Signal-sequence Trap in Mammalian and Yeast Cells: A Comparison

Membrane associated and secreted proteins are translated as precursors containing a signal peptide that allows protein-insertion into the membrane of the endoplasmic reticulum and is co-translationally removed in the lumen. The ability of the signal peptide to direct a polypeptide into the secretory pathway is exploited in methods developed to select cDNAs encoding such proteins. Different strategies are known in which cDNA libraries can be screened for signal peptides by the ability of the latter to rescue the translocation of signal sequence-less proteins. In one method, a cDNA library is tested for interleukin 2 receptor α chain translocation to the membrane in COS cells, in another one for invertase secretion from yeast. In this work, we compared the two systems by testing six mouse signal peptides in COS and yeast cells. All of them were functional in the mammalian system, whereas only three of them in yeast. Two other sequences needed the 5′ cDNA sequence flanking the ATG codon to be removed in order to enable protein translocation. Although the structure of signal sequences and the functioning of the secretory machinery are well conserved from prokaryotes to eukaryotes, it seems evident that not all signal peptides can be interchanged between different proteins and organisms. In particular, signal peptides that are functional in the mammalian system do not necessarily lead to protein translocation in yeast. Received: 9 March 2001     

Definition of Epitopes for Monoclonal Antibodies Developed Against Purified Sodium Channel Protein: Implications for Channel Structure

To test sodium channel structural models, we defined the epitopes for nineteen independently cloned monoclonal antibodies previously generated against purified, detergent-solubilized, adult rat skeletal muscle sodium channel protein using channel proteolysis, synthetic peptides, and fusion proteins. All identified epitopes were continuous and unique to the skeletal muscle subtype α-subunit. Of the nineteen independent clones, seventeen had epitopes located either in the origin of the amino-terminus or in the interdomain 2–3 region while only two antibodies had epitopes located in the mid-portion of the interdomain 1–2 region. No immunogenic regions were identified on the α-subunit's extracellular regions, interdomain 3–4 segment, or carboxyl-terminus or on channel β-subunits. While immune tolerance may explain the lack of immunogenicity of extracellular regions, the lack of immunogenicity of most of the channel's cytoplasmic mass may be due to segment inaccessibility from organization of these regions as globular domains, to insertion of parts of these regions into the membrane phase, or to interaction with other protein elements. The definition of monoclonal antibody epitopes allows us to reinterpret previously reported monoclonal antibody competition studies, providing independent support for our model of sodium channel cytoplasmic domain structure. In addition, these data suggest additional testable hypotheses concerning the interactions of the sodium channel amino- and carboxyl-termini with each other as well as with other protein elements. Received: 4 March 1998/Revised: 15 May 1998     

Renal Cortical Basolateral Na+/HCO− 3 Cotransporter: IV Characterization and Localization with Polyclonal Antibodies

We have previously partially purified the basolateral Na⁺/HCO⁻ ₃ cotransporter from rabbit renal cortex and this resulted in a 400-fold purification, and an SDS-PAGE analysis showed an enhancement of a protein band with a MW of approximately 56 kDa. We developed polyclonal antibodies against the Na⁺/HCO⁻ ₃ cotransporter by immunizing Dutch-belted rabbits with a partially purified protein fraction enriched in cotransporter activity. Western blot analysis of renal cortical basolateral membranes and of solubilized basolateral membrane proteins showed that the antibodies recognized a protein with a MW of approximately 56 kDa. The specificity of the purified antibodies against the Na⁺/HCO⁻ ₃ cotransporter was tested by immunoprecipitation. Solubilized basolateral membrane proteins enriched in Na⁺/HCO⁻ ₃ cotransporter activity were incubated with the purified antibody or with the preimmune IgG and then reconstituted in proteoliposomes. The purified antibody fraction caused a concentration-dependent inhibition of the Na⁺/HCO⁻ ₃ cotransporter activity, while the preimmune IgG failed to elicit any change. The inhibitory effect of the antibody was of the same magnitude whether it was added prior to (inside) or after (outside) reconstitution in proteoliposomes. In the presence of the substrates (NaHCO₃ or Na₂CO₃) for the cotransporter, the inhibitory effect of the antibody on cotransporter activity was significantly blunted as compared with the inhibition observed in the absence of substrates. Western blot analysis of rabbit kidneys showed that the antibodies recognized strongly a 56 kDa protein band in microsomes of the inner stripe of outer medulla and inner medulla, but not in the outer stripe of outer medulla. A 56 kDa protein band was recognized in microsomes of the stomach, liver, esophagus, and small intestine but was not detected in red blood cell membranes. Localization of the Na⁺/HCO⁻ ₃ cotransporter protein by immunogold technique revealed specific labeling of the cotransporter on the basolateral membranes of the proximal tubules, but not in the brush border membranes. These results demonstrate that the polyclonal antibodies against the 56 kDa basolateral protein inhibit the activity of the Na⁺/HCO⁻ ₃ cotransporter suggesting that the 56 kDa protein represents the cotransporter or a component thereof. These antibodies interact at or near the substrate binding sites. The Na⁺/HCO cotransporter protein is expressed in different regions of the kidneys and in other tissues. Received: 27 January 1996/Revised: 23 July 1996     

Effect of Antipeptide Antibodies Directed against Three Domains of Connexin43 on the Gap Junctional Permeability of Cultured Heart Cells

Cell-to-cell communication can be blocked by intracellular injections of antibodies raised against gap junction proteins, but the mechanism of channel obstruction is unknown. Binding to connexins could lead to a conformational change, interfere with regulatory domains or cause a steric hindrance. To address these questions, the effects on cell-to-cell communication of affinity purified polyclonal antibodies raised against peptides reproducing the intracellular sequences 5–17, 314–322 and 363–382 of rat connexin43 were investigated in cultured rat ventricular cells. The antibodies against sequence 363–382 were characterized by immunoblotting and immunocytochemistry. Characterization of antibodies 5–17 and 314–322 has been previously reported. In a first series of experiments, the effect on gap junctional communication was assessed by injecting a junction-permeant fluorescent dye into cells adjacent to one cell previously microinjected with antibodies. In a second series, junctional permeability was quantitatively determined on records of fluorescence recovery after the photobleaching of 6-carboxyfluorescein-loaded cells. Antibodies 5–17 marked a 43 kDa band on immunoblots, but did not immunolabel gap junctions and had no functional effect. Antibodies 314–322 recognized the 43 kDa protein and labeled the intercalated disks, but failed to interfere with junctional permeability. Antibodies to the nearby sequence 363–382, for which all immunospecific tests had been positive, caused a delayed diffusional uncoupling in 50% of the microinjected cells. It is suggested that the blocking of junctional communication by antibodies results from interference with a regulatory domain of the connexin. Received: 25 July 1995/Revised: 21 December 1995     

Identification of Isoforms of G Proteins and PKC that Colocalize with Tight Junctions

Recent evidence suggests that the formation and permeability of tight junctions are actively regulated by second-messenger-generating systems involving G proteins and protein kinase C (PKC). A possible specific target for these regulatory proteins is the tight junction protein ZO-1. An extensive immunocytochemical study was performed in cultured epithelial monolayers of MDCK and Caco-2 cells to identify which isoforms of G proteins and PKC are present at or near the zonula occludens complex. Antibodies against α-subunits of each one of the four major subfamilies were used for the localization of the G proteins. For the PKC localization, antibodies against eight different isoforms were used. In confluent monolayers, Gα₁₂ and PKC ζ, were the only isoforms of these proteins present at the cell borders. In subconfluent monolayers, Gα₁₂ and PKC ζ were found at the plasma membrane only along the areas of lateral cell-cell contact. These isoforms formed a pattern of distribution very similar to the ZO-1 protein. The present findings indicate that Gα₁₂ and PKC ζ may be part of the zonula occludens complex and may locally regulate formation and permeability of tight junctions. Received: 29 July 1995/Revised: 13 October 1995     

Cobra Venom Cytotoxin Free of Phospholipase A2 and Its Effect on Model Membranes and T Leukemia Cells

Membrane-active toxins from snake venom have been used previously to study protein-lipid interactions and to probe the physical and biochemical states of biomembranes. To extend these studies, we have isolated from Naja naja kaowthia (cobra) venom a cytotoxin free of detectable phospholipase A₂ (PLA₂). The amino acid composition, pI (10.2), and net charge of the cytotoxin compares well with membrane-active toxins isolated from venoms of other cobras. The cytotoxin, shown by a spin label method, associates with PLA₂ in buffers at pH values between 7.0 and 5.0, but not at pH 4.0. It is suggested that cytotoxin and PLA₂ (pI close to 4.8) associate electrostatically in the native venom. The effect of the cytotoxin on model phospholipid membranes was studied by EPR of spin probes in oriented lipid multilayers and ¹H-NMR of sonicated liposomes. The cytotoxin did not significantly affect the packing of lipids in pure phosphatidylcholine (PC) membranes and in PC membranes containing 10 mol% phosphatidic acid (PA) or cardiolipin (CL). However, the cytotoxin induced an increase in membrane permeability and formation of nonbilayer structures in PC membranes containing 40 mol% of PA or CL. The purified cytotoxin was cytocidal to Jurkat cells, but had little effect on normal human lymphocytes. However, both Jurkat cells and normal lymphocytes were killed equivalently when treated with 10⁻⁹ m PLA₂ and 10⁻⁵ m cytotoxin in combination. From its effect on model membranes and Jurkat cells, it is suggested that purified cytotoxin preferentially targets and disrupts membranes that are rich in acidic phospholipids on the extracellular side of the plasma membrane. Received: 20 March 1996/Revised: 25 September 1996     

Formation of Flat Lamellar Intramembrane Lipid Structures in Microorganisms

Analysis of freeze-fracture replicas and thin sections of cells of the bacteria Sulfobacillus thermosulfidooxidans and Anaerobacter polyendosporus showed that their cytoplasmic membranes contain some regions in the form of flat lamellar inverted lipid membranes a few tenths of nanometers to a few microns in size. The specific features of these membrane structures are as follows: (i) they contain no familiar intramembrane particles commonly present on freeze-fracture replicas; (ii) in cross thin sections, intramembrane structures are bifurcate on the periphery and look like thylakoids; and (iii) the leaflets of intramembrane structures in S. thermosulfidooxidans cells are corrugated. These structures were revealed in bacterial cells cultivated under normal growth conditions. The data obtained suggest the occurrence of a complex type of compartmentalization in biological membranes. Received: 17 July 2000/Revised: 22 November 2000     

Fusion of Human Sperm to Prostasomes at Acidic pH

Prostasomes are membranous vesicles (150–200 nm diameter) present in human semen. They are secreted by the prostate and contain large amounts of cholesterol, sphingomyelin and Ca²⁺. In addition, some of their proteins are enzymes. Prostasomes enhance the motility of ejaculated spermatozoa and are involved in a number of additional biological functions. The possibility that they may fuse to sperm has never been proved. In this work, we studied the fusion of sperm to prostasomes by using various methods (relief of octadecyl Rhodamine B fluorescence self-quenching, fluorescence microscopy and flow cytometry) and we found that it occurs at acidic pH (4–5), but not at pH 7.5 pH-dependent fusion relies on the integrity of one or more proteins and is different from the Ca²⁺-stimulated fusion between rat liver liposomes and spermatozoa that does not require any protein and occurs at neutral pH. We think that the H⁺-dependent fusion of prostasomes to sperm may have physiological importance by modifying the lipid and protein pattern of sperm membranes. Received: 19 June 1996/Revised: 4 September 1996     

Interactions of Primary Amphipathic Vector Peptides with Membranes. Conformational Consequences and Influence on Cellular Localization

The conformations of two peptides produced by the combinations of a nuclear localization sequence and a sequence issued from the fusion protein gp41 of HIV 1 have been analyzed both in solution and in membranes or in membrane mimicking environments. Both are shown to be nonordered in water, α-helical when incorporated into SDS micelles where the helical domain concerns the hydrophobic part of the peptides. Interactions with lipids induce the formation of β-sheet and the lipid-peptide interactions are governed by the nature of the lipid polar headgroups. A monolayer study shows that replacement of the sequence separating the two sequences with an arginine favors the lipid-peptide interactions which may contribute to the understanding of the different, nuclear and membrane associated, cellular localizations of the peptides. Received: 10 October 1997/Revised: 15 January 1998     

Isolation and Characterization of Five Actin cDNAs from the Cestode Diphyllobothrium dendriticum: A Phylogenetic Study of the Multigene Family

Five cDNAs (pDidact2–pDidact6), representing different actin genes, were isolated from a Diphyllobothrium dendriticum cDNA library, and the DNA as well as the putative amino acid sequences were determined. The corresponding Didact2 and Didact4 genes code for peptides 376 amino acids long, with molecular weights 41,772 and 41,744 Da, respectively, while the deduced Didact3 protein is 377 amino acids long and weighs 41,912 Da. The pDidact5 and -6 cDNAs lack nucleotides corresponding to three to six amino acids at the amino-terminus. Two of the five cDNAs contain the conventional AATAAA as the putative polyadenylation signal, one has the common variant ATTAAA, whereas the hexanucleotide AATAGA is found 15 and 18 nucleotides, respectively, upstream of the poly(A) site in two of the cDNAs. Phylogenetic studies including 102 actin protein sequences revealed that there are at least four different types of cestode actins. In this study three of these types were found to be expressed in the adult D. dendriticum tapeworm. Structurally the cestode actin groupings differ from each other to an extent seen only among the metazoan actins between the vertebrate muscle and cytoplasmic isoforms. In the phylogenetic trees constructed, cestode actins were seen to map to two different regions, one on the border of the metazoan actins and the other within this group. It is, however, difficult to say whether the cestode actins branched off early in the metazoan evolution or if this position in the phylogenetic tree only reflects upon differences in evolutionary rate. Received: 19 June 1996 / Accepted: 20 August 1996     

Heart and T-Lymphocyte Cell Surfaces Both Exhibit Positive Cooperativity in Binding a Membrane-Lytic Toxin

Cobra venom cytotoxins (CTX) have been shown to disrupt cells as different as immunocytes, skeletal myocytes, erythrocytes and tumor cells. Nevertheless, even subpopulations of tumor cells are differentially susceptible to CTX by an order of magnitude. In the present study, our objective was to compare CTX-specific binding with cytolytic potency for two disparate cell types in vitro. We investigated the lytic activity of cytotoxin-III from Naja naja atra (NNA, fraction D) using heart cells and human leukemic T-cells (CEM cells). For both cell types, 50% cytolysis, assessed by tetrazolium dye conversion, occurred with μm concentrations of toxin (EC₅₀= 2.2 μm). We examined the binding of radiolabeled CTX III to both heart cells and CEM cells and found the apparent dissociation constant (K _Dapp) to be 0.69 μm and 0.75 μm, for CEM and heart cells respectively. The B _max for the CEM cells was 1.0 fmoles/cell and that for heart cells was 5.2 fmoles/cell, both exhibiting positive cooperativity between the sites (Hill coefficients 1.4, T-cells; 1.6, heart). Relatively modest dissociation constants plus high numbers of binding sites per cell are consistent with a model of CTX binding to plasma membranes by interaction with phospholipids in the bilayer. Our results suggest that the lytic activity of this cytotoxin follows its binding to a population of sites on the cells in a cooperative fashion. Received: 8 May 1995/Revised: 17 November 1995     

Cardiac Ryanodine Receptor Activity is Altered by Oxidizing Reagents in Either the Luminal or Cytoplasmic Solution

The location of reactive cysteine residues on the ryanodine receptor (RyR) calcium release channel was assessed from the changes in channel activity when oxidizing or reducing reagents were added to the luminal or cytoplasmic solution. Single sheep cardiac RyRs were incorporated into lipid bilayers with 10⁻⁷ m cytoplasmic Ca²⁺. The thiol specific-lipophilic-4,4′-dithiodipyridine (4,4′-DTDP, 1 mm), as well as the hydrophilic thimerosal (1 mm), activated and then inhibited RyRs from either the cis (cytoplasmic) or trans (luminal) solutions. Activation was associated with an increase in the (a) mean channel open time and (b) number of exponential components in the open time distribution from one (∼2 msec) to three (∼1 msec; ∼7 msec; ∼15 msec) in channels activated by trans 4,4′-DTDP or cis or trans thimerosal. A longer component (∼75 msec) appeared with cis 4,4′-DTDP. Activation by either oxidant was reversed by the thiol reducing agent, dithiothreitol. The results suggest that three classes of cysteines are available to 4,4′-DTDP or thimerosal, SHa or SHa* activating the channel and SHi closing the channel. SHa is either distributed over luminal and cytoplasmic RyR domains, or is located within the channel pore. SHi is also located within the transmembrane domain. SHa* is located on the cytoplasmic domain of the protein. Received: 17 March 1998/Revised: 26 October 1998     

H+-ATPase-Mediated Cytoplasmic pH-Responses Associated with Elevation of Cytoplasmic Calcium in Cultured Rabbit Nonpigmented Ciliary Epithelium

Studies were conducted to test whether an increase of cytoplasmic calcium concentration influences H⁺-ATPase activity in cultured rabbit nonpigmented ciliary epithelium (NPE). Cytoplasmic calcium concentration or cytoplasmic pH was measured by a fluorescence ratio technique in cells loaded with either Fura-2 or BCECF. Cytoplasmic calcium was increased in three ways; by exposure to BAY K 8644 (1 μm), by exposure to a mixture of epinephrine (1 μm) + acetylcholine (10 μm) or by depolarization with potassium-rich solution. In each case cytoplasmic pH increased significantly. In all three cases 100 nm bafilomycin A₁, a specific H⁺-ATPase inhibitor, significantly inhibited the pH increase. These results suggest an increase of cytoplasmic calcium might initiate events that lead to activation of proton export from the cytoplasm by a mechanism involving H⁺-ATPase. This notion is supported by the observation that the pH increase was suppressed when either verapamil or nifedipine was used to prevent the cytoplasmic calcium increase in cells exposed to potassium-rich solution. Protein kinase C activation might also be involved in the mechanism of H⁺-ATPase stimulation since staurosporine suppressed the pH response to potassium-rich solution. A transient rise of cytoplasmic calcium concentration was observed when cytoplasmic acidification was induced by exposure to high pCO₂. This suggests a rise of cytoplasmic calcium might represent part of a physiological mechanism to stimulate H⁺-ATPase-mediated protein export under acid conditions. Received: 11 August 2000/Revised: 29 March 2001     

Gene structure of a chlorophyll a/c-binding protein from a brown alga: Presence of an intron and phylogenetic implications

A Laminaria saccharina genomic library in the phage EMBL 4 was used to isolate and sequence a full-length gene encoding a fucoxanthin-chlorophyll a/c-binding protein. Contrary to diatom homologues, the coding sequence is interrupted by an intron of about 900 bp which is located in the middle of the transit peptide. The deduced amino acid sequence of the mature protein is very similar to those of related proteins from Macrocystis pyrifera (Laminariales) and, to a lesser extent, to those from diatoms and Chrysophyceae. Seven of the eight putative chlorophyll-binding amino acids determined in green plants are also present. Alignments of different sequences related to the light-harvesting proteins (LHC) demonstrate a structural similarity among the three transmembrane helices and suggest a unique ancestral helix preceded by two β-turns. The β-turns are conserved in front of the second helices of the chlorophyll a/c proteins more so than in chlorophyll a/b proteins. Phylogenetic trees generated from sequence data indicate that fucoxanthin-chlorophyll-binding proteins diverged prior to the separation of photosystem I and photosystem II LHC genes of green plants. Among the fucoxanthin-containing algae, LHC I or II families could not be distinguished at this time. Received: 14 February 1996 / Accepted: 4 April 1996     

Freeze-Fracture Analysis of Plasma Membranes of CHO Cells Stably Expressing Aquaporins 1-5

Several studies suggest that aquaporin water channels can be identified in membranes by freeze-fracture electron microscopy. For this report, Chinese Hamster ovary cells were stably transfected with cDNAs encoding aquaporins 1–5. Measurement of the osmotic water permeability of the cells confirmed that functional protein was expressed and delivered to the plasma membrane. By freeze-fracture electron microscopy, a 20% increase in intramembrane particle (IMP) density was found in plasma membranes of cells expressing AQP2, 3 and 5, and a 100% increase was measured in AQP1-expressing cells, when compared to mock-transfected cells. On membranes of cells expressing AQP4, large aggregates of IMPs were organized into orthogonal arrays, which occupied 10–20% of the membrane surface. IMP aggregates were never seen in AQP2-transfected cells. Hexagonally packed IMP clusters were detected in ∼5% of the membranes from AQP3-expressing cells. Particle size-distribution analysis of rotary shadowed IMPs showed a significant shift from 13.5 (control cells) to 8.5 nm or less in AQP-expressing cells; size distribution analysis of unidirectionally shadowed IMPs also showed a significant change when compared to control. Some IMPs in AQP expressing cells had features consistent with the idea that aquaporins are assembled as tetramers. The results demonstrate that in transfected CHO cells, AQP transfection modifies the general appearance and number of IMPs on the plasma membrane, and show that only AQP4 assembles into well-defined IMP arrays. Received: 17 March 1998/Revised: 19 June 1998     

Mouse VDAC Isoforms Expressed in Yeast: Channel Properties and Their Roles in Mitochondrial Outer Membrane Permeability

The channel-forming protein called VDAC forms the major pathway in the mitochondrial outer membrane and controls metabolite flux across that membrane. The different VDAC isoforms of a species may play different roles in the regulation of mitochondrial functions. The mouse has three VDAC isoforms (VDAC1, VDAC2 and VDAC3). These proteins and different versions of VDAC3 were expressed in yeast cells (S. cerevisiae) missing the major yeast VDAC gene and studied using different approaches. When reconstituted into liposomes, each isoform induced a permeability in the liposomes with a similar molecular weight cutoff (between 3,400 and 6,800 daltons based on permeability to polyethylene glycol). In contrast, electrophysiological studies on purified proteins showed very different channel properties. VDAC1 is the prototypic version whose properties are highly conserved among other species. VDAC2 also has normal gating activity but may exist in 2 forms, one with a lower conductance and selectivity. VDAC3 can also form channels in planar phospholipid membranes. It does not insert readily into membranes and generally does not gate well even at high membrane potentials (up to 80 mV). Isolated mitochondria exhibit large differences in their outer membrane permeability to NADH depending on which of the mouse VDAC proteins was expressed. These differences in permeability could not simply be attributed to different amounts of each protein present in the isolated mitochondria. The roles of these different VDAC proteins are discussed. Received: 19 June 1998/Revised: 1 April 1999     

Episodic Evolution of Protein Hormones in Mammals

Pituitary growth hormone (GH) and prolactin have been shown previously to display a pattern of evolution in which episodes of rapid change are imposed on a low underlying basal rate (near-stasis). This study was designed to explore whether a similar pattern is seen in the evolution of other protein hormones in mammals. Seven protein hormones were examined (with the common α-subunit of the glycoprotein hormones providing an additional polypeptide for analysis)—those for which sequences from at least four eutherian orders are available with a suitable non-eutherian outgroup. Six of these (GH, prolactin, insulin, parathyroid hormone, glycoprotein hormone α-subunit, and luteinizing hormone β-subunit) showed markedly variable evolutionary rates in each case with a pattern of a slow basal rate and bursts of rapid change, the precise positions of the bursts varying from protein to protein. Two protein hormones (follicle-stimulating hormone β-subunit and thyroid-stimulating hormone β-subunit) showed no significant rate variation. Based on the sequences currently available, and pooling data from all eight proteins, the phase of slow basal change occupied about 85% of the sampled evolutionary time, but most evolutionary change (about 62% of the substitutions accepted) occurred during the episodes of rapid change. It is concluded that, in mammals at least, a pattern of prolonged periods of near-stasis with occasional episodes of rapid change provides a better model of evolutionary change for protein hormones than the one of constant evolutionary rates that is commonly favored. The mechanisms underlying this episodic evolution are not yet clear, and it may be that they vary from one group to another; in some cases, positive selection appears to underlie bursts of rapid change. Where gene duplication is associated with a period of accelerated evolution this often occurs at the end rather than the beginning of the episode. To what extent the type of pattern seen for protein hormones can be extended to other proteins remains to be established. Received: 10 October 2000 / Accepted: 18 December 2000     

Structure of the Transmembrane Cysteine Residues in Phospholamban

Phospholamban, a 52-residue membrane protein, associates to form a pentameric complex of five long α-helices traversing the sarcoplasmic reticulum membrane of cardiac muscle cells. The transmembrane domain of the protein is largely hydrophobic, with only three cysteine residues having polar side chains, yet it functions as a Ca²⁺-selective ion channel. In this report, infrared spectroscopy is used to probe the conformation of the three cysteine side chains and to establish whether the free S-H groups form intrahelical hydrogen bonds in the pentameric complex. Vibrational spectra of a transmembrane peptide were obtained which corresponded to the transmembrane domain of wild-type phospholamban and three peptides each containing a cysteine ⇒ alanine substitution. The observed S-H frequencies argue that each of the sulfhydryl groups is hydrogen-bonded to an i-4 backbone carbonyl oxygen. Electrostatic calculations on a model of phospholamban based on molecular dynamics and mutagenesis studies, show that the sulfhydryl groups may significantly contribute to the electrostatic potential field of the protein. Received: 22 July 1996/Revised: 10 October 1996