Modulating Enzyme Function via Dynamic Allostery within Biliverdin Reductase B

Abstract

The biliverdin reductase B (BLVRB) class of enzymes catalyze the NADPH-dependent reduction of multiple flavin substrates and are emerging as critical players in cellular redox regulation. However, the role of dynamics and allostery have not been addressed, prompting studies here that have revealed a position 15 Å away from the active site within human BLVRB (T164) that is inherently dynamic and can be mutated to control global micro-millisecond motions and function. By comparing the inherent dynamics through nuclear magnetic resonance (NMR) relaxation approaches of evolutionarily distinct BLVRB homologues and by applying our previously developed Relaxation And Single Site Multiple Mutations (RASSMM) approach that monitors both the functional and dynamic effects of multiple mutations to the single T164 site, we have discovered that the most dramatic mutagenic effects coincide with evolutionary changes and these modulate coenzyme binding. Thus, evolutionarily changing sites distal to the active site serve as dynamic "dials" to globally modulate motions and function. Despite the distal dynamic and functional coupling modulated by this site, micro-millisecond motions span an order of magnitude in their apparent kinetic rates of motions. Thus, global dynamics within BLVRB are a collection of partially coupled motions tied to catalytic function.

Keywords: NMR; allostery; coupling; dynamics; enzyme; reductase.

FIGURE 1

The general reaction mediated by BLVRB and the global impact mediated by coenzyme binding. (A) The catalytic reaction of BLVRB comprises the NADPH-dependent reduction of flavins. For flavins, “R” corresponds to different moieties that define the flavin such as FAD or FMN. For NADPH/NADP⁺, only the nicotinamide moiety is shown, and “R” corresponds to the remaining molecule. (B) X-ray crystal structure of WT BLVRB, accession number 1HDO (Pereira et al., 2001), along with a blow-up of the active site with residues S111 and H153 and residue T164 that is 15 Å away from the coenzyme. ¹⁵N-HSQC of apo BLVRB (black) and holo BLVRB (red) of the amides for (C) S111, (D) H153, and (E) T164.

FIGURE 2

Solution ensembles identify structural heterogeneity. (A) Structural ensembles were used to calculate a continuum of potentially sampled conformations using MovieMaker (Maiti et al., 2005). (B) Backbone RMSDs calculated both with (black line) and without (grey-dashed line) NMR data. (C) R2 relaxation rates extracted from the first 50 Hz refocusing time point of R2-CPMG relaxation dispersion data. (D) R1 relaxation rates of apo BLVRB.

FIGURE 3

BLVRB active site mutations S111A and H153A. (A) Michaelis-Menten kinetics monitored for BLVRB WT, S111A, and H153A. (B) Pre-steady-state kinetics for BLVRB WT, S111A, and H153A. (C) CSPs between apo BLVRB WT and apo BLVRB S111A with the line indicating 0.1 ppm that is the average plus ½ of the standard deviation. Residues for which their corresponding resonances were not found are also shown (gray). (D) CSPs between apo BLVRB WT and apo BLVRB H153A with the line indicating 0.075 ppm that is the average plus ½ of the standard deviation.

FIGURE 4

Evolutionary changes of BLVRB position 164 and comparative dynamics of human and lemur BLVRB family members. (A) Sequence comparison of six mammalian BLVRBs for residues 151–169 shown from top (most recent) to bottom as a function of their most recent evolutionary branch point as described by Hallstrom and Janke (2010). Position 164 is highlighted (yellow). (B) Superposition of human and lemur BLVRBs (PDB accession numbers 1HDO and 6OQG, respectively). (C) Lemur BLVRB comprises 16 amino acid changes mapped onto human BLVRB (magenta) with several of these amino acids highlighted that include position 164 (human in black, lemur in magenta). (D) R2 relaxation rate at 50 Hz CPMG refocusing for human (black) and lemur (magenta) as a function of residue. (E) Individually extracted rates of exchange (kex) for lemur BLVRB amides are shown as balls for rates less than 1000 s⁻¹ (blue), from 1000 to 3000 s⁻¹ (yellow), and higher than 3000 s⁻¹ (red). (F) Individually extracted rates of exchange (kex) for the same residues for human BLVRB amides are shown as balls for rates less than 1000 s⁻¹ (blue), from 1000 to 3000 s⁻¹ (yellow), and higher than 3000 s⁻¹ (red). Only residue human T15 is not shown as the exchange contribution was too low to fit (Supplementary Figure S1).

FIGURE 5

Functional changes induced by point mutations to human BLVRB T164. (A) A panel of mutations was screened for their dissociation constants (K_D) to NADP⁺ using ITC, as previously described. Mutations that were selected for further analysis are colored for WT BLVRB (black), T164A (orange), T164I (purple), and T164S (blue). (B) Steady-state kinetics for BLVRB WT (black), T164A (orange), T164I (purple), and T164S (blue). (C) Pre-steady-state kinetics BLVRB WT, T164A, T164I, and T164S at the lowest FAD concentration used.

FIGURE 6

CSPs and R1 relaxation rates of BLVRB T164 mutants. (A) CSPs between apo BLVRB WT and apo BLVRB T164A with the line indicating 0.08 ppm that is the average plus ½ of the standard deviation. (B) CSPs between apo BLVRB WT and apo BLVRB T164I with the line indicating 0.07 ppm that is the average plus ½ of the standard deviation. (C) CSPs between apo BLVRB WT and apo BLVRB T164S with the line indicating 0.07 ppm that is the average plus ½ of the standard deviation. (D) R1 relaxation rates for BLVRB WT previously published (Paukovich et al., 2018). (E) R1 relaxation rates for BLVRB T164A. (F) R1 relaxation rates for BLVRB T164I. (G) R1 relaxation rates for BLVRB T164S. All CSPs above the line are mapped onto the X-ray crystal structure (PDB accession 1HDO) and all data were collected at 900 MHz at 20°C.

FIGURE 7

Mutations to BLVRB T164 induce long-range dynamic changes. R2-CPMG dispersion profiles for BLVRB WT (black), T164A (orange), T164I (purple), and T164S (blue) at position 164 are shown with the amide of distal changes mapped onto the X-ray crystal structure shown as either green spheres (R2-CPMG profiles that are differentially altered upon mutation to T164 along with S111 that is not differentially perturbed) or red spheres (R2-CPMG profiles that exhibit similar responses upon mutation to T164).

FIGURE 8

Evolution and mutagenesis “dial” dynamics and function in BLVRB. (A) R2-CPMG dispersion profiles are shown for human BLVRB WT (black) and lemur BLVRB WT (magenta) along with human BLVRB WT and human BLVRB T164S (blue). (B) Amides that exhibit changes to their R2-CPMG dispersion profiles (blue spheres) are shown together with residues differences between human and lemur BLVRB homologues (magenta). (C) Steady-state kinetics of human WT BLVRB, lemur WT BLVRB, and human BLVRB T164S.