Poster Abstracts

Chad Altobelli, UCSF, Arkin Lab

Abstract: Mutations in VCP can cause a degenerative disease called Mutli-System Proteinopathy (MSP1) which primarily affects muscle tissue but also impacts neurological and skeletal health. On a molecular level, these mutations disrupt the conformational landscape of VCP, which is tightly regulated by the nucleotide binding. Mutations destabilize the ADP-bound “down” conformation, and cause VCP to sample the “up” conformation usually associated with ATP binding. This dysregulation disrupts homeostatic protein-protein interaction networks. We are working to develop compounds which can compensate for MSP1 mutations. To this end, I’ve developed a minimally-perturbing FRET system to report on the conformation of VCP’s N-domain by combining a FlAsH-EDT2 fluorophore and iodoacetamide-TAMRA. After scaling up this approach, we plan to screen the SMDC compound libraries to discover compounds that can correct for VCP mutations and restore the normal conformational equilibrium.

Amanda Caceres, UC Davis, Heffern Lab

Iron is an  essential  metal  in  a  variety  of  biological  processes  including  oxygen  transport,  mitochondrial respiration, DNA replication and repair, and cell signaling. Disruptions in iron homeostasis are associated with endocrine and metabolic disorders, including diabetes. Our lab has previously shown that high dietary sugar intake alters metal misregulation in human subjects.  To gain mechanistic insight to these observations, I initiated short-term in vivo studies evaluating the effects on mice given water supplemented with high glucose for four weeks. At this short time frame, the mice are not expected to progress to diabetic states, allowing us to assess early changes that result from elevated sugar intake. The livers of the glucose-supplemented  mice  showed  upregulation  in  both  fatty  acid  synthesis  enzymes and regulators of lipid homeostasis, suggesting a metabolic shift toward triglyceride synthesis. Interestingly, glucose supplementation also seems to induce changes in hepatic iron regulation, even though dietary iron content was equal between the two groups. Analysis of hepatic and serum iron markers suggests increased uptake of transferrin bound iron. Cell-based models mimicking the high iron state were utilized to determine whether the anti-diabetic drug metformin was capable of restoring both metabolic and iron regulation. In total, results show that high glucose diet is sufficient to cause hepatic fat production and iron overload with metformin able to rescue iron overload and metabolic dysfunction.

Kristen Campbell, UC Davis, Beal Lab

Adenosine Deaminases Acting on RNA (ADARs) are an important class of RNA editing enzymes that catalyze the hydrolytic deamination of adenosine (A) to inosine (I) in double-stranded RNA (dsRNA). Since inosine is effectively read as guanosine (G) during translation, ADARs can produce A to G transitions in dsRNA. Site-directed RNA editing (SDRE) is a promising therapeutic tool wherein we can exogenously supply guide RNAs to direct endogenous human ADARs to reverse disease causing mutations in specific RNA transcripts. Guide RNA (gRNA) modifications at locations that contact the ADARs’ active site are often used to improve editing efficiency. However, little is known about rate enhancing chemical modifications in the gRNA at the dsRNA binding domain’s (dsRBDs) interface. Analysis of a previously solved crystal structure of ADAR2 bound to dsRNA suggested positions at this interface would be sensitive to gRNA modification. In this work, gRNAs with altered 2’-riboses at observed contacts were synthesized and subsequently tested to determine their effects on the editing rate of therapeutically relevant ADAR targets. We have found that changing a single 2’-OH at specific positions on the gRNA with a 2’-F drastically increases the editing efficiency of ADAR2 in vitro in two different sequences which was also validated in cellulo. Clashes with 2’-OMe modifications and N235 of ADAR2 were modelled on a newly solved crystal structure with a 2’-F modified gRNA, indicating that -fluoros are well-accommodated in the dsRBD region. This work shows the importance 2’-ribose modification patterns that help catalyze ADAR’s editing event.

Claire Caputo, UC Davis, Segal Lab

ADNP syndrome, also called Helsmoortel-Van Der Aa syndrome, is a rare neurodevelopmental disorder caused by mutations in the ADNP gene encoding for activity-dependent neuroprotective protein (ADNP). ADNP is a multifunctional, regulatory protein that is especially critical for brain development. It is essential for embryonic neural tube formation and plays a critical role in neuron differentiation and protection. ADNP is also a transcription factor involved in chromatin remodeling and participates in the epigenetic regulation of over 400 human genes, impacting nearly every cell type in the body with highest expression in the brain, gastrointestinal system, lungs, reproductive system, and heart. Heterozygous mutations cause haploinsufficiency and/or dominant negative gene expression and results in global cognitive and motor delays, such as intellectual disability, impaired mobility, and autistic behavior. There is currently no approved drug or standardized treatment for ADNP syndrome. This project proposes an innovative activating antisense oligonucleotide (ASO) therapeutic to restore ADNP levels by increasing translation efficiency. Healthy ADNP mRNA contain upstream open reading frames (uORFs) in the 5’ untranslated region (UTR) that are thought to reduce translation initiation from the primary open reading frame (pORF) as a natural inhibitory mechanism of gene expression regulation. I hypothesize that treatment with an uORF-targeting ASO will decrease uORF translation and increase pORF translation, thereby increasing the amount of ADNP. The first aim of this project is to characterize the uORF-mediated regulation of ADNP translation with reporter constructs. The second aim is to demonstrate the impact of ASO treatment on ADNP translation in healthy and ADNP syndrome modeled human induced pluripotent stem cell (iPSC)-differentiated neurons. 

Sarah Casebeer, UCSF, Zaro Lab

Ligand-receptor interactions are critical to the immune system’s ability to distinguish self from non-self. ‘Don’t eat me’ (DEM) ligand-receptor pairs specifically downregulate macrophage phagocytosis of self-cells. The most well characterized DEM signaling axis is between ligand CD47, which is ubiquitously expressed on mammalian cells, and receptor SIRPα, which is found on macrophages. CD47 can also interact in cis with SLAMF7 on hematopoietic cells; this interaction similarly downregulates macrophage phagocytosis of CD47-expressing cells by preventing homotypic SLAMF7 ‘eat me’ signaling. Intriguingly, homologs of CD47 are found in poxvirus genomes. While previous work has postulated their role in downregulating innate immune response during infection, the mechanism by and pathways through which they do so are unknown. In this work, I demonstrate that viral CD47 homologs interact with human SLAMF7 but do not interact with SIRPα, and discuss ongoing efforts to evaluate their impact on macrophage phagocytosis.

Jennifer Cordoza, UCSC, McKinnie Lab

The Micromonospora genus has provided a myriad of chemically interesting and biomedically-relevant secondary metabolites, including gentamicin aminoglycoside antibiotics and calicheamicin enediyne cytotoxins. Another interesting molecule from this genus is TLN-05220 (1), a polyketide-peptide hybrid that was first isolated from Micromonospora echinospora ssp. challisensis NRRL 12255, and displays antibiotic activity comparable to the last resort therapeutic vancomycin. In addition to its angular hexacyclic polyketide core, 1 contains an unusual heterocycle piperazinone moiety (2), which is present in other small molecules such as the HIV integrase inhibitor dolutegravir, and the bioactive secondary metabolite class of the bravomicins. Despite being present in several molecules, the biosynthesis of 2 has yet to be discovered, and given the antibiotic potency of 1, we sought to determine the biosynthetic pathway for 2. Previous collaborative stable isotopic labelling studies determined that alanine, glycine, and serine are present in 1. Bioinformatic studies lead us to investigate tln1 and tln5, which both annotate as pyridoxal-5’-phosphate (PLP)-dependent enzymes, and tln4 and tlnD, which annotate as asparagine-like synthetases. Using in vitro enzyme assays with synthetic standards and liquid chromatography-mass spectrometry, we determined that Tln1 is a PLP-dependent alanine epimerase that produces the noncanonical amino acid D-alanine from L-alanine. Additionally, we found that Tln5 is a PLP-dependent beta-substitution enzyme that uses the subsequent D-alanine with O-phospho-L-serine to form a new C-N bond and create a pseudodipeptide. While most PLP-dependent beta substitution enzymes prefer L-amino acids, Tln5 diverges from current literature to prefer a D-amino acid over a more prevalent L-amino acid counterpart. We also tested numerous non-native nucleophiles with Tln5 to find the enzyme tolerates small, branched and slightly polar amino acids and amines to form several non-native pseudodipeptides. Future work will assay Tln4 and TlnD for the cyclization to form 2 and determine how 2 is ligated to form 1. Understanding how this uncommon heterocycle is formed and incorporated will provide bioinformatic insight into novel antibiotic secondary metabolites derived from Micromonospora along with assisting in future biocatalytic engineering to produce new polyketide-peptide antibiotic derivatives. 

Natalie Dugan, UC Davis, Beal Lab

The Adenosine Deaminase Acting on RNA (ADAR) family of enzymes catalyze the hydrolytic deamination of adenosi

The Adenosine Deaminase Acting on RNA (ADAR) family of enzymes catalyze the hydrolytic deamination of adenosine (A) to inosine (I) in double stranded RNA (dsRNA). Due to this requirement, guide RNAs (gRNAs) can be designed to utilize endogenous ADARs for Site Directed RNA Editing (SDRE) for therapeutically relevant sequences. However, the ability to edit difficult to edit sites is important in expanding the scope of disease-relevant sequences that can be targeted with SDRE. EMERGe is a high-throughput screen designed to assess guide strand sequences. Out of the fourteen sequences found with the screen, one sequence was verified in a therapeutically relevant context. 

ne (A) to inosine (I) in double stranded RNA (dsRNA). Due to this requirement, guide RNAs (gRNAs) can be designed to utilize endogenous ADARs for Site Directed RNA Editing (SDRE) for therapeutically relevant sequences. However, the ability to edit difficult to edit sites is important in expanding the scope of disease-relevant sequences that can be targeted with SDRE. EMERGe is a high-throughput screen designed to assess guide strand sequences. Out of the fourteen sequences found with the screen, one sequence was verified in a therapeutically relevant context. 

Jingxin Fu, UC Davis, Chen Glyco Group

Glycosylation plays a critical role in the functions and the therapeutic efficacies of glycoproteins, influencing their binding affinity, stability, pharmacokinetics, and half-life. The manufacturing of therapeutic proteins, however, often results in glycan heterogeneity, leading to variability in drug performance and potential immunogenicity. Applying carbohydrate-active enzymes (CAZymes) in glycan engineering is an effective approach but is hindered by poor expression of mammalian glycosyltransferases (GTs) in heterologous systems, low enzyme stability, and high production costs associated with mammalian expression platforms. To address these challenges, we developed an in vitro glycoengineering platform to enable remodeling of N-glycans on therapeutic proteins. Through protein engineering and computational design, we significantly improved the stability and the soluble expression in E. coli of key N-acetylglucosaminyltransferases (GlcNAcTs), galactosyltransferases (GalTs), and sialyltransferases (STs). Notably, high-level soluble expressions in E. coli of GlcNAcTs for generating complex N-glycan branches were achieved. The enzymes have been successfully applied to engineer therapeutic proteins with up to tetra-antennary N-glycans. This platform offers a cost-effective and scalable approach for producing homogeneous glycoproteins with desired N-glycan structures, paving the way for the development of glycoprotein-based drugs with improved properties by designing their N-glycan structures.

Trevor Gannalo, UC Davis, BMCDB Graduate Group

Aflatoxin B1 is a carcinogenic metabolite known to contaminate staple foods across the world. In this study, we seek to develop a novel enzymatic method for bioremediation of Aflatoxin. We first employ innovative protein computational tools, such as RFDiffusion, ProteinMPNN, and Rosetta, to redesign the structure and amino acid character of several Glycosyl Hydrolase active sites with the goal of promoting Aflatoxin binding. We then experimentally characterize our B-Galactosidases by measuring enzyme expression levels and assaying activity using test strips, colorimetric measurements, and LC-MS. From this approach, we demonstrate successful soluble expression and activity on the natural substrate using our first three designed enzymes, with future plans to test 8 additional designs. While we did not observe activity on Aflatoxin, the tolerance of B-Galactosidase to major active site changes indicates the potential for protein engineering efforts. Additionally, we successfully modeled the proposed reaction using Density Functional Theory using the Gaussian program, which indicated the chemical feasibility of this reaction. Given these encouraging results, we plan to continue designing and testing enzymes for this reaction, while also expanding our methodology to include de novo protein design, guided by the ground and transition states in the DFT-modeled reaction.

Virginia Garda, UCSF, Arkin and Kao Labs

"Lysosomes are the most acidic organelles in cells and play a critical role in maintaining cellular homeostasis by breaking down waste products. Defects in lysosomal pathways lead to the progression of neurodegenerative diseases (NDs), which are characterized by the apparition of protein aggregates in affected areas of the brain. An increase in lysosomal pH has been observed in several models of NDs, which might contribute to the breakdown of protein homeostasis. Studying mechanisms of lysosomal pH regulation can thus lead to further understanding of mechanisms that lead to NDs. Identifying small molecules that can lower lysosomal pH can give us tools to be able to study the effects of rescuing this phenotypic change in models of NDs and can also uncover novel mechanisms for lysosomal pH regulation. With this goal, we conducted a phenotypic high throughput screen using a genetically encoded lysosomal pH sensor with a 7500-compound bioactive library of small molecules. We found several compounds that can acidify the lysosomes and are currently working on validating them and characterizing other effects they have on lysosomal activity.  

Jason Guerrero, UCSC, Sanchez Lab

Eukaryotic cells are often depicted in literature as having compartmentalized subcellular organelles composed of lipids and proteins. These analytes are essential for maintaining the integrity and function of the cell. However, the localization of small molecules (metabolites) within a single cell—whether they are partitioned into specific organelles or dispersed uniformly without a discernible pattern—is often overlooked. Addressing this gap in knowledge requires the development of the ExMS method, which will provide insight into the spatial distribution of small molecules within a cell's microenvironment. This spatial information will enhance our understanding of complex biological processes, potentially overlooked disease mechanisms, and offer a much more accurate representation of a single cell.

Matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI-MSI) is a label-free, soft ionization analytical technique that allows for the spatial mapping of small molecules, lipids, and proteins in the form of a 2D ion image. However, despite its capabilities, the lateral resolution of MALDI is limited both physically and instrumentally. The lateral resolution in MALDI-MSI is determined by the stepping of the stage, laser focus (5µm - 10µm), and the size of the matrix crystals using a common sprayer apparatus (HTX TM Sprayer). These spatial limitations present challenges when attempting single-cell imaging, given the typical dimensions of mammalian cells (~10µm), ultimately resulting in an insufficient number (~2) of data points per cell. This limitation prevents the creation of a proper map of (sub)cellular metabolites.

Drawing inspiration from expansion microscopy (ExM), a sample preparation method that facilitates super-resolution imaging of biological tissue via crosslinking in osmotically swollen hydrogels, we are developing a method in collaboration with Dr. Lydia Kisley at Case Western Reserve University to mechanically (equibiaxially) stretch human high-grade serous ovarian cancer (HGSOC) cell lines that have been 2D-loaded onto a hydrogel. We achieve this by utilizing sulfo-SANPAH, a heterobifunctional crosslinker that binds to the hydrogel and the fibronectin protein, facilitating the adhesion of this highly aggressive and adaptive cell line. Proper cell adherence is necessary for the cells to withstand the physical stress of being stretched.

While we are optimizing the visualization of cell expansion using live-cell fluorescence, we are also creating a metabolomics library of small molecules present in OVCAR8 cells. This involves using both reverse-phase liquid chromatography (RPLC) and hydrophilic interaction liquid chromatography (HILIC) to capture the full range of polar and non-polar compounds in our biological system. Creating this library will help us pinpoint reproducible ions in ESI, which we will then track in MALDI as reference points to achieve high ion efficiency in our sample preparation.

The implementation of single-cell methodology for investigating disease and physiological processes offers significant potential to advance the effectiveness of disease studies. It expands the toolkit for therapeutic development and the identification of disease-related biomarkers by enabling the chemical measurement of small molecules.

Valerie Jin, UC Berkeley, Francis Lab

Tyrosinase-mediated oxidative coupling (OC) enables site-specific conjugation by introducing a cysteine mutation at any desired site on a protein’s surface, allowing it to be linked to a tyrosine residue on another protein. Unlike conventional antibodies, nanobodies and single-chain variable fragments (scFvs) are smaller in size, allowing for improved tumor penetration. Bispecific constructs can bridge immune cells, such as T cells or NK cells, with tumor cells, leading to enhanced immune-mediated tumor destruction.

Tyrosinase OC provides a versatile synthetic approach to generating bispecifics with diverse protein architectures, such as scFv-nanobody conjugates or bispecific nanobodies. Additionally, this method enables precise control over bispecific geometry, allowing us to investigate how structural variations impact therapeutic efficacy and anti-tumor activity. By exploring the relationship between bispecific geometry and functional performance, we aim to optimize their design for enhanced anti-tumor responses in cancer immunotherapy.

Hanzhang Jin, UC Davis, Chen Glyco Group

The accumulation, aggregation, and fibrillization of amyloid-β (Aβ) peptides play a central role in the pathology of Alzheimer’s disease (AD), although the regulatory mechanisms governing these aggregation states and their bioactivities remain unclear. Aβ is produced through the neuronal processing of amyloid precursor protein (APP), a transmembrane protein with signaling receptor functions. Given the importance of glycans on glycoproteins and glycopeptides in various biological processes and their potential applications in immunotherapy, understanding their structural roles in modulating Aβ aggregation is critical. However, the heterogeneity of glycoproteins and glycopeptides due to variations in glycosylation sites and glycan forms complicates their study. To address this, efficient chemoenzymatic synthetic strategies were developed to produce homogeneous amyloid-β peptides containing O-GalNAc tyrosine, which is underestimated compared with other O-glycosylated residues and considered pathologically related. Chemically synthesized monosaccharide-modified amino acid building blocks were incorporated into glycopeptides through solid-phase peptide synthesis (SPPS), followed by enzymatic glycosylation using the stepwise one-pot multienzyme (StOPMe) strategy. The resulting glycopeptides were used to identify and characterize polypeptide N-acetylgalactosaminyltransferases (ppGalNAcTs) involved in this specific glycosylation, as well as to investigate glycopeptide interactions with lectins and glycan-binding antibodies. This research aims to investigate O-GalNAc-Tyr glycosylation by focusing on its presence and functional significance in Aβs and enhance the understanding of the role of glycosylation in Aβ aggregation and highlight the potential of synthetic glycopeptides for biomedical applications in neurodegenerative diseases and immunotherapy.

Alexandra Kent (Postdoc), UC Berkeley, Cate Lab

Incorporation of β-amino acids into peptides imparts proteolytic stability, unique architectures, and membrane permeability as demonstrated by β-amino acid containing natural products. Additionally, cyclic β-amino acids can act as turn inducers, and multiple incorporations can generate foldamers with rigid helical structures. However, incorporation efficiency and number of residues incorporated is highly dependent on the stereochemistry of the cyclic β-amino acid. We sought to improve incorporation of cyclic β-amino acids through structure guided mutations of the E. coli ribosome exit tunnel that allow for more space for sterically constrained monomers to progress through. We generated and purified two different mutant ribosomes with base changes in the 23S rRNA sequence, A2062C and A2062U. Both mutant ribosomes exhibit similar activity levels to wild type ribosomes in bioluminescence-based translation assays. Because proline is also a cyclic, sterically constrained amino acid, we challenged these mutant ribosomes with polyproline sequences of various lengths as an initial proxy for cyclic β-amino acids. Using liquid-chromatography mass-spectrometry of peptide products, we determined that ribosomes with an A2062U mutation improve translation of polyproline containing peptides. Conversely, A2062C ribosomes show a decrease in ability to translate polyproline containing sequences. A2062U and wild type ribosomes are also able to incorporate cyclic β-amino acids. Initial cryo-EM structures of cyclic β-amino acids in the PTC of the ribosome have provided some insight into their arrangement. We are currently obtaining further cryo-EM structures of these complexes to shed light on the difference in reactivities between ribosomes harboring A2062C and A2062U mutations. 

Kiyozumi Shuhei, UC Davis, Chen Glyco Group

 Human milk oligosaccharides (HMOs) play beneficial roles for breast-fed infants. Over 150 different HMO structures have been identified.1 The detailed structure-function relationship of HMOs, however, has not been fully explored.1 Access to individual HMOs in sufficient amounts for research and clinical studies is critical. We have developed highly efficient stepwise one-pot multienzyme (StOPMe) methods to synthesize complex HMOs from a Cbz-tagged lactoside, readily obtained from inexpensive lactose via simple chemical derivatization.2 Moreover, optimized enzymes required for synthesis are obtained in large quantities. The synthesized HMOs are readily purified via C-18 cartridge.3 An automated platform is being developed for large-scale synthesis of HMOs regardless of their structural complexity. The programmable automated reactor system provides precise controls of reaction time, pH, and amounts of enzyme being added, further improving the efficiency of the StOPMe process for multigram-scale synthesis of HMOs.

Jordan Kleinman, UCSF, Fujimori Lab

The oxazolidinones are a class of fully synthetic ribosome-binding antibiotics employed as a last line of defense against multidrug resistant gram-positive pathogens. Although these compounds were initially understood to act as global inhibitors of protein synthesis, recent work has shown that they in fact inhibit elongation in a nascent peptide dependent manner. Specifically, the clinical frontrunner compound linezolid exhibits context-specific ribosome stalling when alanine is present at the -1 position of a growing nascent peptide (Ala -1). Similar sequence-specific stalling has been observed in numerous natural product antibiotics, such as erythromycin and chloramphenicol. Bacteria take advantage of this feature, relying on context-specific stalling by an antibiotic to induce expression of resistance features, thereby mitigating the activity of said antibiotic. Thus, an in-depth understanding of the structural features which contribute to this context-specific stalling is critical to the design of improved antimicrobial compounds. The oxazolidinones are an ideal candidate for this exploration due to the minimal nature of their existing context-specificity and the simplicity of their structure compared to most natural product antibiotics. We use a combination of ribosome profiling, toeprinting, inverse toeprinting, and cryo-EM to explore how structural variation to the oxazolidinone core effects their sequence-specificity via interactions with nascent peptide and ribosomal rRNA on the bacterial ribosome. In particular, ribosome profiling elucidates a broader stalling profile with slight Ile -1 specificity for the second FDA-approved oxazolidinone, tedizolid. Cryo-EM shows structural characteristics of this modification, especially due to tedizolid’s shorter C5 substituent vs that of linezolid.

Isabel Lee, UCSF, Seiple Lab

Approximately two million people are infected with antibiotic-resistant strains of bacteria annually due to the growing resistance of both Gram-positive and Gram-negative bacteria to antibiotics. Of this group of individuals, more than 35,000 die as a result of these infections. An increasing prevalence of antibiotic resistance mechanisms, especially in bacteria found in hospitals, has rendered many antibiotics ineffective. With growing knowledge of antibiotic resistance mechanisms, the design of novel antibiotic analogs that can interfere with these specific mechanisms is now possible. From the development of a modular synthesis of Group A streptogramins in the Seiple Lab, a small library of group A streptogramin analogs was synthesized by modifying the scaffold of virginiamycin M2. One promising hit compound showed improved antibiotic activity against multiple bacteria-resistant strains, including Vat-resistant S. aureus, which contained a modified C4 sidechain compared to that of the parent molecule. However, of this library, only one C4 analog was synthesized. This hit compound provides a good starting point in designing novel C4-modified analogs that make increased ribosomal binding interactions and can overcome other resistance mechanisms, specifically ABC-F and cfr resistance. 

Bryant Luu, UC Davis, Atsumi Lab

"Breastmilk is an ideal and highly specialized food source for infants. After lipids and lactose, human milk oligosaccharides (HMOs) are the 3rd most abundant nutrient in breast milk, that serve as prebiotics, immune system modulators, and supporters of neurocognitive development.

The study and inclusion of HMOs in infant formula and therapeutics is limited by our inability to cheaply synthesize large, pure quantities of HMOs. Strategies such as chemoenzymatic synthesis have achieved stepwise production of complex HMOs but require multiple synthetic steps and costly cofactors. As such, new production strategies are necessary to generate these structures. One such strategy, microbial biosynthesis, has the potential to produce HMOs efficiently from cheaper starting materials, allowing for cheaper production and greater accessibility of these molecules.

To more cost effectively produce HMOs, we developed a biosynthetic pathway leveraging inexpensive starting materials. We find that the model organism Escherichia coli can be engineered to produce HMOs from inexpensive starting materials, using recombinant heterologous genes and balancing metabolite flux through native metabolic pathways. We also characterize HMOs’ prebiotic effect on multiple infant-associated Bifidobacterium species. Increased availability and understanding of HMOs will allow for their use in infant formula and as potential food additives, so that people can live improved, healthier lives."

Jeus Madrigal, UCSF, Seiple Lab

The resorcinolic macrolide (RM) class of natural products is characterized by a resorcinolic ester core linked to a 10-14-membered macrocyclic lactone, encompassing over 50 members with diverse biological activities. Among the most notable is radicicol, a potent natural inhibitor of the 90-kDa heat-shock protein (Hsp90). Here we present an enantioselective, modular synthetic strategy to access diverse RMs,  focusing on the modification of the C15 position, covalent warheads, and rigidity of the macrocycle to explore their effects on Hsp90 inhibitory activity and paralog selectivity. We synthesized 27 RM derivatives and evaluated their inhibitory activity against Hsp90𝛼 and Hsp90β paralogs and KRas. Compound 12A exhibited the highest potency with IC50 values of 14 µM and 37 µM against Hsp90𝛼 and Hsp90β, respectively. Additionally, compound 1C displayed high selectivity for Hsp90𝛼. These findings indicate that targeted modifications at C15 can yield novel RMs with potential as biological probes and therapeutic leads, offering a new avenue for the development of Hsp90 or KRas inhibitors with reduced resistance and improved specificity.

Chandrima Majumdar (Postdoc), UC Berkeley, Cate Lab

The Escherichia coli ribosome has been shown to accept non-α-amino acid substrates with diverse chemistries such as β2 and β3-amino acids, aminobenzoic acids and α- or β-hydroxy acids. These substrates can allow for the recombinant synthesis of sequence defined, non-protein or hybrid polymers. In addition, peptides constructed with β-amino acids can have unique conformations and improved proteolytic stability, making them valuable synthetic targets for medicinal chemistry applications. Although it has been shown that these molecules are substrates for the E. coli ribosome, the efficiency of their incorporation into peptides remains low and are largely confined to in vitro systems. However, it has recently been reported that a series of β3 bromo-substituted phenylglycine monomers were site-specifically incorporated into green fluorescent protein (GFP) in cells, and the efficiency of incorporation was dependent upon the specific regioisomer used. We aim to understand the structural basis for the differences in incorporation efficiency of the ortho-, meta- and para- regioisomers of bromo-phenylglycine by obtaining cryo-EM structures of the WT ribosome complexed with these monomers. These structures are expected to reveal differences in accommodation of the monomers and the orientation of the reactive groups that would affect the ability of these monomers to undergo peptide bond formation. Additionally, we carried out in vitro translation reactions to evaluate the differences in in vitro translation activity and correlate the observed differences in reactivity with our structural findings. Taken together, the structural and biochemical approach will guide the design of new monomers that may serve as substrates for the ribosome, enabling access to more diverse genetically encoded materials. 

Melody Malek, UC Davis, David Lab

Oxidative DNA damage, caused by products of cellular respiration and inflammatory responses, ionizing radiation, or environmental toxicant exposure is believed to be a major underlying cause of cancer. The DNA base guanine is particularly susceptible to oxidative damage, given its low redox potential, the foremost product of which is 8-oxo-7,8-dihydroguanine (OG). The OG lesion is particularly insidious, given its dual coding potential to base pair with C or A. The Hoogsteen face of OG preferentially base pairs with adenine’s Watson-Crick face. The BER glycosylase human MutY (MUTYH) catalyzes the cleavage of the N-glycosidic bond to the undamaged, yet incorrectly templated, adenine base. The failure of MUTYH to perform BER is detrimental, and leads to G:C to T:A transversion mutations. As such, biallelic inheritance of variants of MUTYH is associated with a mutator phenotype and an increased susceptibility to cancer in a syndrome known as MUTYH associated polyposis (MAP). Many of these inherited cancer-associated variants surround the two metal cofactors MUTYH harbors: a [4Fe-4S]2+ (Fe-S cluster) and a mononuclear zinc ion. Recent work in our lab demonstrates a functional interplay between the Fe-S cluster and active site, and it has been suggested that the Zn2+ plays an important role in regulating MUTYH activity by engaging its two functional domains. In order to understand the importance of the Fe-S cluster and the zinc linchpin motif, we have historically relied on using the recombinant protein—characterizing the kinetics, DNA-binding affinity, and capacity to repair OG:A lesions. However, being that these variants could behave differently within the context of the cellular environment (where MUTYH’s protein partners and the entire BER machinery are present), we can understand how coordinated repair occurs if we utilize mammalian cell-based repair assays. Briefly, we generate stable cell lines of the MAP variants, as well as a synthetic OG:A-containing plasmid. After the plasmid is transfected in the cells, we can assess OG:A repair capacity by the variants via flow cytometry using a GFP-based mammalian cell reporter assay developed by the David Lab. This study aims to understand how perturbations near the Fe-S cluster and zinc linchpin motif impact catalysis, and our results corroborate recent findings in our lab which suggest that allosteric cross-talk occurs between the cluster and the active site. 

Aashrita Manjunath, UC Davis, Beal Lab

Adenosine Deaminases Acting on RNA (ADARs) are members of a family of RNA editing enzymes in metazoans that catalyze the conversion of adenosine to inosine in double-stranded RNA (dsRNA). Inosine is read as guanosine by cellular machinery, so ADARs effectively catalyze an A-to-G edit. ADARs’ selective activity on dsRNA presents the ability to correct aberrant adenosine mutations in the transcriptome through a process known as Site Directed RNA Editing (SDRE). SDRE as a therapeutic tool relies on antisense guide oligonucleotides (ASOs) to create a duplex RNA substrate at discrete, disease-relevant sites. As these targets do not always feature classical editing motifs that ADARs prefer, rational design of ASOs is crucial to recruiting ADARs to these non-canonical targets. Specifically, the base directly adjacent to the target adenosine in the 5’ direction (5’-G) is of particular consequence to ADARs’ ability to efficiently catalyze the editing event. Our work is focused on the incorporation of chemical modifications to the ASO at the position paired opposite the 5’ adjacent site to assist in recruiting ADARs to disfavored, yet therapeutically relevant targets. To this end, we have overcome editing deficiencies in ADARs’ least preferred target sequence contexts both in vitro and in cellula by incorporating nucleoside analogs to improve RNA-protein interactions while increasing nuclease resistance via backbone modifications to the antisense guide. X-Ray crystallography confirms favorable structural motifs induced by these analogs that enable the editing event at 5’-G sites. Expanding the chemical toolbox used to improve ADAR editing broadens the scope of disease targets that ADARs can edit, maximizing the therapeutic potential of A-to-I editing. 

Jenna Manske, UC Berkeley, Hartwig Lab

Enzymatic halogenation is a promising alternative to traditional synthetic methods for the installation of halogens at the site of C(sp3)–H bonds. This method enables precise and stereoselective modification of complex molecules with simple reagents under mild conditions. However, there are few halogenases that operate on free-standing complex molecules and can accommodate a diverse set of non-native substrates. Additionally, directed evolution of halogenases can suffer from low selectivity for halogenation over oxygenation of non-native substrates. We harness an anchoring group strategy wherein we tether non-native substrates to an indole moiety present on the native substrate of WelO5*, a non-heme alphaKG-dependent halogenase. Rational mutagenesis affords enzyme variants with high selectivity and activity towards chlorination of C(sp3)–H bonds of non-native substrates with varied structures. Furthermore, we demonstrate that cleavage of the anchoring group gives free halogenated natural products

Vineet Mathur, UCSF, Renslo and Jun Lab

Despite the versatility and advantages of conventional liposomes (CL), only 4 liposome-based treatments for cancer therapies are available due to limitations with tissue targeting and cellular uptake. Traditional liposomes are up taken via endocytic pathways which is a relatively inefficient cellular process. Furthermore, payloads often accumulate in lysosomes where they immediately degrade. Thus, payloads delivered via endocytosis are limited to small molecule chemotherapeutics with known liposomal compatibility. 

Alternatively, fusogenic liposomes (FL) are a class of liposomes that deliver their payloads directly to the cytosol via membrane fusion. This fusogenic characteristic affords far greater delivery efficiencies compared to endocytic liposomes. To date FLs have been successfully demonstrated through the in vitro delivery of payloads with poor cell permeability such as proteins and nanoparticles, however, their application with clinically relevant payloads, such small molecule therapeutics and oligonucleotides has been studied sparsely. Additionally, standard FLs fuse rapidly and ubiquitously and therefore have poor tissue targeting and pharmacokinetic (PK) properties. As a result, the goal of this work is to develop a novel liposomal delivery platform that can transform from a stable endocytic liposome to a highly efficient fusogenic liposome in target microenvironments providing a stable, selective, and high efficiency drug delivery system. 

Erin McCann, UC Berkeley, Francis Lab

RNA-based therapeutics have gained significant interest in recent years due to their potential to treat a wide range of diseases. RNA treatments possess a broad array of applications because there are several types of RNA, including small interfering RNA (siRNA), microRNA (miRNA) and antisense oligonucleotides (ASOs) which can be used to modulate gene expression, as well as mRNA which can be used to encode therapeutic proteins. Though there is much excitement around the potential of RNA-based therapies, delivery remains a significant challenge because RNA molecules are highly susceptible to nuclease degradation, and they are too large and anionic to passively cross cell membranes. Effective RNA delivery requires a carrier to transport it into the cytoplasm and protect it from nuclease degradation. Bacteriophage MS2 virus-like particles (VLPs) show promise as RNA delivery vehicles. These self-assembling protein capsids can encapsulate various cargo and have been used as a delivery platform for small molecule drugs and peptides. Recently, a double mutant, T71K/G73R, of MS2 (KR MS2) was found to have enhanced internalization into mammalian cells, laying the groundwork for further optimization of MS2 as a delivery platform. In this work we have successfully encapsulated both siRNA and mRNA within KR MS2 using in vitro and in vivo encapsulation methods. Efforts to explore functionality of the delivered RNA are currently ongoing. We aim to elucidate the mechanism of RNA release from the KR MS2 VLP to guide future engineering and/or modification of the capsid to optimize for RNA delivery.

Alex McGill, UC Davis, Atsumi Lab

Plastic pollution and increasing atmospheric carbon dioxide levels are two environmental challenges facing the planet. Microorganisms have been used by humans for thousands of years for a variety of purposes relating to food, fermentation, and medicine. Today, microorganisms can be engineered and deployed to produce chemical compounds that would normally be synthesized via petroleum-based chemistry, including precursors to biodegradable polymers. Unlike traditional synthesis methods for producing bio-degradable polymer precursors, microbial systems can directly utilize atmospheric CO2 to generate these compounds. Cyanobacteria are prokaryotic microorganisms that use photosynthesis to convert CO2 into organic compounds that are used for energy and biomass. Synechococcus elongatus PCC 7942 is a well-studied cyanobacterium that has been engineered for the microbial production of chemicals from CO2 directly. Interestingly, this microorganism has a unique and incomplete TCA cycle where α-ketoglutarate cannot be converted into succinate. Here, we aimed to extend this cyanobacterium’s incomplete TCA cycle to produce bio-degradable polymer precursors. To achieve this, three separate biosynthetic pathways were installed into the genome and tested for precursor production. Upon identification of a biosynthetic pathway capable of producing the precursor, additional genes were installed to increase carbon flux towards the TCA cycle, and this resulted in over a 2.5-fold improvement in production. Host genes were also knocked out to direct carbon flux towards the target, and this also resulted in improved production. This study demonstrates the microbial production of a biodegradable-polymer precursor, addressing both plastic pollution and rising atmospheric CO2 levels simultaneously. 

Lauren Orr, UC Berkeley, Nomura Lab

We have discovered a fumarate-based covalent fragment, which when appended to various protein-targeting ligands, is able to induce the target's ubiquitin-dependent degradation.  Using a functional genetic FACS-based screen, we identified that a BRD4 degrader with this covalent fragment depends on DCAF16, a cysteine-rich cullin RING ligase 4 (CRL4)-associated E3 ligase.  Using proteome-wide chemoproteomics, we confirmed that the fumarate directly engages DCAF16, and mutagenesis studies revealed that the fumarate likely binds covalently to C173, which is a different cysteine than other known DCAF16-targeting covalent degraders.  We synthesized drug derivatives using this fumarate which were able to degrade BRD4, CDK4/6, BTK, SMARCA2/4, and AR, as well as the AR truncation variant ARv7, which is currently regarded as undruggable.  Overall, our study bolsters the significance of DCAF16 as an important tool for targeted protein degradation, since its inherent flexibility and possession of multiple cysteines makes it a prime candidate for inducing proximity to diverse neosubstrate proteins.

William Rackear, UC Berkeley, Z. Zhang Lab

Ankylosing Spondylitis (AS) is a debilitating autoimmune disease that affects more than 200,000 people in the United States each year. Despite the well-documented characteristic symptoms – fused vertebrae, hunched posture, and difficulty breathing – there is currently no cure. However, with a comprehensive investigation of its molecular basis, potential treatments can be developed. Studies have identified strong correlations between AS and the expression of the MHC-I protein corresponding to the allele, HLA-B*27:05. Population data indicate that HLA-B*27:05 is involved in presenting self-peptides that stimulate immune attack of healthy cells. Crystal structures of HLA-B*27:05 reveal an unpaired cysteine residue directly adjacent to its peptide binding groove at position 67, providing an opportunity for a small-molecule targeted therapy. We hypothesize that preventing peptide binding to HLA-B*27:05 through treatment with a small molecule could prevent this autoimmune attack. In this work, we have identified a small molecule scaffold that can bind to Cys67 and act as a pharmacological chaperone to support the formation of the MHC-I complex in the absence of a peptide ligand. As a result, we are able to disrupt problematic antigen presentation and subsequently suppress unwanted T-cell activation.

Laura Rodriguez-Velandia, UCSC, Sanchez Lab

Microbial drug discovery has been revolutionized by instrumentation with increased resolving power and the development of computational tools that assist in sample and analyte prioritization. One such tool, IDBac, facilitates the rapid identification of natural products and their producing bacterial strains. We hypothesize that integrating trapped ion mobility (TIM) as an additional dimension of separation will increase the detection and prioritization of natural products and the prioritization of their producing source. As a proof of concept, we used liquid chromatography coupled to TIM-MS for the rapid identification of surfactin, a known cyclic lipopeptide produced by a known source: Bacillus subtilis. We are currently applying IDBac and TIM to identify unknown bacterial strains associated with the banana slug and their metabolite production. 

Maria Sajimon, UCSC, Raskatov Lab

Alzheimer's disease is a growing global challenge that imposes tremendous burden to the society and economy. Though recently approved anti amyloid immunotherapies show effectiveness in clearing abeta and modest slowing in cognitive decline, the removal of cerebral abeta can cause serious adverse events. Therefore decreasing detrimental effects of abeta in the brain without physically removing is an unmet need in the field. Site-specific deamidation can show offer modulation of toxic effects and aggregation of abeta by passing the need of removing from the brain. Here we show that deamidation of Asparagine 27 can reduce toxicity and decrease microglial expression, pronounced decrease in aggregation and oligomers. This approach of modulating abeta properties represent first novel approach of altering detrimental effects of abeta without removing it from the brain.

Robert Shepherd, UCSC, Sanchez Lab

Natural products (NPs) have played a key role in drug discovery and design. The ‘tried-and-true’ method of NP discovery via bioactivity-guided fractionation, while effective, can be expensive and often results in the rediscovery of known biologically active compounds. Advances in ‘dereplication’ have decreased NP rediscovery rates. Data-driven methods for prioritizing ‘talented’ NP-producing microbial strains prior to lengthy fermentation, extraction, and fractionation procedures could positively impact the dereplication process. In this work, I showcase the use of matrix assisted laser desorption/ionization mass spectrometry (MALDI-MS) to pseudo-phylogenetically cluster a variety of bacteria based on MALDI protein spectral ‘fingerprints’. Additionally, the use of MALDI tandem mass spectrometry (MS/MS) directly from colonies spread onto target plates allowed for the rapid annotation of bioactive molecules produced by the same strains. In total, two MALDI-MS campaigns, spanning 35 strains total, identified at least 7 previously discovered bioactive compounds, and yielded MS/MS data suggestive of unreported NPs. Ongoing efforts are focused on isolating and identifying these compounds and comparing the dereplication similarities and differences between MALDI-MS/MS data and  LC-ESI-MS/MS data.

Alex Solivan, UC Berkeley, Schepartz Lab

Macrocycles and atropisomers are among the most hotly pursued scaffolds in medicinal and process chemistry due to their ability to impart tremendous conformational control on overall molecular architecture. These interactions can become mutually reinforcing and can serve to dramatically improve molecular recognition, especially in medium sized molecules (>500 Da) such as peptides. Here, we present the first example of a biocompatible peptide macrocyclization strategy that simultaneously generates a stable class-III atropisomer in the macrocyclic backbone. This strategy relies on an aqueous Friedländer macrocyclization reaction between a N-terminal β-keto tertiary-amide and the sidechain of an internal kynurenine residue. The resulting tertiary amide-quinoline bond is rotationally hindered by ortho-substitutions on the quinoline, which gives rise to stable conformational isomers, each of which we characterized with MicroED, NMR, and molecular dynamics simulations. Atropisomerization of these compounds is defined by a flip in the orientation of a tertiary amide carbonyl and the concurrent reorganization of intramolecular H-bonding networks. Together, the local stereochemical axis and the long-range H-bonding patterns elicit strong control over global conformation of the atropisomeric peptide macrocycles. In summary, we present a unified chemical strategy for the introduction of aryl-amide atropisomeric units during peptide macrocyclization, with fascinating implications for macrocyclic peptide conformational control. 

Nora Spulin, UC Davis, Heffern Lab

Advanced glycation end-products (AGEs) are heterogeneous compounds formed through non enzymatic reactions between reducing sugars and proteins, lipids, or nucleic acids. This process occurs when reducing sugars react with amino groups of lysine, arginine, or N-terminal amino acid residues – leading to irreversible AGE formation. These post-translational modifications (PTMs) have been associated with a host of pathological conditions, including diabetes, cardiovascular, and neurological diseases; however, the factors controlling their formation and accumulation are not fully understood. Research has revealed a potential link between AGE formation and oxidative stress. Interestingly, oxidative stress has also been linked to the dyshomeostasis of redox-active metals in similar diseases where AGE accumulation is observed. My research seeks to determine the potential connection between disrupted metal homeostasis and AGE formation, aiming to elucidate their biochemical potential as biomarkers. Here, we present our initial efforts to determine how iron influences the glycation states of a model protein, Bovine Serum Albumin (BSA). Utilizing fluorescence spectrophotometry, we found that AGE formation is elevated in vitro with iron, evidenced by strong fluorescent intensity over a prolonged period. Further analysis of the modified proteins was investigated utilizing Matrix Assisted Laser Desorption Ionization-Time of Flight-Mass Spectrometry (MALDI-TOF-MS) of trypsin digested samples. MALDI-TOF-MS analysis revealed unique fingerprinting regions between glycated BSA samples with and without metals, suggesting that metal presence significantly modifies peptide fragmentation patterns during AGE formation. This analytical approach provides valuable information into metal-induced AGE formation patterns. Future work seeks to correlate specific peptide sequences to identify the sites and types of glycations.  Understanding these metal-dependent mechanisms will be essential for decoding the biological functions of AGEs and their potential as biomarkers. 

Nhu Tong, UC Davis, Atsumi Lab

Industrialization and petroleum-based manufacturing of building materials have resulted in alarming emissions of greenhouse gases such as carbon dioxide. Therefore, research into sustainable methods of producing construction materials should be highly prioritized as a renewable alternative for petroleum, which prompted the emergence of engineered living materials (ELMs), a field founded on the integration of polymer science and synthetic biology. Within ELMs, microbes are encapsulated in polymeric matrix to generate biologically programmable materials, in which the living component can be genetically engineered to modulate the function of its surrounding matrices. The ability of hydrogel polymers to hold water for extended time makes cell encapsulation and survival feasible for long-term microbial chemical production. ELMs are also remarkable in the fact that the living components can be programmed to respond to external stimuli and self-regenerate the encapsulating material. 

Various studies have demonstrated biosynthesis by embedded heterotrophs such as Escherichia coli and yeast in hydrogels. However, these organisms need carbohydrate substrates such as glucose. This feedstock can be generated by an autotroph such as the model cyanobacterium Synechococcus elongatus PCC 7942 (7942). Cyanobacteria have garnered much attention for biosynthesis due to their photosynthetic capabilities and utilization of CO2 as a renewable carbon source. Since there is limited information on the ability to encapsulate cyanobacteria in polymeric materials, in this study, 7942 is used for establishing ELMs. To develop an ELM with cyanobacteria, we first screened for polymeric materials biocompatible with 7942. Once this assessment was accomplished, an established 2,3-butanediol production strain of 7942 was used to demonstrate 7942’s ability to biosynthesize chemicals inside an ELM. The results illustrate a remarkable biocompability of bovine serum albumin-polyethylene glycol diacrylate (BSA-PEGDA) with 7942, as evident in the healthy green color of encapsulated 7942 for prolonged time. Furthermore, this enables the biosynthesis of 2,3-butanediol by embedded cyanobacteria. Using optimized culturing conditions, we will then genetically engineer 7942 to produce sucrose, which will be utilized as a feedstock to sustain a coculture of 7942 and a heterotrophic organism. A comprehensive understanding of engineered living materials using 7942 will have far-reaching applications as building materials and sources of bioremediation, which would address both CO2 emissions and more sustainable manufacturing practices. 

Kyle Vanderschoot, UC Davis, Neumann Lab

HIV disrupts lipid metabolism, contributing to renal dysfunction in acute kidney injury, chronic kidney disease, and HIV-associated nephropathy. While antiretroviral therapy (ART) slows viral replication, prolonged use induces lipid dysregulation and nephrotoxicity. Viral-driven metabolic reprogramming, cholesterol-rich lipid raft exploitation, and inflammation-driven lipid synthesis pathways may drive these changes. Using MALDI MSI, we investigate lipidomic alterations in SIV-infected and ART-treated rhesus macaque kidneys, hypothesizing that these changes vary across nephrotic functional units. MALDI MSI enables spatial analysis of lipidomic shifts within kidney histological substructures. Prior studies link HIV/SIV infection to lipidemia and long-term renal complications. Phospholipids, critical for membrane function and immune signaling, showed significant alterations across nephrotic units. A bisecting k-means segmentation algorithm identified functional units, while PCA revealed distinct lipid profiles in glomeruli from uninfected, SIV-infected, and ART-treated groups. ART mitigated some virus-induced lipid upregulation but maintained a unique profile. Key lipid species globally affected by infection included PC(36:4), SM(36:1;O2), PA(44:10), PS(38:5), and PI(34:1). Segmented glomeruli from SIV-infected kidneys were marked by SM(d34:1), HexCer(d30:1), and PC(38:4), while SIV/ART-related changes included PC(36:2), PE(44:10), and PE(O-40:7). ART-specific alterations were evident with PA(38:5), PA(42:10), CerP(d44:2), PA(36:3), PC(36:5), and PC(38:7). Despite ART reducing some disruptions, it generates unique lipid signatures linked to membrane dynamics, immune modulation, and apoptotic pathways. Our findings highlight lipidomic profiling as a tool for understanding SIV-associated kidney injury and ART’s unintended effects on lipid metabolism.

Anna Vernier, UC Davis, Olson Lab

The serotonergic psychedelic and psychoplastogen N,N-dimethyltryptamine (DMT) has great therapeutic potential for various neuropsychiatric disorders, such as depression, anxiety, post-traumatic stress disorder, and substance use disorder, due to its ability to rapidly promote structural and functional neuroplasticity. However, its hallucinogenic effects greatly limit the potential patient population; therefore, decoupling these effects from its therapeutic properties may enable the development of a fast-acting, efficacious, accessible, and scalable treatment for neuropsychiatric disorders. Previously, DMT analogs such as iso-DMT and AAZ-A-154 have been shown to be non-hallucinogenic while maintaining DMT’s psychoplastogenic qualities. This work aims to elucidate the basis of DMT’s hallucinogenic effects by modulating its structure and evaluating its activity using the conformation-based biosensor assay for hallucinogenic potential, psychLight, to inform future drug development.

Celine Wang, UC Berkeley, Z. Zhang Lab

β-Lactones are highly strained electrophiles that have previously been exploited by both natural products and synthetic compounds to target both catalytic and noncatalytic serines, threonines, or glutamic/aspartic acids. In this project, we have synthesized complete sets of stereochemically differentiated β-lactone probes and characterized their cellular targets. We show that compounds with identical chemical scaffolds but diasteromeric electrophiles capture distinct subsets of cellular proteins. We further utilize these electrophiles in the design of chemical ligands against drug-resistant mutant BTK(C481S). Unexpectedly, we discover that our ligands retained activity with Cys481 in wildtype BTK but reacted with the catalytic Lys430 residue in the mutant kinase. The ability of β-lactones to engage multiple nucleophiles with stereochemically encoded selectivity highlight their versatility as privileged electrophiles that expand the chemical tractability of the proteome.

Paloma Whitworth, UC Berkeley, Francis Lab

Amphiphiles are a broad class of molecules which act to solubilize and emulsify otherwise immiscible chemicals and are essential formulants of most products including detergents, pharmaceuticals, and herbicides. Despite their ubiquity, commonly used amphiphiles such as alkylphenol ethoxylates are linked to adverse human health effects such as endocrine disruption and contribute to ecological disruption through processes like eutrophication. These concerns have prompted the development of biologically derived surfactants. Previously, our group developed one such surfactant based on an intrinsically disordered protein (IDP) appended to an engineered hydrophobic peptide chain. These protein-based surfactants can self-assemble at micromolar concentrations and deliver otherwise insoluble drugs to human cancer cells.

To expand the chemical diversity and properties of protein-derived surfactants, we aim to engineer hybrid protein-based surfactants. These amphiphiles will combine the exceptional water solubility of the IDP sequence with the chemical diversity of synthetic hydrophobic tails. Using a tyrosinase-mediated oxidative coupling strategy, a library of hydrophobic molecules will be covalently attached to the terminal tyrosine residue of the IDP to create a diverse class of hybrid surfactants. We aim to tune these hybrid surfactants to different applications including the encapsulation of hydrophobic cargo for drug delivery.

Miranda Wu, UC Berkeley, Z. Zhang Lab

Phosphatidylinositol 3-kinases (PI3Ks) play a crucial role in cell signaling pathways, regulating cell proliferation, growth, and metabolism. Thus, PI3K activity is highly regulated, yet mutations of PI3Ks are among the most frequent mutations in human cancers. The H1047R hotspot mutation on PI3Kα is the most common mutation, increasing kinase activity through an activating conformational change of the membrane binding motif found at the C-terminus. Currently, many small molecules that inhibit PI3K disrupt catalytic activity by outcompeting ATP, offering little mechanistic opportunity for selective mutant inhibition. We hypothesize that by exploiting an allosteric pocket directly adjacent to the H1047R residue, we can selectively target the mutant arginine. Expanding upon previous work on selective engagement of K-Ras(G12R) via an arginine reactive warhead, we are designing small molecules containing arginine reactive warheads and specificity to the allosteric pocket on PI3Kα to explore how modification of the mutant arginine residue disrupts PI3Kα(H1047R) activity. Unlike current inhibitors, this strategy is an irreversible, covalent modification directly at the site of mutation.

Deborah Zhuang, UC Berkeley, Francis Lab

Cyanobacteria are arguably one of the most successful organisms on Earth, inhabiting a wide range of ocean, fresh water, soil, and even desert environments on every continent. The cyanobacterial phycobilisome consists of two principal light-collecting moieties within its structure. It has a central core made up of allophycocyanin (APC) disk-like protein-pigment complexes that are surrounded by rods of phycocyanin (PC) disks, and occasionally other hypsochromic light-absorbing antenna complexes as well. The ways in which the energies of the specific chromophores are tuned by the proteins of this system to achieve its high efficiency and directional energy transfer are not fully understood. Early examinations of fully intact PC structures isolated from native organisms using time-resolved spectroscopy have revealed estimates for EET rates between individual chromophores, but while this has provided insight into the systems as they occur in nature, complex combinations of decay pathways are occurring simultaneously and competitively and thus are difficult to differentiate in fully intact systems. 

We have recombinantly expressed a fully functioning PC complex, and selectively created minimal chromophore sets to study their individual contributions to the overall PC spectrum. Structural and computational analysis of this protein system has provided a greater understanding of how the protein environment serves to alter the photophysics of each of these chromophores. Introduction of a quencher into various positions within PC confirms the ability of the protein environment to tune the directionality of energy transport in this assembly. Further explorations of the role of key interactions with the pigments were explored, showing the deeper insight this approach provides to better understanding structural impacts on the EET mechanism in the phycobilisome. 

As another approach, we have constructed a site-saturation variant library of APC to determine in a high-throughput manner the sequence-structure-function relationship between the protein environment and the spectral output of the chromophores. This has revealed an array of mutants which show comparable spectral properties to the wild-type protein, through which we have performed molecular dynamics simulations to provide structural insight into red-shifted emission, enhanced chromophore coupling, or higher emission intensities of select mutants.