Congratulations, USEF students!

The 2019 University of Utah Science & Engineering Fair (USEF) was a success! Congratulations to the 700+ students who participated. Student awards have been announced.

There are 6 projects that will advance to ISEF, the International Science & Engineering Fair:

Divyam Goel - West High

Bacterial infections in Cystic Fibrosis patients lead to decreased lifespan. In particular, Pseudomonas aeruginosa infections are linked to higher mortality. This study examined the use of phages as a clinical treatment method. However, instead of growing biofilms on a simple matrix and measuring the effect of phages in dispersing biofilms, the study tested the effectivity of phages in preventing biofilm growth by growing the biofilms with phage. The effects of two antibiotics – trimethoprim and chloramphenicol – were also tested to compare potency of treatment. A phage was isolated for one of three organisms used – Two P. Aeruginosa and a MRSA. The phage was also characterized with a simple one-step growth curve and titer calculation. Two 96 well microtiter plate assays were used, with one utilizing the P.A. 27853 phage isolate and another using an unisolated phage stock. This allowed the creation of poly-microbial (multi-specie) biofilms. The results showed that the phages and antibiotics generally increased biofilm growth instead of preventing it. Possible reasons for these observations are discussed. Following the microtiter assays, plastic Simulated Anatomic Models (SAMs) of a pediatric CF patient’s lung were 3D printed. With minor adaptations, the phage examination was repeated. Data show that the phage was extremely successful in preventing biofilms on the lungs. These results are also discussed.

Christopher Li - West High

Melanoma, a tumor of melanocytes located in the epidermis, is responsible for most skin cancer deaths. Familial melanoma is strongly linked to germline mutations in the CDKN2A (p16) tumor suppressor gene, which regulates cell cycle and cell senescence as well as suppresses reactive oxygen species (ROS). Mutations in p16 can affect these vital tumor-suppressor functions by affecting binding site pathways. p16 binding to adenosine monophosphate (AMP) is a key component of the AMP-activated kinase (AMPK) pathway as greater AMP levels feed into the cell senescence and cell proliferation regulation functioning of AMPK. As part of the AMPK pathway, p16 binds inosine monophosphate dehydrogenase (IMPDH), which is related to rate-limiting AMPK’s purine nucleotide synthesis. The cycling adenosine monophosphate (cAMP) pathway is important in suppressing tumorigenesis through DNA damage repair. To gain insight into how mutations change p16 function as related to its interactions with cAMP and AMP, 12 known familiar pathogenic p16 mutations and their effect on cAMP and AMP binding were investigated. The amino acid sequence of the p16 wild-type and the mutations were used as input into 3D protein modeling COACH, BSP-SLIM, BioLIP and RaptorX metaservers targeting binding probability at prospective protein binding locations. The binding locations and strength for the mutants were compared to wild-type p16. IMPDH binding activity to p16 was characterized by immunoprecipitation. Structural modeling findings indicate that when compared to the wild-type p16, 10 of the 12 (83%) mutations disrupt AMP binding pocket or shift binding location. cAMP binding results indicate weak binding strength to p16 in each mutant as well as the wild-type. Immunoprecipitation data shows 3 mutations having lost IMPDH binding capacity to p16. Metaserver data indicates that these 3 mutations also lost AMP binding. The changes observed in AMP-binding pockets suggest an important role for p16 binding in AMP-mediated signaling in melanoma. IMPDH binding capacity changes were observed in mutations which also impaired AMP binding, indicating the potential direct function of IMPDH in regulation of AMP levels in the AMPK pathway. This study emphasizes the importance of understanding p16 mutations in melanoma predisposition, potentially identifying courses of gene therapy, specifically in cAMP, AMP, and IMPDH targeted-treatment.

Madeline Joklik-McLeod - Juan Diego

Overarching goal: This proposal will test a novel tumor suppressor-proapoptotic hybrid (called p53-Bad) for gene therapy of ovarian cancer. p53 tumor suppressor will be fused to Bad (a proapoptotic protein) to trigger “apoptotic collapse” at the mitochondria by activating apoptotic pathways and inhibiting pro survival pathways. Previous work in vitro has shown that mitochondrially targeted p53 can kill cancer cells in vitro via the extrinsic apoptotic pathway [1-3]. Our overarching hypothesis is that potent amplification of the p53 effect at the mitochondria can be achieved by fusing it with Bad, an apoptotic mitochondrial protein which can directly bind to and inactivate antiapoptotic proteins. A fusion ensures that both proteins are at the mitochondria at the same time, and can act together.

Tarun Martheswaran - Waterford Academy

Dengue Fever is a debilitating viral disease of the tropics, transmitted by female Aedes mosquitos, causing sudden fever, acute pains in the joints and hemorrhaging. The World Health Organization (WHO) estimated almost 96 million clinical cases with more than 25,000 deaths annually from this disease worldwide. To date, no vaccine has been developed for Dengue Fever due to the existence of four virus serotypes that are extremely difficult to develop an immunity against. Early detection of Dengue is proving to be the only viable options of mitigating the transmission of this disease and potentially containing it. The purpose of my project is to innovate a novel approach to detect the outbreak of Dengue disease using the SIR compartment system incorporating climate changes and lag times, as well as utilizing Desolve R software and Ordinary Differential Equations Mathematical modeling triggered by a passive-active Dengue disease monitoring system that was mandated by WHO and the Center for Disease Control (CDC) worldwide. Using actual 2005 Dengue disease outbreak data from Singapore, as well as 10 years worth of Singaporean climate data, the average climate, rain, humidity and vector-human population were calculated. Thousands of simulations with varying susceptible human populations, infected vector populations, and mosquito bite rates were done to produce a quartic function relationship between the climate, lag time and bite rate. The resulting simulation data were tested on an actual 2014 Singapore Dengue outbreak. Statistical testing using Pearson’s values showed a significant correlation between the test data and actual Dengue outbreaks, with Test 1 producing an r value of 0.5891 and p value of 0.018, and Test 2 producing an r value of 0.5537 and p value of 0.0004. This concludes that we can use this novel model to accurately predict Dengue disease outbreaks and as a tool for early detection with proactive vendor control measures in place, which ultimately reduces the number of deaths associated with this disease. To date, no similar robust early detection modeling has been developed. This experimentation can serve to further control the disease worldwide as well as aid in the search for a vaccine for this life-threatening virus.

Anisa Habib and Tejita Agarwal - West High

With a dramatic 2.4-fold increase over the past several decades (1973-2002), thyroid cancer is the fastest growing cancer in the United States. This is often attributed to improved screening practices, but this drastic increase and the variation across regions indicate that there may be other factors at play. Identified risk factors of thyroid cancer include geographical location, race, age, female sex, hereditary conditions, and unstable iodine levels in the body. From 1951 to 1992, the United States Atomic Energy Commission conducted a series of nuclear weapon tests at the Nevada Test Site, releasing radioactive substances, including I-131, from 1952 to 1958. Theories describing I-131 as a causal risk factor of thyroid cancer incidence have been difficult to support largely due to problems of limited available data which causes studies to assume a short latency period, government suppression, and the ecological nature of this topic. The project expands on our previous work and aims to quantify differences in age, period, and cohort trends in thyroid cancer risk across geographic regions by sex in the US. We incorporate historical information about screening innovations, I-131 exposure, smoking, and migration in order to elucidate mechanisms that may precipitate observed differences and explain regional variations. R Studio software was used to clean and merge our collected data files. Poisson regression was used to construct incidence models by exposure level and sex by birth cohort, period, and age. Sex-specific smoking levels and state migration rates were controlled for. An interaction term was used to test birth-cohort specific difference in exposure-based risk. The differences in risk across each exposure level were statistically significant. For females with low exposure as the reference variable, medium exposure and high exposure Poisson estimates were 0.158663 and 0.270037 respectively. For males with low exposure as the reference variable, the medium exposure and high exposure Poisson estimate were 0.210239 and 0.28563 respectively. The trend analysis found a significant slope for females [0.1321339 (p-value: <2e-16)] and significant slope for males [0.139853 (p-value: <2e-16)], indicating a dose-response, meaning the effect of exposure increases with exposure level. The results of this project provide evidence that I-131 exposure may contribute to the observed increase in thyroid cancer incidence, that the effect may be dose-dependent, and that increasing levels of thyroid cancer may be due to more than just improved screening practices. As our world becomes more technologically advanced and begins to develop more nuclear technologies, it is important to remember the health and safety risks that accompany the environmental and economic benefits. Individuals exposed to radioactive iodine may need to undergo increased screening and clinicians should be aware of the potential increased risk. Further research may reveal additional information about groups of the population affected most.

Sanjana Kargi and Dua Azhar - Beehive Science & Technology

The p53 protein, known for its role in preventing cancer, is also, in some cases, the cause of it when mutated. Over the past few decades, finding a way to understand the protein more and then develop a way to reactivate it has been the goal. By utilizing artificial intelligence and machine learning algorithms, we can predict the transcriptional activity of a p53 protein in silico, and in turn, reduce the workload and time that is taken to do so in vitro. The tumor suppressor p53’s 2D and 3D structural properties can be analyzed to predict whether it is active or inactive. By using a machine learning dataset we can train a compute to assist in the prediction. ¬†Using Correlation-based Feature Selection, we can select the features that influence p53’s transcriptional activity the most and fit an Adaptive Boosted Decision Tree. We found that when using up to 300 of the 5409 features available, we can reach a 99.7% accuracy in the prediction of p53’s transcriptional state. p53 mutations are present in around half of all human cancers and is one of the most common causes of cancer. By identifying which p53 mutations are harmful, we can more accurately identify cancer in its early stages and work to correct the mutation. We hope to continue improvements on our model so that it will be functional with a more diverse data set.