SummaryProtein design provides an ultimate test of our knowledge about proteins and allows the creation of novel enzymes for biotechnological applications. While progress has been made in designing proteins that mimic native proteins structurally1–3, it is more difficult to design functional proteins4–8. In comparison to recent successes in designing non-metalloproteins4,6,7,9,10, it is even more challenging to rationally design metalloproteins that reproduce both the structure and function of native metalloenzymes5,8,11–20, since protein metal binding sites are much more varied than non-metal containing sites, in terms of different metal ion oxidation states, preferred geometry and metal ion ligand donor sets. Because of their variability, it has been difficult to predict metal binding site properties in silico, as many of the parameters for metal binding sites, such as force fields are ill-defined. Therefore, the successful design of a structural and functional metalloprotein will greatly advance the field of protein design and our understanding of enzymes. Here, we report a successful, rational design of a structural and functional model of a metalloprotein, nitric oxide reductase (NOR), by introducing three histidines and one glutamate, predicted as ligands in the active site of NOR, into the distal pocket of myoglobin. A crystal structure of the designed protein confirms that the minimized computer model contains a heme/non-heme FeB center that is remarkably similar to that in the crystal structure. This designed protein also exhibits NOR activity. This is the first designed protein that models both the structure and function of NOR, offering insight that the active site glutamate is required for both iron binding and activity. These results show that structural and functional metalloproteins can be rationally designed in silico.

Summary Rational design of functional enzymes with high turnovers is a significant challenge, especially those with complex active site and difficult reactions, such as in respiratory oxidases. Introducing 2 His and 1 Tyr into myoglobin resulted in designed enzymes that reduce O2 to H2O with > 1000 turnovers and minimal release of reactive oxygen species. This also showed that presence and positioning of Tyr, not Cu, are critical for activity.

A structural and functional model of bacterial nitric oxide reductase (NOR) has been designed by introducing two glutamates (Glu) and three histidines (His) in sperm whale myoglobin. X-ray structural data indicate that the three His and one Glu (V68E) residues bind iron, mimicking the putative Fe B site in NOR, while the second Glu (I107E) interacts with a water molecule and forms a hydrogen bonding network in the designed protein. Unlike the first Glu (V68E), which lowered the heme reduction potential by ∼110 mV, the second Glu has little effect on the heme potential, suggesting that the negatively charged Glu has a different role in redox tuning. More importantly, introducing the second Glu resulted in a ∼100% increase in NOR activity, suggesting the importance of a hydrogen bonding network in facilitating proton delivery during NOR reactivity. In addition, EPR and X-ray structural studies indicate that the designed protein binds iron, copper, or zinc in the Fe B site, each with different effects on the structures and NOR activities, suggesting that both redox activity and an intermediate five-coordinate heme-NO species are important for high NOR activity. The designed protein offers an excellent model for NOR and demonstrates the power of using designed proteins as a simpler and more welldefined system to address important chemical and biological issues.biomimetic models | heme-copper oxidase | metalloprotein | protein design | protein engineering R ational design of proteins that mimic both structure and function of more complex native enzymes has been a long soughtafter goal, as the process is an ultimate test of our knowledge and an excellent means to develop advanced biocatalysts (1-3). Although designed proteins that model the structure of native enzymes have been known for a while (4-10), successful designs of proteins that mimic both the structure and function of native enzymes have been reported only recently (11-16). While being able to design such functional proteins is laudable, the impact of such an achievement would be greater if the designed proteins can be used to address fundamental issues in chemistry and biology that are difficult to tackle by other methods. One primary example is the roles of conserved glutamates and metal ions in bacterial nitric oxide reductase (NOR) (17)(18)(19).NO is critical for all life (20). Bacterial denitrification is a crucial part of the nitrogen cycle in nature that involves a four-step, five-electron reduction of nitrate (NO 3 − ) to dinitrogen (N 2 ) (17, 19). Bacterial NOR is a membrane-bound protein that catalyzes one step of this process, namely, the two-electron reduction of NO to N 2 O (17, 19). With no crystal or solution structure available for bacterial NOR to date, sequence alignments and homology modeling (21, 22) have indicated that NOR is structurally homologous to the largest subunit (subunit I) of hemecopper oxidases (HCOs) (23), enzymes that catalyze reduction of O 2 to water. The active sites of both NOR and HCO contain a proximal histidine-coo...

Denitrifying NO reductases are evolutionarily related to the superfamily of heme-copper terminal oxidases. These transmembrane protein complexes utilize a heme-nonheme diiron center to reduce two NO molecules to N2O. To understand this reaction, the diiron site has been modeled using sperm whale myoglobin as a scaffold and mutating distal residues Leu-29 and Phe-43 to histidines, and Val-68 to a glutamic acid to create a nonheme FeB site. The impact of incorporation of metal ions at this engineered site on the reaction of the ferrous heme with one NO was examined by UV-vis absorption, EPR, resonance Raman, and FTIR spectroscopies. UV-vis absorption and resonance Raman spectra demonstrate that the first NO molecule binds to the ferrous heme, but while the apoproteins and CuI- or ZnII-loaded proteins show characteristic EPR signatures of S = 1/2 six-coordinate heme {FeNO}7 species observable at liquid nitrogen temperature, the FeII-loaded proteins are EPR silent at ≥ 30 K. Vibrational modes from the heme [Fe-N-O] unit are identified in the RR and FTIR spectra using 15NO and 15N18O. The apo- and CuI-bound proteins exhibit ν(FeNO) and ν(NO) that are only marginally distinct from those reported for native myoglobin. However, binding of FeII at the FeB site shifts the heme ν(FeNO) by +17 cm-1 and the ν(NO) by -50 cm-1 to 1549 cm-1. This low ν(NO) is without precedent for a six-coordinate heme {FeNO}7 species and suggests that the NO group adopts a strong nitroxyl character stabilized by electrostatic interaction with the nearby nonheme FeII. Detection of a similarly low ν(NO) in the ZnII-loaded protein supports this interpretation.

The intracellular level of the ubiquitous bacterial secondary messenger, cyclic di-(3′,5′)-guanosine monophosphate (c-di-GMP) represents a balance between its biosynthesis and degradation, the latter via specific phosphodiesterases (PDEs). One class of c-di-GMP PDEs contains a characteristic HD-GYP domain. Here we report that an HD-GYP PDE from Vibrio cholerae contains a non-heme diiron-carboxylate active site, and that only the reduced form is active. An engineered D-to-A substitution in the HD dyad caused loss of c-di-GMP PDE activity and of two iron atoms. This report constitutes the first demonstration that a non-heme diiron-carboxylate active site can catalyze the c-di-GMP PDE reaction and that this activity can be redox regulated in the HD-GYP class.