Adrenodoxin allosterically alters human cytochrome P450 11B enzymes to accelerate substrate binding and decelerate release

Abstract

Two human mitochondrial membrane CYP11B enzymes play a pivotal role in steroidogenesis. CYP11B1 generates the major glucocorticoid cortisol, while CYP11B2 catalysis yields the primary mineralocorticoid aldosterone. Catalysis by both requires electron delivery by a soluble iron-sulfur adrenodoxin redox partner. However recent studies have shown that adrenodoxin/CYP11B interaction alone allosterically increases substrate and inhibitor affinity as exhibited by decreased dissociation constant (K _d) values. The current study moves beyond such equilibrium studies, by defining adrenodoxin effects on the rates of P450 ligand binding and release separately. Stopped-flow data clearly demonstrate that adrenodoxin interaction with the P450 proximal surfaces increases ligand binding in both P450 CYP11B active sites by increasing the on rate constant and decreasing the off rate constant. As substrate entry and exit from the sequestered P450 active site requires conformational changes on the distal side of the P450 enzyme, a likely explanation is that adrenodoxin binding allosterically modulates CYP11B conformational changes. The 93% identical CYP11B enzymes can bind and hydroxylate each other's native substrates differing only by a hydroxyl. However, CYP11B1 exhibits monophasic substrate binding and CYP11B2 biphasic substrate binding, even when the substrates are swapped. This indicates that small differences in amino acid sequence between human CYP11B1 and CYP11B2 enzymes are more functionally important in ligand binding and could suggest avenues for more selective inhibition of these drug targets. Both protein/protein interactions and protein/substrate interactions are most likely to act by modulating CYP11B conformational dynamics.

Fig. 1. Cytochrome P450 11B enzyme ligands. (A) The CYP11B1-mediated 11β-hydroxylation of 11-deoxycortisol results in the major glucocorticoid hormone cortisol. (B) Human CYP11B2 generates the major mineralocorticoid hormone aldosterone. Similar to CYP11B1, CYP11B2 performs an 11β hydroxylation to convert 11-deoxycorticosterone to corticosterone, but subsequently performs an 18-hydroxylation and 18-oxidase reaction to generate aldosterone. (C) LCI699 is an FDA-approved cushing syndrome drug displaying low nanomolar affinity for both CYP11B1 and CYP11B2 and serves as a heme iron-ligating inhibitor.

Fig. 2. Determination of rate constants for CYP11B1 binding to its major substrate 11-deoxycortisol. CYP11B1 (1 μM) preincubated with various concentrations of adrenodoxin (0, 1, 10, or 40-fold excess) was mixed with 11-deoxycortisol (5, 10, 20, or 50 μM before mixing) and spectra recorded. (A) Changes in the ΔA₃₉₁–A₄₁₉ nm representing substrate binding are shown for CYP11B1 with constant (10-fold) excess adrenodoxin mixed with four different 11-deoxycortisol concentrations. (B) Changes in the ΔA₃₉₁–A₄₁₉ nm representing substrate binding are shown for CYP11B1 with constant (10 μM) 11-deoxycortisol mixed with four different adrenodoxin concentrations. All CYP11B1 binding curves were a single phase, best fit by a single exponential equation (solid lines). The remaining adrenodoxin and 11-deoxycortisol conditions are shown in Fig. S2 (ESI†). (C) Plots of the rate constants (k_obs) from panels A, B, and the additional experiments in Fig. S2 (ESI†) are plotted vs. 11-deoxycortisol concentration for the various adrenodoxin concentrations. (D) The same plot of the rate constant (k_obs) against 11-deoxycortisol concentration is shown for an artificial adrenodoxin/CYP11B1 fusion enzyme.

Fig. 3. Determination of rate constants for CYP11B2 binding to its major substrate 11-deoxycorticosterone. CYP11B2 (2 μM) preincubated with various concentrations of adrenodoxin (0, 1, 10, or 40-fold excess) was mixed with 11-deoxycorticosterone (5, 10, 20, or 50 μM before mixing) and spectra recorded. (A) Changes in the ΔA₃₉₁–A₄₁₉ nm representing substrate binding are shown for CYP11B2 in the presence of constant (10-fold excess) adrenodoxin with four different 11-deoxycorticosterone concentrations. (B) Changes in the ΔA₃₉₁–A₄₁₉ nm representing substrate binding are shown for CYP11B2 with constant (10 μM) 11-deoxycorticosterone and varying adrenodoxin concentrations. The remaining adrenodoxin and 11-deoxycorticosterone conditions are shown in Fig. S3 (ESI†). All traces were best fit to a biexponential equation (solid lines) yielding rate constants for both the slow and fast phases (k_fast and k_slow). For these experiments, the percent fast phase ranged from 40–59%. (C) The k_fast values for CYP11B2 in the presence of various adrenodoxin concentrations were plotted against 11-deoxycorticosterone concentration and fit to a line to determine k_on (slopes) and k_off (y-intercepts) in Table 1. (D) The k_slow values for CYP11B2 in the presence of various adrenodoxin concentrations were plotted against 11-deoxycorticosterone concentration and fit to a line but yielded little change systematic change in the rates. (E) The corresponding k_fast values for the adrenodoxin/CYP11B2 fusion enzyme plotted against 11-deoxycorticosterone concentration yielded the k_on and k_off rates in Table 1. (F) The k_slow values for the adrenodoxin/CYP11B2 fusion enzyme did not significantly vary with substrate concentration. In all experiments, data was collected for at least 15 seconds, but only the first 5 seconds are shown here to facilitate visualization of both phases.

Fig. 4. Determination of rate constants for CYP11B1 binding its minor substrate 11-deoxycorticosterone. CYP11B1 (1 μM) preincubated with various concentrations of adrenodoxin (0, 1, 10, or 40-fold excess) was mixed 1 : 1 with 11-deoxycortisol (10, 20, 35, or 50 μM before mixing). (A) Changes in the ΔA₃₉₁–A₄₁₉ nm representing substrate binding are shown for CYP11B1 in the presence of constant (10-fold excess) adrenodoxin with four different 11-deoxycorticosterone concentrations. (B) Changes in the ΔA₃₉₁–A₄₁₉ nm representing substrate binding are shown for CYP11B1 with constant (10 μM) 11-deoxycorticosterone and varying adrenodoxin concentrations. The remaining adrenodoxin and 11-deoxycorticosterone conditions are shown in Fig. S4 (ESI†). All experiments were readily fit to a single exponential (solid lines). (C) Plots of the resulting k_obs plotted against 11-deoxycorticosterone concentration were fit to a line to determine the k_on (slopes) and k_off (y-intercepts) in Table 1.

Fig. 5. Determination of rate constants for CYP11B2 binding to its minor substrate 11-deoxycortisol. CYP11B2 (2 μM) preincubated with various concentrations of adrenodoxin (0, 1, 10, or 40-fold excess) was mixed 1 : 1 with 11-deoxycorticosterone (10, 20, 35, or 50 μM before mixing). (A) Changes in the ΔA₃₉₁–A₄₁₉ nm representing substrate binding are shown for CYP11B2 in the present of constant (10-fold excess) adrenodoxin with four different 11-deoxycortisol concentrations. (B) Changes in the ΔA₃₉₁–A₄₁₉ nm representing substrate binding are shown for CYP11B2 with constant (10 μM) 11-deoxycortisol and varying adrenodoxin concentrations. The remaining adrenodoxin and 11-deoxycorticosterone conditions are shown in Fig. S5 (ESI†). All experiments fit best to the biexponential equation (solid lines) yielding rate constants for both the slow and fast phases (k_slow and k_fast). For these experiments, the percent fast phase ranged from 44–62%. (C) These k_fast values plotted against 11-deoxycortisol concentration for CYP11B2 with various amounts of adrenodoxin were fit to a line (solid line) to determine the k_on (slopes) and k_off (y-intercepts) in Table 1. (D) The k_slow values were also plotted against 11-deoxycorticosterone concentration yielding k_on (slopes) and k_off (y-intercepts) that are fairly constant.

Fig. 6. Determination of rate constants for LCI699 binding to CYP11B1 and CYP11B2 to evaluate LCI699 as a potential way to measure the formation of substrate-free enzyme (“trapping compound”). CYP11B1 or CYP11B2 (4 μM) was mixed with 100 μM LCI699 and absorbance spectra were recorded, resulting in a type II shift with increasing absorbance a 424 nm and decreasing absorbance at 420 nm (Fig. S7, ESI†). The Δ424–420 nm was plotted against time and fit to determine the binding rate constants. (A) The data for CYP11B1 binding LCI699 was best fit to the one-phase association equation (solid line) yielding a k_obs of 65.1 (61.6–68.9) s⁻¹. (B) The data for CYP11B2 binding LCI699 was best fit to the two-phase association equation (solid line) yielding a k_fast value of 47.3 (41.1–54.4) s⁻¹ and a k_slow value of 4.9 (4.2–5.6) s⁻¹.

Fig. 7. Experiments using LCI699 as a trapping agent to quantitate substrate k_off. Either CYP11B1 or CYP11B2 (2 μM) was pre-equilibrated with a saturating amount (70 μM) of the substrates 11-deoxycortisol or 11-deoxycorticosterone and either 0, 1, 10, or 40-fold excess adrenodoxin. The CYP11B/substrate/adrenodoxin sample was mixed 1 : 1 with 100 μM LCI699 and absorbance spectra was recorded. These spectra revealed an increase in absorbance at 424 nm and decrease at 390 nm consistent with conversion from the substrate-bound to inhibitor-bound form (Fig. S8, ESI†). The ΔA₃₉₀–A₄₂₄ plotted against time are shown for (A) CYP11B1 with 11-deoxycorticosterone and (B) CYP11B2 with 11-deoxycorticosterone. The remaining traces are shown in Fig. S9 (ESI†). The fit of this data was used to determine the substrate k_off rates in Table 2.

Fig. 8. CYP11B2/adrenodoxin structure (PDB 7M8I). Adrenodoxin (dark grey ribbons) binds to the proximal side of CYP11B2 (cyan ribbons), while substrates and inhibitors (here illustrated by fadrozole in light grey sticks) bind to the distal side of the heme (black sticks). Substrate entry and exit is thought to occur via conformational changes in the P450 secondary structure elements at the far top in this orientation. Thus adrenodoxin is thought to act allosterically or at a distance to modulate substrate on and off rate constants.