The de novo construction of synthetic cells using non-living components is the approach used by bottom-up synthetic biologists. Building minimal cells and controlling each aspect of their design not only gives us the opportunity to understand real cells and their origins, but also provides alternative routes to novel biotechnologies. Giant unilamellar vesicles (GUVs) are used extensively as scaffolds to construct synthetic cells owing to their compatibility with existing biological components. Microfluidic-based approaches for GUV production show great potential for encapsulating large biomolecules required for mimicking life-like functions (Yandrapalli et al. Micromachines, 2020; Love et al. Angew Chemie, 2020) and recent advances have also given researchers access to high-throughput on-line analysis (Robinson, Adv. Biosyst. 2019; Yandrapalli and Robinson, Lab Chip, 2019; Bhatia, Soft Matter, 2020). Here we present our latest results on how microfluidics can further aid in bottom-up synthetic biology.

First, we will show a microfluidic design that is able to produce surfactant-free pure lipid GUVs in a high-throughput manner (Yandrapalli et al. Commun Chem, 2021). The major advancement is that the lipid membranes are produced in the absence of block co-polymers or surfactants that can affect their biocompatibility (as is commonly overlooked). The design can produce homogenously sized GUVs with tuneable diameters from 10 to 130 µm. Encapsulation is uniform and we show that the membranes are oil-free by multiple tests including measuring the diffusion of lipids via FRAP measurements. Next, we present how we modified this device to encapsulate two sub-populations of nano-sized vesicles for the purpose of establishing enzymatic cascade reactions across membrane-bound compartments, therefore mimicking eukaryotic cells (Shetty et al. doi:10.26434/CHEMRXIV.14593518.V1.). The final synthetic cell comprises three coupled enzymatic reactions, which propagate across three separate compartments in a specific direction due to selective membranes pores. Not only does microfluidics provide a high degree of control over the intra-vesicular conditions such as enzyme concentrations, buffers, and the number of inner compartments, but the monodispersity of our synthetic cells allows us to directly compare the effects that compartmentalisation has on the biochemical reaction rates.  This work demonstrates the effectiveness of microfluidics for the bottom-up assembly of synthetic cell constructs.

Microfluidic double emulsions have been widely applied to construct cell-mimicking chassis. However, the excess oil in the double emulsion shells often lead to instability of the membrane structures and can be incompatible with the contents of the vesicle or of the interface, such as membrane proteins. This is even more prominent for asymmetric bilayers, which are required to support the vesicles' biological functions.

Herein, a new strategy is introduced to construct giant unilamellar vesicles with asymmetric membranes from ultrathin-shell double emulsions, based upon shell rupture. Two types of reconstruction processes, linked to the oil gathering in an 'oil cap within the membrane, were observed depending upon the oil cap either being removed or remaining.

Calcein leaking and fluorescence quenching assays were used to evaluate the unilamellarity and asymmetry of lipid bilayer structures. We subsequently engineered polymer-lipid hybrid vesicles to demonstrate the versatility of the approach, and show our ability to construct an asymmetric membrane system for mimicking cells, with the potential application for controllable matrix release.

There have been numerous innovations within the field of bottom-up synthetic biology over the past decade. By manipulating the self-assembly of amphiphilic molecules in aqueous media, researchers have fabricated cell mimetics with applications ranging from biosensing to drug delivery. Unfortunately, the incorporation of higher-order complexity into these cell-like chassis is non-trivial for commonly used ‘’bulk’’ approaches such as extrusion and tip sonication. This is mainly due to the poor control offered by these methods over the conditions of self-assembly, including temperature, pressure, and concentration. Several impressive microfluidic platforms have been established to address this shortfall, capable of generating micrometre scale compartmentalised vesicles predominantly via droplet mediated templating. For the pharmaceutical industry however, structures greater than 200 nanometres can pose significant problems for downstream applications. The microfluidic generation of higher-order assemblies on the nanoscale remains relatively under-researched despite these size constraints, and thus presents an exciting new opportunity for scientific investigation. I will introduce some recent work from our group where we harness microfluidic techniques to produce a variety of complex membraneous nanostructures, including cubosomes and multilamellar vesicles. I will also discuss some future research directions for the microfluidic generation of biomimetic nanoparticles and propose some potential applications.  

Synthetic biology aims to engineer new organisms using engineering principles by modifying their genetic makeup consisting of DNA, RNA, proteins and metabolites.  Proteins are of significant interest since they are the machinery of life. Their interactions  induce control over the general behaviour of the cells and organisms.  Thus, the study of protein-protein interaction (PPI) has broad applications in the pharmaceutical, food and agriculture industries and in the development of environmentally friendly renewables.  There are many methods to study PPIs, some of which are yeast two-hybrid systems and affinity purification coupled to mass spectrometry.  These are generally coupled with other systems – e.g., NMR, isothermal titration, and colorimetric methodologies to study their kinetics and interactions. Yet, these techniques are slow, which are not ideal for studying weaker interactions. Furthermore, studying PPIs in a cell environment increases the complexity given the requirement to culture cells and to provide maintenance to ensure their high viability.  In our work, we present the first cell-free system to study PPIs in a microfluidic device.  We use a D² (droplet-digital) microfluidic device since microfluidics provide conditions that are close to the native state of the protein and use a droplet-digital system since it provides the advantages of control and throughput. Our device consists of several components – (1) on demand electrofusion of droplets of different proteins and inhibitors, (2) on demand picoinjection of different concentrations of fluorescence substrate, (3) kinetics analysis of enzymes interactions using fluorescence readout. As proof of concept, we are studying the interaction of different cellulases enzymes (mainly cellobiohydrolases and betaglucosidases) from different thermophilic organisms in droplets using cell free system. Our goal is to test different combinations of those enzymes to improve cellulose breakdown and study how those combinations could affect the hydrolysis under ionic liquid environment.  We anticipate that this device could be used for other applications related to PPI – e.g., studying how new drugs could affect target proteins related to neurodegenerative diseases (e.g., Alzheimer), develop novel biocontrol agents (e.g. amylase inhibitors for fungal control in plants) and discovering new metabolic pathways to produce biofuels. 

Artificial cells are cell-like entities constructed from the bottom-up using molecular building blocks, which resemble real biological cells in form and function. They are used both as simplified models of biological cells, and as smart soft-matter microdevices with a range of potential applications in industrial and clinical biotechnology. However, due to the lack of cellular infrastructure and absence of spatial organisation, the capabilities of artificial cells have not matched their biological counterparts. In this talk, I will present work from our group which aims to address this gap.

We have developed a series of microfluidic technologies that allow us (i) to build cells of defined sizes, lamellarity, level of compartmentalisation, and internal architecture and (ii) to manipulate them in order to recapitulate various membranous motifs found in biology (e.g. double membranes, gap junctions, and tunnelling nanotubes). By deploying molecular bioengineering principles and by transplanting cellular machinery, we can programme our cells to possess the behaviours that are the hallmarks of life: communication, signalling, motility, sense/response, and biosynthesis. To further enhance artificial cell functionality, we use living cells and organelles as discrete functional modules that are embedded inside artificial cells. The resultant ‘hybrid’ cells are composed of a synthetic host and a living organelle, which enjoy a mutually beneficial relationship, and can be considered a novel living/synthetic cellular bionic material.

The vast surface area of the distal lung, sparsely populated with resident immune cells, is the often the site of emerging (COVID-19) or established (Tuberculosis) infection diseases. For example, severe COVID-19 result in microvascular thrombosis in the lung and systemic endothelialitis, but the underlying dynamics of damage to the vasculature and whether it is a direct consequence of endothelial infection or an indirect consequence of immune cell mediated cytokine storms is still unclear. In contrast, the minimum infectious dose in tuberculosis can be as low as one, which makes the early stages of infection difficult to study even in animal models. Yet heterogenous outcomes in the early stages of infection significantly alter the course of infection and may explain why the proportion of exposed individuals who develop clinical tuberculosis is low. In my talk, I will describe efforts to develop the lung-on-chip as model systems for respiratory infectious diseases and showcase the ability of these systems to enable new experiments that would not be possible in vivo.

In the case of COVID-19, we used a vascularised human lung-on-chip model [1] where we find that rapid infection of the underlying endothelial layer leads to the generation of clusters of endothelial cells with low or no CD31 expression, a progressive loss of endothelial barrier integrity, and a pro-coagulatory microenvironment. These morphological changes do not occur if endothelial cells are exposed to SARS-CoV-2 apically. Viral RNA persisted in individual cells, which generated a response skewed towards NF-KB mediated inflammation and an antiviral interferon response which was transient in epithelial cells but persistent in endothelial cells. Perfusion with Tocilizumab, an inhibitor of trans IL-6 signalling slows the loss of barrier integrity but does not prevent the formation of endothelial cell clusters with reduced CD31 expression. Further work is ongoing to understand the role of cell-cell communication in the inflammation observed.

Prior to this, we used a a murine lung-on-chip model for tuberculosis to recreate a level of pulmonary surfactant deficiency that would be lethal in vivo [2]. Using long-term time-lapse imaging, we measured the growth rates of small bacterial microcolonies in the induvial host cells at the air-liquid interface and showed that pulmonary surfactant secreted by epithelial cells dramatically reduced bacterial growth in both epithelial cells and macrophages, whereas deficient levels of surfactant led to unimpeded growth. These insights suggest a greater role for alveolar epithelial cells in early tuberculosis than previously assumed.

Integrin α_vβ₃, often referred to as the vitronectin receptor, is a cell-ECM adhesion protein highly expressed on activated endothelial cells found in newly formed blood vessels and, as a result, has been found to be crucial for angiogenesis. The constant pro-angiogenic signaling across the tumour microenvironment results in tumour vasculature expressing increased levels of α_vβ₃ compared to resting endothelial cells in normal tissue, therefore presenting α_vβ₃ as a potential site for the targeting of anti-cancer therapeutics ¹.  Inhibition of α_vβ₃ signaling with RGD peptides or monoclonal antibodies has been observed to induce apoptosis in newly formed tumour vessels and reduce overall tumour growth however, little investigation into exploiting α_vβ₃ upregulation as a drug carrier target has been performed ^2,3. This study uses self-assembled, perfusable vasculature networks grown within microfluidic devices to investigate the effectiveness of α_vβ₃ targeting ⁴. Tumour vasculature was recreated by conditioning networks with media taken from tumour cell cultures (TCM) which resulted in greater rates of angiogenesis compared to healthy networks. Flow cytometry revealed that growing endothelial cells in TCM resulted in increased rates of α_vβ₃ expression - as what would be expected in tumour vasculature. α_vβ₃-targeted liposomes were perfused through the networks and their accumulation within the vasculature quantified. Results observed increased accumulation in tumour conditioned networks compared to healthy networks – indicating that α_vβ₃ may be a suitable target for drug-loaded liposomes and allow for increased local drug delivery to the tumour site.

Plant roots grow in highly heterogeneous environments where they need to locally adapt their architecture to optimally absorb nutrients and water, interact with pathogenic and symbiotic microbes, and cope with diverse abiotic stress conditions. We investigate plastic adaptation of root development and intercellular communication in response to changing environmental conditions. To enable microscopic studies of root-environment interactions, we are developing RootChips. This technology integrates microfluidic perfusion and imaging platforms that combine on-chip root cultivation, microscopic access and precise control over the root microenvironment. Here I will summarize recent developments in the field to control the root microenvironment using microfluidics. The production of microfluidic devices using 3D printing technology will be discussed. I will also present our latest efforts to investigate how roots perceive environmental complexity, respond to local physical stimuli and develop under asymmetric conditions.

Biologically active chemistry is mostly concerned with the design of compounds able to translocate biological membranes; complex organelles of interwoven lipids, sugars and proteins. Historically, much work has focused on simple diffusion across the lipid matrix as a major transport route. Current in vitro assays to delineate a small molecules structural dependency on different transport routes have been limited, where typically the model interface is oversimplified. Droplet interface bilayers (DIBs) are formed at the contact of two lipid monolayer coated water in oil droplets and have shown vast potential within biomimetic synthetic biology. DIB technology is highly applicable within microfluidics and bespoke chip design, in particular as a chassis for the study of permeant translocation across biomimetic membranes. Up till now, permeation studies within DIBs have mostly utilised fluorescent microscopy, thus limiting the application of DIB technology in divergent physicochemical space typical of small molecule drug discovery. Here, we address this technological bottleneck by presenting a novel, label free approach enabled by custom chip design principles. Our method is highly implementable in multiple applications and can perform in situ measurement with a data interval as low as 0.02 s. Our platform has enabled us to undertake structure - function relationship studies across compositionally varied bilayers unlocking the complementarity of DIB technology to current widely used assays.

Droplet interface bilayers (DIBs) have recently started to be used as human-mimetic artificial cell membranes.  DIBs are bilayer sections created at the interface of two aqueous droplets, such that one droplet can be used as a donor compartment and the other as an acceptor compartment for the quantification of molecular transport across the artificial cell membrane. However, synthetic phospholipids are overwhelmingly used to create DIBs instead of naturally derived phospholipids, even though the diverse distribution of phospholipids in the latter is more biomimetic. We present the first systematic study of the role of temperature in DIB formation, which shows that the temperature at which DIBs are formed is a key parameter for the formation of DIBs using naturally derived phospholipids in a microfluidic platform.  The phospholipids that are most abundant in mammalian cell membranes (phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS), and phosphatidylinositol (PI)) only form DIBs when the temperature is above the phase transition temperature (Tm).  Similarly, DIB formation usually only occurs above the highest Tm of a single phospholipid in a bespoke formulation. We show a new phenomenon wherein the DIB “melts” without disintegrating for bilayers formed predominantly of phospholipids that occupy cylindrical spaces. We also demonstrate differences in DIB formation rates as well as permeability of biomimetic membranes. Given the difficulties associated with making DIBs using naturally derived phospholipids, we anticipate this work will illuminate the role of phospholipid phase transition in mono- and bilayer formation and lay the foundation for DIBs to be used as human-mimetic artificial cell membranes.

The miniaturization of manufacturing processes using microfluidics (MF) has been increasingly used in chemistry, biology and engineering, due to advantages in terms of cost, size and control. The fabrication of small, uniformly-sized polymer microspheres (MS) using MF has been done to a small extent using materials such as glass capillaries, polytetrafluoroethylene (PTFE), and once using polydimethylsiloxane (PDMS). Here we show a new MF platform for the manufacture of polycaprolactone (PCL) MS for drug delivery using PDMS. This platform uses a flow-focusing geometry for droplet formation and takes advantage of a highly viscous PCL solution in dichloromethane (DCM), as the inner phase (IP), and a 2% polyvinyl alcohol (PVA) solution in water, as the outer phase (OP), to achieve a jetting regime. This forms MS of sizes around 30 micrometers, which are 8 times smaller than prior MF methods using PDMS, and twice as small as methods using other materials. Due to the hydrophobicity of PDMS, surface chemistry modification was necessary to make the channels in the device hydrophilic. High-speed imaging and Scanning Electron Microscopy (SEM) were used for the measurement of MS size. A 50% increase in the flow rate of the OP reduced the size of the MS by 30%; and a 5-fold increase of the OP flow rate compared to the IP flow rare reduces the MS size by 57%. Our new microfluidic design increases the throughput of uniformly-sized MS and reduces by a factor of 4 the starting materials and the fabrication time, when compared to batch production methods. We also include data to show how these MS interact with artificial cell membranes, using a platform of droplet interface bilayers (DIBs). These DIBs are lipid bilayers formed by with droplets containing a buffer system similar to real cells. This platform allows us to model a biological system, studying transport and quantification of drug delivery across artificial membranes, as well that allow us to have a new platform for the development of new materials for drug releasing particles.

Encapsulation of single cells in monodisperse water-in-oil microdroplets offers powerful means to perform quantitative biological studies on a single cell basis, within large cell populations. For example, to understand the interaction of living cells with their environments, we need to look into what guides individual behaviour and movement. For this purpose, we have developed a versatile image-based sorter that can deterministically trap single cells in microdroplets [1]. We have used recent advances in deep learning for real-time object detection to provide rapid and acurate classifications for micro-swimmers of varying appearances. The tool was applied to the study of motility for unicellular microflagellates and compare two species of green algae, C. reinhardtii - a freshwater biflagellate, and P. octopus - a marine octoflagellate, to reveal their stereotyped behaviours and emergence of distinct motility macrostates. In such mobile single-cells, behavioural responses can be tracked with high-speed imaging, using movement as a dynamic read-out of behaviour and physiology. These comprehensive datasets will allow us to query and catalogue single-cell behavioural actions at unprecedented resolution.

Fungal-bacterial interactions (BFIs) are highly prevalent in nature and are critically important in a variety of fields such as agriculture, biotechnology, and medicine [1]. An important feature of BFIs is the ability of bacteria to control and exploit their eukaryotic hosts [2]. The most intriguing case of bacteria controlling host reproduction is the endosymbiosis between the zygomycete Rhizopus microsporus and its bacterial endosymbiont Burkholderia rhizoxinica [3, 4]. In this agriculturally relevant symbiosis [5], host reproduction through spores relies exclusively on the presence of endobacteria [4]. However, there is a considerable lack of knowledge about the molecular basis of this interaction. Through a combination of genomic and functional studies, we show that B. rhizoxinica transcription activator-like effectors (BATs) are essential for the establishment of a stable symbiosis. Utilising novel microfluidics devices [6] in combination with fluorescence microscopy we report induction of septa biogenesis in R. microsporus. This leads to trapping of BAT-deficient endobacteria in infected hyphae. Considering that the survival rate of trapped bacteria is significantly reduced, endosymbionts incapable of secreting BAT proteins may elicit a protective response from the fungus. The occurrence of septa is one of the most surprising results, as Zygomycetes generally lack septate hyphae. The impact of endobacteria on fungal physiology offers a broader view on the dynamic interactions between bacteria and fungi.

1.               Frey-Klett, P., et al., Bacterial-fungal interactions: hyphens between agricultural, clinical, environmental, and food microbiologists. Microbiology and Molecular Biology Reviews, 2011. 75(4): p. 583-609.

2.               Dale, C. and N.A. Moran, Molecular interactions between bacterial symbionts and their hosts. Cell, 2006. 126(3): p. 453-465.

3.               Partida-Martinez, L.P. and C. Hertweck, Pathogenic fungus harbours endosymbiotic bacteria for toxin production. Nature, 2005. 437(7060): p. 884-888.

4.               Partida-Martinez, L.P., et al., Endosymbiont-dependent host reproduction maintains bacterial-fungal mutualism. Curr Biol, 2007. 17(9): p. 773-777.

5.               Scherlach, K., et al., Symbiotic cooperation in the biosynthesis of a phytotoxin. Angew Chem Int Ed Engl., 2012. 51(38): p. 9615-9618.

6.               Stanley, C.E., et al., Probing bacterial-fungal interactions at the single cell level. Integr Biol, 2014. 6(10): p. 935-945.

Various microfabrication techniques have been applied to produce microfluidic devices, including wet etching, reactive ion etching, hot embossing, and conventional machining. Despite obvious advantages like the convenience of use and the possibility for fast prototyping, steoreolithography (SLA) 3D printing has not yet been widely applied to produce microfluidic devices. In recently years, the price of SLA printers dropped massively, and the rapid expansion of 3D printing in medical and dental applictions made a spectrum of biocompatible resins accessible. Taking advantage of this, we developed and 3D printed a device using a heat-resistant hydrophilic resin to study bacterial dispersal by active swimming. Our device, named the “bacterial bridge”, consists of  two sampling wells connected by two verticle capillaries and a bridge-like structure that can establish along it a stable liquid film of 0.12mm in width and several centimeters in length. The elevation created by the two capillaries between the sampling wells and the liquid film can efficiently reduce passive flow. The “bridge” device thus only allows motile bacteria to pass from one sampling well to the other by active flagella-propelled swimming, and prevents the dispersal of non-motile cells. Our 3D-printed device can be quickly produced at low material cost (less than 0.5 USD per piece) and autoclaved for repeated use. It can be wide applied as an abiotic control with stable physicochemical properties and fixed structue to study dispersal and interactions of microorganisms in water-unsaturated complex enviorenments like the soil. Besides the “bridge” device, we have made several other micro-fluidic devices by 3D printing. Our work demonstrates the potential for much wider applications of modern SLA 3D printing in microbial ecology research, especially for fast prototyping and producing customized devices, which is highly valued under the current pandemic circumstances with disrupted supply chains around the world.

Absorbance measurement is a useful analytical tool that can be used along with colorimetric assays to measure a wide range of analytes. However, since sensitivity is directly proportional to path length, sensitive measurements in microfluidic channels are inherently challenging. This is especially true for droplet microfluidics where the lensing at the droplet/carrier interface further constrains path length. Conventional sensitivity enhancement approaches such as extended path length geometries and cavity-enhanced systems could still suffer from the lensing effect, high cost optics and high power consumption.[1, 2] This limits their applicability in low resources settings such as in-the-field or point-of-care analysis. Hence there is an urgent need to design low cost flow cells that can deliver enhanced measurement sensitivities but at lower power consumption.

In this study, we developed multipass flow cells assembled with squared PTFE tube with parallel mirrors on both sides, laser diode and detector. The devices were integrated with T-junction droplet microfluidic chip for in-line droplet absorbance spectroscopy, where reagents and samples were introduced into droplets for quantifying phosphate level via phosphomolybdeum blue reaction. They featured affordable low power components, and was made using simple fabrication techniques making it accessible to a wide range of researchers.  In testing they allowed multiple reflections of light in the detection chambers, which significantly increased the optical path length by 8 times and reached low limit of detection of phosphate at around 0.52 uM.

The optimised flow cell was used to determine phosphate concentrations in water samples collected from a tidal chalk river which could not be measured with a simple single-pass flow cell. We envision this flow cell design to be broadly applied for detection of trace species on droplet microfluidics to achieve high sensitivity and low limit of detection, especially for environmental and biomedical monitoring.

A minority of cells with unique phenotypic or genotypic characteristics can drive disease progression and response to therapeutic agents. Therefore, single-cell analysis is critical to the accurate characterization of disease states. A major challenge for the development of next-generation devices for the evaluation of individual cells is the integration of all steps of analysis into a single platform. Such integration would greatly reduce the time and cost of diagnosis while increasing accuracy. Major challenges to the realization of this goal include a need for high volumetric throughput, selectivity for the cells of interest, and parallel and sensitive analysis. We discuss methods that leverage electrochemical and electrokinetic processes at arrays of bipolar electrodes to address each of these challenges. Specifically, we will discuss selective capture of cells by dielectrophoresis at these arrays of these wireless electrodes and the integration of biomolecular assays to achieve sensitive and rapid analysis.

Fossils are preserved remains of life in the past. Through dating methods, fossils can be arranged chronologically, crucial in reconstructing evolutionary history and understanding responses to environmental changes that can aid modern conservation efforts and addressing climate change [1]. Amino acid racemization (AAR) allows direct dating of calcium carbonate-based biominerals (e.g. bone and teeth) over quaternary timescales (~2.5 million years), beyond the limits of radiocarbon dating (~50, 000 years) [2]. AAR as a dating technique measures the D and L ratio of intra-crystalline amino acids trapped within fossil samples, which has not been exposed to the external chemical environment.  However, current AAR methodology relies on specialist laboratory equipment, relatively large samples sizes and lengthy processing times [3]. By exploiting the advantages of miniaturization, we aim to develop an integrated microfluidic device that allows on-site preparative and analytical processes for AAR dating. Here, we demonstrate the development of a microfluidic device for the first phase of AAR which involves sample preparation to isolate the intracrystalline amino acids trapped within mammoth tooth enamel. Using a microfluidic device, sample size was reduced from ~30 mg to 1 mg and oxidative treatment time was improved by ~97%. Result showed that reduction of the first phase of AAR to a microscale did not significantly affect the yield and composition of the extracted intracrystalline amino acids, especially the four key amino acids (Asx, Glx, Ala and Phe) and showed good agreement to the corresponding macroscale conventional method. This has the potential for both carrying out sample treatment on site, outside specialist labs, and as an opportunity to use a less destructive sampling procedure of precious fossil samples.

Clonostachys rosea is a filamentous fungi used as a bio-control agent (BCA) with natural fungicidal properties. While already widely applied in agriculture, many of the underlying metabolic mechanisms are still unknown. Studying filamentous fungi in high-throughput fashion using laboratory automation systems remains a challenging task due to their complex life cycle with multicellular states, and hyphal growth. Microfluidics, especially droplet-based systems, is a robust approach to study single cell organisms or conidia (fungal spores) since they allow the study of their germination behavior in an in-vivo like environment and the identification of enzymes secreted by the filamentous fungi. However, current microfluidic methods are not well adapted to host filamentous fungi due to their hyphal growth, which can cause the droplets to easily burst after long droplet incubation times and under high-field dielectrophoretic (DEP)sorting conditions (e.g., 1.3kVpp). Moreover, their polydispersity in droplet size after incubation also requires careful tuning of the sorting conditions.[1]

In this work, we have developed a low voltage electrostatics based co-planar binary sorter that reliably sorts a C. rosea filamentous fungi droplet library (> 87.6% +- 2.7) using a field as low as 12.5 VRMS. The combination of a low voltage sorter, the use of an optical fiber based detection system, a portable mini-spectrometer, and open-source software suite drastically reduce the footprint of this system compared to gold-standard techniques.[2] Using our system, a mutant library of C. rosea was incubated for at least 36 h under several test conditions and subsequently screened for production of chitinases and b-1,3-glucanases. After sorting, we will test recovered temperature resistant BCA candidates for their activity against a plant pathogen (Botrytis cinerea). Since both of these enzymes are responsible for plant pathogen cell wall breakdown, we could measure reduction in growth of B. cinerea, while the wild type C. rosea showed less growth and activity under the test conditions.[3] We anticipate this accessible system adds to the available toolkit for researchers to study and screen other relevant industrial or agricultural filamentous fungi.

We have developed a label-free and passive method for the early isolation of activated T-cells and the first technique that enables the isolation prior to the display of cellular surface markers. T-cells are white blood cells whose activation is a critical step in adaptive immune response. The isolation of activated T-cells from naive cells is important for their study and downstream use.

The conventional methods of isolating activated T-cells rely on the use of antibodies specific to surface markers. However, it can take many hours and as much as a full day after activation for the markers to be displayed; greatly limiting fast detection and selection. In contrast, changes in cell metabolism occur within minutes after activation.

The presented work is based on our recently developed sorting platform dubbed "Sorting by Interfacial Tension" (SIFT) that sorts droplets based on pH. After bead activation (CD3/CD28) and a brief incubation on chip, droplets containing activated T-cells display a lower droplet pH due to proton secretion associated with increased glycolysis. Under specific surfactant conditions, a change in pH can lead to a concurrent increase in droplet interfacial tension. The sorting of activated T-cells on chip is hence achieved as flattened droplets are displaced as they encounter a micro-fabricated trench oriented diagonally with respect to the direction of flow.

The pH of droplets containing cells was studied as a function of time of activation. Longer activation times led to a lower droplet pH with a change in droplet pH observed in as little as 15 minutes of activation. Droplets containing activated cells (compared to naive cells or an empty droplet) have a lower pH due to the enhanced metabolism, and hence higher interfacial tension, leaving the trench at a different lateral position. This leads to an enrichment of activated T-cells from an initial mixed population of activated (15 minutes) and naive T-cells. Preliminary results led to an enrichment from 28% to 75% of activated T-cells. Moreover, since the pH change is correlated to the level of activation, the technique allows the isolation of T-cells with the highest activation, an important subset of cells for study and potentially immunotherapy.

This novel label-free technique for the early detection, enrichment and isolation of activated T-cells in minutes rather than hours can have broad usage as a biotechnology tool and for the study and selection of T-cells for immunology and immunotherapy.

Early and accurate diagnosis of infectious disease is critical to the delivery of timely and appropriate treatment that improves patient care, reduces the economic healthcare burden, and prevents disease transmission. However, access to gold standard detection of nucleic acids (DNA and RNA) is typically limited to high resource hospitals with extensive laboratory facilities. In order to bring these intricate analyses out of the lab and to the point of care, diagnostics must balance robustness, ease of use, accuracy, and cost. Here, we present a novel detection device that is designed to meet these needs by integrating the simplicity and scalability of commonly used paper-based lateral flow immunoassays (LFIAs), such as pregnancy tests, with highly sensitive nucleic acid amplification. Our paper-based platform that has potential to be scaled via roll-to-roll manufacturing for affordable and reliable molecular diagnostic devices at the point of care. This fully-integrated microfluidic rapid and autonomous analytical device (microRAAD) is completely automated from sample-in to result-out [1]; minimizing sample preparation and time-critical steps to enable robust infectious disease detection even in remote settings.

Leveraging a number of innovations in fluidic control [2], low-power heating [3], and novel assay designs, our sample-to-answer device can enable detection of pathogens from diverse sample matrices at clinically relevant pathogen levels and has recently been demonstrated to detect HIV from blood. Inside microRAAD, we incorporate heat-stable amplification reagents, thermally-actuated wax valves for fluidic control, and low-resistance silver ink resistors with a temperature control circuit for USB-powered heating. We perform loop-mediated isothermal amplification (LAMP) and visualize these products via LFIA, resulting in an integrated, sample-to-answer test that runs in under 90 minutes. Human-centered design studies were performed to ensure usability and readability of test results by minimally trained users [4]. To run the test, the user simply adds their sample and a wash buffer, seals the test, and connects the device to a cellphone or battery. Based on preliminary results in the laboratory, the microRAAD device has potential to enable early and rapid detection of highly infectious diseases at the point-of-care.

Traditional, ex situ nutrient monitoring efforts require the use of instruments such as autoanalyzers to perform reagent-based analytical techniques on sea water samples to determine nutrient concentrations. The time and cost of sample retrieval and preparation often leads to poor spatial and temporal resolution of nutrient measurements in remote environments. Microfluidic technologies enable in situ sensors to perform the same reagent-based analysis in these environments, with a comparable limit of detection (LOD) and repeatability to the bench top instruments [1].

Here, we present a fully automated in situ phosphate analyzer based on an inlaid microfluidic absorbance cell [2]. The cell is made from opaque polymethyl(methacrylate) (PMMA) inlaid into clear PMMA to attenuate any non-directional or scattered light. It uses embedded microprisms to direct light from an 880 nm LED through the interrogation channel, and then up into a photodiode to measure the resultant intensity. The sensor uses three independently controlled stepper motors to actuate four syringes: two for reagents, one for a standard and one for the sample. There are 10 solenoid valves in a separate oil filled pressure tolerant compartment to ensure reliable and repeatable fluid control. Power is supplied either by the optional separate battery back, or through an external source. The fluids and waste are all housed in a separate perforated container and connected by Teflon tubing.

Using colorimetric absorbance spectrophotometry to determine phosphate concentrations in sea water, the sensor achieves a practical LOD of less than 100 nM. Furthermore, an in situ verification with an on-board standard showed that the measurements were repeatable, with a relative standard deviation of less than 1.5 %. We determined the temperature dependence of the colorimetric reaction by carrying out four sequential calibrations, at intervals between 5 °C and 20 °C, with phosphate standards ranging from 0.2 to 10 µM. Next, we fixed the sensor to a jetty at a depth of 1.5 m to capture the phosphate fluxes during the diurnal tidal cycle in the Bedford Basin. To verify and bench-mark the sensor measurements, we used a Niskin bottle to retrieve samples from alongside the instrument, and then prepared them for analysis on an autoanalyzer. Finally, we mounted the phosphate analyzer to the Stella Maris multi-sensor platform 100 m off shore at a depth of 10 m for a multi-month deployment, in the Bedford Basin in Nova Scotia, Canada.

Owning to robust, portable, and pump-free design, microfluidic paper analytical devices (μPADs) are increasingly adopted for performing complex multiplex assays in a variety of fields. The current methods for detecting multiple targets in μPADs require patterning the membranes to create hydrophobic barriers, a technique introduced by Martinez et al[1]. Although these barriers created using wax printing, inkjet printing, photolithography, or by chemical modification of paper[2] efficiently utilize membrane surface area and consume less reagent volume, their fabrication requires expensive equipment and skilled personnel, posing a challenge for scale-up. In addition, the generation of non-uniform colorimetric signal due to convection of rehydrated signals create difficulty in signal quantification. To overcome the limitations of traditional μPADs, we have developed a new device called barrier-free μPAD (BF-μPAD) which can detect multiple targets without requiring any physical or chemical membrane modification. The device is fabricated by stacking two paper membranes of different wicking rates. The bottom membrane (called as detection layer) has lesser wicking rate and stores multiple dried reagents. The top membrane (called as distribution layer) has higher wicking rate and acts as a fluid distributor for the bottom membrane. In one embodiment, an 8cm x 2cm device assembly can perform up to 20 different tests in 30 seconds. To demonstrate the multiplexing feature of BF-μPAD, we have selected four salivary analytes; thiocyanate, glucose, nitrite, and protein. Chemistries for their colorimetric detection were deposited on the detection layer and fluid samples containing different concentrations of the analytes were added to the distribution layer. A user only requires a smartphone to visualize and interpret signals, making it a desirable point of care tool at low resource settings. A direct comparison of the limit of detection between conventional μPAD and BF-μPAD shows that BF-μPAD improves the limit of detection by ~3.7x and produces perfectly uniform colorimetric signals. Barrier-free detection in BF-μPAD is enabled by the generation of unique flow patterns, which were modeled in COMSOL Multiphysics using the Richards equation. In contrast to μPADs, large-scale manufacturing of BF-μPADs would only require a commercially available benchtop robotic dispensing system to handle microliter volume, thus significantly reducing fabrication costs and complications.

Total alkalinity (TA) is one of the four parameters which characterize the oceanic carbonate system. TA data with high spatial coverage and high temporal frequency can contribute to better measurements, modelling, and understanding of the carbon cycle in water, providing insights into problems from global climate change to ecosystem functioning.  To provide this data, we present an autonomous sensor capable of in situ measurements of TA, based on a generic oceanographic lab-on-chip platform which has been implemented for a number of chemical assays.

This sensor implements a benchtop TA assay on a small portable device. The system is based a generic ocean chemical lab-on-chip hardware platform with integrated microfluidics, optics, pumps, valves, and electronics.  The TA sensor samples seawater, mixes it with reagents, degasses the resulting solution, and performs an optical measurement.  It can carry multiple calibration materials on board, allowing for routine re-calibration and quality checks in the field. The scientific applications of this sensor require reliable accuracy (~0.1%-0.5%) during deployments lasting months in harsh and varying conditions. 

This sensor has been tested and demonstrated both in the lab and in the field.  Field deployments have included a carbon storage experiment in the North Sea, an ecosystem study in a Mediterranean seagrass meadow, a coral reef health study in Australia, and most recently a carbon uptake study off the coast of Antarctica.  The results from lab and sea trials show that the technology is in a strong position to be able to meet the scientific needs for TA measurements.

To close, I’ll discuss some of the general engineering challenges that we’ve faced in creating microfluidic technology that has to survive at sea.

At least one-quarter of the ≥20 million people receiving antiretroviral drugs for human immunodeficiency virus (HIV) treatment and prevention have drug concentrations outside the therapeutic range and are at risk of treatment failure or adverse reactions. Regular antiretroviral drug measurement could help people living with HIV who receive antiretroviral therapy (ART) to suppress viral replication and also help people receiving pre-exposure prophylaxis (PrEP) to prevent HIV infection. However, the gold standard for ARV measurement is liquid chromatography tandem mass spectrometry (LC-MS/MS) which is slow, centralized, and expensive. Developing and implementing a rapid point-of-care test could help to monitor and improve medication dosing and ART/PrEP outcomes.

In this talk, I will describe the REverSe TRanscrIptase Chain Termination (RESTRICT) assay for rapid measurement of nucleotide analog drugs – used in over 90% of ART and all approved PrEP regimens – based on their inhibition of DNA synthesis by HIV reverse transcriptase (RT) enzyme. Guided by a probabilistic model, RESTRICT uses readily available nucleic acid synthesis reagents and user-friendly sample preparation techniques to detect clinically relevant antiretroviral drug concentrations in <1 hour. Using DNA templates designed to account for the chemical structure of nucleotide analogs, we selectively measure clinically relevant concentrations of the drugs tenofovir diphosphate (TFV-DP), emtricitabine triphosphate (FTC-TP), and azidothymidine triphosphate (AZT-TP) with agreement between experiment and theory.

We completed a pilot evaluation using 18 clinical samples from clients at the Madison HIV Clinic in Seattle. Blood samples are diluted in water, DNA templates, nucleotides, RT, and intercalating dye added, and results measured with a fluorescence reader—stronger fluorescence indicating higher RT activity. We compared RESTRICT assay results to TFV-DP concentrations from matched dried blood spot samples measured by liquid chromatography tandem mass spectrometry (LC–MS/MS) using ≥ 700 fmol/punch TFV-DP as a threshold for adequate adherence (≥ 4 doses/week). Among 18 adults enrolled, 4 of 7 participants receiving PrEP had TFV-DP levels ≥ 700 fmol/punch by LC–MS/MS. RESTRICT fluorescence correlated with LC–MS/MS measurements (r = − 0.845, p < 0.0001). Median fluorescence was 93.3 (95% confidence interval [CI] 90.9 to 114) for samples < 700 fmol/punch and 54.4 (CI 38.0 to 72.0) for samples ≥ 700 fmol/punch. When calibrated to an a priori defined threshold of 82.7, RESTRICT distinguished both groups with 100% sensitivity and 92.9% specificity.

RESTRICT represents a new class of activity-based assays for therapeutic drug monitoring and precision dosing that could be extended to other diseases that are treated with nucleotide analogs or enzyme inhibitors.

Blood typing, donor compatibility testing, and hematocrit analysis are blood tests performed to ensure the compatibility of the donor red blood cells (RBCs), plasma, or platelets with the transfusion recipient and to determine the need for transfusion. These tests are performed routinely prior to blood transfusions [1] and organ transplants [2], for the management of trauma patients [3], and for pretransfusion testing for conditions such as anemia, leukemia or surgical patients. Where centralized laboratories are available, blood samples are collected, transported, centrifuged to separate red blood cells (RBCs) from plasma, and then batched prior to evaluation using automated instruments. Nevertheless, the centralized/batched model can add minutes to hours to the test time, which is adequate for some patients but can be life-threatening for critical-care patients who need rapid testing results [4].

An alternative to centralized testing is the use of portable tests that can be performed directly on whole blood at the "point of care" (POC). Unfortunately, it is widely understood that user mishandling and misinterpretation of results is an Achilles' heel for these techniques [5–8], which has limited their use in medical settings. In sum, there are substantial limitations to both centralized and POC blood tests, which often leads healthcare practitioners to rely on blood products from “universal blood donors” (i.e. group O RBCs and group AB plasma). However, these products are in chronic short-supply and inventory levels are often below target levels worldwide [9].

As a step towards providing rapid results at the bedside, we developed a point-of-care hemagglutination system relying on digital microfluidics (DMF) a fluid handling technique that enables the manipulation of discrete liquid droplets on an array of electrodes [10]. We also developed a unique, automated hemagglutination readout algorithm. We validated our algorithm for rapid (<6 min) blood typing, donor compatibility testing, and hematocrit analyses on whole blood samples using an inexpensive shoebox-sized instrument. The system was then transported to a clinical setting, where it was operated by an inexperienced user, with results found to be 100% concordant with the gold standard technique. These results suggest great promise for our DMF platform to deliver rapid, reliable results in a format well suited for the trauma center and other settings where every minute counts.

Quantifying the concentration of dissolved nutrients in water is traditionally achieved by spot sampling and subsequent laboratory analysis. The process is manually intensive and intrinsically limits the frequency and/or number of data points that can be obtained therefore the short-lived episodic events could be missed. By using sensors to monitor in situ, however, much larger data sets can be obtained which can be reported in real-time and with less overall capital expenditure. The majority of current commercial ammonium sensors are based on ion-selective electrodes, produced for example by Xylem YSI and RSHydro. These are user-friendly and feature high measurement frequencies but, as is commonly found for electrochemical-based sensors, their calibration drifts over time requiring frequent manual recalibration. Moreover, they have also been shown to be highly susceptible to interference from changes in ionic concentration. We report here the development of a novel droplet microfluidics based ammonium sensor, and its deployment to a sequential batch bioreactor enriched with polyhydroxyalkanoates (PHAs) accumulating bacteria. PHAs are microbially produced polyesters which are biodegradable, biocompatible, and have tuneable mechanical and thermal properties and their microbial production has garnered vast research interest [1]. The sensor miniaturizes the widely adopted indophenol blue method in droplets (LOD of 0.11 mg/L), can perform high frequency and accurate measurement and provide long period of monitoring with minimal reagent consumption (1700 nL per analysis). Driven by a specially designed persistaltic pump based on a phased-flow droplet generation method [2], the sensor can autonomously collect samples (via filter) from the bioreactor, produce droplet trains, regulate temperature for completion of reaction and provide colorimetric measurements (5 per minute) via an inline spectrophotometer, thus removing human intervention. The preliminary data obtained over a four day period shows high resolution monitoring and reveals the fast feeding behavior of the bacteria, which could lead to the design of more efficient bioreactors and feedback controlled bioprocessing. These type of sensors could form the next generation of in situ sensors, to address wide applications in monitoring for wastewater treatment, environmental monitoring, aquaculture and others, especially when transient and episodic events are involved.

Capillary-driven flow microfluidics offers the ability of loading samples into microchannels without the requirement for external pumping mechanisms. Microfluidics devices can be coated with specific reagents and a simple dipping into a fluid sample could trigger a reaction in between reagents and analytes of interest. There is a great potential to explore capillary-driven flow microfluidics and utilize it for point-of-care (POC) diagnostics e.g. for the prevention of antimicrobial resistance (AMR) in healthcare by loading bacterial/pathogenic samples into microchannels, incubating and reading the results near the patient’s bedside [1]. Several such devices have been developed for detection of biomarkers and AMR testing [1-3] along with smartphone detections which offer an alternative approach for portable point-of-need imaging requirements [4]. However, bacterial sample loading into microchannels could be challenging because of the surface properties of the microchannels and clogging of samples at the inlets. Recently, highly branched poly(N-isopropyl acrylamide) incorporating Nile red has been shown to provide a fluorescence signal upon binding to bacteria [5]. This paper showcases a capillary-driven flow microfluidics device (Chip-and-Dip) for loading of bacterial/pathogenic samples into microchannels for antimicrobial testing. This device offers the capability of capturing cells into microchannels that can be further treated with reagents to generate a colorimetric/fluorescent signal. The Chip-and-Dip device, fabricated with inexpensive materials and coated with these reagents, works by simply dipping the reagents-coated microfluidics chip into a sample. Here, we show a successful coating of microchannels with fluorescently labelled polymer and loading of Enterococcus Faecalis spiked in milk samples.