Abstract: Despite intensive study, plant lysine catabolism beyond the 2-oxoadipate (2OA) intermediate remains unvalidated. Recently we described a missing step in the D-lysine catabolism of Pseudomonas putida in which 2OA is converted to D-2-hydroxyglutarate (2HG) via hydroxyglutarate synthase (HglS), a DUF1338 family protein. Here we solve the structure of HglS to 1.1 Å resolution in substrate-free form and in complex with 2OA. We propose a successive decarboxylation and intramolecular hydroxylation mechanism forming 2HG in a Fe(II)- and O 2 -dependent manner. Specificity is mediated by a single arginine, highly conserved across most DUF1338 proteins. An Arabidopsis thaliana HglS homolog coexpresses with known lysine catabolism enzymes, and mutants show phenotypes consistent with disrupted lysine catabolism. Structural and biochemical analysis of Oryza sativa homolog FLO7 reveals identical activity to HglS despite low sequence identity. Our results suggest DUF1338-containing enzymes catalyze the same biochemical reaction, exerting the same physiological function across bacteria and eukaryotes.

Abstract: Structural features of proteins are critical for mediating the diffusion of important molecules. In hemoproteins that bind diatomic gases, protein structure is essential for controlling the capture of gases as well as their escape. A tunnel network in an H-NOX sensor protein directs gases to the heme site where gas binding to the heme cofactor initiates H-NOX-mediated signaling events. H-NOX sensor proteins have unique interior topologies that differ from globins and appear to be connected with their divergent biological roles. Indeed, these tunnels stand in stark contrast to myoglobin and hemoglobin, which have structurally distinct pathways for gas entry and exit. Hence, alternative topologies in hemoproteins appear to control gas binding in diverse ways. A deeper understanding of how protein structural features influence function is essential for determining protein mechanisms and enabling the prediction of chemical properties across protein families. To further examine the importance of the gas tunnels, H-NOX variants were made in which amino acids that line the tunnels were replaced with larger amino acids (tryptophans) to block each branch of the tunnel network. X-ray crystallography was used to show that the amino acid changes did not perturb the overall protein structure. Additionally, an experimental method that measures the rates at which diatomic gas binds to and is released from the heme was used to examine these altered proteins. We found that H-NOX gas-binding properties were changed upon blocking of the tunnels, further suggesting that the tunnels play a functional role in facilitating gas flux. Together, our results indicate that blocking the tunnels may enhance gas trapping near the heme following gas diffusion into the protein interior. This result is expected because the routes for gas escape have been eliminated. To map pathways for gas diffusion in H-NOX proteins, X-ray crystal structures were obtained of a bacterial NO-binding H-NOX protein following exposure of the crystals to xenon gas. Xenon has similar properties, such as size and polarity, to NO and O 2 , but possesses a distinctly different signal in X-ray diffraction experiments, allowing xenon to be unambiguously identified in protein crystal structures. The structures obtained show that xenon in the protein forms a continuous pathway through the tunnel network ( Fig. P1 ). These findings support the hypothesis that tunnels are the most favorable routes for gas movement between the exterior of the protein and the buried heme site. Through structural analysis of H-NOX proteins, we identified a network of potential gas channels that extends from the surrounding solution (the solvent) to the interior heme-binding pocket ( Fig. P1 ). Importantly, these putative tunnels do not contain observable water molecules and are lined with hydrophobic (nonpolar) amino acids that are favorable for the movement of gases. Together, these observations suggested that the tunnels could serve as routes for diatomic gases between the solvent and the heme. Fig. P1. An X-ray crystal structure of an H-NOX protein following xenon pressurization shows an interior tunnel network. The tunnels mediate gas diffusion to the buried heme site and influence the gas-binding properties of the heme. Xenon atoms are in light blue. The heme cofactor is in gray. Investigations have been underway to understand the structural features that control gas binding in H-NOX proteins ( 3 ). Biochemical methods have been used to elucidate important determinants of gas binding in combination with three-dimensional structure determination (X-ray crystallography). These studies have shown that critical amino acids located around the heme enable the protein to bind O 2 at the heme cofactor by providing stabilizing interactions with O 2 ( 3 ). Although much insight has been gained into the roles of amino acids near the heme in influencing diatomic gas affinity, little is known about how other structural elements outside the heme site contribute to the range of gas-binding properties seen in the H-NOX family. Moreover, because little is known about how gases travel to the buried heme from the external environment in H-NOX proteins, we considered specifically whether protein structure could influence gas binding by providing internal pathways for gas migration. The globins are the most thoroughly studied family of gas-binding hemoproteins. Members of the globin family, such as myoglobin and hemoglobin, reversibly bind O 2 at the heme group for O 2 transport, delivery, and storage. Detailed studies have shown that gas-binding pockets located around the heme are critical for the efficient trapping of O 2 at the heme site and tuning O 2 affinity to match O 2 tensions in surrounding cellular environments ( 2 ). H-NOX proteins possess a distinct structure and, as mentioned above, are involved in sensing and signaling ( 3 ). Some H-NOX proteins bind NO and O 2 like the globins, whereas others do not bind O 2 , allowing for selective NO signaling in aerobic (O 2 -containing) environments ( 3 ). NO signaling has been studied most extensively in mammals where soluble guanylate cyclase, an H-NOX–containing protein, is involved in processes such as the dilation of blood vessels and communication between neurons ( 3 ). In bacteria, NO and O 2 signaling by H-NOX proteins regulates the formation of high-order bacterial assemblies referred to as “biofilms” ( 4 ), beneficial bacterial colonization of host organisms ( 5 ), and possibly bacterial movement in response to these gaseous stimuli ( 3 ). Proteins have evolved a variety of structures that provide the basis for specific biological functions. For example, enzymes (proteins that catalyze biochemical reactions) possess pockets and channels that mediate the entry and binding of reactants, sequester reaction intermediates, and facilitate the exit of reaction products. Protein function can be extended and diversified with the support of small, tightly bound molecules known as cofactors. Proteins that bind an iron-containing cofactor called heme (commonly referred to as hemoproteins or heme proteins) are found in organisms from bacteria to humans. The structural features of heme proteins work in concert with the heme cofactor to tune heme chemistry for various functions (e.g., electron transfer, catalysis, and diatomic gas binding) ( 1 ). In this study, we investigated a recently discovered family of hemoproteins known as heme nitric oxide/oxygen binding (H-NOX) proteins, which bind diatomic gases (e.g., nitric oxide, NO; molecular oxygen, O 2 ; carbon monoxide, CO) and function in gas-mediated biological signaling ( 1 ). Although an understanding of how diatomic gas molecules bind to the heme of H-NOX proteins has been developed, it remains unclear how the gases are directed to the internal heme cofactor from the external environment. Using structural and biochemical techniques, we found that gases can move to the heme through a network of internalized gas tunnels in the protein. This mechanism is quite different from that of other families of hemoproteins and suggests that alternative topologies in hemoproteins control gas binding in diverse ways.

Abstract: Human variation in susceptibility to hypoxia-induced pulmonary hypertension is well recognized. High-altitude residents who do not develop pulmonary hypertension may host protective gene mutations. Methods and Results— Exome sequencing was conducted on 24 unrelated Kyrgyz highlanders living 2400 to 3800 m above sea level, 12 (10 men; mean age, 54 years) with an elevated mean pulmonary artery pressure (mean±SD, 38.7±2.7 mm Hg) and 12 (11 men; mean age, 52 years) with a normal mean pulmonary artery pressure (19.2±0.6 mm Hg) to identify candidate genes that may influence the pulmonary vascular response to hypoxia. A total of 140 789 exomic variants were identified and 26 116 (18.5%) were classified as novel or rare. Thirty-three novel or rare potential pathogenic variants (frameshift, essential splice-site, and nonsynonymous) were found exclusively in either ≥3 subjects with high-altitude pulmonary hypertension or ≥3 highlanders with a normal mean pulmonary artery pressure. A novel missense mutation in GUCY1A3 in 3 subjects with a normal mean pulmonary artery pressure encodes an α 1 -A680T soluble guanylate cyclase (sGC) variant. Expression of the α 1 -A680T sGC variant in reporter cells resulted in higher cyclic guanosine monophosphate production compared with the wild-type enzyme and the purified α 1 -A680T sGC exhibited enhanced sensitivity to nitric oxide in vitro. Conclusions— The α 1 -A680T sGC variant may contribute to protection against high-altitude pulmonary hypertension and supports sGC as a pharmacological target for reducing pulmonary artery pressure in humans at altitude.