The partial retro-inverso modification: a road traveled together

In the mid-1970s, Dr. Murray Goodman was interested in a reversed peptide bond as a surrogate to understand the functional role of the amide bond in aspartame, a dipeptide sweetener. Very soon, realizing the breath and potential of this modification, Murray expanded this activity into a full program and I was fortunate to be part of it. Together we formulated new concepts such as the partially modified retro-inverso and end-group modified retro-inverso transformations, tested hypotheses, generated novel nomenclature, developed synthetic routes, characterized the preferred conformations of the unique building blocks employed in this modification, the gem-diaminoalkyl and the C2-substituted malonyl residues, and studied the biological activity of retro-inverso isomers of bioactive peptides. In the early 1980s several laboratories initiated extensive research targeted at the retro-inverso modification. The revival of this field led to new applications, new methods of synthesis, and new insights on the conformational and topological properties of the retro-inverso modification. Among the fields that embraced the retro-inverso concept were immunology as pertains to subjects such as synthetic vaccines, immunomodulators, and diagnostic tools, and drug delivery field as pertains to targeted and nontargeted cell permeation vectors loaded with bioactive cargo. Doctor Murray Goodman's sudden death leaves behind not only family, friends, and colleagues, but also an impressive record of scientific achievements among which is the revival of the modern era of the retro-inverso transformation. Murray's numerous contributions, excellent leadership, enthusiastic promotion, and outstanding teachings in this field will carry and illuminate his memory far into the future.     

Functional consequences of retro-inverso isomerization of a miniature protein inhibitor of the p53–MDM2 interaction

Peptide retro-inverso isomerization is thought to be functionally neutral and has been widely used as a tool for designing proteolytically stable d-isomers to recapitulate biological activities of their parent l-peptides. Despite success in a wide range of applications, exceptions amply exist that clearly defy this rule of thumb when parent l-peptides adopt an α-helical conformation in their bound state. The detrimental energetic effect of retro-inverso isomerization of an α-helical l-peptide on its target protein binding has been estimated to be 3.0–3.4 kcal/mol. To better understand how the retro-inverso isomer of a structured protein works at the molecular level, we chemically synthesized and functionally characterized the retro-inverso isomer of a rationally designed miniature protein termed stingin of 18 amino acid residues, which adopts an N-terminal loop and a C-terminal α-helix stabilized by two intra-molecular disulfide bridges. Stingin emulated the transactivation peptide of the p53 tumor suppressor protein and bound with high affinity and via its C-terminal α-helix to MDM2 and MDMX—the two negative regulators of p53. We also prepared the retro isomer and d-enantiomer of stingin for comparative functional studies using fluorescence polarization and surface plasmon resonance techniques. We found that retro-inverso isomerization of l-stingin weakened its MDM2 binding by 720 fold (3.9 kcal/mol); while enantiomerization of l-stingin drastically reduced its binding to MDM2 by three orders of magnitude, sequence reversal completely abolished it. Our findings demonstrate the limitation of peptide retro-inverso isomerization in molecular mimicry and reinforce the notion that the strategy works poorly with biologically active α-helical peptides due to inherent differences at the secondary and tertiary structural levels between an l-peptide and its retro-inverso isomer despite their similar side chain topologies at the primary structural level.¹     

Evidence for an arginine residue at the allosteric sites of spinach leaf ADPglucose pyrophosphorylase

The covalent modification of spinach leaf ADPglucose pyrophosphorylase leads to inactivation of both activator-stimulated and -unstimulated activity. Inactivation can be prevented if either the activator 3PGA or the inhibitor Pi are present during the modification. Pi proved to be more effective at protecting the enzyme from inactivation as it afforded 50% protection at 51 µM compared to 50% protection by 405 µM 3PGA. Partial modification of the enzyme using [¹⁴C]-phenylglyoxal leads to a decrease in bothV _max,A _0.5 and a decrease in the ability of the 3PGA to stimulate the enzyme's activity. Modification increased the enzyme's susceptibility to inhibition by Pi and completely abolished the cooperative binding of Pi seen in the unmodified enzyme in the presence of 3PGA. Thus, phenylglyoxal appears to interfere, with the normal allosteric regulation of ADPglucose pyrophosphorylase from spinach leaf. Greater than 90% of the enzyme's activity is lost when 7.2 mol [¹⁴C]-phenylglyoxal are bound per mole of tetramer and this label is present in both the larger and small subunits. In addition, inactivation appears to involve two different arginine residues having different rates of modification.     

Binding site amino acid residues of jack fruit (artocarpus integrifolia) seed lectin: chemical modification and protein difference spectral studies

The effect of chemical modification of amino acid residues essential for sugar binding in the α-D-galactoside specific jack fruit (Artocarpus integrifolia) seed lectin and the protection of the residues by specific sugar from modification were studied. Citraconylation or maleylation of 75 % of its lysyl residues or acetylation of 70 % of the tyrosyl residues completely abolished sugar binding and agglutination without dissociation of subunits. 1-O-methyl α-D-galactoside could protect its essential lysyl and tyrosyl groups from modification. Tryptophan could not be detected in the protein. Difference absorption spectra on binding of the above sugar confirmed the role of tyrosine residues and showed an association constantK = 0.4 × 10³ M⁻¹. Data suggests that the lectin could be immobilized without any loss of sugar binding activity     

Structural comparison between retro-inverso and parent peptides: Molecular basis for the biological activity of a retro-inverso analogue of the immunodominant fragment of VP1 coat protein from foot-and-mouth disease virus

Antibodies induced against intact foot-and-mouth disease Virus (FMDV) particles bind to the retro-inverso analogue of fragment 141–159 of the viral coat protein VP1 of FMDV, variant A, equally well as to the parent peptide. A conformational investigation of this retro-inverso peptide was carried out by nmr spectroscopy and restrained molecular modeling in order to identify the structural basis for the antigenic mimicry between these retro-inverso and parent peptides. In 100% trifluoroethanol a well-defined left-handed α-helical region exists from residue 150 to residue 159, which is consistently present in all conformational families obtained from restrained modelling. A less-defined left-handed helical region is present in the tract 144–148, which is also consistent for all structures. Conformational flexibility exists about Gly149, which leads to two types of structures, either bent or linear. In the bent structures, a three-residue inverse tight turn is found, which can be classified as an inverse γ-turn centered at Gly149. The overall structural features of the retro-inverso peptide are shown to be similar to those of the parent L-peptide. The two molecules, however, are roughly mirror images because they share inherently chiral secondary structure elements. By comparing these conformational conclusions with the x-ray structure of the Fab complex of a corresponding VP1 antigenic fragment, a rationale is proposed to account for the topological requirements of specific recognition that are implied by the equivalent antigenic activity of the natural and retro-inverso compounds. © 1997 John Wiley & Sons, Inc. Biopoly 41: 569–590, 1997.     

Limitations of Peptide Retro-inverso Isomerization in Molecular Mimicry

A retro-inverso peptide is made up of d-amino acids in a reversed sequence and, when extended, assumes a side chain topology similar to that of its parent molecule but with inverted amide peptide bonds. Despite their limited success as antigenic mimicry, retro-inverso isomers generally fail to emulate the protein-binding activities of their parent peptides of an α-helical nature. In studying the interaction between the tumor suppressor protein p53 and its negative regulator MDM2, Sakurai et al. (Sakurai, K., Chung, H. S., and Kahne, D. (2004) J. Am. Chem. Soc. 126, 16288–16289) made a surprising finding that the retro-inverso isomer of p53(15–29) retained the same binding activity as the wild type peptide as determined by inhibition enzyme-linked immunosorbent assay. The authors attributed the unusual outcome to the ability of the d-peptide to adopt a right-handed helical conformation upon MDM2 binding. Using a battery of biochemical and biophysical tools, we found that retro-inverso isomerization diminished p53 (15,–29) binding to MDM2 or MDMX by 3.2–3.3 kcal/mol. Similar results were replicated with the C-terminal domain of HIV-1 capsid protein (3.0 kcal/mol) and the Src homology 3 domain of Abl tyrosine kinase (3.4 kcal/mol). CD and NMR spectroscopic as well as x-ray crystallographic studies showed that d-peptide ligands of MDM2 invariably adopted left-handed helical conformations in both free and bound states. Our findings reinforce that the retro-inverso strategy works poorly in molecular mimicry of biologically active helical peptides, due to inherent differences at the secondary and tertiary structure levels between an l-peptide and its retro-inverso isomer despite their similar side chain topologies at the primary structure level.     

X-ray structures of new dipeptide taste ligands

The molecular basis of sweet taste was investigated by carrying out the crystal state conformational analysis by X-ray diffraction of the following dipeptide taste igands:N-3,3-dimethylbutyl-aspartyl-phenylalanine methyl ester, I (N-DMB-Asp-Phe-OMe), its sodium salt (N-DMB-Asp-Phe-ONa), II , aspartyl-D -2-aminobutyric acid-(S)-α-ethylbenzylamide, III (Asp-D -Abu-(S)-α-ethylbenzylamide), aspartyl-N′-((2,2,5,5-tetramethylcyclopentanyl)-carbonyl)-(R)-1,1-diamino-ethane, IV (Asp-(R)-gAla-TMCP), and aspartyl-D -valine-(R)-α-methoxymethylbenzyl amide, V (Asp-D -Val-(R)-α-methoxymethylbenzylamide). With the exception of the sodium salt II , all compounds are sweet-tasting, showing in some cases considerable potency enhancement with respect to sucrose. The results of this study confirm the earlier model that an ‘L-shape’ molecular array is essential for eliciting sweet taste for dipeptide-like ligands. In addition, it was established that (i) substitution of the N-terminal group does not inhibit sweet taste, if its zwitterionic character is maintained; (ii) a hydrophobic group located between the stem and the base of the L-shape could be responsible for sweetness potency enhancement, as found in I, III and IV ; in fact, the extraordinary potency of the N-alkylated analogue I would support a model with an additional hydrophobic binding domain above the base of the ‘L’; (iii) removal of the methyl ester at the C-terminus of compound I with the salt formation gives rise to the tasteless compound II ; (iv) for the first time all possible side-chain conformers (g⁻,g⁺andt) for the N-substituted aspartyl residue were observed; and (v) a retro-inverso modification, incorporated at position 2 of the dipeptide chain, confers greater flexibility to the molecule, as demonstrated by the contemporary presence of six conformationally distinct independent molecules in the unit cell and yet sweet taste properties are maintained, as found in IV . © 1998 European Peptide Society and John Wiley & Sons, Ltd.     

Proton NMR studies of partially modified retro–inverso peptides

High-resolution ¹H-nmr analyses are presented for partially modified retro-inverso derivatives of peptides with emphasis on enkephalins. Studies in DMSO-d₆ reveal the unique as well as common characteristics of the non-amino acid residues incorporated into these modified peptides. The complete assignment of 1,1-diaminoalkyl and malonyl, as well as of amino acid, residues provides the basis for the exploration of the conformational features induced in peptides by such topochemical modifications.     

Conformational studies of retro–inverso peptides: The crystal and molecular structure of the hydantoin from H-Ala-G-Ala-mGly-OBzl

The crystal structure of the hydantoin 1-[(S)-1′–aminoethylmalonyl benzyl ester]-(S)-4-methylimidazolidin-2.5-dione (1) derived from the peptide H-Ala-gAla-mGly-OBzl, Having the retro-inverso modification of the Ala-Gly bond, has been determined by x-ray diffraction analysis. The crystals are orthorhombic, space group P2₁2₁2₁ with a = 6.539. b = 14.721, c = 17.101 Å, z = 4. The structure was solved by direct methods and refined with anisotropic thermal factors to a final R value of 0.067 for the 947 observed reflections. Reversal of the Ala-Gly amide bond perturbs the folding tendency of the backbone shown by the parent peptide t-BuCO-Ala-Gly-NHiPr. The gem-diamino residue, gAla, and the malonyl moieties are found in the helical and the extended conformations, respectively. Intramolecular hydrogen bonding is not observed. The molecules in the crystal are held together by the formation of two intermolecular hydrogen bonds of the N? H?O?C type with N?O distances of 2.86 and 3.17 Å respectively. 1995 John Wiley&Sons. Inc. © 1995 John Wiley & Sons, Inc.     

Synthesis of peptide nucleic acid oligomers carrying 5-methylcytosine derivatives by postsynthetic substitution

The preparation of the thymine peptide nucleicacid (PNA) monomer carrying a 2-nitrophenyl group in position4 is described. This monomer is incorporated into PNAoligomers and reacted with amines to yield PNA oligomerscarrying 5-methylcytosine derivatives. During thedeprotection-modification step two side reactions weredetected: degradation of PNA oligomer from the N-terminal residue and modification of N ⁴-tert-butylbenzoyl cytosine residue. Protection of the N-terminal position and the use of N ⁴-acetyl group for the protection of cytosine eliminate these side reactions.     

Complex I of Rhodobacter capsulatus and its role in reverted electron transport

The activities of NAD⁺-photoreduction and NADH/decyl-ubiquinone reductase in membrane preparations of Rhodobacter capsulatus changed to the same extent under different conditions. These results indicated that NADH:ubiquinone oxidoreductase (complex I) catalyzes the electron transport in the downhill direction (respiratory chain) and in the uphill direction (reverted electron flow). This conclusion was confirmed by the characterization of a complex-I-deficient mutant of R. capsulatus. The mutant was not able to reduce NAD⁺ in the light. Since this mutant was not able to grow photoautotrophically, we concluded that complex I is the enzyme that catalyzes the reverted electron flow to NAD⁺ to provide reduction equivalents for CO₂ fixation. Complex I is not essential for the reverted electron flow to nitrogenase since the mutant grew under nitrogen-fixing conditions. As shown by immunological means, NuoE, a subunit of complex I from R. capsulatus having an extended C-terminus, was modified depending on the nitrogen source present in the growth medium. When the organism used N₂ instead of NH₄ ⁺, a smaller NuoE polypeptide was synthesized. The complex-I-deficient mutant was not able to modify NuoE. The function of the modification is discussed. Received: 28 February 1997 / Accepted: 28 August 1997     

New Screening Method for the Rapid Identification of Blocked Mutants and Their Products: Application to Erythromycin and Oleandomycin Producing Strains

γ-Glutamyltransferase from fruiting bodies of Lentinus edodes was further tested for its activation by chaotropic ions such as SCN^?, NO₃^?, Cl^?, Br^?, I^?, F^? and C1O₄^?. The thiocyanate ion increased the Km value for γ-glutamyl-p-nitroanilide without affecting the V_max value of the reaction, whereas other anions as represented by NO₃^? and Br^? increased the V_max without affecting the Km. Jhe inactivation of the enzyme by the SH group-orienting reagents, iodoacetamide and hydrogen peroxide, was stimulated by SCN^? but not by the other anions.

The activator anions protected the enzyme against its inactivation by chemical modification with 2,3-butanedione in borate. Their efficiency was parallel to the activator potency of the respective anions, except for SCN^? which provided less protection than expected from its activation potency. These dissociable effects of activator anions might be explained by two different mechanisms; binding of SCN^? to a basic group to bring about a significant change in protein conformation and binding of other anions by electrostatic and hydrophobic forces to an arginyl residue located near the active site of the enzyme.     

Studies onAspergillus niger glutamine synthetase: Regulation of enzyme levels by nitrogen sources and identification of active site residues

The specific activity of glutamine synthetase (L-glutamate: ammonia ligase, EC 6.3.1.2) in surface grownAspergillus niger was increased 3–5 fold when grown on L-glutamate or potassium nitrate, compared to the activity obtained on ammonium chloride. The levels of glutamine synthetase was regulated by the availability of nitrogen source like NH ₄ ⁺ , and further, the enzyme is repressed by increasing concentrations of NH ₄ ⁺ . In contrast to other micro-organisms, theAspergillus niger enzyme was neither specifically inactivated by NH ₄ ⁺ or L-glutamine nor regulated by covalent modification. Glutamine synthetase fromAspergillus niger was purified to homogenity. The native enzyme is octameric with a molecular weight of 385,000±25,000. The enzyme also catalyses Mn²⁺ or Mg²⁺-dependent synthetase and Mn²⁺-dependent transferase activity. Aspergillusniger glutamine synthetase was completely inactivated by two mol of phenyl-glyoxal and one mol of N-ethylmaleimide with second order rate constants of 3.8 M^-1 min^-1 and 760 M^-1 min^-1 respectively. Ligands like Mg. ATP, Mg. ADP, Mg. AMP, L-glutamate NH ₄ ⁺ , Mn²⁺ protected the enzyme against inactivation. The pattern of inactivation and protection afforded by different ligands against N-ethylamaleimide and phenylglyoxal was remarkably similar. These results suggest that metal ATP complex acts as a substrate and interacts with an arginine ressidue at the active site. Further, the metal ion and the free nucleotide probably interact at other sites on the enzyme affecting the catalytic activity.     

Synthesis of peptide nucleic acid oligomers carrying 5-methylcytosine derivatives by postsynthetic substitution

Summary The preparation of the thymine peptide nucleic acid (PNA) monomer carrying a 2-nitrophenyl group in position 4 is described. This monomer is incorporated into PNA oligomers and reacted with amines to yield PNA oligomers carrying 5-methylcytosine derivatives. During the deprotection-modification step two side reactions were detected: degradation of PNA oligomer from theN-terminal residue and modification ofN ⁴-tert-butylbenzoyl cytosine residue. Protection of theN-terminal position and the use ofN ⁴-acetyl group for the protection of cytosine eliminate these side reactions.     

Computer simulation of the conformational properties of retro–inverso peptides. I. Empirical force field calculations of rigid and flexible geometries of N-acetylglycine-N′- methylamide,bis(acetamido) methane,and N,N′- dimethylmalonamide and their corresponding Cα-methylated analogs

Rigid and flexible geometry calculations are described for N-acetylglycine-N′-methylamide, N-acetylalanine-N′-methylamide, and their retro-inverso analogs, bis(acetamido) methane, 1,1-bis(acetamido) ethane, N,N′-dimethylmalonamide, and N,N′-dimethyl-2-methyl-malonamide. The significance of relaxing all degrees of freedom, especially angular flexibility is demonstrated. The flexible geometry approach yields energy maps similar to those from rigid geometry, but the energy barriers between minima are substantially reduced, leading in general, to more probable transitions and a higher volume of accessible conformational space. Whereas the glycine and alanine derivatives exhibit their lowest energy minima in the C region, the gem-diaminoalkyl and malonyl residues show their lowest minima in the “α-helical” regions. With respect to the effect of side chains (H versus CH₃), the greatest conformational influence appears with the gem-diaminoalkyl residues. These results indicate significantly different conformational behavior of retro peptides and the implications of these pairwise incorporations of retro-inverso residues in peptide chains, are discussed.     

Amino group modification of (Na++K+)-ATPase

The effects of three amino group reagents on the activity of (Na⁺+K⁺)-ATPase³ and its component K⁺-stimulatedp-nitrophenylphosphatase activity from rabbit kidney outer medulla have been studied. All three reagents cause inactivation of the enzyme. Modification of amino groups with trinitrobenzene sulfonic acid yields kinetics of inactivation of both activities, which depend on the type and concentration of the ligands present. In the absence of added ligands, or with either Na⁺ of Mg²⁺ present, the enzyme inactivation process follows complicated kinetics. In the presence of K⁺, Rb⁺, or Tl⁺, protection occurs due to a change of the kinetics of inactivation toward a first-order process. ATP protects against inactivation at a much lower concentration in the absence than in the presence of Mg²⁺ (P ₅₀ 6 µM vs. 1.2 mM). Under certain conditions (100 µM reagent, 0.2 M triethanolamine buffer, pH 8.5) modification of only 2% of the amino groups is sufficient to obtain 50% inhibition of the ATPase activity. Modification of amino groups with ethylacetimidate causes a nonspecific type of inactivation of (Na⁺+K⁺)-ATPase. Mg²⁺ and K⁺ have no effects, and ATP only a minor effect, on the degree of modification. The K⁺-stimulatedp-nitrophenylphosphatase activity is less inhibited than the (Na⁺+K⁺)-ATPase activity. Half-inhibition of the (Na⁺+K⁺)-ATPase is obtained only after 25% modification of the amino groups. Modification of amino groups with acetic anhydride also causes nonspecific inactivation of (Na⁺+K⁺)-ATPase. Mg²⁺ has no effect, and ATP has only a slight protecting effect. The K⁺-stimulatedp-nitrophenylphosphatase activity is inhibited in parallel with the (Na⁺+K⁺)-ATPase activity. Half-inactivation of the (Na⁺+K⁺)-ATPase activity is obtained after 20% modification of the amino groups.This article is No. 52 in the series Studies on (Na⁺+K⁺)-Activated ATPase.     

The Characteristics of Arylamine N-Acetyltransferase in Pseudomonas aeruginosa

N-Acetyltransferase (NAT), responsible for bioactivation and detoxification of arylamines, has been demonstrated to be widely distributed in many organisms ranging from humans to microorganisms. Using high performance liquid chromatography (HPLC) to analyze NAT activity in bacteria, the authors found that Pseudomonas aeruginosa exhibited high NAT activity with 2-aminofluorene (2-AF) as substrate. Characteristics of this bacterial NAT were further investigated. The N-acetylation catalyzed by this enzyme is an acetyl coenzyme A (AcCoA)-dependent reaction. As the concentration of AcCoA in the reaction mixture was increased, the apparent K _m and V _max for 2-AF increased. The K _m and V _max were 0.504 ± 0.056 mM and 31.92 ± 3.23 nmol/min/mg protein, respectively, for the acetylation of 2-AF with 0.5 mM AcCoA. The optimum pH for the enzyme activity was estimated to be around 8.5. It was active at a temperature range from 5°C to 55°C, with maximum activity at 37°C. The enzyme activity was inhibited by divalent metal ions including Cu⁺⁺, Fe⁺⁺, Zn⁺⁺, Ca⁺⁺, Co⁺⁺, Mn⁺⁺, and Mg⁺⁺, suggesting that a sulfhydryl group is involved in the N-acetylation activity. The three chemical modification agents, iodoacetamide, phenylglyoxal, and diethylpyrocarbonate, all exhibited a dose-, time-, and temperature-dependent inhibition effect. Preincubation of the NAT with AcCoA provided significant protection against the inhibition of iodoacetamide and diethylpyrocarbonate, but only partial protection against the inhibition of phenylglyoxal. These results indicate that cysteine, histidine, and arginine residues are essential for this bacterial enzyme activity, and the first two are likely to reside on the AcCoA binding site, but arginine residue may be located only near the AcCoA binding site. Our data demonstrate that P. aeruginosa possesses highly active N-acetyltransferase which shares a similar catalytic mechanism as that of higher organisms. These findings are very helpful for further investigating the role of arylamine NAT in this bacterial species. Received: 29 August 1997 / Accepted: 23 December 1997