In vivo selection in non-human primates identifies AAV capsids for on-target CSF delivery to spinal cord.

Systemic administration of adeno-associated virus (AAV) vectors for spinal cord gene therapy has challenges including toxicity at high doses and pre-existing immunity that reduces efficacy. Intrathecal (IT) delivery of AAV vectors into cerebral spinal fluid can avoid many issues, although distribution of the vector throughout the spinal cord is limited, and vector entry to the periphery sometimes initiates hepatotoxicity. Here we performed biopanning in non-human primates (NHPs) with an IT injected AAV9 peptide display library. We identified top candidates by sequencing inserts of AAV DNA isolated from whole tissue, nuclei, or nuclei from transgene-expressing cells. These barcoded candidates were pooled with AAV9 and compared for biodistribution and transgene expression in spinal cord and liver of IT injected NHPs. Most candidates displayed increased retention in spinal cord compared with AAV9. Greater spread from the lumbar to the thoracic and cervical regions was observed for several capsids. Furthermore, several capsids displayed decreased biodistribution to the liver compared with AAV9, providing a high on-target/low off-target biodistribution. Finally, we tested top candidates in human spinal cord organoids and found them to outperform AAV9 in efficiency of transgene expression in neurons and astrocytes. These capsids have potential to serve as leading-edge delivery vehicles for spinal cord-directed gene therapies.

Production of cholesterol-like molecules impacts Escherichia coli robustness, production capacity, and vesicle trafficking.

The economic viability of bioprocesses is constrained by the limited range of operating conditions that can be tolerated by the cell factory. Engineering of the microbial cell membrane is one strategy that can increase robustness and thus alter this range. In this work, we targeted cellular components that contribute to maintenance of appropriate membrane function, such as: flotillin-like proteins, membrane structural proteins, and membrane lipids. Specifically, we exploited the promiscuity of squalene hopene cyclase (SHC) to produce polycyclic terpenoids with properties analogous to cholesterol. Strains producing these cholesterol-like molecules were visualized by AFM and height features were observed. Production of these cholesterol-like molecules was associated with increased tolerance towards a diversity of chemicals, particularly alcohols, and membrane trafficking processes such as lipid droplet accumulation and production of extracellular vesicles. This engineering approach improved the production titers for wax-esters and ethanol by 80- and 10-fold, respectively. Expression of SHC resulted in the production of steroids. Strains engineered to also express truncated squalene synthase (tERG9) produced diplopterol and generally did not perform as well. Increased expression of several membrane-associated proteins, such as YqiK, was observed to impact vesicle trafficking and further improve tolerance relative to SHC alone, but did not improve bio-production. Deletion of YbbJ increased lipid droplet accumulation as well as production of intracellular wax esters. This work serves as a proof of concept for engineering strategies targeting membrane physiology and trafficking to expand the production capacity of microbial cell factories.

A systematic framework for using membrane metrics for strain engineering.

The cell membrane plays a central role in the fitness and performance of microbial cell factories and therefore it is an attractive engineering target. The goal of this work is to develop a systematic framework for identifying membrane features for use as engineering targets. The metrics that describe the composition of the membrane can be visualized as "knobs" that modulate various "outcomes", such as physical properties of the membrane and metabolic activity in the form of growth and productivity, with these relationships varying depending on the condition. We generated a set of strains with altered membrane lipid composition via expression of des, fabA and fabB and performed a rigorous characterization of these knobs and outcomes across several individual inhibitory conditions. Here, the knobs are the relative abundance of unsaturated lipids and lipids containing cyclic rings; the average lipid length, and the ratio of linear and non-linear lipids (L/nL ratio). The outcomes are membrane permeability, hydrophobicity, fluidity, and specific growth rate. This characterization identified significant correlations between knobs and outcomes that were specific to individual inhibitors, but also were significant across all tested conditions. For example, across all conditions, the L/nL ratio is positively correlated with the cell surface hydrophobicity, and the average lipid length is positively correlated with specific growth rate. A subsequent analysis of the data with the individual inhibitors identified pairs of lipid metrics and membrane properties that were predicted to impact cell growth in seven modeled scenarios with two or more inhibitors. The L/nL ratio and the membrane hydrophobicity were predicted to impact cell growth with the highest frequency. We experimentally validated this prediction in the combined condition of 42 °C, 2.5 mM furfural and 2% v/v ethanol in minimal media. Membrane hydrophobicity was confirmed to be a significant predictor of ethanol production. This work demonstrates that membrane physical properties can be used to predict the performance of biocatalysts in single and multiple inhibitory conditions, and possibly as an engineering target. In this manner, membrane properties can possibly be used as screening or selection metrics for library- or evolution-based strain engineering.

Streamlined assessment of membrane permeability and its application to membrane engineering of Escherichia coli for octanoic acid tolerance.

The economic viability of bio-production processes is often limited by damage to the microbial cell membrane and thus there is a demand for strategies to increase the robustness of the cell membrane. Damage to the microbial membrane is also a common mode of action by antibiotics. Membrane-impermeable DNA-binding dyes are often used to assess membrane integrity in conjunction with flow cytometry. We demonstrate that in situ assessment of the membrane permeability of E. coli to SYTOX Green is consistent with flow cytometry, with the benefit of lower experimental intensity, lower cost, and no need for a priori selection of sampling times. This method is demonstrated by the characterization of four membrane engineering strategies (deletion of aas, deletion of cfa, increased expression of cfa, and deletion of bhsA) for their effect on octanoic acid tolerance, with the finding that deletion of bhsA increased tolerance and substantially decreased membrane leakage.

Distinguishing between metabolically active and dormant bacteria on paper.

Switching between metabolically active and dormant states provides bacteria with protection from environmental stresses and allows rapid growth under favorable conditions. This rapid growth can be detrimental to the environment, e.g., pathogens in recreational lakes, or to industrial processes, e.g., fermentation, making it useful to quickly determine when the ratio of dormant to metabolically active bacteria changes. While a rapid increase in metabolically active bacteria can cause complications, a high number of dormant bacteria can also be problematic, since they can be more virulent and antibiotic-resistant. To determine the metabolic state of Escherichia coli and Salmonella Typhimurium, we developed two paper-based colorimetric assays. The color changes were based on oxidoreductases reducing tetrazolium salts to formazans, and alkaline phosphatases cleaving phosphates from nitrophenyl phosphate salt. Specifically, we added iodophenyl-nitrophenyl-phenyl tetrazolium salt (INT) and methylphenazinium methyl sulfate to metabolically active bacteria on paper and INT and para-nitrophenyl phosphate salt to dormant bacteria on paper. The color changed in less than 60 min and was generally visible at 103 CFU and quantifiable at 106 CFU. The color changes occurred in both bacteria, since oxidoreductases and alkaline phosphatases are common bacterial enzymes. On one hand, this feature makes the assays suitable to a wide range of applications, on the other, it requires specific capture, if only one type of bacterium is of interest. We captured Salmonella or E. coli with immobilized P22 or T4 bacteriophages on the paper, before detecting them at levels of 102 or 104 CFU, respectively. Determining the ratio of the metabolic state of bacteria or a specific bacterium at low cost and in a short time, makes this methodology useful in environmental, industrial and health care settings.

An AAV capsid increases transduction of striatum and a ChAT promoter allows selective cholinergic neuron transduction.

Adeno-associated virus (AAV) vectors are currently the most efficient option for intracranial gene therapies to treat neurodegenerative disease. Increased efficacy and safety will depend upon robust and specific expression of therapeutic genes into target cell-types within the human brain. In this study, we set out with two objectives: (1) to identify capsids with broader transduction of the striatum upon intracranial injection in mice and (2) to test a truncated human choline acetyltransferase (ChAT) promoter that would allow efficient and selective transduction of cholinergic neurons. We compared AAV9 and an engineered capsid, AAV-S, to mediate widespread reporter gene expression throughout the striatum. We observed that AAV-S transduced a significantly greater area of the injected hemisphere primarily in the rostral direction compared with AAV9 (CAG promoter). We tested AAV9 vectors packaging a reporter gene expression cassette driven by either the ChAT or CAG promoter. Specificity of transgene expression of ChAT neurons over other cells was 7-fold higher, and efficiency was 3-fold higher for the ChAT promoter compared with the CAG promoter. The AAV-ChAT transgene expression cassette should be a useful tool for the study of cholinergic neurons in mice, and the broader transduction area of AAV-S warrants further evaluation of this capsid.

In vivo selection in non-human primates identifies superior AAV capsids for on-target CSF delivery to spinal cord.

Systemic administration of adeno-associated virus (AAV) vectors for spinal cord gene therapy has challenges including toxicity at high doses and pre-existing immunity that reduces efficacy. Intrathecal delivery of AAV vectors into the cerebral spinal fluid (CSF) can avoid many of the issues of systemic delivery, although achieving broad distribution of the vector and transgene expression throughout the spinal cord is challenging and vector entry to the periphery occurs, sometimes initiating hepatotoxicity. Here we performed two rounds of in vivo biopanning in non-human primates (NHPs) with an AAV9 peptide display library injected intrathecally and performed insert sequencing on DNA isolated from either whole tissue (conventional selection), isolated nuclei, or nuclei from transgene-expressing cells. A subsequent barcoded pool of candidates and AAV9 was compared at the DNA (biodistribution) and RNA (expression) level in spinal cord and liver of intrathecally injected NHPs. Most of the candidates displayed enhanced biodistribution compared to AAV9 at all levels of spinal cord ranging from 2 to 265-fold. Nuclear isolation or expression-based selection yielded 4 of 7 candidate capsids with enhanced transgene expression in spinal cord (up to 2.4-fold), while no capsid obtained by conventional selection achieved that level. Furthermore, several capsids displayed lower biodistribution to the liver of up to 1,250-fold, compared to AAV9, providing a remarkable on target/off target biodistribution ratio. These capsids may have potential for gene therapy programs directed at the spinal cord and the selection method described here should be useful in clinically relevant large animal models.

Allelic variation of Escherichia coli outer membrane protein A: Impact on cell surface properties, stress tolerance and allele distribution.

Outer membrane protein A (OmpA) is one of the most abundant outer membrane proteins of Gram-negative bacteria and is known to have patterns of sequence variations at certain amino acids-allelic variation-in Escherichia coli. Here we subjected seven exemplar OmpA alleles expressed in a K-12 (MG1655) ΔompA background to further characterization. These alleles were observed to significantly impact cell surface charge (zeta potential), cell surface hydrophobicity, biofilm formation, sensitivity to killing by neutrophil elastase, and specific growth rate at 42°C and in the presence of acetate, demonstrating that OmpA is an attractive target for engineering cell surface properties and industrial phenotypes. It was also observed that cell surface charge and biofilm formation both significantly correlate with cell surface hydrophobicity, a cell property that is increasingly intriguing for bioproduction. While there was poor alignment between the observed experimental values relative to the known sequence variation, differences in hydrophobicity and biofilm formation did correspond to the identity of residue 203 (N vs T), located within the proposed dimerization domain. The relative abundance of the (I, δ) allele was increased in extraintestinal pathogenic E. coli (ExPEC) isolates relative to environmental isolates, with a corresponding decrease in (I, α) alleles in ExPEC relative to environmental isolates. The (I, α) and (I, δ) alleles differ at positions 203 and 251. Variations in distribution were also observed among ExPEC types and phylotypes. Thus, OmpA allelic variation and its influence on OmpA function warrant further investigation.