Deep learning-assisted concentration gradient generation for the study of 3D cell cultures in hydrogel beads of varying stiffness.

The study of dose-response relationships underpins analytical biosciences. Droplet microfluidics platforms can automate the generation of microreactors encapsulating varying concentrations of an assay component, providing datasets across a large chemical space in a single experiment. A classical method consists in varying the flow rate of multiple solutions co-flowing into a single microchannel (producing different volume fractions) before encapsulating the contents into water-in-oil droplets. This process can be automated through controlling the pumping elements but lacks the ability to adapt to unpredictable experimental scenarios, often requiring constant human supervision. In this paper, we introduce an image-based, closed-loop control system for assessing and adjusting volume fractions, thereby generating unsupervised, uniform concentration gradients. We trained a shallow convolutional neural network to assess the position of the laminar flow interface between two co-flowing fluids and used this model to adjust flow rates in real-time. We apply the method to generate alginate microbeads in which HEK293FT cells could grow in three dimensions. The stiffnesses ranged from 50 Pa to close to 1 kPa in Young modulus and were encoded with a fluorescent marker. We trained deep learning models based on the YOLOv4 object detector to efficiently detect both microbeads and multicellular spheroids from high-content screening images. This allowed us to map relationships between hydrogel stiffness and multicellular spheroid growth.

Droplet-based methodology for investigating bacterial population dynamics in response to phage exposure.

An alarming rise in antimicrobial resistance worldwide has spurred efforts into the search for alternatives to antibiotic treatments. The use of bacteriophages, bacterial viruses harmless to humans, represents a promising approach with potential to treat bacterial infections (phage therapy). Recent advances in microscopy-based single-cell techniques have allowed researchers to develop new quantitative methodologies for assessing the interactions between bacteria and phages, especially the ability of phages to eradicate bacterial pathogen populations and to modulate growth of both commensal and pathogen populations. Here we combine droplet microfluidics with fluorescence time-lapse microscopy to characterize the growth and lysis dynamics of the bacterium Escherichia coli confined in droplets when challenged with phage. We investigated phages that promote lysis of infected E. coli cells, specifically, a phage species with DNA genome, T7 (Escherichia virus T7) and two phage species with RNA genomes, MS2 (Emesvirus zinderi) and Qß (Qubevirus durum). Our microfluidic trapping device generated and immobilized picoliter-sized droplets, enabling stable imaging of bacterial growth and lysis in a temperature-controlled setup. Temporal information on bacterial population size was recorded for up to 25 h, allowing us to determine growth rates of bacterial populations and helping us uncover the extent and speed of phage infection. In the long-term, the development of novel microfluidic single-cell and population-level approaches will expedite research towards fundamental understanding of the genetic and molecular basis of rapid phage-induced lysis and eco-evolutionary aspects of bacteria-phage dynamics, and ultimately help identify key factors influencing the success of phage therapy.

Phenotyping single-cell motility in microfluidic confinement.

The movement trajectories of organisms serve as dynamic read-outs of their behaviour and physiology. For microorganisms this can be difficult to resolve due to their small size and fast movement. Here, we devise a novel droplet microfluidics assay to encapsulate single micron-sized algae inside closed arenas, enabling ultralong high-speed tracking of the same cell. Comparing two model species - Chlamydomonas reinhardtii (freshwater, 2 cilia), and Pyramimonas octopus (marine, 8 cilia), we detail their highly-stereotyped yet contrasting swimming behaviours and environmental interactions. By measuring the rates and probabilities with which cells transition between a trio of motility states (smooth-forward swimming, quiescence, tumbling or excitable backward swimming), we reconstruct the control network that underlies this gait switching dynamics. A simplified model of cell-roaming in circular confinement reproduces the observed long-term behaviours and spatial fluxes, including novel boundary circulation behaviour. Finally, we establish an assay in which pairs of droplets are fused on demand, one containing a trapped cell with another containing a chemical that perturbs cellular excitability, to reveal how aneural microorganisms adapt their locomotor patterns in real-time.

Deep learning guided image-based droplet sorting for on-demand selection and analysis of single cells and 3D cell cultures.

Uncovering the heterogeneity of cellular populations and multicellular constructs is a long-standing goal in fields ranging from antimicrobial resistance to cancer research. Emerging technology platforms such as droplet microfluidics hold the promise to decipher such heterogeneities at ultra-high-throughput. However, there is a lack of methods able to rapidly identify and isolate single cells or 3D cell cultures. Here we demonstrate that deep neural networks can accurately classify single droplet images in real-time based on the presence and number of micro-objects including single mammalian cells and multicellular spheroids. This approach also enables the identification of specific objects within mixtures of objects of different types and sizes. The training sets for the neural networks consisted of a few hundred images manually picked and augmented to up to thousands of images per training class. Training required less than 10 minutes using a single GPU, and yielded accuracies of over 90% for single mammalian cell identification. Crucially, the same model could be used to classify different types of objects such as polystyrene spheres, polyacrylamide beads and MCF-7 cells. We applied the developed method for the selection of 3D cell cultures generated with Hek293FT cells encapsulated in agarose gel beads, highlighting the potential of the technology for the selection of objects with a high diversity of visual appearances. The real-time sorting of single droplets was in-line with droplet generation and occurred at rates up to 40 per second independently of image size up to 480 × 480 pixels. The presented microfluidic device also enabled storage of sorted droplets to allow for downstream analyses.