Thermal analysis of biochemical systems

Abstract

Scientists, both academic and industrial, develop two main types of drugs: 1) small molecule drugs, which are usually chemically synthesized and are taken orally and 2) large molecule, biotherapeutic, or protein-based drugs, which are often synthesized via ribosome transcription in bacteria cells and are injected. Historically, the majority of drug development, revenue, and products has come from small molecule drugs. However, recently biotherapeutic drugs have become more common due to their increased potency and specificity (the ability to chemically bond to the targeted protein of interest). Researchers now estimate that as much as 50% of current drug development activities (pre-market approval) are focused on these protein-based drugs. There are several well-documented steps necessary in the development of a new large molecule drug. One critical element during the end of the biotherapeutic drug discovery phase and the beginning of the manufacturing phase is known as preformulation or formulation development. During this stage scientists systematically test the effects of adding various excipients (non-protein additives added to enhance the protein stability, solubility, activity of the drug, etc.) to the potential large molecule drug. Differential scanning calorimetry (DSC) is a common technique used to perform these formulation studies. In a classic DSC experiment, a protein is heated from 20-80°C and the heat absorbed while the protein unfolds is measured. Many researchers prefer the use of a DSC instrument because of its label-free nature, meaning that no fluorescent or radio-labeled tag is necessary to perform the measurement. The heat absorbed during the unfolding event(s) is directly measured. However, current commercial DSC instruments suffer from high protein consumption (especially when compared to other labeled techniques), low sensitivity, and slow throughput. The aim of this thesis is to address two of the three areas mentioned above: high protein consumption and slow throughput. Since many formulation development studies are performed at therapeutic or high protein concentrations, one can reduce the experimental cell volume and thereby reduce the amount of protein material consumed. However, since there is less sample, less heat is produced. While in the literature there are several heat transfer models that describe how a DSC instrument literature there are several heat transfer models that describe how a DSC instrument functions, there are surprisingly few heat transfer models that detail how ambient temperature disturbances impact the thermal measurement. To better describe this behavior, a simplified state-space thermal model was created to predict the disturbance rejection of a custom DSC instrument. This model was verified experimentally using linear stochastic system identification techniques. To reduce sample throughput, the prototype calorimeter cell was made from disposable materials. Because the majority of protein systems are thermodynamically irreversible, at elevated temperatures the protein solution often aggregates and needs to be cleaned before a subsequent experiment can be run. This cleaning process constitutes a significant portion of the overall time to run an experiment. This thesis documents a fully functional DSC instrument that, while not completely disposable, has been designed, built, and tested with disposable microfluidic materials. Future work would then solve the technical hurdles of repeatably loading disposable microfluidic cells into the DSC instrument.