Department of Biochemistry (Professor)
Department of Biochemistry (Professor)
Mitochondria are small but complex organelles with a disproportionately large impact on human health. Changes in mitochondrial enzyme activities, respiratory capacity, genome sequence and superoxide generation play important roles in the pathogenesis of heart failure, cancer, neurodegenerative disorders such as Parkinson's, Alzheimer's and Huntington's disease and in aging and longevity. The best current inventory of mammalian mitochondrial resident proteins consists of 1098 proteins (Pagliarini, et al. 2008). Surprisingly, nearly 300 of these proteins are uncharacterized (Pagliarini, et al. 2008). This includes many that are highly conserved throughout eukarya, a strong indication that they perform a fundamentally important function. The genes that encode the mitochondrial proteome are heavily represented amongst known human disease genes, with about 20% of predicted human mitochondrial proteins implicated in one or more hereditary diseases (Andreoli, et al. 2004; Elstner, et al. 2008). Presumably, the quarter of the mitochondrial proteome that is uncharacterized contains many others that await discovery. Making this connection would be greatly facilitated by an understanding of the genetic connections, biochemical properties and physiological functions of these proteins. Therefore, elucidating the functions of these uncharacterized, conserved mitochondrial proteins will not only explain important aspects of mitochondrial biology, but will also provide a framework for identifying new human disease genes.
As a first step toward this goal, we bioinformatically identified many mitochondrial protein families that are pan-eukaryotic and unstudied. Using yeast mutants generated in our lab, we have analyzed the loss-of-function growth phenotypes for each. We have also determined the subcellular and sub-mitochondrial localization for each protein. Such higher throughput, standardized studies have then been used to guide specific hypotheses for a subset of proteins. These projects, all focused on proteins for which no function had ever been described, have progressed to various levels of understanding; four are published.
In summary, our goal to functionally annotate the mitochondrial proteome has enabled discovery of biochemical functions important for mitochondria, elucidation of the genetic basis of two human diseases and catalysis of future studies with a direct impact on common human diseases. In addition to the four projects described above, we are actively pursuing the functions of a number of additional protein families. We are only scratching the surface, however. The majority of the protein families we identified initially await a concerted effort and we are confident that important discoveries will follow.
We have a growing but incomplete understanding of the mechanisms whereby the body senses its nutrient status and responds to adapt cellular and organismal behavior accordingly. The resulting energetic efficiencies are of obvious evolutionary importance as organisms faced a variety of challenging environmental situations, including prolonged exertion, episodic food shortage and competition for resources. In modern human societies, however, these adaptations often have negative health consequences. One clear example is the propensity of most mammals to store excess ingested calories, primarily as fat, in anticipation of an ensuing period of food scarcity. That period rarely comes in our society today and the result is obesity, with all of its attendant comorbidities.
We are interested in the functions of PASK, an evolutionarily conserved serine/threonine kinase, in coupling nutrient status with metabolic state, energy storage and growth. PASK-/- mice were resistant to high-fat diet induced obesity, hepatic steatosis and insulin resistance. This phenotype appears to be due to hypermetabolism in PASK-/- mice in vivo as measured by indirect calorimetry and in isolated skeletal muscle. These findings suggest an important physiological role of PASK in regulating metabolism and controlling energy balance in mammals (Hao, et al. 2007).
A major focus of our lab is to understand the mechanisms by which PASK controls both cellular and organismal energy metabolism. Specifically, we are working to understand how PASK regulates mitochondrial metabolism in skeletal muscle (and probably in many cell types). We are also working to understand how PASK controls hepatic lipid metabolism (as described in Hao, et al. 2007). We have strong in vivo data, using genetics and pharmacology, that PASK plays a key role in mediating the effects of a Western diet to promote dyslipidemia and disease.
Our biochemistry textbooks are filled with examples of allosteric regulation of metabolic enzymes by small molecule metabolites. Herein, I will use the term "allostery" to refer specifically to that subset of allosteric interactions wherein a small molecule regulates a protein by binding at a site other than an enzyme active site. These interactions are a fundamental component of homeostatic metabolic control, but also enable the cell to be agile in response to a changing environment. Discovery of allosteric interactions came as a result of biochemists directly testing their metabolic enzyme for allosteric regulation using an assay of enzyme activity. Unfortunately, these biochemical studies fell out of fashion, and the discovery of new allosteric interactions is now relatively rare.
The age of molecular biology has enabled an explosion in the discovery of proteins and protein families, including regulatory proteins that are the subject of most biomedical research and drug discovery efforts today. We hypothesize that if the old biochemical techniques were applied to these "new" proteins, many of them would be found to be allosterically regulated as well (Lindsley, et al. 2006). In addition to providing a more accurate depiction of regulatory networks, a comprehensive understanding of the metabolite interactome for a given protein would also enable discovery of substrates and products for unstudied enzymes.
In spite of the importance, the difficulty of discovering new allosteric interactions has prevented rapid and systematic progress. By definition, the affinity of regulatory interactions must be in a similar range as the cellular concentration of the metabolite. For many of the most relevant metabolites, such as ATP, this concentration is in the low millimolar range. Existing techniques for the discovery of small molecule-protein interactions with such weak affinity are limited and difficult. For this purpose, we developed and validated a screening platform—Mass spectrometry Integrated with equilibrium Dialysis for the discovery of allostery Systematically (MIDAS)—that enables the discovery of small molecule-protein interactions with up to 20mM affinities (Orsak, et al. 2012). MIDAS has proven effective in confirming known interactions and in discovering new interactions that have bona fide regulatory impact. For example, we discovered that saturated fatty acids allosterically inhibit two enzymes driving carbohydrate metabolism, glucokinase and glycogen phosphorylase, perhaps as a mechanism of sparing glucose when fatty acids are abundant.
Because the methodology and validation was recently published (Orsak, et al. 2012), I will touch upon it only briefly. The formative principle was to develop a method for detection of small molecule-protein binding that was not limited by the fast off-rates that plague conventional binding methods. The equilibrium method that we employed was equilibrium dialysis, wherein a protein of interest is confined by a semi-permeable membrane and small molecules are allowed to equilibrate across the membrane. An interaction with the protein will cause a metabolite to be retained in the protein compartment and this increased concentration is detected by chromatography-coupled mass spectrometry. To date, we have used a test set of 138 metabolites analyzed against five proteins—all enzymes and four with known allosteric regulators. We detected 16 known interactions, including the majority of substrates, products and allosteric regulators for those five proteins. We also found 13 new interactions, eight of which we independently demonstrated to affect enzyme activity (Orsak, et al. 2012).
In addition to continuing to exploit MIDAS in the discovery of novel small molecule-protein interactions, we plan to expand the platform on two fronts. First, we will elaborate our small molecule test set from 138 to ~300 metabolites. In addition to simply enabling a more comprehensive analysis, we will expand into additional classes of compounds. Second, as indicated above, a long-term objective of this project is to perform nearly proteome-wide analysis of metabolite binding. High-throughput protein expression and purification has already been accomplished on this type of scale (Fasolo, et al. 2009). However, we still need to adapt MIDAS to analysis of small protein volumes in multi-well plates. We believe that the milestone of nearly proteome-wide and metabolome-wide interaction identification is eminently achievable. If it were to be accomplished, the impacts would be many and varied. We might discover substrates and products of heretofore unidentified enzymes; we might find allosteric regulators for proteins that we know well; and the resulting networks of interactions would help us understand how cells monitor their metabolic environment and make the appropriate adaptations.