Regular article
Distinct cysteine sulfhydryl environments detected by analysis of Raman S-H markers of Cys→Ser mutant proteins1

https://doi.org/10.1006/jmbi.2001.4476Get rights and content

Abstract

Very little is known about the character or functional relevance of hydrogen-bonded cysteine sulfhydryl (S-H) groups in proteins. The Raman S-H band is a unique and sensitive probe of the local S-H environment. Here, we report the use of Raman spectroscopy combined with site-specific mutagenesis to document the existence of five distinguishable hydrogen-bonded states of buried cysteine sulfhydryl groups in a native protein. The 666 residue subunit of the Salmonella typhimurium bacteriophage P22 tailspike contains eight cysteine residues distributed through the elongated structure. The tailspike cysteine residues display an unusual Raman S-H band complex (2500-2600 cm−1 interval) indicative of diverse S-H hydrogen-bonding interactions in the native trimeric structure. To resolve specific Cys contributions to the complex Raman band we characterized a set of tailspike proteins with each cysteine replaced by a serine. The mutant proteins, once folded, were structurally and functionally indistinguishable from wild-type tailspikes, except for their Raman S-H signatures. Comparison of the Raman spectra of the mutant and wild-type proteins reveals the following hydrogen-bond classes for cysteine sulfhydryl groups. (i) Cys613 forms the strongest S-H…X bond of the tailspike, stronger than any heretofore observed for a protein. (ii) Cys267, Cys287 and Cys458 form robust S-H…X bonds. (iii) Moderate S-H…X bonding occurs for Cys169 and Cys635. (iv) Cys290 and Cys496 form weak hydrogen bonds. (v) It is remarkable that Cys287 contributes two Raman S-H markers, indicating the population of two distinct hydrogen-bonding states. The sum of the S-H Raman signatures of all eight mutants accurately reproduces the composite Raman band of the wild-type tailspike. The diverse cysteine states may be an outcome of the folding and assembly pathway of the tailspike, which though lacking disulfide bonds in the native state, utilizes transient disulfide bonds in the maturation pathway. This Raman study represents the first detailed assessment of local S-H hydrogen bonding in a native protein and provides information not obtainable directly by other structural probes. The method employed here should be applicable to a wide range of cysteine-containing proteins.

Introduction

Cysteine thiol or sulfhydryl (S-H) groups are the most chemically reactive sites in proteins at physiological conditions. 1 The oxidation of two cysteine S-H bonds to form the cystine disulfide linkage (S-S) is a key step in the regulation of oxidoreductase chemistry and is important in the stabilization of many protein native states. 2 The cysteine thiolate ion (S) may function as a potent nucleophile in active-site chemistry of thiol proteases, as well as in reactions with exogenous modifying agents. 3 The cysteine sulfhydryl may function as either a hydrogen bond donor (e.g. S-H…OH2) or acceptor (e.g. HOH…S-H) group. 4 Alternatively, the sulfhydryl may release its proton to facilitate sulfur ligation at metal coordination sites in proteins. 5, 6 Sulfhydryls are also vulnerable to aberrant oxidations, which can compromise protein function or stability. 7, 8

Sulfhydryl hydrogen bonding in proteins is not well understood, primarily because such interactions are difficult to detect experimentally. Owing to the low electron density of hydrogen, S-H…X interactions (where X indicates any suitable hydrogen bond acceptor group) cannot be detected in protein crystal structures. Therefore, the presence or absence of an S-H…X hydrogen bond is usually inferred from the interatomic distance between the atoms S and X. The validity of such an inference is difficult to assess, particularly in view of the greater tendency of S than O (or N) to participate in non-linear hydrogen bonds. Other factors contribute to the paucity of information on protein S-H…X interactions. For example, the feasibility of hydrogen bonding in a protein structure is usually deduced after optimizing the geometries of all covalent bonds except those involving hydrogen; also, electrostatic interactions are typically ignored in the refinement process. Even in cases where hydrogen bond structural constraints are considered, those involving a sulfur center are weighted less heavily than those involving oxygen or nitrogen. Structures refined in such a manner will tend to underestimate the extent of cysteine S-H hydrogen bonding. Although neutron crystallography is, in principle, more effective than X-ray crystallography in locating hydrogen atoms, the neutron technique is little used for proteins because of the requirements of extraordinarily large crystals and time-consuming data collection protocols.

Baburina and co-workers have demonstrated the utility of Fourier transform infrared (FTIR) spectroscopy in a systematic analysis of yeast pyruvate decarboxylase Cys→Ser mutants. They were able to use this technique to determine the specific ionization state of each of the four cysteine residues of the protein. 9

Raman spectroscopy provides a convenient alternative for investigating cysteine sulfhydryl sites in proteins. 10 The Raman spectrum monitors the S-H group directly and is applicable to proteins in both crystalline and solution states. Previous studies of thiol model compounds demonstrate that the sulfhydryl Raman marker (S-H bond stretching vibration) is a highly sensitive probe of local S-H interaction. 11, 12, 13 The Raman band diagnostic of the S-H environment occurs within a spectral range (2500–2600 cm−1) that is well removed from all other Raman bands of a protein (300–1800 and 2800–3600 cm−1 intervals). Spectral interferences from solvent water molecules and from possible overtone or combination bands of a protein are also much less problematic in Raman than in FTIR spectroscopy. Because no other chemical groups contribute to the sulfhydryl region of the Raman spectrum, the method offers an interference-free probe of cysteinyl structure and local environment. 10, 13 In previous applications, the active site of Escherichia coli thioredoxin, 14, 15 the dynamics of viral capsid assembly 16 and the redox chemistry of γII crystallin 17 have been investigated. Here, we exploit Raman spectroscopy to identify and characterize an unexpected range of cysteine S-H environments in the trimeric P22 tailspike protein.

The 666 amino acid subunit of the P22 tailspike contains eight cysteine residues, Cys169, Cys267, Cys287, Cys290, Cys458, Cys496, Cys613 and Cys635, 18 all of which are reduced in the native protein. 19 The X-ray crystal structure shows that the six Cys residues proximal to the N terminus are located in the so-called “parallel β-helix” domain, which is an extended coil-like conformation stabilized by parallel β-strand secondary structure (Figure 1). 20, 21 All sulfhydryl groups of the tailspike are unreactive toward cysteine modifying agents in solution 22 and all are inaccessible to water molecules in the crystal structure. 21 Near the C-terminal end of the protein, the three polypeptide chains wrap around each other to form a unique prism of interdigitated β-strands. 21 On the C-terminal side of the interdigitated region each subunit forms a five-stranded antiparallel β-sheet, which combine to form a triangular β-sheet prism, as shown in Figure 1.23, 24 Cysteine residues 613 and 635, from each chain, form a distinctive ring of six cysteine residues in this C-terminal β-sheet prism (Figure 1(a)).

The tailspike folding and subunit assembly pathway is guided by unusual cysteine chemistry: intersubunit disulfide bonds involving a few of the cysteine residues are formed transiently in a trimeric folding intermediate, both in vivo and in vitro. 22, 25 There is evidence to suggest, but not conclusively demonstrate, that Cys496, Cys613 and Cys635 (Figure 1) are involved in the transient disulfide bond formation. 25, 26, 28 The P22 tailspike provides the first and, as yet, only example of a protein that utilizes a cysteine-disulfide linked folding intermediate en route to a completely reduced and highly stable native state. 25, 27

It is reasonable to assume that cysteine residues that play different roles in the folding process may occupy distinct final environments in the native protein. Here we report that the S-H region of the tailspike Raman spectrum is unusually complex, suggesting several distinct sulfhydryl environments. To determine the specific contribution of each cysteine sulfhydryl to the native-state Raman signature and to deduce the corresponding S-H hydrogen-bonding environments, we have constructed eight single-site Cys→Ser mutants and systematically analyzed their respective Raman S-H signatures. The Raman analysis reveals the contributions of individual S-H…X hydrogen bonds to the stability of the native state. The results provide information about S-H interactions that currently cannot be obtained by any other structural technique.

Section snippets

Raman signature of the recombinant tailspike protein

The Escherichia coli-expressed recombinant tailspike of phage P22 has been fully characterized with respect to composition, morphology and biological activity. 26 No differences were apparent between the recombinant tailspike and the tailspike isolated by phage infection of Salmonella. The Raman spectrum of the recombinant protein, shown in Figure 2, is indistinguishable from the Raman signature reported previously for the tailspike isolated from P22-infected Salmonella. 19, 28 The spectral

Raman signature of the native tailspike

The Raman spectrum of the tailspike is atypical. The sulfhydryl region (2500-2600 cm−1) is distinguished by a rich pattern of S-H stretching bands (Figure 2, inset), indicating that cysteine sulfhydryl groups populate a variety of local environments. 11, 12 Conversely, Raman amide I (1667 cm−1) and amide III (1238 cm−1) markers of the native tailspike (Figure 2) are remarkably sharp, diagnostic of the prevalent parallel β-strand secondary structure that characterizes the robust parallel β-helix

Tailspike protein and reagents

P22-derived tailspike protein was isolated from phage infection of Salmonella typhimurium, using a previously published method. 46 Recombinant tailspike proteins, including wild-type and the eight Cys→Ser mutants, were overexpressed in E. coli and purified as described by Haase-Pettingell et al. 26 SDS, EDTA, Tris, mono and dibasic phosphate salts, l-cysteine and deuterium oxide (99.9 % 2H2O) were purchased from Sigma Chemical Co. (St. Louis, MO). Centricon protein concentrators were obtained

Acknowledgements

Support of this research by NIH grants GM50776 (to G.J.T.) and GM17980 (to J.K.) and by Amgen, Inc. is gratefully acknowledged. S.W.R. and P.L.C. thank Drs James M. Benevides (UMKC) and Stacy A. Overman (UMKC) for their assistance with data collection and analyses and recommendations to improve the text. We also thank Dr Marilyn D. Yoder (UMKC) for reviewing the manuscript prior to submission and carrying out calculations with HBPLUS (MacDonald & Thornton, 1994).

References (47)

  • J.M Sturtevant et al.

    Thermostability of temperature-sensitive folding mutants of the P22 tailspike protein

    J. Biol. Chem.

    (1989)
  • W Parker et al.

    Protein structures in SDS micelle-protein complexes

    Biophys. J.

    (1992)
  • C Haase-Pettingell et al.

    Prevalence of temperature sensitive folding mutations in the parallel β-coil domain of the phage P22 tailspike endorhamnosidase

    J. Mol. Biol.

    (1997)
  • J Jenkins et al.

    Structure and evolution of parallel β-helix proteins

    J. Struct. Biol.

    (1998)
  • T.N Petersen et al.

    The crystal structure of rhamnogalacturonase A from Aspergillus aculeatusa right-handed parallel β-helix

    Structure

    (1997)
  • M.D Yoder et al.

    Unusual structural features in the parallel β-helix in pectate lyases

    Structure

    (1993)
  • W Huang et al.

    Crystal structure of chondroitinase B from Flavobacterium heparinum and its complex with a disaccharide product at 1.7 Å resolution

    J. Mol. Biol.

    (1999)
  • J Pande et al.

    The calf γ-crystallinsa Raman spectroscopic study

    Exp. Eye Res.

    (1991)
  • M.-H Yu et al.

    Surface amino acids as sites of temperature-sensitive folding mutations in the P22 tailspike proteins

    J. Biol. Chem.

    (1988)
  • T.E Creighton
  • N Darby et al.

    Disulfide bonds in protein folding and stability

    Methods Mol. Biol.

    (1995)
  • E Shaw

    Cysteinyl proteinases and their selective inactivation

    Advan. Enzymol. Relat. Areas Mol. Biol.

    (1990)
  • J.O Konz et al.

    Effects of oxygen on recombinant protein expression

    Biotechnol. Prog.

    (1998)
  • Cited by (84)

    • Design and synthesis of a novel “turn-on” long range measuring fluorescent probe for monitoring endogenous cysteine in living cells and Caenorhabditis elegans

      2021, Analytica Chimica Acta
      Citation Excerpt :

      Until now, a variety of selective and sensitive assays for the measurement of Cys have been reported. Among them, there were many effective methods such as high performance liquid chromatography (HPLC) [17], mass spectrometry [18–20], photoelectrochemical sensor [21], microelectrode biosensors [22], and Raman spectroscopy [23]. As a springing up technique, the fluorescent probes are attaching increasing attention due to several advantages, such as high sensitivity, rapid response, and good bio-compatibility, thus the researchers are exploiting various fluorescent probes to detect Cys in physiological systems.

    • Changes of myofibrillar protein structure improved the stability and distribution of baicalein in emulsion

      2021, LWT
      Citation Excerpt :

      The hydrophobic groups exposed by protein increased the hydrophobic interaction between BN and MP (Liu, Ma, Gao, & McClements, 2017). Additionally, the change of sulfhydryl groups increased the number of hydrogen bond donors and receptors, thus contributing to the stable attachment of BN in MP (Raso, Clark, Haase-Pettingell, King, & Thomas, 2001; Wu et al., 2019). Notably, the change of ζ-potential indicated that MP was in a more stable dispersion state, which enhanced the interaction between BN and MP (Fig. 3).

    • Formation mechanism of BAMLET by 2D Raman correlation analysis

      2018, Journal of Molecular Structure
      Citation Excerpt :

      The results agree well with those of our previous results on the pH-induced transition from the native to the acid state of ALA, which demonstrated that a Ca2+ ion was released below pH 3.5 [24]. Fig. 2(e) and (f) show that bands for Trp and the α-helix structure changed in intensity mostly above pH 4.0, while below pH 3.0, the bands for Phe, Tyr, the β-sheet, and protonated/deprotonated carboxylic acid, which appeared at 1591, 1601, 1611, 1620, 1670, 1706, and 1724 cm−1 [25,28–31], changed greatly. These components are associated with formed the ALA/OA complex.

    View all citing articles on Scopus
    1

    Edited by P. E. Wright

    View full text