The actin cytoskeleton is essential for maintaining mechanical integrity of cells and tissues and for providing structural support during dynamic processes including migration, endocytosis and cytokinesis. From a molecular perspective, it consists of (1) actin monomers polymerized in double helical filamentous structures and (2) an ensemble of regulatory proteins that regulate shape and function of actin structures. From a mechanics perspective, the cytoskeleton is a dynamic entity that can generate force while being subject to various load perturbations. Though molecular understanding of actin networks is extensive, our understanding of how molecular signaling is converted to force output and how force input feeds back into molecular activity remains limited. The goal of this dissertation is to investigate how the interplay between molecular and mechanical attributes of the actin cytoskeleton results in desired cellular activity and physiological phenotypes.

We first focus on the leading edge of migrating cells where nucleation of branched actin structures is involved in membrane protrusion. In chapter 2, we investigate the effect of the biochemical composition of these structures on the observed dynamic properties of network growth. To do so, we reconstitute branched actin network assembly using a minimal set of essential proteins (i.e. nucleation promoting factors, ARP2/3 and capping proteins) and evaluate their role over a broad range of concentrations. We find that in the absence of opposing force, changes in the nanomolar range of soluble protein concentration significantly modulates architectural and kinetic properties of nucleating actin structures.

In cells, branched actin networks do not just transmit forces in the form of protrusion but also resists opposing load imposed by the membrane and other physical barriers. In chapter 3, we use atomic force microscopy to study the impact of external force on the biochemical composition and mechanical properties of reconstituted branched actin structures. Interestingly, we find that mechanical loading alters network density and composition, which in turn modulates its bulk mechanical properties and renders it stiffer, more powerful and efficient.

Central to assembly and function of actin networks is the activity of actin binding proteins. We next extend our investigation to ask whether forces on actin filaments can influence actin binding protein (ABP) localization and activity in the cytoskeleton. Despite sharing the same cytoplasm, ABPs in cells spatially segregate and differentially regulate actin structures. In the context of the leading edge of migrating cells, cofilin binds and severs filaments in the lamellipodia, whereas tropomyosin is secluded as it binds and stabilizes filaments in the lamellum. In chapter 4, we hypothesize that these proteins are mechanosensitive and show that cofilin preferentially binds to network structures subject to compression whereas tropomyosin favors relaxed structures.

Lastly, in chapter 5, we explore the sensitivity of calponin homology domain-containing proteins to the mechanical state of actin filaments. We focus our study on wild type and mutated versions of the utrophin actin binding domain, which is used as a universal actin marker. Using a multiscale biophysical approach, we show that mutant utrophin can selectively bind highly stressed actin filaments in vitro and in cells. We use this mutant to develop a ratiometric actin mechanosensor for mapping physiological forces in-vivo which provides a new tool for exploring mechanoregulation of cellular processes. Overall, the findings in this dissertation provide direct evidence for the importance of mechanical perturbations in regulating structure and function of the actin cytoskeleton.

Macrophages comprise a compartment of immune cells capable of performing a variety of functions necessary for fighting disease. One of the key roles of macrophages, which exist throughout the body in both tissue-resident and monocyte-derived forms, is to act as sentinels able to detect the presence of pathogens. Once a pathogen is recognized, a macrophage can destroy it via a process of internalization and digestion known as phagocytosis. To recognize foreign objects and potential pathogens, macrophages rely on a diverse set of surface receptors that are able to bind to pathogen-associated ligands. Binding between macrophage surface receptors and their associated ligands on the surface of a pathogen leads to the formation of a cellular interface, which represents a critical checkpoint at which the decision whether to phagocytose or not is made. In this dissertation I will examine this phagocytic interface and highlight the importance of molecular biophysical properties – including size, density, and mobility – in forming such interfaces and determining outcomes of phagocytosis.

In order to precisely modulate the biophysical characteristics of target surface properties and record the resulting outcomes of phagocytosis, it is necessary to develop a robust in vitro phagocytosis assay. This is the focus of the first part of this dissertation, in which I outline a novel protocol for making reconstituted phagocytic target particles and measuring amounts of their phagocytosis by macrophages. Reconstituted target particles are made by coating glass beads with lipid bilayers and conjugated proteins of interest. The composition of the lipid bilayers can be varied to control the surface properties of the targets – including surface protein identity, density, and fluidity. Automated microscopy and analysis of multi-well plates containing macrophages and target particles allows me to easily quantify phagocytic efficiency for targets with different surface properties in a relatively high-throughput manner. As a demonstration of this assay, I compare rates of phagocytosis for antibody opsonized targets and find that targets coated with fluid lipid bilayers and antibody are more efficiently phagocytosed than targets coated with non-fluid bilayers and antibodies.

Next, I use the phagocytosis assay to determine the effect of antigen size on antibody-dependent phagocytosis. Macrophages are able to detect objects opsonized with antibody with a class of receptors known as Fc receptors (FcRs) that can bind to the Fc region of antibodies. There have a been a number of proposed mechanisms by which FcR binding is thought to lead to phagocytic signaling, one of which relies on dephosphorylation of the intracellular signaling motifs of FcRs by the phosphatase the CD45 phosphatase. The tall extracellular motif of CD45 (~20 nm) raises the question of whether it will be excluded from cellular interfaces with small intermembrane gaps. I find that the inclusion or exclusion of CD45 from the phagocytic interface varies with antigen size and that phagocytosis is more efficient for antibodies targeted to short antigens (< 10nm) than it is for antibodies targeted to long antigens. This finding has direct implications for designing antibodies for use in cancer immunotherapies.

The surface of target cells is composed of a dense array of proteins and only a small subset of these will act as antigens to which antibodies bind. The remaining proteins, which I refer to as ‘bystander proteins’, do not directly contribute to phagocytic signaling, but are, nevertheless, present on targets and at phagocytic interfaces. The final chapter of this dissertation uses the reconstituted in vitro phagocytosis assay to examine the effect of bystander proteins. I find that phagocytosis is inhibited for antibody opsonized targets with bystander proteins when compared to targets opsonized with the same density of antibody in the absence of bystander proteins. Using a combination of live cell microscopy and computational simulations, I find that phagocytic inhibition is caused by bystander proteins posing a steric barrier to interface formation. I demonstrate that this barrier can be partially overcome with the addition of a second, long binding protein (P-Selectin) to the target surface.

The biophysical insights presented in this dissertation add to our current understandings of the mechanisms that control macrophage target recognition and will be useful for designing more effective immune targeted therapies.

Controlled activation of immune cells protects the body from pathogens and disease while limiting damage to healthy self-cells Immune cells collect information through pattern-recognition receptors (PRRs) displayed on the cell membrane. While some PRRs respond to soluble ligands – small molecules, proteins, nucleotides, and sugars – signals are often transduced through direct physical contact – surface to surface – between an immune-cell and a potential target cell. Here, the cell-membrane is more than a passive platform for receptor display. Rather, the physical environment of the cell membrane influences the two-dimensional spatial organization of PRRs, which in turn controls their activation.

The goal of the first part of this dissertation is to describe a simple, protein size-dependent physical mechanism that can control protein organization at membrane-membrane interfaces. I begin by reconstituting a synthetic membrane-membrane interface in vitro using giant unilamellar vesicles decorated with adhesion proteins, and I use this model system to show that a size difference of ~5 nm relative to the membrane interface is sufficient to drive the exclusion of a tall non-binding protein from the interface. Combining in vitro measurements with Monte Carlo simulations, I find that non-binding protein exclusion is also influenced by lateral crowding, binding protein affinity, and, to a lesser extent, thermally-driven membrane height fluctuations that transiently limit access to the interface. This simple, sensitive, and highly effective means of passively segregating proteins has implications for signaling at many cell-cell contacts.

Next, I explore the consequences of size-dependent protein segregation at the interface between a macrophage and a target cell. Phagocytosis is mediated through activation of Fc receptors upon ligation with a surface-bound antibody, but it remains unclear how receptor-antibody binding triggers phosphorylation of the receptor, or how Fc receptors can accommodate the variable binding geometries formed by antigens with diverse structure, shape, and size. To determine the influence of antigen height on antibody-dependent phagocytosis, I reconstitute a minimal model of a target cell surface using lipid-bilayer coated glass beads and protein antigens with height-specific antibody binding sites. I find that phagocytosis is dramatically impaired for antigens that position an antibody > 10 nm from the target surface. The finding that close contact (< 10 nm) between a macrophage and target cell is a requirement for triggering of Fc receptors suggests a new design principle for therapeutic monoclonal antibodies to stimulate immune effector cells against cancer and infectious disease.

Not all pathogens and diseases are recognized and cleared by the immune system. For those that result in chronic infections, especially in low-resource areas, advances in mobile technology can be used to aid in diagnosis and treatment. The last part of this dissertation describes the development of CellScope Loa, a mobile phone-based microscope for quantifying the blood-borne parasitic worm Loa loa at the point-of-care. Efforts to eliminate onchocerciasis and lymphatic filariasis in Central Africa through mass drug administration have been suspended because of ivermectin-associated serious adverse events, including death, in patients infected with the filarial parasite Loa loa. To safely administer ivermectin (IVM) for onchocerciasis or lymphatic filariasis in regions co-endemic with L. loa, a strategy termed “test and (not) treat” has been proposed whereby those with high levels of Loa loa microfilariae (>30,000/ml) that put them at risk for life-threatening serious adverse events are identified and excluded from mass drug administration. To address this, I worked on the development of an automated mobile-phone microscope for quantifying Loa loa directly in whole blood. The device has been used in Cameroon to identify and exclude individuals with >20,000 mf per milliliter of blood (at-risk for SAEs) from ivermectin treatment in a “test and treat” pilot, where IVM was safely delivered to more than 15,000 patients in a “test and treat” pilot.

There are innumerable systems of interest in medicine and biological research for which the fundamental imaging technology exists to examine the entities of interest, but no sufficiently robust and efficient technique for isolating and preparing the sample has been devised. The lack of sample preparation methodology often serves as a barrier to healthcare access and scientific discovery. Here, I describe three instances of such barriers and the progress made to overcome them.

First, I discuss the poor access to medical diagnostics in endemic regions for human schistosomiasis and other neglected enteric and urinary helminthic diseases. These pathogens have a profound impact on morbidity and mortality in the regions they affect, despite the existence of safe and effective therapeutics. In wealthy regions, detecting the causative pathogens has been successfully performed via microscopy in medical laboratories for a century. However, in regions with poor access to laboratory equipment like pipettes, centrifuges, and microscopes, the available diagnostic technologies have poor sensitivity and specificity, requiring drug administration decisions to be made at a population level.

To address the need for a point-of-care test that can diagnose schistosomiasis on an individual basis, I present a novel, inexpensive, specialized capillary which accepts urine from a syringe and in a single step, captures schistosome eggs from solution and spreads them in a monolayer. The monolayer can then be imaged using an inexpensive mobile phone microscope, and schistosome eggs can be identified and counted using a machine vision algorithm running on the phone’s processor. In this way, with minimal training, users can take urine from a patient and receive a quantitative diagnosis for Schistosoma Haematobium infection in under a minute, for less than one US dollar. I present our assessment of the new capillary for capturing parasite eggs and the new algorithm for detecting eggs, both in the lab with reconstituted samples, and in the hands of clinical personnel in field studies of school children in Ghana and Côte D’Ivoire. Because of the promising results of these studies, we have invested in infrastructure to mass produce the diagnostic, with the ultimate goal of providing rapid diagnostic capabilities to the entire region affected by S. Haematobium.

Next, I address a barrier to imaging which manifests within the Fletcher Lab and other groups who are interested in probing the biophysical characteristics and interactions of transmembrane proteins. For many transmembrane proteins of interest, it is difficult to decouple the behavior of the specific protein from the vast complexity of a living cell. In other words, the noise due to other cellular processes drowns out the signal of the molecules of interest under a given condition chosen by the researcher. The desire to limit this noise leads many experimentalists to pursue in vitro reconstitution of transmembrane proteins in an artificial lipid bilayer, such as a black lipid membrane or a supported lipid bilayer, as a platform for their experiments. In these systems, the composition of lipids, proteins, and other biomolecules can be controlled. One such system, which is homeomorphic with the plasma membrane of a cell, is a giant unilamellar vesicle (GUV). Past results from our lab demonstrated the ability to form GUVs with a great deal of control over contents of the lumen using a microfluidic jet. Here, I present a new technique for forming GUVs by replacing the microfluidic jet with an ultrasonic transducer, which preserves the loading capabilities of the microfluidic jet and confers additional potential advantages over the state of the art.

In the new system, a black lipid membrane is formed, with the desired luminal solution of the GUV on one side and the extravesicular solution on the other. This membrane is placed such that it is orthogonal to the axis and incident to the focal point of a high-intensity focused ultrasound transducer. By activating the piezoelectric transducer with a short burst of RF power, the system forms a region of highly concentrated Reynolds stress near the focal point and the resulting acoustic streaming results in a tightly focused jet of moving liquid. This jet deforms the planar membrane until a GUV pinches off, filled with the solution on the proximal side. Compared to the microfluidic jet, this method eliminates the complications associated with bringing a capillary into close proximity with the planar lipid bilayer and potential clogging of capillaries with small diameters. The acoustic jetting technique is also more amenable to encapsulating micron-scale particles, small unilamellar vesicles, and viscous solutions. I discuss some of the operating conditions for the acoustic transducers and other elements in the system, and their effects on the resulting GUVs, and I review practical limitations to this technique that may prevent its wide adoption except in niche applications.

Finally, I address a barrier to sample preparation at the cutting edge of biological imaging: transmission electron cryo-microscopy (cryo-EM). Cryo-EM requires a thin sample of biomolecules encased in vitreous ice, through which electrons can be transmitted. As the imaging instrumentation has improved significantly in recent years, the inability to consistently form thin uniform samples of biomolecules has become a bottleneck in the cryo-EM workflow. Currently, the vast majority samples are prepared by pipetting a volume of aqueous sample onto a TEM grid, pressing a piece of filter paper against the droplet to wick off the majority of the liquid, and plunging the grid and remaining solution into liquid ethane to freeze it. This blotting technique is effective, but often results in regions of thick ice therefore poor electron transmission, dried areas of the grid, and significant protein denaturation at the air-water interface.

To improve our understanding of the fluidic phenomena at play during the blotting technique, and to inform the development of new techniques, I examine the liquid lubrication layer during and after blotting. I found that the capillary pressures within the absorbent filter paper cause air to enter the lubrication layer in a tortuous and unrepeatable fashion while the filter paper is still in contact with the sample. This propagation of air allows proteins to denature at the newly formed interfaces and creates dewetted regions on the surface which serve as boundaries to prevent the complete relaxation of the water surface when the filter paper is removed. The resulting lubrication layer therefore has varying thickness which can persist to the ice stage. With these insights in mind, we propose directions for the development of cryo-EM sample preparation techniques which avoid the irreproducibility and sample damage inherent in filter paper blotting.

Breast epithelia exist in a constant state of interaction with their surrounding environment. Morphogenesis is the developmental process by which breast cells grow into their surrounding matrix and form the ducts and milk-producing lobules. When morphogenesis breaks down, breast cancer occurs. Traditionally, biologists think of cancer through the framework of genetic mutations. Significant work in the past few decades demonstrated that the mechanical environment plays a critical role in determining growth and malignancy of breast cancers independent of genetic mutations. For example, increasing the stiffness of the extracellular matrix drives phenotypic malignancy through cell-generated contraction. However, the role of forces felt by the tissue due to external causes remains unclear. This dissertation describes experiments that reveal a critical role for external forces in branching morphogenesis and tumorigenesis. The experiments make use of a simple method to apply external compression to mammary epithelial cells embedded in biologically relevant gels.

In branching, compression mechanically aligned collagen fibers and directed multicellular branch growth along these fibers. Fiber alignment sensing required fascin activity, but did not require RhoA-mediated contraction. Contraction served a separate purpose of generating fiber alignment in collagen networks. These findings suggest that migrating cells sense fiber alignment through fascin-mediated filopodia formation rather than through RhoA-mediated contraction.

In tumorigenesis, compression encouraged malignant cells to form normal-looking acini, a process called `phenotypic reversion.' A transient compressive force at the one-cell state was sufficient to induce reversion without genetic manipulations or pharmacological treatments. Time-lapse microscopy of the malignant cells revealed that compression restored coherent rotation of malignant cell doublets, a behavior associated with the formation of a phenotypically normal structure. Blocking E-Cadherin eliminated compression sensitivity, indicating that cell-cell communication was required for force-induced reversion.

As external forces altered the structure of a growing multicellular colony, changes in multicellular structure could affect epithelial mechanics. The mechanical properties of multicellular epithelial structures were measured using an atomic force microscope. Hollow lumen structures were softer than filled lumen structure. The increased stiffness associated with lumen filling could contribute to malignancy during cancer progression.

Taken together, this work demonstrates the importance of external forces during morphogenesis and tumorigenesis of the mammary gland. External compression can direct multicellular migration and encourage malignant cells to re-enter the `normal' morphogenetic program. Cell-cell communication plays an important role in multicellular mechanosensing, contributing to both morphogenesis and mechanosensing in mammary epithelial structures. Further studies of multicellular mechanosensing incorporating the data and techniques presented here could identify new molecular targets for breast cancer treatment and prevention.

The global health implications of improper disease screening and diagnosis are tremendous, contributing to unnecessary disease burden and ongoing transmission of infectious diseases. While tools have been developed to help trained physicians and scientists diagnose disease in centralized healthcare facilities, many patients present with symptoms in resource-constrained settings. The limitations in infrastructure and skill present are often incompatible with existing diagnostic tools. The work in this dissertation centers on the engineering design, validation, and deployment of optical diagnostics targeted to these resource constraints. The approaches presented here take advantage of advances in mobile consumer electronics that have put high quality optics, automation, and data transmission capabilities into a compact and widely available package. In Chapter 1, I describe the limitations of current diagnostic tools for two diseases, oral cancer and tuberculosis, and outline the potential for development of novel tools around mobile devices.

In Chapter 2, I present work that is a prerequisite for the rational design of reproducible and quantitative imaging with a mobile phone. Mobile phones and their cameras have seen rapid changes in specifications and performance, but they remain consumer devices with several corresponding limitations for medical use. I characterize the imaging capabilities of mobile phones across time as part of a custom portable microscope, detailing sampling limitations, aberrations, and the consequences of unwanted image processing. I conclude that mobile phones, through appropriate optical design and workflow management, can perform sufficiently for diagnostic imaging applications.

In Chapter 3, I demonstrate a multi-color fluorescence imaging device based on a mobile phone for read-out of a sandwich immunoassay on a microfluidic chip. Microfluidic devices, systems that manipulate small volumes of liquid, have the potential to enable rapid and appropriate diagnostics at the point of care through reducing volume for patient samples and lowering cost. One of the limitations of microfluidic systems is that measuring the output of the device traditionally requires a large conventional microscope and scientific camera for visualization and quantification. Multi-color fluorescence on a mobile phone offers a powerful and portable alternative.

Current diagnostics for tuberculosis, a disease that kills over 1 million people a year globally, have advanced dramatically in the recent decade, but the devices that detect the DNA of tuberculosis have not succeeded in extending testing capabilities to peripheral healthcare settings where they are most needed. New nucleic acid amplification tests are promising for this setting; they eliminate the need for cycling temperature and provide a more rapid time to answer. In Chapter 4, I demonstrate that it is possible to control and measure the amplification of DNA in real-time with this new class of assay using a mobile phone camera. Demonstration of this capability required development of an optical system, custom software, sample handling geometry, and thermal management. I also include initial validation of a primer set contributed by collaborators that is sensitive across a range of strains of tuberculosis. Ongoing work is focused on combining these capabilities into an integrated tuberculosis diagnostic.

In Chapter 5, I present a field deployment of a diagnostic microscope for oral cancer built around a tablet computer. Oral cancer is the single largest cause of cancer mortality for men in India and other high burden countries. Cultural habits that increase the risk of developing oral cancer combine with delays in diagnosis leading to poor prognoses for patients. To improve the diagnostic process, it is important to screen patients early in the disease progression and refer them for a full diagnosis and care. In this work, I adapt both a manual microscope and automated slide-scanning device to the needs of oral cancer screening by brush biopsy. Through collaboration with clinical and corporate partners, these devices have been deployed and data collection has begun. I present initial images generated from patients and classified by pathologists that demonstrate the diagnostic workflow in practice.

Affecting care in a global health setting is a complex challenge, but through work presented in this dissertation I contribute a combination of novel diagnostic method development, detailed device characterization, and field data collection that demonstrate the potential of global health diagnostics based on consumer electronics.

The reductionist approach to modern cell biology aims to identify the individual molecules and interactions that give rise to complex biological activity. A complementary constructionist approach, known as reconstitution, aims to recapitulate biological structures and functions from basic building blocks in order to show which components are essential and how biophysical constraints, such as membrane boundaries, influence organization and activity. My dissertation research has applied and extended this `bottom-up' approach to study the role of membrane mechanics in the formation of cellular filopodia and to develop new tools for reconstituting processes encapsulated within membranes and for engineering cell-like devices.

We investigated the mechanics of actin-membrane interactions by studying dendritic actin networks grown on the surface of giant unilamellar vesicles. In this minimal system, we observed the formation of parallel filament protrusions arising from the highly branched dendritic actin network, notably in the absence of bundling proteins. We confirmed through a simple theoretical model that a lipid bilayer can drive the emergence of bundled actin filament protrusions from branched actin filament networks, thus playing a role normally attributed to actin-binding proteins. This revealed a critical role for the membrane in organizing actin filaments at the plasma membrane.

This work motivated the development of a technique for encapsulating protein contents in the lumen of lipid vesicles in order to emulate the biophysical boundary conditions of real cells. We demonstrated the use of a microfluidic jet to form lipid vesicles with controlled contents by deforming a planar bilayer. These vesicles mimic an essential organizational feature of cells - encapsulation within a lipid membrane - and provide a platform for more complex cellular reconstitution. Subsequently, we adapted this technique to a pulsed inkjet-based device, enabling greater control of vesicle size and improved throughput. Using this inkjet-based device for vesicle formation, we were able to control membrane properties such as asymmetric lipid composition and insertion of membrane proteins, which are essential for numerous cellular processes. We demonstrated the applicability of this technique by reconstituting SNARE-mediated membrane fusion in a geometry that mimics exocytosis.

In summary, this work has provided new insight into the role of lipid bilayer mechanics on the reconstitution of cellular protrusions and developed a novel technique that enables formation of lipid vesicles with controlled contents and membrane properties. Further development of this technique will enable advanced reconstitution experiments and construction of functional cell-like devices for medical and biomaterials applications.