Modern drug discovery approaches commonly utilize computational methods to find or generate molecules that show favorable binding to a specified macromolecular target. Millions of molecules can be rapidly tested via a virtual screen to filter out bad binders early before purchasing or synthesizing the molecules for experimental assays. Virtual screening involves the generation of poses for a library of ligands, and ranking using simple energy functions with limited flexibility and generally, no water solvation. Top scored poses are used to rank and prioritize ligands with respect to their theoretical interaction energies with the target. The simplified electrostatics approximations and lack of solvation limit the accuracy of the virtual screening pose energy rankings, increasing incidents of choosing the incorrect pose as the most favorable one. Here we adapt an accelerated molecular dynamics (MD) method, developed for peptide folding, to enhance and refine the virtual screening pose energy ranking process. The method re-ranks the poses using the full molecular mechanics energy function and allows for receptor flexibility as well as solvation, increasing the accuracy of the predicted pose rankings relative to the simplified virtual screening protocols. ligand poses from the virtual screen are subjected to short MD equilibration and are used to populate a structure reservoir which then can be used alongside another MD simulation of the poses to allow for rapid exchanges between the conformations via Monte Carlo (MC) jumps. This allows us to circumvent the sampling time constraint of normal MD which would be too slow in generating the same poses that were already predicted by the virtual screen. Relative populations of poses can be obtained with regards to their binding interaction stability. This method provides a more accurate way to rank ligand poses to either corroborate or refine the original virtual screening rankings.

Recently we have shown that this new MD approach with implicit solvation can predict the most favorable pose amongst the ones generated in a DOCK virtual screen for systems where DOCK also chose the same pose as the favorable one. The approach was also able to select the correct pose in instances where DOCK was not able to predict it as being the most favorable. Since MD is more accurate in its sampling dynamics, the pose rankings and favorability observed during the MD simulations can be used as a guide to determine whether the initial DOCK screening was accurate or if the rankings could benefit from MD refinement. Although this approach showed success with implicit solvent, ongoing work will focus on using explicit solvation to account for systems that may have critical water molecules mediating ligand binding, as well as developing an additional approach to account for residue protonation states in cases where shifted pKa’s in the active site are ambiguous. Further in the future this MD method will be adapted for use in estimating relative binding affinities for different types of ligands  rather than just different conformations of the same one. This will allow for the comparison of very distinct ligands which is not feasible with current free energy methods that rely on ligands having similar scaffolds.

Kindra Becker

B.A. in English Literature, Master of Library and Information Science, University of South Carolina

Program/Dates Supported by NIGMS: Chemistry/ August 2022-present

Title: Synthesis of Fluorophores for APEX2 Proximity Labeling in Mycobacterium tuberculosis

The Microbial infection is closely tied to nutrient acquisition. Bacteria rely on iron acquisition from the host for survival, the host iron is sequestered as Fe³⁺ by bacterially produced metallophores termed siderophores. While biding Fe³⁺ efficiently, siderophores also present high affinity toward other metal ions such as Ga³⁺, Sc³⁺, and In³⁺ which share similar size and coordination chemistry with high-spin Fe³⁺. Previous studies have shown these Fe³⁺ mimics display growth inhibitory activity.  The over-arching goal of my project is to investigate how the metallophore-mediated transport of Ga³⁺, Sc³⁺, and In³⁺ takes advantage of the iron siderophore acquisition mechanism and identify which cytoplasmic targets are affected that result in the observed growth-inhibitory activity. To address these questions, we aim to develop metallophore probes designed to incorporate a ⁶⁷Ga or ¹²⁴I label for metabolic tracking and photoaffinity probes incorporating a fluorescent diaziridine photoaffinity tag, for the isolation and identification of proteins involved in active trans-membrane transport. These metallophore probes have the potential to be useful tools to uncover siderophore conjugate mechanism of action and further improve their efficacy.

Computer aided drug design is an integral part of modern drug discovery, particularly for hit identification. A hit is typically defined as a molecule which has a micromolar affinity for a protein target and is a starting point for optimization. This stage of drug development is particularly amenable to the use of computational techniques as they can search large areas of chemical space in a quick and efficient manner. My research aims to further improve this process via the development of new routines and algorithms in DOCK6. 

One such method for hit identification is the fragment by fragment construction of molecules within the constraints of a binding site, called de novo design. Currently DOCK6 can bias the construction of molecules towards those which are more similar to a given reference molecule, such as an experimentally validated inhibitor. This is achieved through the use of similarity-based scoring functions which assess the similarity of growing molecules to the reference molecule. Those which are most similar are advanced to subsequent stages of growth. However this inadvertently creates a bias for molecules which are bulkier at early stages of growth, and leads to ill-informed decisions. I believe that by reconstructing the reference molecule in parallel to the growth of new molecules, a higher degree of similarity can be achieved. This dynamic scoring will further focus the search for new hits in areas of chemical space which are likely to be successful, while still identifying novel chemical structures.

Jason M. Withorn

BS Biochemistry and Chemical Biology from Wayne State University in Biology, Pepperdine University

Program/Dates Supported by NIGMS: Chemistry/ July 2021-present

Title:  Functional characterization of the NosP associated histidine kinase, NahK, and its role in biofilm formation in Pseudomonas aeruginosa

Active Trainees 

John Bickel

BS in Chemistry, Truman State University

Program/Dates Supported by NIGMS: Chemistry/ 2020-2021

Title:  Strategies for biasing de novo design in DOCK6

A significant number of pH sensitive polymers are biodegradable. For this matter, they have the potential to play relevant roles in life. Biodegradable polymers/plastics have been championed in the last four decades owing to a myriad of opportunities they present including environmentally friendly and biocompatibility. Generally, there are two classes of biodegradable polymers which are natural and synthetic. Synthetic polymers offer more advantages than natural polymers because synthetic tool offers handles to modify and functionalize polymers to suit certain applications. To an extent, with synthetic tools, we can bridge the gap between natural and synthetic polymers. Notable synthetic biodegradable polymers are poly-L-lactic acid (PLLA), polyglycolic acid (PGA)/polylactic acid (PLA), polycaprolactone (PCL). These polymers are mostly synthesized by condensation techniques which makes it somewhat impractical to prepare polymers with incredibly high mechanical strength. For this reason, we sought to develop methodology which would enable us to prepare very long high molecular weight polymers which are readily degradable. Our group has developed a type of polymerization called alternating ring opening metathesis polymerization (AROMP). Together with my colleague, Jingling Zhang, we have successfully prepared high molecular weight polymers using AROMP. What is interesting is that these polymers have ester and acetal functionalities in their backbone which are readily degradable under basic and acid conditions respectively. This would possibly have applications in drug delivery and lithographic patterning.

Synthetic tools provide handles to have access to natural polymers. Multivalent sugars have shown to have better efficacy compared to monovalent sugars. Using polymerization method such as ring opening metathesis polymerization (ROMP) and norbornene as the backbone scaffold. Our group has developed several sugar polymers. These glycopolymers, mannose, fucose, GlcNAc, have shown to induce acrosomal exocytosis of male sperm. Galactose and fucose polymers have also shown to bind to cholera toxins. The mechanism of action in these cell binding processes are not well understood. The glycopolymers used in these studies are based on polynorbornene scaffold which tend to lack sequence-definition. Nature’s heavy reliance on sequence have caused us to reason that having sequence-control in our glycopolymers’ backbone would enhance their efficacies. Therefore, my goal is to use AROMP to prepared and sequence sequence-defined copolymers of sugars.

 

---

Danielle Cervasio

BS, summa cum laude,  SUNY Cortland, 2017

Program/Dates Supported by NIGMS: Chemistry/ 2018-2021

Title:  Translational applications of small-molecule, astrocyte-targeted probes

Astrocytes are the most abundant non-neuronal cell type in the brain, providing support, nutrient buffering, and instructions to neighboring cells, particularly neurons, in order to maintain normal physiological function and homeostasis. Importantly, astrocytes play a large role in the brain by responding to insult, communicating with neurons via gliotransmitters, and forming protective barriers around endothelial cell layers. Because astrocytes are present abundantly in all cortical areas of the brain, performing essential metabolic, structural, and homeostatic roles, it is not surprising that they are implicated in a slew of disease and disorders, including, but not limited to, Alzheimer’s, Parkinson’s, glioblastoma, stroke, seizure, infection, and depression. Thus, it is becoming increasingly important to consider these cells for therapeutic intervention of such disease, seeing that they are important players in maintaining normal brain function. In leveraging organic cation transporters (OCTs) on the surface of astrocytes, we have developed a system by which a permanently positive N-heterocyclic amine can be attached, via organic chemistry, to diverse small molecules like fluorophores, photoactivatable markers, or calcium indicators and delivered specifically to astrocytes over neurons and other glia (Fig. 1a and b). These data have inspired us to think about more translational applications, including the potential to deliver drugs or transcription activators to astrocytes for therapeutic intervention. Gene-targeted therapies are currently in the clinic, but there exists a lack of regulated systems. We currently have some control over where, but no control over when therapeutic genes are expressed. I plan to synthesize and assay a doxycycline modified with a permanently positive N-heterocycle to exploit the powerful tetracycline-inducible gene expression system (Tet-On/Off) in order to control gene expression in astrocytes, exclusively. In adding this modification, I hypothesize that we will be able to target astrocytes for therapeutic intervention, decrease concentrations of drug needed by increasing blood brain barrier (BBB) penetration, and reduce off target effects of current constitutively active therapies as seen in Glial cell-derived neurotrophic factor (GDNF) or Neurturin (NTN) gene therapy for Parksinson’s disease.  

 

 

---

Jiayuan Fu

BS, Beijing Normal University, 2018

Program/Dates Supported by NIGMS: Chemistry/ 2019-2021

Title:  Characterization of a Novel Nitric Oxide/Heme Sensing Protein and its Role in
Regulating Biofilm Formation in Burkholderia Thailandensis

Serine/Theonine protein Kinase one (SMG-1) is involved in nonsense mediated mRNA decay (NMD) which is one of the DNA Damage Response (DRR) pathways in mammalian cells.  The DDR serves to detect, fix, or ameliorate errors that may arise during DNA replication, and if left unchecked can lead to cell cycle arrest, regularion of DNA replication, and the failure to repair of DNA damage.  DDR plays an important role in protecting cancer cells from chemotherapeutics, and the underlying hypothesis of this project is that inhibition of SMG-1 will senstive cancer cells to cancer therapies as well as the immune response.  However, efforts to evaluate this hypothesis using genetic knockouts in breast cancer cells have been hindered by the essentiality of SMG-1.

Proteolysis targeting chimeras (PROTACs) utilize the cells natural ubiquitin proteasome system for protein degradation and maintaining homeostasis in the eukaryotic cells, in order to force the initiation of a degradation cascade on a target. In order to determine if SMG-1 is a suitable target for PROTAC development, we will utilize the Achilles TAG (aTAG) tool that was developed by C4 Therapeutics to knockdown the SMG-1 protein. A fusion protein of the aTAG with SMG-1 will be created using CRISPR/Cas9, and PROTACs directed at the aTAG will be used to degrade the aTAG/SMG-1 fusion protein. Targeted protein degradation with PROTACs will be used to chemically knockdown SMG-1 and evaluate SMG-1 as a potential therapeutic target.

Neurons were the big players in neuroscience research until recent efforts identified that astrocytes too are active players in brain function. Astrocytes are the most abundant type of non-neuronal cells in the brain. In addition to maintaining blood-brain barrier integrity and providing nutrients to the surrounding cells, astrocytes form a specialized structure with pre- and post-synaptic neurons known as the tripartite synapse where they take part in the uptake and release of neurotransmitters. Astrocytes have been implicated in many neurological diseases  like Parkinson’s disease (PD), Alzheimer’s disease (AD), Amyotrophic lateral sclerosis (ALS), glioblastoma and depression. Accordingly, to understand how the brain circuitry functions it is essential to study astrocytes.

However, only a limited number of strategies are available for astrocyte visualization and manipulation. These existing methods lack specificity and versatility. A new class of methyl-pyridinium tagged fluorescent markers has been developed in our lab that shows specific targeting and visualization of astrocytes. The positive charge of the methyl-pyridinium targeting moiety of this probe allows active transport of diverse cargo such as Ca²⁺ sensors, small molecule drugs and transcriptional activators to astrocytes through the astrocyte resident Organic Cation Transporter (OCT).

I am currently synthesizing a methyl-pyridinium tagged cyanodoxycycline to activate transcription of specific genes in astrocytes using the tetracycline-inducible gene expression (Tet-On) system. My work is also focused on the drug tamoxifen to create a traceless astrocyte delivery strategy. These molecules will help us to evaluate the ability of the targeting moiety to deliver drugs specifically to astrocytes. This strategy will be beneficial for exploring the therapeutic aspects of astrocytes in various neurological diseases and will improve our understanding of the functions of astrocytes in the brain.

Kirsten Martin

BS in Chemistry

Program/Dates Supported by NIGMS: Chemistry / 2019-2021

Title:   In Situ Excitation of Luminescent Lanthanide Complexes: Towards Enabling In Vivo Bioimaging

 The acrosome reaction (AR) in spermatozoa is an essential process in mammalian fertilization. Although the mechanism behind the activation of AR is not well-understood, it is known that the glycoproteins surrounding the egg cell’s zona pellucida (ZP) can mediate the activation of AR through a series of carbohydrate binding events with the receptors on the sperm plasma membrane. Previous research in the Sampson lab has demonstrated that AR in mouse sperm can be induced successfully in vitro with the aid of glycopolymers. These glycopolymers, synthesized by ring-opening metathesis polymerization (ROMP), possess norbornene backbones and resemble terminal residues of the sugar moieties present on the ZP. These glycopolymers are hypothesized to be potential inducers of AR in human sperm. To determine the efficacy of AR induction by glycopolymers, a flow cytometry assay is currently under development, which will allow for efficient sorting and quantitative analysis of dead, live, and acrosome-reacted cells. With the aid of this assay, promising AR inducers can be identified and will allow for the distinction of sperm that are able to participate in fertilization and those that are not. The development of this flow cytometry assay will ultimately provide the clinical community with a new method of assessing male infertility that can predict AR activation, thus improving current techniques used for diagnosis and treatment.

My goal of my proposed research is to discover small molecule inhibitor for mutated fibroblast growth factors 2 in intrahepatic cholangiocarcinoma patients (ICC). ICC is a deadly liver cancer that includes very few treatment options. A large group of patients that presents all types of liver malignancies present ICC. Without adequate treatment, life expectancy after diagnosis is very short. Currently, there are no specific inhibitors for the mutational active FGFR2. Therefore, the goal of this project is to design small molecular candidates through virtual screening to bind to the binding site of the mutated FGF2.

By utilizing crystallographic structures of hyperactive FGFR2, structural models were prepared for docking simulations. Because mutated FGFR2 does not exist, additional models will need to be created for simulation. Then, large-scale virtual screening of commercially available compound libraries will be used to identify potential drug-like ligands. After virtual screening, results were ranked ordered via relevant scoring functions. Then, subsets of the most promising ligands will be purchased and tested for activity.

Mycobacteria, including the human pathogen Mycobacterium tuberculosis (Mtb) that causes tuberculosis (TB), have a complex and impervious double membrane that distinguishes them from Gram negative and Gram positive bacteria. Unfortunately, Mtb infects one fourth of the global population. More disturbingly, new cases of Mtb are increasingly resistant to frontline therapies. By elucidating mycobacterial membrane biogenesis pathways, we can reveal new drug targets and identify proteins that control membrane permeability.

The Jessica Seeliger lab focuses on the LprG & Rv1410c proteins, which form an operon and likely work together to transport lipids to the outer membrane of mycobacteria, more specifically referred to as the mycomembrane. Loss of the LprG-Rv1410c operon leads to attenuation in vivo and a growth defect under nutrient limitation in vitro. We recently used a computational docking screen in search of commercially available small molecule inhibitors of LprG. We discovered molecules that inhibit mycobacterial growth in an LprG-dependent manner with micromolar affinity, indicating that LprG can be specifically inhibited. Additionally, we found that a single residue in the cavity of LprG mediated interaction with one of our most potent compounds. Overall, despite having a large and generally hydrophobic pocket, LprG can be realistically inhibited and rational drug design is feasible.

Understanding LprG-Rv1410c mediated lipid transport is also of interest because these proteins are predicted to be important for intrinsic resistance to multiple TB antibiotics. It has also been shown experimentally that the loss of the LprG-Rv1410c operon leads to increased susceptibility to certain antibiotics. Therefore, targeting or inhibiting LprG-Rv1410c may promote the efficacy of antibiotics that normally cannot cross the mycomembrane efficiently.

In future studies, we will determine how LprG & Rv1410c mechanistically mediate lipid transport, in order to obtain a model of mycobacterial lipid transport to the mycomembrane. Additionally, we will investigate how LprG & Rv1410c affect membrane permeability in order to better understand drug diffusion across the mycomembrane. Ultimately, this work will better define mechanisms of mycobacterial membrane biogenesis and drug uptake.