Home  /  Research

# Current research activities are listed below.

A. Further success from previous JIMEM-funded research.

Our success with enzymatic Si-C bond formation in 2016 has led to a 3-year $1M research contract with a major chemical company to explore new biological catalysts for silicon chemistry. Our previous JIMEM research also generated results for a successful NIH RO1 grant and patents that formed the basis of Caltech spin-out company Aralez Bio, founded in 2019. Aralez Bio uses tryptophan synthase enzymes developed at Caltech to prepare non-canonical amino acids for a wide array of research and therapeutic applications. Caltech postdoc Dr. Tina Boville won a Cyclotron Road Fellowship to establish Aralez Bio in the Bay Area. They are already supplying products to customers. B. Summary of current JIMEM-funded research. Using directed evolution, we developed a new biocatalytic approach to synthesizing medicinally important β-lactams from readily available carboxylic acids. We first engineered cytochrome P450-based enzymes that perform enantioselective C‒H amidation, an unprecedented reaction in the field of C‒H functionalization, to yield β-lactams. These enzymes were further engineered to differentiate among several C‒H bond and direct amidation precisely to a desired position while simultaneously preventing other side reactions. Our technology provides access to chiral lactams, a particularly attractive motif in pharmaceutical chemistry, and showcases the selectivity these new-to-nature enzymes can exert in synthesis. This work was published in Science in 2019. C. Synthesizing β-lactams through asymmetric, non-benzylic C-H amination. β-Lactams, historically one of the most effective classes of antibiotics (Figure 1A), are susceptible to microbial resistance. Widely used β-lactam antibiotics including penicillins are produced primarily by fermentation. On the other hand, non-natural β-lactams must be synthesized by lengthy processes. To supply the core for non-natural antibiotics, half of the penicillins produced are degraded to an intermediate which is used to produce semi-synthetic β-lactams. This laborious method often produces mixtures of stereoisomers that may have different, even opposing, clinical effectiveness. Even some of the newest and most reliable methods for β lactam synthesis are still only diastereoselective and often require chiral starting materials, product-specific chiral auxiliaries, precious metal catalysts, or complicated multistep processes. To address the need for a continuing supply of non-natural β-lactam antibiotics, we proposed a biocatalytic strategy for enantioselective β-lactam synthesis (Figure 1B). Prior to the current JIMEM funding, we had engineered a heme enzyme for enantioselective β-lactam synthesis through benzylic C‒H amidation, a reaction not known in nature. We discovered that a cytochrome P450 hemoprotein variant (named E10FA), originally derived from enzymes engineered for benzylic C‒H amination using a sulfonyl nitrene precursor, provided 1.5% yield for a desired β-lactam product with 90% enantiomeric excess (ee). We improved the activity and selectivity of this variant by directed evolution, and eventually obtained variant LSsp3, after four rounds of mutagenesis and screening. With six active-site residues mutated from its parent enzyme (Figure 1C), LSsp3 is a highly efficient catalyst for β-lactam synthesis with up to 98% yield, 96% ee, and 1 million total turnover number (TTN) for the model substrate (Figure 1D). This enzyme successfully converted a large range of substrates bearing aromatic substituents, substituents adjacent to carbonyl group, allylic C‒H bonds, or heteroatom substituents (Figure 1E). Figure 1. Overview of β-lactam synthesis via C‒H amidation. (A) Medicinal importance of β-lactams. (B) Reaction scheme for enzymatic enantioselective β-lactam synthesis. (C) Active-site structure of a variant closely related to the parent E10FA shows location of beneficial mutations (PDB ID: 5UCW). (D) Directed evolution of E10FA to LSsp3 increases activity (total turnover number) and enantioselectivity. (E) Reaction scope. Our interest in developing a broad synthetic platform for preparing new β-lactams led us to further engineer LSsp3 to go beyond benzylic C‒H amidation. With the current JIMEM funding, we engineered variants LSβ and LSγ for selective synthesis of β- and γ-lactams, respectively, from substrates that bear more than one set of target C‒H bonds (Figure 2A). Both these variants were evolved from LSsp3 by site-saturation mutagenesis on the active site and screening for improved regioselectivity (Figure 2B). These variants can completely overrule the inherent reactivities of the C‒H bonds to provide lactams with specific ring sizes and still maintain remarkably high turnovers and enantioselectivities (Figure 2C). We have engineered a lactam synthase to access chiral benzylic β-lactams and a toolbox of lactam synthases to convert a given substrate into different lactams through a catalyst-controlled and selective C‒H amidation. In order to deliver the right bond in the right place, these enzymes are ready to overturn the reactivity trends due to bond strength, inductive effects, steric accessibility, or ring strain. Our work demonstrates that the macromolecular structure of biological machineries can be effectively repurposed to achieve new-to-nature catalytic functions, providing a general strategy to yield a valuable motif in numerous bioactive compounds as well as address a great challenge in synthetic chemistry. Figure 2. Development of regioselective lactam synthases for β- and γ-lactam synthesis (Cho et al., Science, 2019). (A) Selectivity of LSβ and LSγ-catalyzed lactam synthesis. (B) Evolutionary trajectory of variants LSsp3, LSβ, and LSγ. (C) Regiodivergent lactam synthesis was achieved starting from a single substrate using three lactam synthase variants LSsp3, LSβ, and LSγ . Engineering Non-Natural Fluorinases for the Synthesis of Therapeutic and Diagnostic Organofluorines Using a combination of chemistry-inspired rational engineering and evolution in the laboratory, we have developed two new biocatalytic approaches to incorporating fluorine into organic molecules. We have engineered new enzymes for in vivo and in vitro synthesis of organosilicon compounds, a class of molecules with medical applications such as positron emission tomography (PET) imaging through site-specific 18F-labeling. Ours is the first biocatalyst capable of carbon-silicon bond formation. The technology provides access to organosilicon compounds that are not readily accessible otherwise, opening new opportunities for fluorine-incorporation and labeling studies that are biologically relevant. This breakthrough was published in 2016 in Science. We have also developed a method for introducing trifluoromethyl groups to organic molecules biocatalytically. This transformation is unknown in nature and will find applications in the preparation of pharmaceuticals and agrochemicals. Inhalation Dosimetry for Ultrafine Particles This project had, as the primary focus, the development new low cost sensors capable of assessment of the dose to different regions of the airways when fine particles are inhaled. Present ambient air quality measurements focus on exposure rather than dose. The primary exposure measurement today is PM2.5, the mass concentration of particles smaller than 2.5 µm aerodynamic diameter. Air quality standards are written in terms of this metric, which also serves as the primary measure of fine particle exposure in many research studies. Many particles in the size range, when inhale remain suspended in the air and are exhaled without depositing. Thus, PM2.5 measurements report an amount particulate matter that is larger than deposited in the airways. Nonetheless, a statistical links between PM2.5 exposure and number of adverse health outcomes are well-established. What is not clear, is an appropriate measure of dose in the effort to understand the origins or causes of adverse health outcomes. Numerous studies have suggested alternative metrics. In studies of the potential health consequences of workplace exposures in the burgeoning nanotechnology industry have suggested that deposited surface area may be a more relevant measure of dose to sensitive regions of airways, at least for some varieties of engineered nanoparticles. A number of instruments have been developed of lung- deposited surface area, using the rate relationship between particle charging and surface area to infer the area of particles that may deposit in a system in which deposition efficiency is intentionally biased to match the deposition profile of the respiratory tract. Several other studies have measured the aerosol number concentration in order to infer the number of fine particles that might deposit within the airways, and have found strong associations with short- term responses, on the order of hours or less, particularly with respect to a heart rate variability. The preponderance of data on fine particle exposures report PM2.5; lesser amounts provide other measures of exposure. Optical dust sensors (low cost optical particle counters) are increasingly being used to provide a surrogate for PM2.5 though they are insensitive to ultrafine particles, even in the rare instances when they account for significant aerosol mass. Condensation particle counter measurements of the number concentration are fairly common in air quality studies. The working hypothesis of this study is that data that enables estimation of the dose in terms of these different metrics that is delivered to different regions of the airways will enhance the ability of epidemiological and other health effects studies to examine physiological mechanisms behind adverse health outcomes and to discriminate between different sources of airborne fine particles. Furthermore, the power of such studies will be greatly enhanced if instruments can be made sufficiently small, unobtrusive, and low cost that they can be deployed as personal dosimeters in cohort studies, or in dense networks in community health studies. All three measures of dose of inhaled fine particles can be provided today using a combination of particle size distribution data obtained by commercially available differential mobility analyzers (DMAs) and established lung deposition models that predict the inhaled-particle deposition efficiency inmdifferent regions of the human airways as a function of particle size. Present DMAs are, however, far to large, complex, and expensive to satisfy the latter constraints. This study has examined how this class of measurements could be adapted to meet the needs of the health effects research community. Using atmospheric simulations of major pollution events and validated models of DMAs and other aerosol instruments, we have demonstrated that biases in estimation of regional airway deposition using size distribution data are minimal, even when the instruments have much lower size resolution than is needed for many atmospheric science application. Moreover, we have examined the biases that associated with present-day fine particle exposure measurements. Only size-resolved measurements provide a useful correlation with all three metrics. The ability to make measurements at relatively low size resolution relaxes many design constrains, opening the door to dramatic reductions in instrument cost. By applying these simplified design specifications to the opposed migration aerosol classifier (OMAC), a new form of differential mobility analyzer that allows instrument miniaturization, this project has focused on the development of fine aerosol particle dosimetry instruments that will enable the aforementioned improvements in the health effects of fine particle exposures. Over the past year, we have demonstrated that an OMAC that was designed based on computational simulations performs as predicted, and explored several alternative approaches to the design of the low cost, size-resolving fine particle sensor. In side-by-side comparisons with a conventional, high resolution scanning mobility particle sizer (SMPS), a new, low-size-resolution OMAC effectively captured rapid transients in exposure and dose of fine airborne particles. Additional refinements are being implemented in new low-cost designs, that will further simplify the instruments toward our objective of enabling scientifically valid measurement of exposure and dose of ultrafine particles in personal monitors and community air quality monitors. Toward a rapid test of antibiotic resistance The goal of this work is to answer fundamental scientific questions to develop a point of care (POC) test of antibiotic susceptibility that is amenable to limited-resource settings (LRS). Ideally, an antibiotic susceptibility test (AST) should be able to be performed within a single doctor visit. To achieve such a short (30 min) phenotypic test requires determining a pathogen's susceptibility after only a short antibiotic exposure, performing all of the sample-handling steps and using an ultra-fast assay to acquire the final readout in a very short time frame. We have had great success meeting these requirements. Primarily using clinical isolates and clinical samples provided by our collaborators at UCLA, we have validated digital single molecule counting as a superior way to detect antibiotic susceptibility. We first used digital PCR (dPCR) to test whether assessing DNA replication of the target pathogen via digital single-molecule counting could be used to shorten the required antibiotic exposure time, thus decreasing the overall time of the assay. We have found that partitioning bacterial chromosomal DNA into many small volumes during dPCR enabled AST via (i) precise quantification and (ii) a measure of how antibiotics affect the states of macromolecular assembly of bacterial chromosomes. This digital AST (dAST) determined susceptibility of clinical isolates from urinary tract infections (UTI) after just 15 min of exposure for all four antibiotic classes relevant to UTI. This work was published in Angewandte Chemie (Schoepp et al. 2016). Our work proves our digital AST method can work directly on clinical samples for several types of antibiotics. We also show in this article that we have discovered a new mechanism for rapidly detecting beta-lactam antibiotic susceptibility: chromosome segregation. More recently, we have demonstrated that our dAST method can yield a sample-to-answer result from a clinical urine sample in as few as 30 min! In this work, published in Science Translational Medicine (Schoepp et al. 2017), we demonstrate that we can shorten the time of antibiotic exposure to 15 min, the time for sample prep to 3 min, and that we can then quantify with digital LAMP in 6 min. Thus far, the work being done with JIMEM support is laying a strong foundation to develop a rapid, point-of-care AST and strengthen global antibiotic stewardship. II. Nonconfidential summary of Jacobs-funded research activities We were invited by Joe de Simone and John Rogers to present the work in the PNAS Special Feature devoted to Novel Materials. The lead article in the issue is "Multiplicity of morphologies in poly (L-lactide) bioresorbable vascular scaffolds" by my graduate students Artemis Ailianou and Karthik Ramachandran, with our Abbott collaborators Mary Beth Kossuth and Jim Oberhauser (PNAS 2016 113:11670-11675). Bioresorbable vascular scaffolds (BVSs) support the artery for 6 months but completely dissolve in 2 years, eliminating serious long-term complications. Our research solved a materials science mystery: the first clinically approved BVS is made from a brittle material, poly (Llactide) (PLLA), yet it does not fracture during crimping or deployment. Using X-ray microdiffraction, we discovered multiple, micron-scale morphologies in the crimped BVS, which confer ductility to PLLA and resist fracture upon deployment. Contrary to intuition, the crimping process enhances scaffold strength. Based on feedback from surgeons, there are two main difficulties associated with Absorb: 1) the material is transparent to x-rays that surgeons use to visualize the placement of the scaffold relative to the lesion and 2) the scaffold is thick relative to metal stents, making it more difficult to maneuver into the lesion. Our JIMEM-supported research addresses both of these difficulties using inorganic nanotubes to simultaneously provide radioopacity and increase strength (essential for making a thinner device with adequate hoop strength). The experience we have gained dispersing WS2NT in PLA, provides protocols that preserve PLA molecular weight to maintain control of the release kinetics and disperse inorganic particles (PLA functions as a dispersant in solution and becomes the matrix upon drying). Clinical relevance: Abbott Vascular's bioresorbable vascular scaffold, Absorb, secured FDA approval in June 2016. Approximately 150,000 Absorb devices have been implanted to date. Clinical data show that, when properly implanted, the Absorb device dissolves away, leaving a blood vessel that pulsates, dilates and contracts, without the risk of late stent thrombosis. Pending patent application: US application 2014/0128959 A1 "Biodegradable stent with enhanced fracture toughness" (an Abbott Vascular invention) is supported by data and analysis from our collaborative research. Publications: Ailianou, A.; Ramachandran, K.; Kossuth, M.B.; Oberhauser, J.P.; Kornfield, J.A.; "Multiplicity of morphologies in poly (L-lactide) bioresorbable vascular scaffolds," PNAS 113: 11670-11675, doi:10.1073/pnas.1602311113 (2016). Ph.D. Theses: Ailianou, A., "Development of Semicrystalline Morphology of Poly(L-lactic acid). During Processing of a Vascular Scaffold", defended May 2014. Biolistic delivery of drugs to the cornea was developed under Jacobs support to address the last obstacle to clinical translation of JIMEM research that began in the first round of JIMEM funding (FY07) to develop a "Photochemical Therapy to Arrest Progression of Keratoconus." Clinical translation: Caltech has granted a license to Avedro in the domain of UV-activated therapeutic crosslinking. Results under JIMEM support enabled Kornfield and Schwartz to secure$2M in NIH SBIR support for research on therapeutic cross-linking to preserve vision by stabilizing the shape of the eye. In vivo experiments in both rabbit and guinea pig models show that Eosin Y activated by visible-light (EY/vis) stabilizes ocular tissues (both the cornea and sclera) with very low toxicity. Clinical translation of EY/vis stalled in 2011 for lack of a non-surgical method to deliver EY to the stroma through the epithelium. We believe our recent research on biolistic delivery could open the way to a low-cost, low-risk, non-surgical delivery of EY for photocativated treatment of keratoconus and post-LASIK ectasia. This year's JIMEM proposal pursues an opportunity to apply biolistic delivery to improve safety and efficacy of the first pharmacologic treatment for keratoconus. The FDA granted a orphan drug designation to copper supplementation, so biolistic delivery of copper may provide a faster route to the clinic than EY/green photocrosslinking.

Issued patents (* asterisk indicates IP licensed by Avedro):
US 7,727,544 "Treatment of Myopia"
*US 8,414,911 "Photochemical Therapy to Affect Mechanical and/or Chemical Properties of Body Tissue" (also 5,559,540, 1,502,632, 1,551,590,

Publication:
Mattson, M; Huynh, J; Wiseman, M; Coassin, M; Kornfield, J.A.; Schwartz, D; "An In Vitro Intact Globe Expansion Method for Evaluation of Cross-linking Treatments", Investigative Ophthalmology & Visual Science, 51, 3120-3128, doi:10.1167/iovs.09-04001 (2010).

Ph.D. Theses:
Huynh, J, "Factors Governing Photodynamic Cross-linking Of Ocular Coat", defended May 2011.
Mattson, M.S., "Understanding and Treating Eye Diseases: Mechanical Characterization and Photochemical Modification of the Cornea and Sclera", defended May 2008.

"Molecular Engineering of Surgical Adhesives for Protein-Based Biomaterials: Anchoring Artificial Protein Onlays to the Cornea," with David Tirrell (co-PI) and Dr. Bala Ambati under Jacobs FY08 developed a transparent corneal adhesive that promotes re-epithelialization and enables in situ formation of a corneal onlay.

Clinical translation:
The PEG-aECM gel, developed initially as a corneal adhesive, is a key components of the regenerative tissue scaffolds described below in relation to preventing corneal blindness by preserving corneal clarity and shape after injury.

Publication:
Wang, M.; Kornfield, J.A.; "Measuring shear strength of soft-tissue adhesives", Journal of Biomedical Materials Research Part B-Applied Biomaterials, 100B, 618-623 (2012).

"Polymer-Protein Scaffolds for Accelerated Wound Healing" with David Tirrell and Dr. Bala Ambati, U. Utah (2/10-1/12). Developed a PEG-aECM gel that presents a gradient of cell binding domains and supports adhesion of corneal keratocytes. Established experimental protocols for treatment of superficial corneal injury in a mouse model and for electrospinning protein nanofibers with controlled diameter and alignment. These protocols provided a foundation for the following JIMEM project.

"Composite scaffolds for orderly Corneal Wound Healing" in collaboration with David Tirrell and Dr. Bala Ambati (2/13-2/15). Bilateral corneal blindness deprives more than 5 million people worldwide of their sight. Today, the only way to reverse corneal blindness is a cornea transplant—an option that is essentially unavailable in the developing world. Our goal is to prevent corneal blindness by developing clinically-viable scaffolds that promote orderly corneal wound healing using biochemical and topological cues. Under JIMEM support, we created a transparent composite scaffold that combines oriented nanofibers that guide stromal regeneration by corneal fibroblasts. Preclinical studies in rabbits indicate that the scaffold can be applied in a relatively simple procedure, is well tolerated by the eye, supports normal reepithelialization and the cornea remains clear during two-week duration of our wound healing study (correlated with a reduction in myofibroblasts, neutrophils and macrophage).

Ph.D. Thesis:
Fu, A., "Nanofibrous scaffolds for orderly corneal wound healing", defended May 2015.

Thinner, radiopaque bioresorbable vascular scaffolds for the treatment of coronary heart disease

Bioresorbable vascular scaffolds (BVSs) are a promising new treatment for coronary heart disease (CHD), the leading cause of death in the world (>7 million/year). Unlike permanent metal stents, BVSs are transient implants; they support the artery for the requisite 6 months, at which time the blood vessel tissue has filled the scaffold and the epithelium has covered its luminal surface. Within 2 years, the BVS is completely resorbed and the treated vessels are observed to regain vasomotion and vasodilation (virtually eliminating angina symptoms, which are prevalent and persistent in stented patients). The transient character of the BVS overcomes the most dreaded complication associated with metal stents—late stent thrombosis. Surgeons advocate two main improvements to foster wider adoption and serve a broader cross-section of patients: reduce the BVS profile to enable treatment of smaller arteries and more complex lesions, and increase x-ray opacity to permit visualization during and after implantation.

We proposed to tackle the dual challenge of a thinner yet more radio-opaque BVS by reinforcing it with inorganic nanotubes that have high x-ray scattering power. Specifically, we will focus on tungsten disulfide (WS2) nanotubes (NT) that confer a substantial increase in strength, have shown low toxicity in vitro, and have radio-opacity comparable to platinum, the current standard for radio-opaque markers. The key to a thinner, radio-opaque BVS made from PLLA-WSNT is the control of the nanocomposite morphology during processing: extrusion to create a largely amorphous tubular preform; tube expansion to obtain a uniform wall thickness, laser-cutting to create the struts and rings of the scaffold (Fig. 1A), and crimping onto a balloon catheter (Fig. 1C) prior to deployment via inflation of the balloon (Fig. 1D). We will apply insight into the interplay of these processing steps to create a nanocomposite that has the potential to address both of the most pressing clinical needs in BVS—reduced scaffold profile and increased radio-opacity.

FIGURE: Processing of BVS: (A) struts (along z) and rings (along θ) are created by laser cutting the expanded tube (B); the "as cut" scaffold" is crimped onto a balloon catheter (C); the scaffold is deployed (D) via inflation of the balloon after it is positioned at the lesion.

­­Acoustically Targeted Chemogenetics with Engineered Viral Vectors

Neurological and psychiatric diseases often involve the dysfunction of specific neural circuits in particular regions of the brain. Existing treatments, including drugs and implantable brain stimulators, aim to modulate the activity of these circuits, but are typically not cell type-specific, lack spatial targeting or require invasive procedures. With support from the Jacobs Institute, we have developed an approach to modulating neural circuits noninvasively with spatial, cell-type and temporal specificity. This approach, called acoustically targeted chemogenetics, or ATAC, uses transient ultrasonic opening of the blood brain barrier to transduce neurons at specific locations in the brain with virally-encoded engineered receptors, which subsequently respond to systemically administered bio-inert compounds to activate or inhibit the activity of these neurons. In a recent publication, we showed that we could use this technology to target the mouse hippocampus and noninvasively modulate the formation of contextual memory (Szablowski et al, Nature Biomedical Engineering 2, 475-484, 2018). Now, we are optimizing this technology by engineering viral vectors optimized for delivery across the ultrasound-opened blood-brain barrier and scaling this technology up for use in non-human primates as a stepping-stone to human clinical translation. The latter effort is supported by a new major grant from the NIH BRAIN initiative, led by our lab in collaboration with Doris Tsao and Viviana Gradinaru at Caltech and ultrasound experts at Vanderbilt.

Targeted Cellular Agents

The Shapiro Laboratory is developing methods to use penetrant forms of energy, such as ultrasound and magnetic fields, to control cellular function. This work entails engineering proteins and genetic circuits that can convert these forms of energy into cellular signals such as gene expression. One recent example of our work, funded in part by the Jacobs Institute, showed that bacteria could be engineered to respond remotely to MRI-guided focused ultrasound signals mediated by temperature.

Related Publication on Ultrasonic Control of Cellular Function:
Piraner DI, Abedi MH, Moser BA, Lee-Gosselin A, Shapiro M. G.* Tunable thermal bioswitches for in vivo control of microbial therapeutics. Nature Chemical Biology 13, 75-80 (2017).

FIGURE: Remote control of bacterial agents using focused ultrasound. (a) Illustration of the in vitro focused ultrasound experiment: focused ultrasound is used to heat a target area of a bacterial culture lawn through a tofu phantom (depicted as translucent) under MRI guidance, followed by fluorescent imaging. (b) MRI-based temperature map of the bacterial specimen during steady-state ultrasound application, overlaid on a raw grayscale MRI image of the phantom. (c) Fluorescent image of the region targeted by ultrasound, showing activation consistent with a bacterial construct expressing GFP under the control of TlpA36 and RFP regulated by TcI. (d) Illustration of the in vivo experiment, in which focused ultrasound is used to activate subcutaneously-injected bacterial agents at a specific anatomical site. (e) Representative thresholded fluorescence map of a mouse injected subcutaneously in both left and right hindlimbs with E. coli expressing GFP under the control of TlpA36, following ultrasound activation at only the right hindlimb. Scale bars 2 mm (b and c) and 1 cm (e).

Time-Resolved in situ Proteomic Analysis of MRSA Infection in the Mouse

The objective of our Jacobs Institute research is to elucidate the strategies used by methicillin-resistant Staphylococcus aureus (MRSA) to establish and sustain infection in the skin. Each year in the United States, roughly 90,000 people suffer invasive MRSA infections, and more than 20,000 die. Because the preferred niche of S. aureus is the human host, the pathogen has developed a powerful arsenal of mechanisms for manipulating the mammalian immune system. Staph now accounts for almost one-third of skin and bloodstream infections.

Over the past year, we have discovered a protein that increases the severity of skin infection by MRSA. This protein has not previously been associated with infection. Our current research is directed toward elucidating the mechanism by which the protein contributes to virulence, and to finding ways to treat infection by inhibiting the action of the protein.

FIGURE: Fluorescence image of MRSA infection in the mouse skin. Bacterial cells and proteins can be independently stained to provide unique insight into the nature of the infection.

Electrostatics of Polyelectrolyte Self Assembly

Many biomolecules are charged polymers, or polyelectrolytes – for example, the genetic material RNA and DNA have negatively charged backbones, while many intrinsically disordered proteins have mixed charges along their backbone. The charge on these biomolecules results in a wide range of self-assembly behaviors, from viral assembly to the formation of membraneless organelles, and can be leveraged to produce nanoparticles for the medical delivery of biomolecules. In this reporting period, we used our group's variational theory of charged macromolecules to study the thermodynamics of polyelectrolyte phase separation in the presence of salt, which is always present under physiological conditions. The variational theory is able to describe how chain structure varies under changing solution conditions, and we show how failure to do so in previous theories leads to significant overestimates of the driving force for phase separation. Our work demonstrates how appropriately describing chain structure is critical to accurately predicting the complexation of polyelectrolytes.

FIGURE: Phase boundaries for symmetric solutions of polyecations and polyanions with added salt; dashed lines demarcate the metastability limit. We compare our theory (black), which accounts for changing chain structure over different concentrations, to the commonly-used fg-RPA (green) which assumes a Gaussian chain structure at all length scales, and rods (red). Note that the fg-RPA suggests phase separations can persist even up to the limit where the solution is very nearly all salt (blue dashed line). Inset shows the overly dilute low-concentration branch predicted by the fg-RPA.