Quantifying the deformability of malaria-infected red blood cells using deep learning trained on synthetic cells.

Several hematologic diseases, including malaria, diabetes, and sickle cell anemia, result in a reduced red blood cell deformability. This deformability can be measured using a microfluidic device with channels of varying width. Nevertheless, it is challenging to algorithmically recognize large numbers of red blood cells and quantify their deformability from image data. Deep learning has become the method of choice to handle noisy and complex image data. However, it requires a significant amount of labeled data to train the neural networks. By creating images of cells and mimicking noise and plasticity in those images, we generate synthetic data to train a network to detect and segment red blood cells from video-recordings, without the need for manually annotated labels. Using this new method, we uncover significant differences between the deformability of RBCs infected with different strains of Plasmodium falciparum, providing clues to the variation in virulence of these strains.

The Effect of Geometry and TGF-ß Signaling on Tumor Cell Migration from Free-Standing Microtissues.

Recapitulation of 3D multicellular tissues in vitro is of great interest to the field of tumor biology to study the integrated effect of local biochemical and biophysical signals on tumor cell migration and invasion. However, most microengineered tissues and spheroids are unable to recapitulate in vitro the complexities of 3D geometries found in vivo. Here, lithographically defined degradable alginate microniches are presented to produce free-standing tumor microtissues, with precisely controlled geometry, high viability, and allowing for high cell proliferation. The role of microtissue geometry and TGF-ß signaling in tumor cell migration is further investigated. TGF-ß is found to induce the expression of p-myosin II, vimentin, and YAP/TAZ nuclear localization at the periphery of the microtissue, where enhanced nuclear stiffness and orientation are also observed. Upon embedding in a collagen matrix, microtissues treated with TGF-ß maintain their geometric integrity, possibly due to the higher cell tension observed around the periphery. In contrast, cells in microtissues not treated with TGF-ß are highly mobile and invade the surrounding matrix rapidly, with the initial migration strongly dependent on the local geometry. The microtissues presented here are promising model systems for studying the influence of biophysical properties and soluble factors on tumor cell migration.

Microfabricated Gaps Reveal the Effect of Geometrical Control in Wound Healing.

The geometry (size and shape) of gaps is a key determinant in controlling gap closure during wound healing. However, conventional methods for creating gaps result in un-defined geometries and poorly characterized conditions (cell death factors and cell debris), which can influence the gap closure process. To overcome these limitations, a novel method to create well-defined geometrical gaps is developed. First, smooth muscle cells (SMCs) are seeded in variously shaped micro-containers made out of hyaluronic acid hydrogels. Cell proliferation and cell tension induce fibrous collagen production by SMCs predominantly around the edges of the micro-containers. Upon removal of SMCs, the selectively deposited collagen results in micro-containers with cell-adhesive regions along the edges and walls. Fibroblasts are seeded in these micro-containers, and upon attaching and spreading, they naturally form gaps with different geometries. The rapid proliferation of fibroblasts from the edge results in filling and closure of the gaps. It is demonstrated that gap closure rate as well as closure mechanism is strongly influenced by geometrical features, which points to an important role for cellular tension and cell proliferation in gap closure.

Probing single-cell metabolism reveals prognostic value of highly metabolically active circulating stromal cells in prostate cancer.

Despite their important role in metastatic disease, no general method to detect circulating stromal cells (CStCs) exists. Here, we present the Metabolic Assay-Chip (MA-Chip) as a label-free, droplet-based microfluidic approach allowing single-cell extracellular pH measurement for the detection and isolation of highly metabolically active cells (hm-cells) from the tumor microenvironment. Single-cell mRNA-sequencing analysis of the hm-cells from metastatic prostate cancer patients revealed that approximately 10% were canonical EpCAM+ hm-CTCs, 3% were EpCAM- hm-CTCs with up-regulation of prostate-related genes, and 87% were hm-CStCs with profiles characteristic for cancer-associated fibroblasts, mesenchymal stem cells, and endothelial cells. Kaplan-Meier analysis shows that metastatic prostate cancer patients with more than five hm-cells have a significantly poorer survival probability than those with zero to five hm-cells. Thus, prevalence of hm-cells is a prognosticator of poor outcome in prostate cancer, and a potentially predictive and therapy response biomarker for agents cotargeting stromal components and preventing epithelial-to-mesenchymal transition.

Dysmetabolic Circulating Tumor Cells Are Prognostic in Metastatic Breast Cancer.

Circulating tumor cells (CTCs) belong to a heterogeneous pool of rare cells, and a unequivocal phenotypic definition of CTC is lacking. Here, we present a definition of metabolically-altered CTC (MBA-CTCs) as CD45-negative cells with an increased extracellular acidification rate, detected with a single-cell droplet microfluidic technique. We tested the prognostic value of MBA-CTCs in 31 metastatic breast cancer patients before starting a new systemic therapy (T0) and 3-4 weeks after (T1), comparing results with a parallel FDA-approved CellSearch (CS) approach. An increased level of MBA-CTCs was associated with: i) a shorter median PFS pre-therapy (123 days vs. 306; p < 0.0001) and during therapy (139 vs. 266 days; p = 0.0009); ii) a worse OS pre-therapy (p = 0.0003, 82% survival vs. 20%) and during therapy (p = 0.0301, 67% survival vs. 38%); iii) good agreement with therapy response (kappa = 0.685). The trend of MBA-CTCs over time (combining data at T0 and T1) added information with respect to separate evaluation of T0 and T1. The combined results of the two assays (MBA and CS) increased stratification accuracy, while correlation between MBA and CS was not significant, suggesting that the two assays are detecting different CTC subsets. In conclusion, this study suggests that MBA allows detection of both EpCAM-negative and EpCAM-positive, viable and label-free CTCs, which provide clinical information apparently equivalent and complementary to CS. A further validation of proposed method and cut-offs is needed in a larger, separate study.

Cellular Volume and Matrix Stiffness Direct Stem Cell Behavior in a 3D Microniche.

The central question addressed in this study is whether cells with different sizes have different responses to matrix stiffness. We used methacrylated hyaluronic acid (MeHA) hydrogels as the matrix to prepare an in vitro 3D microniche in which the single stem cell volume and matrix stiffness can be altered independently from each other. This simple approach enabled us to decouple the effects of matrix stiffness and cell volume in 3D microenvironments. Human mesenchymal stem cells (hMSCs) were cultured in individual 3D microniches with different volumes (2800, 3600, and 6000 µm3) and stiffnesses (5, 12, and 23 kPa). We demonstrated that cell volume affected the cellular response to matrix stiffness. When cells had an optimal volume, cells could form clear stress fibers and focal adhesions on soft, intermediate, or stiff matrix. In small cells, stress fiber formation and YAP/TAZ localization were not affected by stiffness. This study highlights the importance of considering cellular volume and substrate stiffness as important cues governing cell-matrix interactions.

Single-cell analysis reveals that stochasticity and paracrine signaling control interferon-alpha production by plasmacytoid dendritic cells.

Type I interferon (IFN) is a key driver of immunity to infections and cancer. Plasmacytoid dendritic cells (pDCs) are uniquely equipped to produce large quantities of type I IFN but the mechanisms that control this process are poorly understood. Here we report on a droplet-based microfluidic platform to investigate type I IFN production in human pDCs at the single-cell level. We show that type I IFN but not TNFα production is limited to a small subpopulation of individually stimulated pDCs and controlled by stochastic gene regulation. Combining single-cell cytokine analysis with single-cell RNA-seq profiling reveals no evidence for a pre-existing subset of type I IFN-producing pDCs. By modulating the droplet microenvironment, we demonstrate that vigorous pDC population responses are driven by a type I IFN amplification loop. Our study highlights the significance of stochastic gene regulation and suggests strategies to dissect the characteristics of immune responses at the single-cell level.

Microfluidic droplets content classification and analysis through convolutional neural networks in a liquid biopsy workflow.

In a recent paper we presented an innovative method of liquid biopsy, for the detection of circulating tumor cells (CTC) in the peripheral blood. Using microfluidics, CTC are individually encapsulated in water-in-oil droplets and selected by their increased rate of extracellular acidification (ECAR). During the analysis, empty or debris-containing droplets are discarded manually by screening images of positive droplets, increasing the operator-dependency and time-consumption of the assay. In this work, we addressed the limitations of the current method integrating computer vision techniques in the analysis. We implemented an automatic classification of droplets using convolutional neural networks, correctly classifying more than 96% of droplets. A second limitation of the technique is that ECAR is computed using an average droplet volume, without considering small variations in extracellular volume which can occur due to the normal variability in the size of the droplets or cells. Here, with the use of neural networks for object detection, we segmented the images of droplets and cells to measure their relative volumes, correcting over- or under-estimation of ECAR, which was present up to 20%. Finally, we evaluated whether droplet images contained additional information. We preliminarily gave a proof-of-concept demonstration showing that white blood cells expression of CD45 can be predicted with 82.9% accuracy, based on bright-field cell images alone. Then, we applied the method to classify acid droplets as coming from metastatic breast cancer patients or healthy donors, obtaining an accuracy of 90.2%.

3D microniches reveal the importance of cell size and shape.

Geometrical cues have been shown to alter gene expression and differentiation on 2D substrates. However, little is known about how geometrical cues affect cell function in 3D. One major reason for this lack of understanding is rooted in the difficulties of controlling cell geometry in a complex 3D setting and for long periods of culture. Here, we present a robust method to control cell volume and shape of individual human mesenchymal stem cells (hMSCs) inside 3D microniches with a range of different geometries (e.g., cylinder, triangular prism, cubic, and cuboid). We find that the actin filaments, focal adhesions, nuclear shape, YAP/TAZ localization, cell contractility, nuclear accumulation of histone deacetylase 3, and lineage selection are all sensitive to cell volume. Our 3D microniches enable fundamental studies on the impact of biophysical cues on cell fate, and have potential applications in investigating how multicellular architectures organize within geometrically well-defined 3D spaces.

A Method for Detecting Circulating Tumor Cells Based on the Measurement of Single-Cell Metabolism in Droplet-Based Microfluidics.

The number of circulating tumor cells (CTCs) in blood is strongly correlated with the progress of metastatic cancer. Current methods to detect CTCs are based on immunostaining or discrimination of physical properties. Herein, a label-free method is presented exploiting the abnormal metabolic behavior of cancer cells. A single-cell analysis technique is used to measure the secretion of acid from individual living tumor cells compartmentalized in microfluidically prepared, monodisperse, picoliter (pL) droplets. As few as 10 tumor cells can be detected in a background of 200 000 white blood cells and proof-of-concept data is shown on the detection of CTCs in the blood of metastatic patients.

Quantitative Single-Cell mRNA Analysis in Hydrogel Beads.

In recent years, technologies capable of analyzing single cells have emerged that are transforming many fields of biological research. Herein we report how DNA-functionalized hydrogel beads can serve as a matrix to capture mRNA from lysed single cells. mRNA quantification free of pre-amplification bias is ensured by using padlock probes and rolling circle amplification followed by hybridization with fluorescent probes. The number of transcripts in individual cells is assessed by simply counting fluorescent dots inside gel beads. The method extends the potential of existing techniques and provides a general platform for capturing molecules of interest from single cells.

Ultrasensitivity by molecular titration in spatially propagating enzymatic reactions.

Delineating design principles of biological systems by reconstitution of purified components offers a platform to gauge the influence of critical physicochemical parameters on minimal biological systems of reduced complexity. Here we unravel the effect of strong reversible inhibitors on the spatiotemporal propagation of enzymatic reactions in a confined environment in vitro. We use micropatterned, enzyme-laden agarose gels which are stamped on polyacrylamide films containing immobilized substrates and reversible inhibitors. Quantitative fluorescence imaging combined with detailed numerical simulations of the reaction-diffusion process reveal that a shallow gradient of enzyme is converted into a steep product gradient by addition of strong inhibitors, consistent with a mathematical model of molecular titration. The results confirm that ultrasensitive and threshold effects at the molecular level can convert a graded input signal to a steep spatial response at macroscopic length scales.

Enhanced transcription rates in membrane-free protocells formed by coacervation of cell lysate.

Liquid-liquid phase transitions in complex mixtures of proteins and other molecules produce crowded compartments supporting in vitro transcription and translation. We developed a method based on picoliter water-in-oil droplets to induce coacervation in Escherichia coli cell lysate and follow gene expression under crowded and noncrowded conditions. Coacervation creates an artificial cell-like environment in which the rate of mRNA production is increased significantly. Fits to the measured transcription rates show a two orders of magnitude larger binding constant between DNA and T7 RNA polymerase, and five to six times larger rate constant for transcription in crowded environments, strikingly similar to in vivo rates. The effect of crowding on interactions and kinetics of the fundamental machinery of gene expression has a direct impact on our understanding of biochemical networks in vivo. Moreover, our results show the intrinsic potential of cellular components to facilitate macromolecular organization into membrane-free compartments by phase separation.

Parallel separations using capillary electrophoresis on a multilane microchip with multiplexed laser-induced fluorescence detection.

Parallel separations using CE on a multilane microchip with multiplexed LIF detection is demonstrated. The detection system was developed to simultaneously record data on all channels using an expanded laser beam for excitation, a camera lens to capture emission, and a CCD camera for detection. The detection system enables monitoring of each channel continuously and distinguishing individual lanes without significant crosstalk between adjacent lanes. Multiple analytes can be determined in parallel lanes within a single microchip in a single run, leading to increased sample throughput. The pK(a) determination of small molecule analytes is demonstrated with the multilane microchip.

Electrokinetic control of fluid transport in gold-coated nanocapillary array membranes in hybrid nanofluidic-microfluidic devices.

The introduction of metallic elements into microfluidic devices that support electrokinetic transport creates several fundamental issues relative to the high conductivity of the metal, which can act as a current shunt, causing profound effects on the transport process. Here we examine the use of Au-coated nanocapillary array membranes (Au NCAMs) as electrically addressable fluid control elements in multi-layer microfluidic architectures. Three alternative methods for fluid injection across Au NCAMs are presented: electrokinetic injection across NCAMs with Au coated on one side (asymmetric NCAM), electrokinetic injection across NCAMs with an embedded Au layer (symmetric NCAM), and field-free electroosmotic flow (EOF) pumping across either type of Au NCAM. Injection efficiency across asymmetric NCAMs depends on the orientation of the asymmetric membrane relative to the driving potential. Efficient injections are enabled when the Au coating is on the receiving side of the membrane, however, some distortion of the injected volume element is observed, especially with large injection potentials. These results for asymmetric membranes agree qualitatively with two-dimensional numerical simulations of injections across a single slit pore, which suggest that the direction-selective transport behavior is related to electrophoretic transport of the anionic fluorescein probe. Reproducible, high quality injections are also achieved in symmetric Au NCAMs having an embedded gold nanoband region within the nanopores. Nanoband Au NCAMs are excellent candidates for a range of applications, including high efficiency electrochemical sensing, electrochemically catalyzed conversion or pretreatment and label free sensing utilizing extraordinary optical transmission. EOF pumping could be an alternative to electrokinetic injections in some applications, however, this approach is only useful for relatively large pore sizes (>400 nm) and presents considerably worse sample spreading via Taylor dispersion.

Nanofluidics in chemical analysis.

Nanofluidic architectures and devices have already had a major impact on forefront problems in chemical analysis, especially those involving mass-limited samples. This critical review begins with a discussion of the fundamental flow physics that distinguishes nanoscale structures from their larger microscale analogs, especially the concentration polarization that develops at nanofluidic/microfluidic interfaces. Chemical manipulations in nanopores include nanopore-mediated separations, microsensors, especially resistive-pulse sensing of biomacromolecules, fluidic circuit analogs and single molecule measurements. Coupling nanofluidic structures to three-dimensional microfluidic networks is especially powerful and results in applications in sample preconcentration, nanofluidic injection/collection and fast diffusive mixing (160 references).

Simultaneous multiselective spectroelectrochemical sensing of the interaction between protein and its ligand using the redox dye Nile blue as a label.

A new binding assay for a protein and its ligand based on a spectroelectrochemical method was demonstrated using avidin-biotin and 17beta-estradiol-antiestradiol antibody. The sensor consists of a selective film coated on an optically transparent electrode (OTE) consisting of indium tin oxide (ITO). Attenuated total reflection (ATR) was used for optical detection. The binding event of the ligand to the protein was detected using the ligand labeled with the electroactive dye Nile blue (NB). The spectroelectrochemical behaviors of NB and the labeled ligand were investigated using various ion-exchange films, such as perfluorosulfonated ionomer (Nafion), Nafion-silica, poly(acrylic acid) (PAA)-silica, poly(styrenesulfonic acid) (PSSA)-silica, and heparin-silica films, which were spin-coated on the ITO electrode. The optical signal was monitored to follow the accumulation of labeled ligand in the film and its electrochemical modulation. The signal from the labeled ligand possesses three modes of selectivity based on charge-selective partitioning, the chosen electrolysis potential, and the particular wavelength for measuring absorbance. The interaction between the labeled ligand and its protein was observed by the decrease in the changes of optical response of the labeled ligand, indicating the specific binding of labeled ligand to the protein.

Electrokinetically driven fluidic transport in integrated three-dimensional microfluidic devices incorporating gold-coated nanocapillary array membranes.

Electrokinetically driven fluid transport was evaluated within three-dimensional hybrid nanofluidic-microfluidic devices incorporating Au-coated nanocapillary array membranes (NCAMs). Gold NCAMs, prepared by electroless gold deposition on polymeric track-etched membranes, were susceptible to gas bubble formation if the interfacial potential difference exceeded approximately 2 V along the length of the gold region. Gold membranes were etched to yield 250 microm wide coated regions that overlap the intersection of two orthogonal microfluidic channels in order to minimize gas evolution. The kinetics of electrolysis of water at the opposing ends of the gold region was modeled and found to be in satisfactory agreement with experimental measurements of the onset of gas bubble formation. Conditions to achieve electrokinetic injection across Au-coated NCAMs were identified, with significant reproducible injections being possible for NCAMs modified with this relatively thin gold stripe. Continuous gold films led to suppressed injections and to a variety of ion enrichment/depletion effects in the microfluidic source channel. The suppression of injections was understood through finite element modeling which revealed the presence of a significant electrophoretic velocity component in opposition to electroosmotic flow at the edge of the Au-dielectric regions.

Characterization and performance of injection molded poly(methylmethacrylate) microchips for capillary electrophoresis.

Injection molded poly(methylmethacrylate) (IM-PMMA), chips were evaluated as potential candidates for capillary electrophoresis disposable chip applications. Mass production and usage of plastic microchips depends on chip-to-chip reproducibility and on analysis accuracy. Several important properties of IM-PMMA chips were considered: fabrication quality evaluated by environmental scanning electron microscope imaging, surface quality measurements, selected thermal/electrical properties as indicated by measurement of the current versus applied voltage (I-V) characteristic and the influence of channel surface treatments. Electroosmotic flow was also evaluated for untreated and O2 reactive ion etching (RIE) treated surface microchips. The performance characteristics of single lane plastic microchip capillary electrophoresis (MCE) separations were evaluated using a mixture of two dyes-fluorescein (FL) and fluorescein isothiocyanate (FITC). To overcome non-wettability of the native IM-PMMA surface, a modifier, polyethylene oxide was added to the buffer as a dynamic coating. Chip performance reproducibility was studied for chips with and without surface modification via the process of RIE with O2 and by varying the hole position for the reservoir in the cover plate or on the pattern side of the chip. Additionally, the importance of reconditioning steps to achieve optimal performance reproducibility was also examined. It was found that more reproducible quantitative results were obtained when normalized values of migration time, peak area and peak height of FL and FITC were used instead of actual measured parameters.

The autofluorescence of plastic materials and chips measured under laser irradiation.

Plastic materials have the potential to substitute for glass substrates used in microfluidic and microTAS systems adding flexibility in materials' choices. Optical quality plastic materials with a low autofluorescence are crucial for optimal detection by fluorescence and laser induced fluorescence techniques. This paper summarizes a series of optical investigations on commercially available plastic chip materials (PMMA, COC, PC, PDMS) and chips made from those materials. Intrinsic optical constants of plastic materials-refractive index for bulk materials-determined by spectroscopic ellipsometry and transmission spectroscopy in the visible range are presented. The laser-induced autofluorescence of materials and chips was assessed at four laser wavelengths, namely, 403, 488, 532 and 633 nm. Considerable bleaching of the autofluorescence was observed under continuous laser illumination. Overall, the longer wavelength laser excitation sources yielded less autofluorescence. PDMS exhibited the least autofluorescence and was comparable to BoroFloat glass. In all cases, chips exhibited slightly higher autofluorescence than the raw plastic materials from which they had been made.