Biological systems are remarkably precise in a lot of different ways. Not only do organisms have the capacity to reproduce, they also have the capacity to defend themselves from external factors. The capability to fight diseases, in particular the immune system, is an integral part of the evolution and natural selection in all plants and animals. For most species there are multiple layers of defense, which are adaptive and provide mechanisms (or adaptive immunological memory) to remember previous attacks and successively improve the response.

From reproduction to defense and maintenance, each organism constantly monitors its internal and external environments at several different levels. Several crucial constituent factors are required to be maintained at close tolerances. A deviation, or a push, away from equilibrium could prove fatal to an individual cell or the whole organism. These deviations also have a shared history with our evolution in the form of diseases like cancer.

In this study, we present some of our efforts to understand the origin and control of this biological noise at four different levels from a physical sciences perspective.

The entire study of this dissertation has its origins linked to a proto-oncogene called c-myc, which is believed to regulate about 10% of mammalian genes. It controls all major decisions of cells, including cell division and cell death, and it is known to be deregulated in most types of cancers. Noisy c-myc transcription can have disastrous effects, thus its expression levels must be controlled very tightly by cells.

At the DNA level, we examine a dynamic feedback mechanism where DNA supercoils during transcription, and dynamic torsional stresses are mechanically coupled with ongoing transcription to control the transcriptional noise. DNA supercoiling has been previously shown to regulate the c-myc proto-oncogene. We have developed genome-wide maps of transcription generated dynamic DNA supercoiling in vivo. We observe, experimentally, that most of the torsional stress is located within about ±2000 bp of transcription start site, and is differentially regulated by topoisomerases I and II.

At the RNA level, we have made an attempt to define the state of the cell using the expression levels of a sub-network of differentially expressed human kinases. Based on this definition, we have been successfully able to cluster together different molecular subtypes in lung cancer cell lines. We were able to identify and confirm previously known deregulated kinases. Many kinase genes are also identified as novel therapeutic targets. Currently we are testing these predictions, and working towards defining the complete state of a cell by getting a digital count of mRNAs at the single cell level.

At the protein level, we studied the dynamics of protein decay to test the hypothesis that protein decay is a one step stochastic process. In several cases we have observed potentially multi-step decay processes in the ubiquitin proteasome system, however more experiments are needed before making any inferences.