A test of the relationship between sequence and structure in proteins: excision of the heme binding site in apocytochrome b5

Structural propensities in the heme binding region of apocytochrome b5. II. Heme conjugates

In the absence of heme cofactor, the water-soluble domain of rat microsomal cytochrome b5 (cyt b5) contains a long flexible region within its 42-residue heme-binding loop. Heme capture induces this region to fold into a well-defined structure containing helices H3-H5, each separated by a turn, with His39 and His63 serving as axial ligands to the heme iron. We have shown that the H4 region of the apoprotein has the greatest tendency for disorder within the isolated binding loop. Here, the effect of the His63-iron bond and proximity of heme plane on the population of helical conformation in H4 and H5 was investigated by synthesis and characterization of a peptide-sandwiched mesoheme construct in which two H4-H5 peptides were covalently attached to a single cofactor. Spectroscopic data indicated that a holoprotein-like bis-histidine coordination state was achieved over a pH range from 7 to 9. Trifluoroethanol titrations of the construct and the analogous free peptide under these pH conditions revealed that heme proximity and iron ligation were insufficient to promote helix formation in H4 and H5. These observations were used to assess the role of disordered regions in heme capture and the loop-scaffold interface in holoprotein folding and stability.

Structural and thermodynamic encoding in the sequence of rat microsomal cytochrome b(5)

The water-soluble domain of rat microsomal cytochrome b(5) is a convenient protein with which to inspect the connection between amino acid sequence and thermodynamic properties. In the absence of its single heme cofactor, cytochrome b(5) contains a partially folded stretch of 30 residues. This region is recognized as prone to disorder by programs that analyze primary structures for such intrinsic features. The cytochrome was subjected to amino acid replacements in the folded core (I12A), in the portion that refolds only when in contact with the heme group (N57P), and in both (F35H/H39A/L46Y). Despite the difficulties associated with measuring thermodynamic quantities for the heme-bound species, it was possible to rationalize the energetic consequences of both types of replacements and test a simple equation relating apoprotein and holoprotein stability. In addition, a phenomenological relationship between the change in T(m) (the temperature at the midpoint of the thermal transition) and the change in thermodynamic stability determined by chemical denaturation was observed that could be used to extend the interpretation of incomplete holoprotein stability data. Structural information was obtained by nuclear magnetic resonance spectroscopy toward an atomic-level analysis of the effects.

Enhancing the thermal stability of mitochondrial cytochrome b5 by introducing a structural motif characteristic of the less stable microsomal isoform

Outer mitochondrial membrane cytochrome b5 (OM b5) is the most thermostable cytochrome b5 isoform presently known. Herein, we show that OM b5 thermal stability is substantially enhanced by swapping an apparently invariant motif in its heme-independent folding core with the corresponding motif characteristic of its less stable evolutionary relative, microsomal cytochrome b5 (Mc b5). The motif swap involved replacing two residues, Arg15 with His and Glu20 with Ser, thereby introducing a Glu11-His15-Ser20 H-bonding triad on the protein surface along with a His15/Trp22 pi-stacking interaction. The ferric and ferrous forms of the OM b5 R15H/E20S double mutant have thermal denaturation midpoints (Tm values) of approximately 93 degrees C and approximately 104 degrees C, respectively. A 15 degrees C increase in apoprotein Tm plays a key role in the holoprotein thermal stability enhancement, and is achieved by one of the most common natural mechanisms for stabilization of thermophilic versus mesophilic proteins: raising the unfolding free energy along the entire stability curve.

Comparison of cytochromes b5 from insects and vertebrates

We report a 1.55 A X-ray crystal structure of the heme-binding domain of cytochrome b(5) from Musca domestica (house fly; HF b(5)), and compare it with previously published structures of the heme-binding domains of bovine microsomal cytochrome b(5) (bMc b(5)) and rat outer mitochondrial membrane cytochrome b(5) (rOM b(5)). The structural comparison was done in the context of amino acid sequences of all known homologues of the proteins under study. We show that insect b(5)s contain an extended hydrophobic patch at the base of the heme binding pocket, similar to the one previously shown to stabilize mammalian OM b(5)s relative to their Mc counterparts. The hydrophobic patch in insects includes a residue with a bulky hydrophobic side chain at position 71 (Met). Replacing Met71 in HF b(5) with Ser, the corresponding residue in all known mammalian Mc b(5)s, is found to substantially destabilize the holoprotein. The destabilization is a consequence of two related factors: (1) a large decrease in apoprotein stability and (2) extension of conformational disruption in the apoprotein beyond the empty heme binding pocket (core 1) and into the heme-independent folding core (core 2). Analogous changes have previously been shown to accompany replacement of Leu71 in rOM b(5) with Ser. That the stabilizing role of Met71 in HF b(5) is manifested primarily in the apo state is highlighted by the fact that its crystallographic Calpha B factor is modestly larger than that of Ser71 in bMc b(5), indicating that it slightly destabilizes local polypeptide conformation when heme is in its binding pocket. Finally, we show that the final unit of secondary structure in the cytochrome b(5) heme-binding domain, a 3(10) helix known as alpha6, differs substantially in length and packing interactions not only for different protein isoforms but also for given isoforms from different species.

A dynamic N-capping motif in cytochrome b5: evidence for a pH-controlled conformational switch

Apocytochrome b5 is a marginally stable protein exhibiting under native conditions a slow conformational exchange in its C-terminal region. The affected elements of secondary structure include a 3(10)-helix containing at its N-terminus a histidine Ncap and a subsequent proline. Participation of the neutral histidine side-chain in backbone amide capping lowers the imidazole pKa. To explore the nature of the conformational exchange in the protein and determine whether it is related to cis-trans isomerization of the His-Pro bond, three octapeptides encompassing the helix were synthesized and studied by NMR spectroscopy. One corresponded to the wild-type sequence, the second was the D-histidine epimer, and the third contained an alanine in place of the proline. It was found that the rates of cis-trans interconversion in the proline-containing peptides were slower than the rates of the conformational exchange in the protein. In addition, the wild-type peptide hinted at a predisposition for Ncap formation when in the trans configuration. Analysis of the pH response of the peptides and protein suggested that at pH near neutral, the conformational exchange detected in the protein involved only species with a trans His-Pro bond and could be approximated with a three-state model by which the terminal helix sampled a locally unfolded state. This state, which contained an uncapped histidine with a normal pKa, partitioned into neutral and protonated populations according to pH. The intrinsic conformational bias of the wild-type peptide and the pH-driven equilibria illustrated how a 3(10)-element could serve as a nucleation site for structural rearrangement.

Retracted article: In-cell protein dynamics

The native intracellular environment of proteins is crowded with metabolites and macromolecules. However, most biophysical information concerning proteins is acquired in dilute solution. To determine whether there are differences in dynamics, nuclear magnetic resonance spectroscopy can be used to measure 15N relaxation in uniformly 15N-enriched apocytochrome b5 inside living Escherichia coli and in dilute solution. Such data can then be used to compare the fast backbone dynamics of the partially folded protein in cells to its dynamics in dilute solution by using Lipari-Szabo analysis. It appears that the intracellular environment does not alter the protein's structure, or significantly change its fast dynamics. Specifically, the cytosol does not change the amplitude of fast backbone motions, but does increase the average timescale of these motions, most likely due to the increase in viscosity of the cytosol.

Insertion of the cytochrome b5 heme-binding loop into an SH3 domain. Effects on structure and stability, and clues about the cytochrome's architecture

Under native conditions, apocytochrome b(5) exhibits a stable core and a disordered heme-binding region that refolds upon association with the cofactor. The termini of this flexible region are in close proximity, suggesting that loop closure may contribute to the thermodynamic properties of the apocytochrome. A chimeric protein containing 43 residues encompassing the cytochrome loop was constructed using the cyanobacterial photosystem I accessory protein E (PsaE) from Synechococcus sp. PCC 7002 as a structured scaffold. PsaE has the topology of an SH3 domain, and the insertion was engineered to replace its 14-residue CD loop. NMR and optical spectroscopies showed that the hybrid protein (named EbE1) was folded under native conditions and that it retained the characteristics of an SH3 domain. NMR spectroscopy revealed that structural and dynamic differences were confined near the site of loop insertion. Variable-temperature 1D NMR spectra of EbE1 confirmed the presence of a kinetic unfolding barrier. Thermal and chemical denaturations of PsaE and EbE1 demonstrated cooperative, two-state transitions; the stability of the PsaE scaffold was found only moderately compromised by the insertion, with a DeltaT(m) of 8.3 degrees C, a DeltaC(m) of 1.5 M urea, and a DeltaDeltaG degrees of 4.2 kJ/mole. The data implied that the penalty for constraining the ends of the inserted region was lower than the approximately 6.4 kJ/mole calculated for a self-avoiding chain. Extrapolation of these results to cytochrome b(5) suggested that the intrinsic stability of the folded portion of the apoprotein reflected only a small detrimental contribution from the large heme-binding domain.

Stabilizing roles of residual structure in the empty heme binding pockets and unfolded states of microsomal and mitochondrial apocytochrome b5

The microsomal (Mc) and mitochondrial (OM) isoforms of mammalian cytochrome b5 are the products of different genes, which likely arose via duplication of a primordial gene and subsequent functional divergence. Despite sharing essentially identical folds, heme-polypeptide interactions are stronger in OM b5s than in Mc b5s due to the presence of two conserved patches of hydrophobic amino acid side chains in the OM heme binding pockets. This is of fundamental interest in terms of understanding heme protein structure-function relationships, because stronger heme-polypeptide interactions in OM b5s in comparison to Mc b5s may represent a key source of their more negative reduction potentials. Herein we provide evidence that interactions amongst the amino acid side chains contributing to the hydrophobic patches in rat OM (rOM) b5 persist when heme is removed, rendering the empty heme binding pocket of rOM apo-b5 more compact and less conformationally dynamic than that in bovine Mc (bMc) apo-b5. This may contribute to the stronger heme binding by OM apo-b5 by reducing the entropic penalty associated with polypeptide folding. We also show that when bMc apo-b5 unfolds it adopts a structure that is more compact and contains greater nonrandom secondary structure content than unfolded rOM apo-b5. We propose that a more robust beta-sheet in Mc apo-b5s compensates for the absence of the hydrophobic packing interactions that stabilize the heme binding pocket in OM apo-b5s.

House fly cytochrome b5 exhibits kinetically trapped hemin and selectivity in hemin binding

We report that cytochrome b(5) (cyt b(5)) from Musca domestica (house fly) is more thermally stable than all other microsomal (Mc) cytochromes b(5) that have been examined to date. It also exhibits a much higher barrier to equilibration of the two isomeric forms of the protein, which differ by a 180 degrees rotation about the alpha-gamma-meso axis of hemin (ferric heme). In fact, hemin is kinetically trapped in a nearly statistical 1.2:1 ratio of rotational forms in freshly expressed protein. The equilibrium ratio (5.5:1) is established only upon incubation at temperatures above 37 degrees C. House fly Mc cyt b(5) is only the second b-hemoprotein that has been shown to exhibit kinetically trapped hemin at room temperature or above, the first being cyt b(5) from the outer membrane of rat liver mitochondria (rat OM cyt b(5)). Finally, we show that the small excess of one orientational isomer over the other in freshly expressed protein results from selective binding of hemin by the apoprotein, a phenomenon that has not heretofore been established for any apocyt b(5).

A rapid test for identification of autonomous folding units in proteins

The structure of a protein is dictated by a large number of weak interactions that cooperatively stabilize the native state. Usually, excised fragments smaller than a domain have little if any residual structure. When autonomous units of structure are found within domains, this challenges common assumptions about the cooperativity of protein structure. Such autonomous folding units (AFUs) are of wide interest and have applications in protein engineering and as simple model systems for studying the determinants of stability and specificity. A new method of identifying AFUs within proteins is presented here. The rapid autonomous fragment test (RAFT) identifies AFUs based on analysis of inter-residue contacts present in the three-dimensional structure of a protein. RAFT is fast enough to mine the entire PDB for AFUs and provide a library of potential small stable folds. We show that RAFT is able to predict whether a protein fragment will be structured if isolated from its parent domain.

A physical basis for protein secondary structure

A physical theory of protein secondary structure is proposed and tested by performing exceedingly simple Monte Carlo simulations. In essence, secondary structure propensities are predominantly a consequence of two competing local effects, one favoring hydrogen bond formation in helices and turns, the other opposing the attendant reduction in sidechain conformational entropy on helix and turn formation. These sequence specific biases are densely dispersed throughout the unfolded polypeptide chain, where they serve to preorganize the folding process and largely, but imperfectly, anticipate the native secondary structure.

Folding of an isolated ribonuclease H core fragment

Based on results from both equilibrium and kinetic hydrogen exchange studies of Escherichia coli ribonuclease HI (RNase H), a fragment of RNase H (eABCD) was designed. The sequence of eABCD contains less than half of the protein's primary sequence and includes the regions that were shown to be the most protected from hydrogen exchange in all previous studies of RNase H. This core fragment of RNase H encodes a well-ordered protein with native-like properties. When isolated from the full-length monomeric protein, the eABCD fragment forms a stable dimer. However, we show indirectly that the monomeric form of eABCD is folded and has an overall secondary structure similar to the dimeric form.

The progressive development of structure and stability during the equilibrium folding of the alpha subunit of tryptophan synthase from Escherichia coli

The urea-induced equilibrium unfolding of the alpha subunit of tryptophan synthase (alphaTS), a single domain alpha/beta barrel protein, displays a stable intermediate at approximately 3.2 M urea when monitored by absorbance and circular dichroism (CD) spectroscopy (Matthews CR, Crisanti MM, 1981, Biochemistry 20:784-792). The same experiment, monitored by one-dimensional proton NMR, shows another cooperative process between 5 and 9 M urea that involves His92 (Saab-Rincón G et al., 1993, Biochemistry 32:13,981-13,990). To further test and quantify the implied four-state model, N <--> I1 <--> I2 <--> U, the urea-induced equilibrium unfolding process was followed by tyrosine fluorescence total intensity, tyrosine fluorescence anisotropy and far-UV CD. All three techniques resolve the four stable states, and the transitions between them when the FL total intensity and CD spectroscopy data were analyzed by the singular value decomposition method. Relative to U, the stabilities of the N, I1, and I2 states are 15.4, 9.4, and 4.9 kcal mol(-1), respectively. I2 partially buries one or more of the seven tyrosines with a noticeable restriction of their motion; it also recovers approximately 6% of the native CD signal. This intermediate, which is known to be stabilized by the hydrophobic effect, appears to reflect the early coalescence of nonpolar side chains without significant organization of the backbone. I1 recovers an additional 43% of the CD signal, further sequesters tyrosine residues in nonpolar environments, and restricts their motion to an extent similar to N. The progressive development of a higher order structure as the denaturant concentration decreases implies a monotonic contraction in the ensemble of conformations that represent the U, I2, I1, and N states of alphaTS.

Identifying the structural boundaries of independent folding domains in the alpha subunit of tryptophan synthase, a beta/alpha barrel protein

Two equilibrium intermediates have previously been observed in the urea denaturation of the alpha subunit of tryptophan synthase (alphaTS) from Escherichia coli, an eight-stranded beta/alpha barrel protein. In the current study, a series of amino-terminal fragments were characterized to probe the elementary folding units that may be in part responsible for this complex behavior. Stop-codon mutagenesis was used to produce eight fragments ranging in size from 105-214 residues and containing incremental elements of secondary structure. Equilibrium studies by circular dichroism indicate that all of these fragments are capable of adopting secondary structure. All except for the shortest fragment fold cooperatively. The addition of the fourth, sixth, and eighth beta-strands leads to distinct increases in structure, cooperativity, and/or stability, suggesting that folding involves the modular assembly of betaalphabeta supersecondary structural elements. One-dimensional NMR titrations at high concentrations of urea, probing the environment around His92, were also performed to test for the presence of residual structure in the fragments. All fragments that contained the first four betaalpha units of structure exhibited a cooperative unfolding transition at high concentrations of urea with significant but reduced stability relative to the full-length protein. These results suggest that the residual structure in alphaTS requires the participation of hydrophobic residues in multiple beta-strands that span the entire sequence.