Rational design of functional heme copper oxidases in myoglobin

Miner, Kyle D.

Permalink

https://hdl.handle.net/2142/29570 Copy

Description

Title

Rational design of functional heme copper oxidases in myoglobin

Author(s)

Miner, Kyle D.

Issue Date

2012-02-01T00:56:04Z

Director of Research (if dissertation) or Advisor (if thesis)

Lu, Yi

Doctoral Committee Chair(s)

Lu, Yi

Committee Member(s)

• Cronan, John E.
• Gennis, Robert B.
• Rienstra, Chad M.

Department of Study

Biochemistry

Discipline

Biochemistry

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Protein Design
• Heme Copper Oxidases
• Enzyme modeling

Abstract

Proteins are involved in nearly every process that occurs in living systems, either as a main participant in the process or preforming a supporting role. It is estimated that approximately half of proteins in living systems are associated with a metal in some fashion. With such a high percentage of proteins interacting with metals, it may not be a surprise that most cellular pathways that have at least one metalloprotein performing one or more steps. Many, if not all, of the most important and complex processes that occur in living systems are performed by a metalloprotein. These processes include photosynthesis, cellular respiration, and nucleic acid repair. The metals in these proteins expand the potential chemistry beyond what can be done with only the 20 naturally occurring amino acids. However, nature uses relatively few metal complexes, such as heme cofactors or iron sulfur clusters or metal ions, considering the number of functions that metalloproteins perform. Also, nature uses a surprisingly small number of protein domains and folds compared the number of possible folds. In metalloproteins, both the metal and protein environment surrounding it play an important role in determining the chemistry that is performed. The protein adjusts the properties of the metal, such as the redox potential or the number of open coordination sites. Many metal ions found in metalloproteins are less reactive outside of a protein environment. Despite many years of study, we are only beginning to understand the functioning of large complex metalloproteins, such as heme copper oxidases (HCOs) in respiration or the oxygen evolving complex in photosystems. Large complex proteins pose two problems with respect to studying function. Large complexes are relatively difficult to isolate in a biologically relevant form and the multiple metal sites can either interfere with spectroscopic analysis or require the use of relatively sophisticated methodology. As an alternative to studying these complex proteins, we have chosen instead to redesign an existing well-studied heme protein, myoglobin, to mimic the bimetallic, heme-CuB site of HCOs. This is the site where molecular oxygen is converted to water as part of cellular respiration. The conversion of oxygen to water is highly difficult as there are many highly reactive intermediates that must stabilized so that the reaction can result in water formation. Such a redesign can be thought of as going from the “bottom up” with respect to the desired function. In the process of building up such a model, we are producing minimalistic versions in order to see what the function of each of the structural features is and how it affects chemistry. This thesis describes the improvement of an existing myoglobin based model system of HCOs, named CuBMb, where an non-native copper site was previously engineered into myoglobin by adding two histidine residues. Along with the native histidine, the resulting site resembles the CuB site found in HCOs. This model protein is purified without metal in the CuB site and therefore it is possible to determine the role of the bound metal and the effect of using other metals. Previous studies of this model have not observed the desired chemistry, production of water from oxygen. However, HCOs have a novel feature found in no other proteins, namely a covalently attached histidine and tyrosine moiety that is critical for function of HCOs in vivo. To roughly mimic this novel feature, a tyrosine was introduced into CuBMb at various locations in the designed heme-CuB site. To guide the selection of the positions to place our tyrosine we used both the amino acid sequence information of HCOs and computer based protein models of myoglobin with a tyrosine. The computer models were compared to reported crystal structures of HCOs. The most similar mutants were made and characterized. The resulting tyrosine containing CuBMbs displayed the ability to produce water from oxygen, despite the absence of the covalent bond between tyrosine and one of the hisitidines used to bind the copper, as in HCOs. Even more unexpectedly the desired activity was observed without the copper in the CuB site.. This result is both interesting and unexpected. To futher improve the observed rate, more features of HCOs such as proton delivery channels and non-natural heme cofactors with similar features compared to heme cofactors found in HCOs were introduced into myoglobin. The proton channels had positive effects on the observed activity. In addition to these interesting results, attempts were made to try and react the tyrosine containing CuBMbs under various conditions to induce formation of a covalent bond analogous to the crosslinked histidine and tyrosine found in HCOs. In the process, a crystal structure of a novel species, where an oxygen species is bound “side-on” to the metal of the heme cofactor instead of the expected “end-on” mode that to our knowledge has never been observed in a heme protein. In summary, to better understand the functioning complex proteins like HCOs, a model protein was previously constructed. Introduction of new structural features, into the model protein (CuBMb), with the purpose of mimicking features similar to those found in found in HCOs caused the model system to become competent to perform the desired chemistry with less features than what is thought to be required in HCOs. As more features were attempted we discovered an interesting oxygen species bound to our protein. These type of results show the advantage of trying to build upto a minimal model. It is possible with such a system to obtain unique proteins and intermediates in addition to what information one is attempting elucidate. One is also able to perform experiments that would be impossible in the native system and obtain useful information. The insights gained by this modeling work will help the designers of the next version of CuBMb overcome the limitations of this version and gain insight into how to generally build and design metalloproteins.

Graduation Semester

2011-12

Permalink

http://hdl.handle.net/2142/29570

Copyright and License Information

Copyright 2011 Kyle Miner

Owning Collections