, 453–457. 10.1038/nbt.2556. [DOI] [PubMed] [Google Scholar]
  69. Modi S.; M. G. Swetha; Goswami D.; Gupta G. D.; Mayor S.; Krishnan Y. A DNA Nanomachine That Maps Spatial and Temporal PH Changes inside Living Cells. Nat. Nanotechnol. 2009, 4, 325–330. 10.1038/nnano.2009.83. [DOI] [PubMed] [Google Scholar]
  70. Glazier R.; Brockman J. M.; Bartle E.; Mattheyses A. L.; Destaing O.; Salaita K. DNA Mechanotechnology Reveals That Integrin Receptors Apply PN Forces in Podosomes on Fluid Substrates. Nat. Commun. 2019, 10, 4507 10.1038/s41467-019-12304-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Iwaki M.; Wickham S. F.; Ikezaki K.; Yanagida T.; Shih W. M. A Programmable DNA Origami Nanospring That Reveals Force-Induced Adjacent Binding of Myosin VI Heads. Nat. Commun. 2016, 7, 13715 10.1038/ncomms13715. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Saha S.; Prakash V.; Halder S.; Chakraborty K.; Krishnan Y. A PH-Independent DNA Nanodevice for Quantifying Chloride Transport in Organelles of Living Cells. Nat. Nanotechnol. 2015, 10, 645–651. 10.1038/nnano.2015.130. [DOI] [PubMed] [Google Scholar]
  73. Woehrstein J. B.; Strauss M. T.; Ong L. L.; Wei B.; Zhang D. Y.; Jungmann R.; Yin P. Sub–100-Nm Metafluorophores with Digitally Tunable Optical Properties Self-Assembled from DNA. Sci. Adv. 2017, 3, e1602128 10.1126/SCIADV.1602128. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Gómez-García P. A.; Garbacik E. T.; Otterstrom J. J.; Garcia-Parajo M. F.; Lakadamyali M. Excitation-Multiplexed Multicolor Superresolution Imaging with Fm-STORM and Fm-DNA-PAINT. Proc. Natl. Acad. Sci. U.S.A.. 2018, 115, 12991–12996. 10.1073/pnas.1804725115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Boschi F.; de Sanctis F. Overview of the Optical Properties of Fluorescent Nanoparticles for Optical Imaging. Eur. J. Histochem. 2017, 61, 2830 10.4081/EJH.2017.2830. [DOI] [PMC free article] [PubMed] [Google Scholar]
  76. Jaiswal J. K.; Mattoussi H.; Mauro J. M.; Simon S. M. Long-Term Multiple Color Imaging of Live Cells Using Quantum Dot Bioconjugates. Nat. Biotechnol. 2003, 21, 47–51. 10.1038/nbt767. [DOI] [PubMed] [Google Scholar]
  77. Cardoso Dos Santos M.; Algar W. R.; Medintz I. L.; Hildebrandt N. Quantum Dots for Förster Resonance Energy Transfer (FRET). TrAC, Trends Anal. Chem. 2020, 125, 115819 10.1016/j.trac.2020.115819. [DOI] [Google Scholar]
  78. Young E. W. K.; Beebe D. J. Fundamentals of Microfluidic Cell Culture in Controlled Microenvironments. Chem. Soc. Rev. 2010, 39, 1036–1048. 10.1039/b909900j. [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Göpfrich K.; Urban M. J.; Frey C.; Platzman I.; Spatz J. P.; Liu N. Dynamic Actuation of DNA-Assembled Plasmonic Nanostructures in Microfluidic Cell-Sized Compartments. Nano Lett. 2020, 20, 1571–1577. 10.1021/acs.nanolett.9b04217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Halley P. D.; Patton R. A.; Chowdhury A.; Byrd J. C.; Castro C. E. Low-Cost, Simple, and Scalable Self-Assembly of DNA Origami Nanostructures. Nano Res. 2019, 12, 1207–1215. 10.1007/S12274-019-2384-X. [DOI] [Google Scholar]
  81. Stahl E.; Martin T. G.; Praetorius F.; Dietz H. Facile and Scalable Preparation of Pure and Dense DNA Origami Solutions. Angew. Chem. 2014, 126, 12949–12954. 10.1002/ange.201405991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  82. Haughton G.; Arnold L. W.; Bishop G. A.; Mercolino T. J. The CH Series of Murine B Cell Lymphomas: Neoplastic Analogues of Ly-1+ Normal B Cells. Immunol. Rev. 1986, 93, 35–51. 10.1111/j.1600-065x.1986.tb01501.x. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

am2c07971_si_001.pdf (2.7MB, pdf)

Articles from ACS Applied Materials & Interfaces are provided here courtesy of American Chemical Society

RESOURCES