(a) Two perspectives of a homodimeric 4-α-helical maquette tethered by a disulfide bond. The tethered dimer is converted to a single chain form (center) by linking helices 2 and 3, allowing the sequence of all four of the helices to be different. Assembly exploits binary patterning with nonpolar amino acids on the interior (purple) and polar amino acids on the exterior (gray). (b) diversity in function is accomplished by securing a variety of cofactors (top) to various sites (bottom) within the maquette using appropriate cofactor anchoring amino acids (green).

Thermal denaturation of α-helical content monitored as a loss of cd at 222 nm. (a) core packing stabilization of the apo-maquette by replacing histidines in a with alanines in D and phenylalanines in E. (b) effects of modifying sequence a to include charge patterning, intrahelical salt bridges (F, g) or an N-cap (H). (c) The impact of one- and two-heme ligation on the thermal stability of a and c, which can only ligate a single heme. All of the cd experiments were performed with 20 μm protein in 150 mm Kcl, 20 mm cHeS at pH 9 with the cd signal amplitude normalized to unity at 5 °c. Thermal melts show substantial reversibility with 85% of heme binding and protein helicity maintained after exposing a to 95 °c for 30 min, then recooling to 25 °c (). Some aggregation or degradation may occur at prolonged near boiling temperatures. All of the cd melt data was performed once with an averaging time of 30 s at each temperature (5 °c to 95 °c in 5J increments). For all melts, 222 nm was monitored. data were collected once, and the error associated with the numbers (± 5°) was taken from the error provided by the fitting program.

(a) In the absence of cofactor, Nmr resonances of a are relatively undispersed. (b) Addition of two equivalents of heme to the His-free E is also relatively undispersed. (c) Addition of one equivalent to a induces partial dispersion, indicative of stable structure. (d) Addition of a second equivalent of heme induces further structuring. Nmr was performed with 200 μm a in 50 mm KH₂Po₄ and 50 mm Kcl, pH 7.9, at 25 °c.

(a) Absorbance (Abs) spectra of 4 μm a equipped with two hemes in 20 mm cHeS, 150 mm Kcl, pH 9.0, 30% (v/v) ethylene glycol. Heme states: blue, ferrous; green, ferric; black, carboxyferrous; red, oxyferrous. Spectra were taken at −15 °c, where oxyferrous states are stable for >1 h. Similar spectra are seen at 15 °c (not shown). dashed red trace shows the one-heme oxyferrous spectrum of a doubled in scale for direct comparison with the two-heme oxyferrous spectrum (solid line). (b) Kinetic measurements were recorded mixing 6 μm ferrous a with oxygen-saturated buffer in Tris buffer (20 mm Tris, 20 mm Nacl, pH 8.0, at 15 °c). Heme state populations (blue, ferrous; red, oxyferrous; green, ferric) were computed by deconvoluting the components of the stop-flow spectra, normalizing the oxyferrous maximum (574 nm). (c) detection of superoxide using chemiluminescent probe mclA (30 μm) during the stopped-flow mixing of 5 μm maquette with air-saturated Tris buffer (pH 8). maquettes shown are the tethered dimer (black), a (blue dashed) and i (red). (d,e) electron-transfer from partially reduced 9.5 μm a with two equivalents of heme to 17 μm cytochrome c. Shown in d are Q-band spectral changes for both proteins from 0 ms (black) to 30 ms (red). Peaks at 558 nm and 548 correspond to the α-bands of ferrous heme in a and cytochrome c, respectively. Shown in e are redox kinetics for holo-a and cytochrome c over 100 ms. The black and red traces show relative absorbance changes at the 558-nm ferrous cytochrome c peak and the 548-nm ferrous a peak, respectively, compared to the isosbestic point at 553 nm. experiments were performed in 50 mm HePeS, 120 mm Kcl, pH 8.0. Statistical analysis for all data is in the online methods.

(a) Transient absorption spectroscopy of single-chain maquettes binding a variety of cofactors. A laser pulse excites Zn chlorin (Znc) bound to c or Zn protoporphyin IX (ZnP) bound to J, and the decay of the resulting triplet state is tracked in the absorption spectra. Normalized triplet populations are plotted for each system and fitted to single exponential decays. The data for all of these traces were fit to a single exponential function. All of these data have an error of 10%. (b) reduction of heme bound in 8 μm J following photoexcitation of ZnP. dashed and bold traces are absorption spectra of the sample before and after 20 min of white light illumination, respectively. experiments were performed in 20 mm cHeS, 150 mm Kcl, pH 9.0, and 0.1 mm aniline. (c) reduction of bound heme following photoexcitation of 8-bromoriboflavin covalently attached to 8 μm K. dashed and bold traces show absorption spectra before and after 60 min of white light illumination, respectively. experiments were performed in 20 mm cHeS, 150 mm Kcl, pH 9.0, and 1 mm edTA. experiments in b and c were done in duplicate.