Structures of biological heme-based sensors of oxygen

Dr. Ann Walker and one of us (M.-A. Gilles-Gonzalez) first met when, by chance, we both chose to eat lunch at the same table at a 1998 meeting on hemeproteins at Asilomar, California. That morning's session had ended with a lecture that had described the de novo design of helical maquettes that could beautifully coordinate protoporphyrin IX [1]. The presenter had left the audience riveted, but with a question, which I believe to have been: “Why are alpha-helical cages ideally suited for heme binding?” That afternoon, Dr. Walker would present the first structure of nitrophorin, an unprecedented fold for a hemeprotein, which showed that a heme could be sandwiched in a beta-barrel [2]. I followed with the surprising structure of the oxygen-sensing domain of Bradyrhizobium japonicum FixL (BjFixL), in which the heme was coordinated to a proximal histidine residue in an alpha helix but lined on the distal side with anti-parallel beta strands [3]. In subsequent years Dr. Walker and I have gravitated toward each other at conferences for many reasons, only one of those being our fondness for unorthodox hemeproteins.

Sensory hemeproteins are biological switches. They are invariably modular, and this is necessary for their function. They not only bind a regulatory heme ligand, such as O2, CO, or NO, but also couple the ligated state to the activity of another domain within the same protein. Over evolutionary time, biological sensors appear to have adapted a broad variety of regulatory modules to bind those cofactors necessary for their functions, heme being only one of them. Since this topic was last reviewed, the breadth of the heme-based sensors, in terms of their module compositions and biological functions, has expanded at a dizzying pace [[4], [5], [6], [7], [8], [9], [10], [11]]. Suffice it to note that many different folds can potentially bind heme, particularly in sensory proteins. So far no cofactor other than heme has been discovered to associate with a globin. The globin fold has, however, been found without any heme cofactor in at least two sensors [12,13]. Such proteins and the heme-based globin-coupled sensors (GCS) more closely resemble the hypothetical universal common ancestor of the hemoglobins, which is thought to be alpha-helical but without a D-helix [14,15].

The definitive undertaking with regard to sensors is to unravel their signal-transduction mechanisms. In the last 15 years many more heme-binding domains have yielded to structural investigations than the full-length proteins from which they originated. For one, the full-length multi-domain proteins appear to adopt multiple conformations. In addition some of those proteins may function together with other factors. They may associate, for example, with regulatory protein partners and macromolecular targets, such as nucleic acids that affect their stability, solubility, and other properties. As daunting as it has been to see full-length O2 sensors at work, the future suddenly looks brighter, partly because of the rapid pace at which cryo-EM methods have advanced. Now seems a propitious time to ponder the structural questions that previous studies have raised but could not fully answer due to technical limitations.

留言 (0)

沒有登入
gif