Correlating Conformational Equilibria with Catalysis in the Electron Bifurcating EtfABCX of Thermotoga maritima

Electron bifurcation (BF) is an evolutionarily ancient energy coupling mechanism in anaerobes, whose associated enzymatic machinery remains enigmatic. In BF-flavoenzymes, a chemically high-potential electron forms in a thermodynamically favorable fashion by simultaneously dropping the potential of a second electron before its donation to physiological acceptors. The cryo-EM and spectroscopic analyses of the BF-enzyme Fix/EtfABCX from Thermotoga maritima suggest that the BF-site contains a special flavin-adenine dinucleotide and, upon its reduction with NADH, a low-potential electron transfers to ferredoxin and a high-potential electron reduces menaquinone. The transfer of energy from high-energy intermediates must be carefully orchestrated conformationally to avoid equilibration. Herein, anaerobic size exclusion-coupled small-angle X-ray scattering (SEC-SAXS) shows that the Fix/EtfAB heterodimer subcomplex, which houses BF- and electron transfer (ET)-flavins, exists in a conformational equilibrium of compacted and extended states between flavin-binding domains, the abundance of which is impacted by reduction and NAD(H) binding. The conformations identify dynamics associated with the T. maritima enzyme and also recapitulate states identified in static structures of homologous BF-flavoenzymes. Reduction of Fix/EtfABCX’s flavins alone is insufficient to elicit domain movements conducive to ET but requires a structural “trigger” induced by NAD(H) binding. Models show that Fix/EtfABCX’s superdimer exists in a combination of states with respect to its BF-subcomplexes, suggesting a cooperative mechanism between supermonomers for optimizing catalysis. The correlation of conformational states with pathway steps suggests a structural means with which Fix/EtfABCX may progress through its catalytic cycle. Collectively, these observations provide a structural framework for tracing Fix/EtfABCX’s catalysis.


Figure S1 .
Figure S1.Detergents shift EtfABCX's oligomeric equilibrium towards the superdimer as shown by dynamic light scattering.Upon the addition of detergent, the hydrodynamic radius (Rh) distribution of EtfABCX shifts from ~8 nm to ~6 nm, representing a shift from supertetramer to superdimer states.Each detergent condition was measured in duplicate and as concentration series.Representative Rh distributions for each detergent are shown for clarity.

Figure S2 .
Figure S2.EtfABCX exhibits time-dependent sensitivity to oxygen-induced aggregation as analyzed with SEC-MALS.A. SEC-MALS traces of EtfABCX specimen as prepared anaerobically or after 10 minutes of air exposure prior to sample injection.B. SEC-MALS traces as shown in A with the addition of an overnight air exposure sample.

Figure S3 .
Figure S3.Anaerobic SEC-SAXS using 0.01% (w/v) DDM-containing running buffer isolates the SAXS signal of EtfABCX superdimer. A. SEC-SAXS trace of a representative EtfABCX specimen (no additives) showing a clearly defined peak for its superdimer and a minimal supertetramer shoulder.B. Merged SAXS profiles from both superdimer peak and supertetramer shoulder showing distinctly isolated scattering curves.Superdimer and supertetramer scattering profiles yielded SAXS MW values of 244 and 474 kDa, respectively (257 kDa expected for superdimer).C.An "X"-shaped EtfABCX supertetramer.D. The calculated profile for an EtfABCX supertetramer overlaid with the experimental scattering obtained from the supertetramer shoulder apparent during SEC-SAXS of NADH-reduced sample.The agreement at low to mid q supports the oligomeric state of the supertetramer.

Figure S4 .
Figure S4.In-line MALS collected during anaerobic SEC-MALS-SAXS show EtfABCX specimen peaks are characterized by species of the expected MW. A. Normalized MALS traces for no additive, NADH, and NAD + conditions for EtfABCX overlaid as collected during SEC-MALS-SAXS.B. MW analysis on the MALS data in (A) shows EtfABCX superdimer peaks contain species within +/-5% of the expected MW for a 9x histidine-tagged superdimer (257 kDa; LS = light scattering signal).C. SEC-MALS of EtfABCX with DT reduction shows steady MW values across its peak (inset) and absence of supertetramer formation.DT-containing datasets are separated from preceding panels for clarity.

Figure S5 .
Figure S5.SAXS profiles along with their respective Guinier and P(r) fits for EtfABCX and EtfAB specimens.Colored lines represent experimental data.Black lines represent calculated fits for P(r) functions.Total estimates (TE) from GNOM are provided.Insets: Guinier fits.Error bars represent the standard deviation of I(q). A. EtfABCX with no additives.B. EtfABCX with NADH.C. EtfABCX with NAD + .D. EtfABCX with DT.E. EtfAB with no additives.F. EtfAB with NADH.G. EtfAB with NAD + .H. EtfAB with DT.I. EtfAB with DT and NAD.

Figure S6 .
Figure S6.Comparison of EtfABCX's SAXS profile with the computed profile for its cryo-EM structure.The black line represents the computed profile for the cryo-EM structure (PDB ID: 7KOE).Inset: residuals.Error bars represent the standard deviation of I(q).

Figure S7 .
Figure S7.EtfABCX's supermonomers exhibit varying spatial correlations, in turn affecting cofactor distancing. A. Flexibly refined model of EtfABCX possessing a "twisted" configuration with respect to supermonomer conformation.B. End-to-end distances between ET-FAD and [4Fe-4S] II cofactors in cryo-EM (left) and refined model (right).ET-FAD and [4Fe-4S] II distances in the refined model resemble those in D-or D-like states.C. Fit of the NMArefined model's calculated profile to the SAXS data.Inset: residuals.Error bars represent the standard deviation of I(q).

Figure S8 .
Figure S8.In-line UV-visible spectra collected during anaerobic SEC-MALS-SAXS.The absorbances at 280, 374, 390, 454, and 636 nm were recorded during measurements of EtfABCX in no additive (A), NADH (B), NAD + (C), and DT (D) conditions, allowing for the monitoring of redox state and coenzyme binding.Similarly, spectra were recorded during measurements of EtfAB in no additive (E), NADH (F), NAD + (G), DT (H), and DT/NAD + (I) conditions.Spectra were normalized to the A280 maxima of the no additive conditions' EtfABCX superdimer or EtfAB heterodimer peaks, respectively.

Figure
Figure S9.UV-visible absorbance spectra of as-purified "no additive" T. maritima EtfABCX and EtfAB as expressed and purified from E. coli.Normalized absorbance peaks centered around 374 and 454 nm correspond to cofactor occupancy.

Figure S11 .
Figure S11.Fits of EtfAB's two-state models to their corresponding experimental data.The overlay of EtfAB's SAXS profiles with that of the fits (black lines) from two-state models as generated with BilboMD.Insets: residuals.A. EtfAB in the no additive condition.B. Same as in A but for EtfAB's NADH condition.C. Fit of the two-state model for EtfAB's NAD + condition.D. Fit of the two-state model for EtfAB's DT condition.E. Fit of the two-state model for EtfAB's DT and NAD + condition.

Figure S12 .
Figure S12.The fit of EtfABCX models to their corresponding experimental data.Insets: residuals.A. No additive condition's data with fit of EtfABCX possessing EtfAB modules in Dlike conformations.B. NADH condition's data with fit of EtfABCX possessing EtfAB modules in B-and D-like conformations.The models and their respective fits shown here were the highest scored from among the possibilities tested.Error bars represent the standard deviation of I(q).

Figure S13 .
Figure S13.Comparison of P(r) functions for EtfABCX's structural models.Overlaid are the P(r) functions calculated from EtfABCX's cryo-EM state, primary analysis of the experimental "no additive" SAXS data, the NMA-refined structure based on the "no additive" SAXS data, and the best-fitting BilboMD-derived model for the "no additive" SAXS data with EtfABs in D-like states.

Figure S14 .
Figure S14.Potential for intermolecular electron transfer between EtfAB .Two partiallyreduced EtfABs may perform intermolecular ET to yield partially or fully reduced forms of the enzyme.The potential for intermolecular ET increases with incubation time.