Identification of an extracellular signal-regulated kinase (ERK) docking site in ribosomal S6 kinase, a sequence critical for activation by ERK in vivo

Glutathione S-transferase (GST)-fusion proteins containing the carboxyl-terminal tails of three p90 ribosomal S6 kinase (RSK) isozymes (RSK1, RSK2, and RSK3) interacted with extracellular signal-regulated kinase (ERK) but not c-Jun-NH2-kinase (JNK) or p38 mitogen-activated protein kinase (MAPK). Within the carboxyl-terminal residues of the RSK isozymes is a region of high conservation corresponding to residues 722LAQRRVRKLPSTTL735 in RSK1. Truncation of the carboxyl-terminal 9 residues, 727VRKLPSTTL735, completely eliminated the interaction of the GST-RSK1 fusion protein with purified recombinant ERK2, whereas the truncation of residues 731PSTTL735 had no effect on the interaction with purified ERK2. ERK1 and ERK2 co-immunoprecipitated with hemagglutinin-tagged wild type RSK2 (HA-RSK2) in BHK cell cytosol. However, ERK did not co-immunoprecipitate with HA-RSK2((1-729)), a mutant missing the carboxyl-terminal 11 amino acids, similar to the minimal truncation that eliminated in vitro interaction of ERK with the GST-RSK1 fusion protein. Kinase activity of HA-RSK2 increased 6-fold in response to insulin. HA-RSK2((1-729)) had a similar basal kinase activity to that of HA-RSK2 but was not affected by insulin treatment. Immunoprecipitated HA-RSK2 and HA-RSK2((1-729)) could be activated to the same extent in vitro by active ERK2, demonstrating that HA-RSK2((1-729)) was properly folded. These data suggest that the conserved region of the RSK isozymes (722LAQRRVRKL730 of RSK1) provides for a specific ERK docking site approximately 150 amino acids carboxyl-terminal to the nearest identified ERK phosphorylation site (Thr573). Complex formation between RSK and ERK is essential for the activation of RSK by ERK in vivo. Comparison of the docking site of RSK with the carboxyl-terminal tails of other MAPK-activated kinases reveals putative docking sites within each of these MAPK-targeted kinases. The number and placement of lysine and arginine residues within the conserved region correlate with specificity for activation by ERK and p38 MAPKs in vivo.     

Liver fibrosis of the turkey on rapeseed products

A method has been developed for selectively preparing the carboxyl-terminal tryptic peptide of proteins by cleavage at arginyl residues. The succinylated protein is digested with trypsin and the peptides produced are maleylated. Maleylated peptides are then submitted to cation-exchange chromatography in urea at low pH and ionic strength. Arginine-containing peptides are retained by the resin. The carboxyl-terminal peptide emerges unretarded and in pure form. This method has been applied to four proteins of known sequence. Yields as high as 88% have been obtained.     

Investigation of substrate activation by 4-chlorobenzoyl-coenzyme A dehalogenase

4-Chlorobenzoyl-coenzyme A (4-CBA-CoA) dehalogenase catalyzes the hydrolysis of 4-CBA-CoA to 4-hydroxybenzoyl-coenzyme A (4-HBA-CoA), using the carboxylate side chain of aspartate 145 to displace the chloride from C(4) of the benzoyl ring. Previous UV-visible, Raman, and 13C NMR studies of enzyme-bound substrate analog or product ligand indicated that the environment of the enzyme active site induces a significant reorganization of the benzoyl ring pi-electrons. This observation was interpreted as evidence for electrophilic catalysis [viz. active-site-induced polarization of electron density away from the ring C(4)] [Taylor, K. L., Liu, R.-Q., Liang, P.-H., Price, J., Dunaway-Mariano, D., Tonge, P. J., Clarkson, J., & Carey, P. R. (1995) Biochemistry 34, 13881]. The recent crystal structure of the dehalogenase-4-HBA-CoA complex reveals two hydrogen bonds contributed to the benzoyl C=O by the backbone amide protons of Gly114 and Phe64 and a possible dipolar interaction with the positive pole of the 114-121 alpha-helix. Residues closely surrounding the benzoyl ring include W137, D145, W89, F64, F82, and H90. In the present study, the mutants D145A, H90Q, W137F, W89F, W89Y, F64L, F82L, and G114A were prepared to examine the effect of amino acid substitution on catalysis and on perturbation of the UV-visible spectral properties of the substrate benzoyl ring. Substitution of the two catalytic residues D145 and H90 inhibited catalysis but not ligand binding or the induction of the red shift in the benzoyl ring absorption. These two residues do not appear to contribute to substrate benzoyl ring binding or polarization. The F64L, F82L, W89F, and W137F mutants retained substantial catalytic activity and the ability to induce the red shift. The W89Y mutant, on the other hand, is inhibited in catalysis and ligand binding, suggesting that hydrophobicity more than packing may be critical for the benzoyl ring binding/activation. The G114A mutant was shown to be strongly inhibited in both substrate binding and activation, indicating that H-bonding and/or interaction with the dipole of the 114-121 alpha-helix may be crucial.     

Structural determinants required for the interaction between Rho GTPase and the GTPase-activating domain of p190

The Rho family small GTP-binding proteins are subjected to regulation by Rho GTPase-activating proteins (GAPs) in the course of transmitting diverse intracellular signals. To understand the mechanism of GAP-catalyzed GTP hydrolysis of Rho GTPases, we have studied the interaction between RhoA and p190, the RasGAP binding phosphoprotein which has been implicated as a Rho-specific GAP, by delineating the structural determinants of RhoA and p190 GAP domain (p190GD) that are involved in their functional coupling. Besides the conserved residues Tyr34, Thr37, and Phe39 in the switch I region of RhoA which are required for p190GD interaction, chimeras made between RhoA and Cdc42, a close relative of RhoA with which p190GD interacts 50-fold less efficiently, revealed that residues outside the switch I and neighboring regions of RhoA, residues 85-122 in particular, contain the major p190GD-specifying determinant(s). Mutation of the unique Asp90 of RhoA in this region mostly abolished p190GD stimulation, whereas the corresponding reverse mutation of Cdc42 (S88D) was able to respond to p190GD-catalysis similarly as RhoA. Further kinetic analysis of these mutants provided evidence that Asp90 of RhoA contributes primarily to the specific binding interaction with p190GD. On the other hand, two charged residues of p190GD, Arg1283 and Lys1321, which are located in the putative G-protein binding helix pocket of GAP domain, were found to be involved in different aspects of interaction with RhoA. The R1283L mutant of p190GD lost GAP activity but retained the ability to bind to RhoA, while K1321A failed to stimulate and to bind to RhoA. These results indicate that residue Asp90 constitutes the second GAP-interactive site in RhoA which is mostly responsible for conferring p190GD-specificity, and suggest that the role of p190GD in the GTPase reaction of RhoA is in part to supply active site residue Arg1283 for efficient catalysis.     

Identification of the binding site on S100B protein for the actin capping protein CapZ

The calcium-binding protein S100B binds to several potential target proteins, but there is no detailed information showing the location of the binding site for any target protein on S100B. We have made backbone assignments of the calcium-bound form of S100B and used chemical-shift changes in spectra of 15N-labeled protein to locate the site that binds a peptide corresponding to residues 265-276 from CapZ alpha, the actin capping protein. The largest chemical-shift changes are observed for resonances arising from residues around the C terminus of the C-terminal helix of S100B and residues Val-8 to Asp-12 of the N-terminal helix. These residues are close to but not identical to residues that have been identified by mutational analysis to be important in other S100 protein-protein interactions. They make up a patch across the S100B dimer interface and include some residues that are quite buried in the structure of calcium-free S100B. We believe we may have identified a binding site that could be common to many S100 protein-protein interactions.     

Functional analysis of the C2A domain of synaptotagmin 1: implications for calcium-regulated secretion

Synaptotagmin 1 is proposed to function as a low affinity calcium sensor for calcium-triggered exocytosis from neural and neuroendocrine cells. Because of the calcium-binding properties of the C2A domain of synaptotagmin 1, calcium-dependent interactions through this domain may modulate neurotransmitter release. We addressed this question by using alanine-scanning mutagenesis to generate a series of mutations within the C2A domain of synaptotagmin 1. The effects of these mutations on synaptotagmin 1 C2A function were analyzed for (1) calcium-dependent phospholipid binding, (2) calcium-dependent binding to syntaxin 1A, a plasma membrane protein critical for vesicle docking or fusion, and (3) calcium-regulated secretion after microinjection into neuroendocrine pheochromocytoma (PC12) cells. Our analyses reveal that a polylysine motif at residues 189-192 confers an inhibitory effect on secretion by recombinant synaptotagmin C2A fragments. The synaptotagmin 1 C2A polylysine motif functions independently of calcium-mediated interactions with phospholipids and syntaxin 1A. Furthermore, alpha-latrotoxin reverses the inhibitory effect of injected recombinant C2A fragments, suggesting that they perturb the cellular calcium-sensing machinery by interfering with synaptotagmin 1 activity in vivo. Our results indicate that novel calcium-independent interactions mediated through the C2A polylysine motif of synaptotagmin 1 function to modulate neurotransmitter release.     

Identification of spike protein residues of murine coronavirus responsible for receptor-binding activity by use of soluble receptor-resistant mutants

We previously demonstrated by site-directed mutagenesis analysis that the amino acid residues at positions 62 and 214 to 216 in the N-terminal region of mouse hepatitis virus (MHV) spike (S) protein are important for receptor-binding activity (H. Suzuki and F. Taguchi, J. Virol. 70:2632-2636, 1996). To further identify the residues responsible for the activity, we isolated the mutant viruses that were not neutralized with the soluble form of MHV receptor proteins, since such mutants were expected to have mutations in amino acids responsible for receptor-binding activity. Five soluble-receptor-resistant (srr) mutants isolated had mutations in a single amino acid at three different positions: one was at position 65 (Leu to His) (srr11) in the S1 subunit and three were at position 1114 (Leu to Phe) (srr3, srr4, and srr7) and one was at position 1163 (Cys to Phe) (srr18) in the S2 subunit. The receptor-binding activity examined by a virus overlay protein blot assay and by a coimmunoprecipitation assay showed that srr11 S protein had extremely reduced binding activity, while the srr7 and srr18 proteins had binding activity similar to that of wild-type cl-2 protein. However, when cell surface receptors were used for the binding assay, all srr mutants showed activity similar to that of the wild type or only slightly reduced activity. These results, together with our previous observations, suggest that amino acids located at positions 62 to 65 of S1, a region conserved among the MHV strains examined, are important for receptor-binding activity. We also discuss the mechanism by which srr mutants with a mutation in S2 showed high resistance to neutralization by a soluble receptor, despite their sufficient level of binding to soluble receptors.     

Molecular cloning of Aralar, a new member of the mitochondrial carrier superfamily that binds calcium and is present in human muscle and brain

We have identified a new calcium-dependent subfamily of mitochondrial carrier proteins with members in Saccharomyces cerevisiae, Caenorhabditis elegans, and various mammalian species. The members of this subfamily have a bipartite structure: a carboxyl-terminal half with the characteristic features of the mitochondrial solute carrier superfamily and an amino-terminal extension harboring various EF-hand domains. A member of this subfamily (that we have termed Aralar) was cloned from a human heart cDNA library. The corresponding cDNA comprises an open reading frame of 2037 base pairs encoding a polypeptide of 678 amino acids. The carboxyl-terminal half of Aralar (amino acids 321-678) has high similarity with the oxoglutarate, citrate, and adenine nucleotide carriers (28-29% identity), whereas the amino-terminal half (amino acids 1-320) contains three canonical EF-hands. Aralar amino-terminal half was shown to bind calcium by 45Ca2+ overlay and calcium-dependent mobility shift assays. The subcellular localization of the protein in COS cells transfected with Aralar was exclusively mitochondrial. Antibodies against Aralar amino-terminal fusion protein recognized a 70-kDa protein in brain mitochondrial fractions. Northern blot analysis showed that the protein was expressed in heart, brain, and skeletal muscle. The domain structure, mitochondrial localization, and presence in excitable tissues suggests a possible function of Aralar as calcium-dependent mitochondrial solute carrier.     

Subcellular localization and topology of the K88 usher FaeD in Escherichia coli

The subcellular localization of the K88 usher FaeD was studied in Escherichia coli whole cells by using isopycnic sucrose density gradient centrifugation of isolated membranes, the detergents Triton X-100 and sodium lauryl sarcosinate and immunoblotting with a specific FaeD antiserum. Cells containing the complete K88 operon, as well as cells containing the subcloned faeD gene in various expression vectors, were used. Most of the FaeD was present in the outer membranes in a detergent-resistant form. Agglutination experiments with E. coli cells expressing FaeD confirmed an outer membrane localization and indicated the presence of FaeD at the cell surface. Automated Edman degradation indicated that the mature FaeD contained 777 amino acid residues and confirmed that FaeD is synthesized with a rather long signal sequence of 35 amino acid residues. Twelve different FaeD-PhoA fusion proteins were prepared and characterized by nucleotide sequencing and immunoblotting. Most of these fusion sites were located in the amino-terminal and carboxyl-terminal regions of FaeD. Six amino-terminal fusion proteins were soluble proteins in the periplasm, whereas the other fusion proteins were associated with the outer membrane. The protease accessibility of FaeD and of the six outer membrane-bound FaeD-PhoA fusion proteins was studied using whole cells, cells with permeabilized outer membranes, and isolated membranes. Collagenase H, kallikrein, trypsin and proteinase K were used. Based on the results of these experiments and computer predictions, a model for the membrane topology of FaeD was developed in which FaeD contains a large central domain containing 24 membrane-spanning segments and two relatively large periplasmic regions, at the amino-terminal and carboxyl-terminal end of the protein, respectively.     

Ser-534 in the hinge 1 region of Arabidopsis nitrate reductase is conditionally required for binding of 14-3-3 proteins and in vitro inhibition

14-3-3 proteins bind to the hinge 1 region of nitrate reductase (NR) and inhibit its activity. To determine which residues of NR are required for 14-3-3-inhibitory interactions, wild-type and mutant forms of Arabidopsis NR were examined in the yeast two-hybrid system and in vitro inhibition assays. NR fragments with or without hinge 1 were introduced into yeast with one of seven Arabidopsis 14-3-3 isoforms (called GF14s). NR fragments (residues 1-562 or 487-562) containing hinge 1 interacted with all GF-14s tested; an NR fragment (residues 1-487) lacking hinge 1 did not. GF14 binding to NR fragments was dependent on Ser-534, since Asp or Ala substitutions at this site blocked the interaction. Revertants with second site substitutions restoring interaction between GF14omega and the Ala- or Asp-substituted NR fragments were identified. One isolate had a Lys to Glu substitution at position 531, which is in hinge 1, and six isolates had Ile to Leu or Phe substitutions at 561 in the heme binding region. Double mutant forms of holo-NR (S534D plus K531E, I561F, or I561L) were constructed and found to be partially inhibited by protein extracts from Arabidopsis containing 14-3-3 proteins. Wild-type NR is phosphorylated and inhibited by these extracts, but S534D single mutant forms are not. These results show that inhibitory NR/14-3-3 interactions are dependent on Ser-534 but only in the context of the wild-type sequence, since substitutions at second sites render 14-3-3 binding and in vitro NR inhibition independent of Ser-534.     

The carboxyl-terminal region of the human papillomavirus type 16 E1 protein determines E2 protein specificity during DNA replication

The mechanism of DNA replication is conserved among papillomaviruses. The virus-encoded E1 and E2 proteins collaborate to target the origin and recruit host DNA replication proteins. Expression vectors of E1 and E2 proteins support homologous and heterologous papillomaviral origin replication in transiently transfected cells. Viral proteins from different genotypes can also collaborate, albeit with different efficiencies, indicating a certain degree of specificity in E1-E2 interactions. We report that, in the assays of our study, the human papillomavirus type 11 (HPV-11) E1 protein functioned with the HPV-16 E2 protein, whereas the HPV-16 E1 protein exhibited no detectable activity with the HPV-11 E2 protein. Taking advantage of this distinction, we used chimeric E1 proteins to delineate the E1 protein domains responsible for this specificity. Hybrids containing HPV-16 E1 amino-terminal residues up to residue 365 efficiently replicated either viral origin in the presence of either E2 protein. The reciprocal hybrids containing amino-terminal HPV-11 sequences exhibited a high activity with HPV-16 E2 but no activity with HPV-11 E2. Reciprocal hybrid proteins with the carboxyl-terminal 44 residues from either E1 had an intermediate property, but both collaborated more efficiently with HPV-16 E2 than with HPV-11 E2. In contrast, chimeras with a junction in the putative ATPase domain showed little or no activity with either E2 protein. We conclude that the E1 protein consists of distinct structural and functional domains, with the carboxyl-terminal 284 residues of the HPV-16 E1 protein being the primary determinant for E2 specificity during replication, and that chimeric exchanges in or bordering the ATPase domain inactivate the protein.     

Calcium and S100B regulation of p53-dependent cell growth arrest and apoptosis

In glial C6 cells constitutively expressing wild-type p53, synthesis of the calcium-binding protein S100B is associated with cell density-dependent inhibition of growth and apoptosis in response to UV irradiation. A functional interaction between S100B and p53 was first demonstrated in p53-negative mouse embryo fibroblasts (MEF cells) by sequential transfection with the S100B and the temperature-sensitive p53Val135 genes. We show that in MEF cells expressing a low level of p53Val135, S100B cooperates with p53Val135 in triggering calcium-dependent cell growth arrest and cell death in response to UV irradiation at the nonpermissive temperature (37.5 degreesC). Calcium-dependent growth arrest of MEF cells expressing S100B correlates with specific nuclear accumulation of the wild-type p53Val135 conformational species. S100B modulation of wild-type p53Val135 nuclear translocation and functions was confirmed with the rat embryo fibroblast (REF) cell line clone 6, which is transformed by oncogenic Ha-ras and overexpression of p53Val135. Ectopic expression of S100B in clone 6 cells restores contact inhibition of growth at 37.5 degreesC, which also correlates with nuclear accumulation of the wild-type p53Val135 conformational species. Moreover, a calcium ionophore mediates a reversible G1 arrest in S100B-expressing REF (S100B-REF) cells at 37.5 degreesC that is phenotypically indistinguishable from p53-mediated G1 arrest at the permissive temperature (32 degreesC). S100B-REF cells proceeding from G1 underwent apoptosis in response to UV irradiation. Our data support a model in which calcium signaling and S100B cooperate with the p53 pathways of cell growth inhibition and apoptosis.     

Identification of N(G)-methylarginine residues in human heterogeneous RNP protein A1: Phe/Gly-Gly-Gly-Arg-Gly-Gly-Gly/Phe is a preferred recognition motif

Three sites of N(G),N(G)-arginine methylation have been located at residues 205, 217, and 224 in the glycine-rich, COOH-terminal one-third of the HeLa A1 heterogeneous ribonucleoprotein. Together with the previously determined dimethylated arginine at position 193 [Williams et al., (1985) Proc. Natl. Acad. Sci. U.S.A. 82, 5666-5670], it is evident that all four sites fall within a span of sequence between residues 190 and 233 that contains multiple Arg-Gly-(Gly) sequences interspersed with phenylalanine residues. These RGG boxes have been postulated to represent an RNA binding motif [Kiledjian and Dreyfuss (1992) EMBO J. 11, 2655-2664]. Dimethylation of HeLa A1 appears to be quantitative at each of the four positions. Arginines 205 and 224 have been methylated in vitro by a nuclear protein arginine methyltransferase using recombinant (unmethylated) A1 as substrate. This suggests A1 may be an in vivo substrate for this enzyme. Examination of sequences surrounding the sites of methylation in A1 along with a compilation from the literature of sites that have been identified in other nuclear RNA binding proteins suggests a methylase-preferred recognition sequence of Phe/Gly-Gly-Gly-Arg-Gly-Gly-Gly/Phe, with the COOH-terminal flanking glycine being obligatory. Taken together with data in the literature, identification of the sites of A1 arginine methylation strongly suggests a role for this modification in modulating the interaction of A1 with nucleic acids.     

The human follitropin alpha-subunit C terminus collaborates with a beta-subunit cystine noose and an alpha-subunit loop to assemble a receptor-binding domain competent for signal transduction

FSH is a member of the pituitary/placental glycoprotein hormone family including luteinizing hormone, thyroid-stimulating hormone, and chorionic gonadotropin. These heterodimeric hormones share a common alpha-subunit and a highly homologous but distinct beta-subunit. The determinant loop of the FSH beta-subunit acts both as a specificity discriminator and as an essential receptor-binding site. The three-dimensional structure of hCG illustrates the proximity of the determinant loop to the carboxyl-terminal residues of the common alpha-subunit. Thus, site-directed mutagenesis was used to mak high-resolution substitutions at this carboxyl-terminal locus. The effects of those substitutions were studied. Twelve single mutations and one composite mutation were made of the region of hFSH alpha 74-92, each residue substituted by alanine. Side chain replacement in this region of FSH proved to be detrimental to binding. hFSH with mutations of either alpha S85A, alpha T86A, alpha K91A, or alpha S92A only retained 10% or less of the hFSH receptor-binding activity, while compared to these, mutants alpha H79A, alpha Y88A, and alpha Y89A retained slightly more binding activity. The single mutant alpha F74A and composite mutant alpha V76A/E77A binding activity was reduced to half of that of wild-type (WT) hFSH. In contrast, mutations of either alpha K75A, alpha T80A, alpha H83A, or alpha H90A did not adversely affect receptor binding, demonstrating the specificity of observed effects. The hFSH and mutant hormones were tested in an in vitro bioassay for stimulation of progesterone production. Those mutations that did not affect receptor binding (alpha K75A, alpha T80A, alpha H83A, and alpha H90A) did not affect signal transduction, measured by progesterone responses. After comparison of wild-type and mutant hFSH activities determined in radioreceptor assays (ID50) and in vitro bioassays (ED50), it became evident that signal transduction correlated with receptor binding.     

Determinants of dietary intake in a sample of white and Mexican-American children

Using a modified method consisting of chromatography on phenyl-Sepharose, Q-Sepharose, and hydroxyapatite, we isolated a highly purified heat shock protein with molecular weight 90 kD (Hsp90) from rabbit liver. The isolated protein was recognized on immunoblot by commercially available monoclonal anti-Hsp90 antibodies. The chromatographic properties, interaction with actin and calmodulin, phosphorylation in the presence of Mg-ATP, and one-dimensional peptide maps of rabbit liver Hsp90 are similar to the corresponding properties of Hsp90 isolated from other sources. In the presence of soluble carbodiimide and N-hydroxysuccinimide, rabbit liver Hsp90 can be cross-linked with calmodulin, troponin C, troponin I, and calponin. The data obtained indicate that Hsp90 may participate in the assembly of regulatory proteins of the actin filament.     

Demonstration of heterodimer formation between S100B and S100A6 in the yeast two-hybrid system and human melanoma

S100B (S100beta) and S100A6 (calcyclin) are two 10-kDa Ca2+- and Zn2+-binding proteins coexpressed in melanoma and cell-cycle regulated. These proteins are members of the S100 subfamily and are thought to exert their function through interaction with intracellular target proteins. In order to search for potential target proteins interacting with S100B, we used a yeast two-hybrid strategy with human S100B as bait to screen a human brain cDNA library. The fusion proteins interacting with the S100B bait were identified as S100B, S100A1, and S100A6. This indicates the potential of S100B to form homodimers and heterodimers with other members of the S100 subfamily. By Northern and Western blotting, S100B and S100A6 were shown to be expressed at high levels in a panel of human melanoma cell lines. S100B and S100A6 were coimmunoprecipitated from melanoma cell lysates in the presence of 100 microM Zn2+. Confocal microscopy demonstrated that both proteins were distributed throughout the cytoplasm and concentrated in the nucleus. The demonstration of an association and colocalization of S100B and S100A6 in melanoma supports the possibility that an S100B/S100A6 heterodimer plays a functional role in these cells.     

Characterization of an autoreduction pathway for the [Fe4S4]3+ cluster of mutant Chromatium vinosum high-potential iron proteins. Site-directed mutagenesis studies to probe the role of phenylalanine 66 in defining the stability of the [Fe4S4] center provide evidence for oxidative degradation via a [Fe3S4] cluster

A number of point mutations of the conserved aromatic residue phenylalanine 66 (Phe66Tyr, -Asn, -Cys, -Ser) in Chromatium vinosum high-potential iron sulfur protein have been examined with the aim of understanding the functional role of this residue. Nonconservative replacements with polar residues have a minimal effect on the midpoint potential of the [Fe4S4]3+/2+ cluster, typically < +25 mV, with a maximum change of +40 mV for Phe66Asn. With the exception of the Phe66Tyr mutant, the oxidized state was found to be unstable relative to the recombinant native, with regeneration of the reduced state. The pathway for this transformation involves degradation of the cluster in a fraction of the sample, which provides the reducing equivalents required to bring about reduction of the remainder of the sample. This degradative reaction proceeds through a transient [Fe3S4]+ intermediate that is characterized by typical g values and power saturation behavior and is prompted by the increased solvent accessibility of the cluster core in the nonconservative Phe66 mutants as evidenced by 1H-15N HMQC NMR experiments. These results are consistent with a model where the critical role of the aromatic residues in the high-potential iron proteins is to protect the cluster from hydrolytic degradation in the oxidized state.