UC Santa Cruz

Mass spectrometry has been shown to be a powerful tool that allows researchers toquery and understand different biological systems through the analysis of the chemistry surrounding the system. Specifically, the fields of proteomics and metabolomics have benefited greatly from both hardware and software innovations that have been developed over the decades. Improvements in mass spectrometry hardware allows for the development of novel acquisition methods to collect data types previously impossible to collect, while improvements in mass spectrometry software are two fold: 1) novel robust software and data analysis pipelines allow for researchers to efficiently analyze increasingly large amounts of various types of data, and 2) improvements in software accessibility increases researcher productivity by offloading menial informatics work to computers. Importantly, novel hardware and software are still being developed daily with ever increasing frequency. This thesis leverages existing mass spectrometry software to create data analysis workflows that allow researchers to query the chemical space of vastly different biological systems, from small molecules native to the cheese rind microbiome to proteins derived from the vaginal microenvironment. First, untargeted metabolomics profiling was used in the cheese rind microbiome in an effort to gain a high level understanding of bacterial-fungal interactions (BFIs) native to this model system. Understanding microbial interactions is essential for intelligently designing synthetic communities to study microbiomes, which in turn can lead to discoveries from human microbiomes that can positively contribute to knowledge of human health. In the cheese rind microbiome, fungi were found to be the major chemical drivers of microbial interactions, and one specific fungal-fungal interaction was scrutinized to attempt to identify the source of selective antifungal bioactivity. Furthermore, preliminary work was done to begin assessment of novel acquisition methods utilizing ion mobility-mass spectrometry. Next, a workflow was developed leveraging mass spectrometry based bottom-up proteomics to annotate putative biomarkers for early-stage ovarian cancer. High-grade serous ovarian cancer has one of the highest rates of cancer related death among women. Cystatin A was identified as a promising hit from a murine xenograft model transfected with OVCAR-8- RFP cells. However, detection of cystatin A from a dilute, complex biofluid proved to be difficult; therefore, a label free detection method called frequency-locked optical whispering evanescent resonator (FLOWER) utilized a cystatin A - Ab functionalized to the surface of a microtoroid resonator, which allowed detection of cystatin A as low as 100 pM. Finally, due to the amount and complexity of multidimensional mass spectrometry data in both metabolomics and proteomics datasets, software such as TIMSCONVERT, BLANKA2, and pyMALDIproc were developed with the goal of being easily integrated into data analysis pipelines utilizing existing software including but not limited to Global Natural Products Social molecular networking, Cardinal MSI, MZmine2, XCMS, MaxQuant, and more. These software also allow researchers to handle newly incorporated data types such as collisional cross section values obtained from ion mobility-mass spectrometry. Overall, the work described in this thesis serves to 1) further the understanding of select biological systems described above and 2) provide researchers with the tools to conduct their own investigations.