Proteases, enzymes that hydrolyze peptide bonds, have numerous implications in the life sciences, ranging from the treatment of various diseases to enabling chemoselective bond cleavage in chemical synthesis in chemical synthesis. The kinetics of proteases can be monitored spectroscopically using nitroanilide substrates, many of which are structurally complex and often require long, multi-step syntheses. This study sought to determine the rate of hydrolysis by various serine proteases via UV-visible (UV-vis) spectroscopy using highly simplified synthetic nitroanilide substrates. Enzymatic cleavage of the amide bond was spectroscopically monitored on time-course. 

Organic solvents have a broad spectrum of polarity and display drastically different properties. Most importantly, tertiary enzyme structure is disrupted in organic solvent. This is a result of the disruption of intermolecular forces that maintain enzyme structure. Enzyme catalysis in organic solvents has recently been explored for a variety of applications. A display of enzyme activity in organic solvents and its limit in various concentrations can provide insight into potential applications in which both a nonaqueous medium and the chemoselectivity of certain enzymes are required. Here, we explore the effect of various concentrations of non-aqueous organic solvent on lipase activity and analyze the relationship between various properties of solvents and enzyme activity. The study was furthered with molecular dynamics simulations on lipase in various solvents using GROMACS.

In late 2019, a novel coronavirus, SARS-CoV-2, was identified in Wuhan, China, and the virus has since spread around the world and has been cause for infection and, in some cases, death, particularly in the elderly or in those with chronic disease. ddRNAi, or DNA-directed RNA interference, is a gene-silencing method that uses DNA constructs to hijack the pre-existing animal cell’s RNA interference pathways. Here, we developed an RNA fragment that may silence a gene in the SARS-CoV-2 genome, which inhibits viral replication, thereby inhibiting the virus’ spread. We designed a novel ddRNAi strand that inhibits gene expression in SARS-CoV-2 infected cells. Our designed siRNA strand does not share any complete similarities with the known human genome and corresponds to the SARS-CoV-2 virus genome at 9841-9859 bp. 

Polyphenols are organic micronutrients that are produced by plants to combat stressors as well as help regulate other plant processes. Many medicinal plants such as herbs, fruits, vegetables, and even oils contain large amounts of these molecules. In addition to aiding plants in their defense, polyphenols have been found to lower the deficits caused by Alzheimer’s disease by weakening the oxidative stress caused by accumulation of beta-amyloid in the brain. In order to increase the polyphenol production in plants to work effectively against AD, we altered the environmental factors of 4 polyphenol producing plants by affecting them with saline solution, UV irradiation, and hydrogen peroxide, respectively, and adding a control. In this experiment, the herb rosemary was used as it is not only rich in polyphenols, but also contains over four different types of these molecules and has a low germination period. To identify the effects of these stress factors on the rosemary plants, we conducted a FOX assay to detect the amount of hydrogen peroxide in the plants and quantified the amount of polyphenols produced through the Folin-Ciocalteu (FC) assay. Later, the rosemary extracts from each growth condition were screened for potential anti-amyloidogenic activity against wild type human amyloid beta 42.

Berberine, a natural alkaloid that is extracted from the roots and stems of plants in the genus Berberis, has been documented to have medicinal potential since 3000 BC, where it was used as an antibacterial agent in ancient Chinese medicine. Berberine and its synthesized analogs have been studied for a wide range of medicinal properties, including DNA binding and antimicrobial activity, along with its potential for the photodynamic treatment of cancer. Berberine has been reported to produce a reactive oxygen species within DNA, causing DNA damage through guanine specific oxidation. Here, we are interested in the photosensitizing ability of berberine and novel analogs. Synthetic modifications to berberine were attempted through borohydride reductions and Grignard additions to carbon 8, oxidative addition of alkyl chains to carbon 13, and radical halogenation of carbon 12. Photosensitizing activity of berberine was studied through an ex-vivo and various in-vivo experiments. An in-vivo cell viability assay and ex-vivo tracking of endoperoxide formation, a known product of reactions with singlet oxygen, by NMR on a time scale yielded inconclusive results.  Antimicrobial activity was also studied through a Kirby Bauer assay with berberine and two berberine analogs, dihydroberberine and 8-methylberberine. 

MS 408, also known as the Voynich Manuscript, has perplexed readers for centuries due to its strange writing and illustrations of plants, symbols, and human figures. The nature of the Voynich Manuscript, along with existing transcriptions of its writing, promote the use of data mining and machine learning techniques to find underlying patterns in its text. Resulting letter frequency analyses reveal that the text in the Voynich Manuscript is closely connected to both Latin and Italian. Comparisons between bigram frequencies from the Voynich Manuscript and those from representative Latin, Italian, Old French, and Old Spanish texts show strong correlations. Essentially, these texts support how the Voynich Manuscript is heavily influenced by Latin or a close derivative of Latin, which is historically plausible.  Ultimately, the resulting conclusions attempt to clarify the mystery surrounding the manuscript and assist ongoing efforts to solve this enigma by forging new connections to help understand the Voynich Manuscript.

In December of 2019, a novel coronavirus, which mainly causes respiratory symptoms in human hosts, was first identified in Wuhan, China. Upon entry into host cells, the main protease is essential for the replication of viral RNA, which is what allows the virus to replicate inside humans. Here, we designed a library of small molecule inhibitors of the main protease that was based on known reactions and driven by the interactions within the active site of key proteins of the main protease in SARS-CoV-2. In our design of covalent inhibitors of the coronavirus protease, we modeled a library of 361 peptidomimetic Michael acceptor small molecules, which are designed to engage the nucleophilic cysteine residue in the active site of the protease in an irreversible 1,4-conjugate addition. We then employed a variety of computational tools to determine binding affinity of our designed compounds when bound to the protease active site, where we determined that cationic side chains are potentially beneficial for inhibition of SARS-CoV-2. 

Polyphenols are compounds characterized by a large presence of phenolic structural units and are naturally occurring in plants. Known for their medicinal applications, polyphenols have antioxidant properties as well as protein aggregation inhibition properties. Previous research has proved that polyphenols may have the ability to inhibit the Amyloid β-42 protein, which  is one of the main causes of Alzheimer’s disease. This has brought our attention to seeking methods of reducing levels of the peptide through reduced production, or aggregation. Our research is focused on determining the most efficient method in which polyphenols can be extracted, and the method in which polyphenols bind to the Aβ-42 protein. Polyphenols can be extracted using many solvents such as acetone, dimethyl carbonate, ethyl acetate, toluene, methanol, ethanol. Using the Folin- Ciocalteu (FC) Assay, we were able to detect and quantify the presence and amount of polyphenols in our extracts. An in-vitro peptide assay of the Amyloid β-42 and polyphenols aimed to measure the increase in absorbance of the dye which would indicate protein aggregation. 

As of April 8th, 2020, approximately four thousand, one-hundred stellar and planetary objects have been discovered and recorded by NASA and Caltech, with additions dynamically added as information is verified. The process to sort through this ever-increasing dataset is time-consuming and ineffective. This paper seeks to solve this issue and simplify the procedure through the use of algorithms which identify any viable exoplanets within their parent star’s “Goldilocks” zone, meaning they have the potential to sustain liquid water, and thus life as we know it. A combination of Python and SQL was used to datamine the extensive public dataset provided by NASA and Caltech. With the exception of types L, O, and T, all major spectral types were analyzed in the study. Type O planets were excluded due to photoevaporation caused by strong ultraviolet emissions, which evaporates the gas and dust surrounding orbiting planets that are necessary for life and any water.1 Spectral types T and L were excluded due to the tidal locking radius for red dwarfs, thus entailing extremely high temperatures and an uninhabitable environment.2 Finally moons were excluded proactively as they were not part of the dataset and, with contemporary technology, it is not possible to accurately verify habitability. After initial examination we concluded that the type A-orbiting planets in the dataset could be immediately excluded from further scrutiny as their semi-major axes were far too large or too small, meaning their surface temperatures would be too extreme for life. After the two month period of research concluded, we found that there were a significant subset of planets hosted by spectral types B, F, G, K, and M were capable of supporting liquid water. Further investigation is indicated in performing more rigorous examination of planetary masses by inferring Keplerian orbital mechanics given the known host star masses, setting the stage for examination of retained atmospheres and the quantum effects of the host star’s irradiation.

Non-nucleoside reverse transcriptase inhibitors (NNRTIs) are used in the treatment of HIV-1 to prevent reverse transcriptase (RT) from transcribing HIV RNA into DNA by binding at a specific “pocket” on HIV-1 reverse transcriptase.  Our research focuses on using menthol and carvone as core scaffolds in the development of an NNRTI with a “butterfly-like” structure similar to first-generation NNRTIs. The presence of a hydrogen bond donor and acceptor in both menthol and carvone and their chiral centers will result in more favorable interactions with the binding pocket. Menthol and carvone are abundantly available natural product monoterpenoids, so using them to develop NNRTIs is largely unique to our research. This is significant as novel menthol-based and carvone-based NNRTIS may have more conformational flexibility, which is a necessary characteristic for increased potency of future NNRTIs. In our work with carvone as a starting material for a potential NNRTI, we have done extensive docking calculations on a number of different analogs and decided to pursue the synthesis of two analogs with the greatest binding affinity and least number of steps. In comparison with the competitive landscape of FDA-approved NNRTIs, the synthetic route of these analogs already has fewer steps, so we will need greater overall yield. In our work, we have also focused on developing a green oxidation of (-)-menthol to (-)-menthone by testing this reaction in a variety of solvent systems. Further, we have looked at the thermodynamic and kinetic factors in the alpha-halogenation-elimination of (-)-menthone to form menthenone.  We have done theoretical computational studies on this reaction considering what product would form from this reaction and empirical studies in that we have attempted this reaction in several different conditions. 

Here we present efforts towards development of a machine learning algorithm for drug discovery with application to small molecule antivirals that target the human immunodeficiency virus (HIV) and the novel coronavirus (SARS-CoV-2).

For SARS-CoV-2 The goal for the machine learning algorithm is to be able to correctly identify possible inhibitors from an intentionally-chosen, unbiased training dataset, from a learning dataset of SARS-CoV-2 inhibitors, demonstrating its ability to correctly identify the requisite three-dimensional structure of a molecule under question. Additionally, a three-dimensional map of the key residues on the SARS-CoV-2 S-glycoprotein receptor-binding domain (RBD), which is the active site responsible for binding to the human ACE2 receptor, was generated to allow for potential screening of molecules by their three-dimensional conformations. Molecules are scored by their binding conformations based on interaction through hydrogen bonding or pi stacking. This three-dimensional mapping function allows for the visualization of potential interactions between the ligand and RBD without having to spend resources on DFT and docking calculations. 

For HIV: Non-nucleoside reverse transcriptase inhibitors (NNRTIs) are a class of antiretroviral drugs that have been used to treat human immunodeficiency virus (HIV). Retroviruses such as HIV use reverse transcriptase to transcribe viral RNA into DNA, which is hence incorporated into the human host cell genomes. This causes the host’s cells to express the viral DNA and synthesize viral proteins to aid in viral replication. The target site for NNRTIs in the reverse transcriptase enzyme contains a variety of hydrophobic amino acids, creating a binding interface for the NNRTI. Two important structural features of NNRTIs are the presence of hydrogen bond donor atoms and aromatic rings, which form hydrogen bonds and pi-stacking interactions with the reverse transcriptase enzyme respectively. Six FDA-approved NNRTIs and 3 novel NNRTIs were analyzed using PaDEL to determine and verify further similarities among molecules belonging to the NNRTI class of drugs. In addition, scripts determining heteroatom positions and aromatic ring positions were written in Python to generate coordinates of these heteroatom and aromatic ring positions in a 3-D environment. The analysis of existing NNRTIs paves the way for the mass screening and identification of other compounds with similar structural features. 

The outbreak of the novel coronavirus disease COVID-19, caused by the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), has resulted in a global pandemic and rapid increase in infection and mortality rates. Thus, there is an urgent need to find potential treatments by discovering drugs that are effective against the virus. By repurposing existing drugs, we attempt to expedite the discovery process by using commercially available compounds. Based on previous studies, we selected emodin, remdesivir, and hydroxychloroquine as candidates for our search. SARS-CoV-2 infects cells by using Angiotensin-Converting Enzyme 2 (ACE-2) as an entry receptor; emodin and hydroxychloroquine were suggested to be inhibitors of this interaction. Remdesivir acts as a ribonucleoside analog and interferes with the viral RNA polymerase of SARS-CoV-2. Using the ZINC database and DINC webserver, we virtually screened commercially available compounds that are structurally similar to these three compounds for their ability to inhibit the activity of the SARS-CoV-2 viral proteins or their interactions with ACE-2. We discovered multiple ligands for each compound that had a higher binding affinity than their original candidates. Notably, emodin interacted with the hydrophilic regions of the spike glycoprotein, which is likely due to the abundance of hydroxyl groups on the ligand. These results provide insight into potential molecules that may be able to disrupt these viral interactions to more effectively treat COVID-19 patients.

As cancer continues to take millions of lives worldwide, the need to create effective therapeutics for the disease persists. The kinesin Eg5 assembly, a motor protein responsible for microtubule movement in mitosis, is a promising target for anticancer drugs since inhibition of this protein would lead to cell cycle arrest. Since its discovery two decades ago, monastrol, a small molecule dihydropyrimidine, has attracted the attention of medicinal chemists with its potency as a kinesin Eg5 inhibitor. High-throughput virtual screening (HTVS) allows hundreds of structures to be evaluated for potential biological activity in a relatively short amount of time, and we hypothesize that this could also be applied to the identification of small molecule Eg5 inhibitors from a targeted library. Here, we employ HTVS in order to screen a library of one hundred analogs of monastrol to determine biologically potent dihydropyrimidine Eg5 inhibitors. Through protein-ligand docking experiments conducted on Swissdock and DINC, it was determined that the analogs with decyl, geranyl, and phenoxyethyl substitutions at C5 exhibited the greatest binding affinities to the allosteric binding pocket of kinesin Eg5. Future studies involving the synthesis and testing of such novel analogs should be conducted to explore the potencies and biological activity.