A novel class of proteins called the heme-based gas sensors (NO, CO and O₂) have now been discovered in a wide variety of organisms, including humans, that regulate many important biological processes. ^1,2,3  However, the exact mechanisms by which these proteins sense their specific gas and send out their regulatory signals have not been elucidated due to a lack of structural and  biochemical information. Our research laboratory studies the mechanism of oxygen sensing in SmFixL, a two-component heme-based O₂ sensor system from Sinorhizobium meliloti, which regulates nitrogen fixation and micro aerobic respiration in the root nodules of alfalfa plants, as a model for this larger family of heme-based oxygen sensors.  

In our laboratory we study a truncated version of  SmFixL_127-505 (43 kd, amino acids 127 to 505). FixL* contains the PAS heme domain and a kinase activity domain, but lacks the transmembrane domain.  Previous studies have shown that the kinase domain of SmFixL is inactive when oxygen or other small molecules such as cynanide are bound to the heme iron of the PAS domain “off-state”, but the kinase domain is activated when oxygen is released from the heme iron (deoxy FixL, the “on state”). The goal of our research is to understand how FixL senses oxygen and transmits this information to the kinase domain by using site-directed mutagenesis studies to probe the role of key conserved amino acids in the heme domain. We consistently get 100 to 200 milligrams of highly pure wild-type FixL_127-505  protein based on Bradford protein concentration assay, UV-vis heme to protein ratio, SDS-PAGE and MALDI mass spectroscopy.

We have several surprising and exciting results with ~14  variant proteins. The expression of the proximal histidine variants H194M, H194C and H194G show that they don't form a normal wild-type hemoprotein spectrum. Instead, they contain a novel hemoprotein with a UV-vis spectrum similar to a six-coordinate heme with a Soret band at ~426 nm (see Figure 1). These H194 variant proteins do not bind to a positively charged DEAE column like the wild-type protein so they are probably not negatively charged at pH 7.8 Tris like wild-type. This tells us that the H194 is crucial for binding heme in FixL and that a change in this residue forms a very different hemoprotein that we are currently studying in more detail.

Next, D195 is strictly conserved in the FixL family and is right next to the proximal histidine 194 and is a nice control to see whether this area is important for oxygen sensing. Although we get much lower yields of D195A than with wild-type, we do see the typical stable oxy spectrum as wild-type suggesting this residue is not important in sensing. Right next to D195 is the strictly conserved residue Y197 that changes position next to the heme in a recent time-resolved crystal structure. We purified and characterized Y197A which quickly oxidized to the met form in our hands (see Figure 1). This suggests that Y197 may be important in stabilizing the oxy state of FixL.

Most importantly, we looked at the strictly conserved residue R200 that moves in to interact with the heme-6-propionate  upon oxygen binding “the off state” and may be the key trigger in signaling the kinase domain. A recent paper from Gilles-Gonzales et al. on BjFixL suggests that R200 is crucial for the kinase activity based on a R200A variant protein. In our lab we have purified and characterized R200A, R200Q, R200E, R200H and R200I (see UV-vis spectra Figure 1). R200Q is very similar to wild-type as far as the stability of its oxy complex and expression which makes sense because this residue is the most conserved. R200A is also similar to wild-type but the oxy complex is less stable. Most interestingly, both R200E and R200H  binding oxygen much more slowly than wild-type presumably due to the difference in charges. R200I is very unstable and loses heme during the protein purification. These results suggest an important role for R200 in the oxygen sensing mechanism of  FixL. Interestingly, when the adjacent residue Y201 is changed to Y201A or its hydrogen bonding partner E234 is changed to E234A, the hemoprotein looks like H194G, suggesting that it is crucial for heme binding. Thus, R200 and Y201 make interact to transmit the oxygen binding signal of FixL. We are currently putting these exciting results together for a communication and one or two full papers.

1) Chan, M. K et al. J. Biol. Inorg. Chem. 2003, 8, 1-11.

2) Gilles-Gonzalez, M.-A. et al. J. Inorg. Biochem., 2005, 99, 1-22.

3) Gilles-Gonzalez, M.-A. et al. Nature, 1991, 350, 170-172.

4) Gilles-Gonzalez, M.-A. et al. J. Mol. Biol. 2006, 360, 80-89.

"Acknowledgment is made to the Donors of the American Chemical Society Petroleum Research Fund for support (or partial support) of this research."