The organic functional group in copper-containing amine oxidases. Resonance Raman spectra are consistent with the presence of topa quinone (6-hydroxydopa quinone) in the active site

Resonance Raman spectroscopy has been used to probe the structure of the organic cofactor in copper-containing amine oxidases from bovine plasma, porcine kidney, pea seedlings, and the bacterium Arthrobacter P1. The enzymes were first derivatized with phenylhydrazine or p-nitrophenylhydrazine; resonance Raman spectra were obtained on the intact derivatized enzymes and on a derivatized active-site peptide isolated from bovine plasma amine oxidase. Spectra of the intact amine oxidase phenylhydrazones are practically identical, consistent with the enzymes examined containing a similar cofactor. Only minor frequency shifts and some intensity variations are detected between the resonance Raman spectra of intact bovine plasma amine oxidase and the isolated peptide. These spectral perturbations are attributable to differences in the micro-environment between the intact, folded protein and the isolated small peptide in aqueous solution. This rules out the possibility that a new structure is formed during the isolation of the derivatized active-site peptide. Importantly, the resonance Raman spectra of the phenylhydrazine and p-nitrophenylhydrazine derivatives of the bovine plasma amine oxidase peptide are identical to the spectra of the corresponding derivatives of topa quinone (6-hydroxydopa quinone). Hence these data provide strong, independent support for the recent identification of topa as the organic functional group in bovine plasma amine oxidase (Janes, S. M., Mu, D., Wemmer, D., Smith, A. J., Kaur, S., Maltby, D., Burlingame, A. L., and Klinman, J.P. (1990) Science 248, 981-987).     

The crystal structure of Pichia pastoris lysyl oxidase

Pichia pastoris lysyl oxidase (PPLO) is unique among the structurally characterized copper amine oxidases in being able to oxidize the side chain of lysine residues in polypeptides. Remarkably, the yeast PPLO is nearly as effective in oxidizing a mammalian tropoelastin substrate as is a true mammalian lysyl oxidase isolated from bovine aorta. Thus, PPLO is functionally related to the copper-containing lysyl oxidases despite the lack of any significant sequence similarity with these enzymes. The structure of PPLO has been determined at 1.65 A resolution. PPLO is a homodimer in which each subunit contains a Type II copper atom and a topaquinone cofactor (TPQ) formed by the posttranslational modification of a tyrosine residue. While PPLO has tertiary and quaternary topologies similar to those found in other quinone-containing copper amine oxidases, its active site is substantially more exposed and accessible. The structural elements that are responsible for the accessibility of the active site are identified and discussed.     

The organic cofactor in plasma amine oxidase: evidence for pyrroloquinoline quinone and against pyridoxal phosphate.

Plasma amine oxidases (EC 1.4.3.6) are classified as containing the organic cofactor pyridoxal phosphate. Biochemical and bioassays on the pig plasma amine oxidase fail to reveal the presence of pyridoxal phosphate and 31P n.m.r. evidence is also inconsistent with pyridoxal phosphate in the enzyme. Resonance Raman spectral studies on phenylhydrazone derivatives of the pig and bovine plasma enzymes have been carried out and comparisons made with the corresponding derivatives of pyridoxal phosphate and pyrroloquinoline quinone (PQQ). The resonance Raman evidence indicates that the cofactor in both plasma amine oxidases is PQQ or a closely related species and not pyridoxal phosphate. The results substantiate earlier reports concerning the identity of the organic cofactor.     

Mechanism-based inactivators of plant copper/quinone containing amine oxidases

Copper/quinone amine oxidases contain Cu(II) and the quinone of 2,4,5-trihydroxyphenylalanine (topaquinone; TPQ) as cofactors. TPQ is derived by post-translational modification of a conserved tyrosine residue in the protein chain. Major advances have been made during the last decade toward understanding the structure/function relationships of the active site in Cu/TPQ amine oxidases using specific inhibitors. Mechanism-based inactivators are substrate analogues that bind to the active site of an enzyme being accepted and processed by the normal catalytic mechanism of the enzyme. During the reaction a covalent modification of the enzyme occurs leading to irreversible inactivation. In this review mechanism-based inactivators of plant Cu/TPQ amine oxidases from the pulses lentil (Lens esculenta), pea (Pisum sativum), grass pea (Lathyrus sativus) and sainfoin (Onobrychis viciifolia,) are described. Substrates forming, in aerobiotic and in anaerobiotic conditions, killer products that covalently bound to the quinone cofactor or to a specific amino acid residue of the target enzyme are all reviewed.     

Evidence for pyrroloquinolinequinone as the carbonyl cofactor in lysyl oxidase by absorption and resonance Raman spectroscopy

The present study investigated the possibility that pyrroloquinolinequinone (PQQ), an aromatic carbonyl recently indicated to be the carbonyl cofactor in bovine plasma amine oxidase, may also be present at the active site of lysyl oxidase. The absorption and resonance Raman spectra of the phenylhydrazones of bovine plasma amine oxidase, of peptides derived from the active site of bovine aorta lysyl oxidase, and of PQQ were very similar, indicating that the carbonyl cofactor of lysyl oxidase is PQQ or a compound which closely resembles PQQ.     

Stereochemical trends in copper amine oxidase reactions

Copper amine oxidases (EC 1.4.3.6) exhibit atypical stereochemical patterns in the reactions they catalyze. Dopamine and tyramine are oxidized with abstraction of the pro-R hydrogen by the porcine plasma amine oxidase, the pro-S hydrogen by pea seedling amine oxidase and a net nonstereospecific proton abstraction by the bovine plasma enzyme. This provides the first example in which a reaction catalyzed by enzymes in the same formal class occurs by all three possible stereochemical routes. To assess the underlying mechanistic significance of this heterogeneity, we have established the stereochemical course of the oxidation of tyramine by five additional copper amine oxidases using 1H NMR spectroscopy. Reactions catalyzed by rabbit and sheep serum amine oxidases are nonstereospecific. These enzymes exhibit rare mirror image binding with differential flux through two opposite and stereospecific reaction pathways. Differential primary kinetic isotope effects are observed for each mode, 8 and 4.6 for pro-S abstraction and 2.6 and 2.7 for pro-R abstraction by the sheep and rabbit amine oxidases, respectively. Tyramine oxidations catalyzed by the soybean and chick pea amine oxidases and porcine kidney diamine oxidase, however, are all stereospecific, occurring with loss of the pro-S hydrogen at C-1. Solvent exchange profiles are consistent within each stereochemical class of enzyme; the pro-R and nonstereospecific enzymes exchange solvent into C-2 of product aldehydes, the pro-S enzymes do not.     

Comparison of Kinetic Properties Between Plant and Fungal Amine Oxidases

Abstract

Kinetic properties of novel amine oxidases isolated from a mold Aspergillus niger AKU 3302 were compared to those of typical plant amine oxidase from pea seedling (EC 1.4.3.6). Pea amine oxidase showed highest affinity with diamines, such as putrescine and cadaverine, while fungal enzymes oxidized preferably n-hexylamine and tyramine. All enzymes were inhibited by carbonyl reagents, copper chelating agents, some substrate analogs and alkaloids, but there were quite significant differences in the sensitivity and inhibition modes. Aminoguanidine, which strongly inhibited pea amine oxidases showed only little effect on fungal enzymes. Substrate analogs such as 1,5-diamino-3-pentanone and l-amino-3-phenyl-3-propanone, which were potent competitive inhibitors of pea amine oxidases, inhibited fungal enzymes much more weakly and non competitively. Also various alkaloids behaving as competitive inhibitors of pea amine oxidases inhibited the fungal enzymes non competitively. Very surprising was the potent inhibition of fungal enzymes by artificial substrates of pea amine oxidases, E- and Z-1,4-diamino-2-butene. The relationships between the different inhibition modes and possible binding at the active site are discussed.     

Confirmation of the presence of a Cu(II)/topa quinone active site in the amine oxidase from fenugreek seedlings

Amine oxidase from etiolated seedlings of fenugreek (Trigonellafoenum—graecum) has been isolated by a purification procedureinvolving three chromato—graphic steps. The homogeneousenzyme is of pink colour with a visible absorption maximum at500 nm. The dimeric enzyme (2 75 kDa) is a slightly acidicprotein (pl 6.8) containing 8% neutral sugars. N—ter—minalamino acid sequence of the enzyme shows a high degree of similarityto other plant and microbial copper—containing amine oxidases.The best substrates of the enzyme are aliphatic diamines andsome polyamines, whereas inhibitors are substrate analogues,copper complexing agents, some alkaloids and several other compounds.Spectrophotometric titra—tions with phenylhydrazines demonstratedone reactive carbonyl group per subunit of the enzyme and redox—cyclicquinone staining after native electrophor—esis indicatedthe presence of a quinone cofactor. Differential pulse polarographyshowed the existence of a copper/quinone—containing activesite. The resonance Raman spectroscopy and the pH—dependentshift of the absorption spectrum of the enzyme p—nitrophenylhydrazoneconfirm unambiguously the identity of the cofactor with topaquinone. EPR spectra of the enzyme are in accordance with thoseof tetragonal cupric complexes as known for other copper—containingamine oxidases. Besides the copper, Mn(II)ions were detectedthat partially occupy another metal site in the enzyme, buttheir catalytical importance is unlikely. Key words: Fenugreek, Trigonella foenum—graceum, amine oxidase, topa quinone     

Towards the development of selective amine oxidase inhibitors. Mechanism-based inhibition of six copper containing amine oxidases.

Four substrate analogs, 4-(2-naphthyloxy)-2-butyn-1-amine (1), 1,4-diamino-2-chloro-2-butene (2), 1,6-diamino-2,4-hexadiyne (3), and 2-chloro-5-phthalimidopentylamine (4) have been tested as inhibitors against mammalian, plant, bacterial, and fungal copper-containing amine oxidases: bovine plasma amine oxidase (BPAO), equine plasma amine oxidase (EPAO), pea seedling amine oxidase (PSAO), Arthrobacter globiformis amine oxidase (AGAO), Escherichia coli amine oxidase (ECAO), and Pichia pastoris lysyl oxidase (PPLO). Reactions of 1,4-diamino-2-butyne with selected amine oxidases were also examined. Each substrate analog contains a functional group that chemical precedent suggests could produce mechanism-based inactivation. Striking differences in selectivity and rates of inactivation were observed. For example, between two closely related plasma enzymes, BPAO is more sensitive than EPAO to 1 and 3, while the reverse is true for 2 and 4. In general, inactivation appears to arise in some cases from TPQ cofactor modification and in other cases from alkylation of protein residues in a manner that blocks access of substrate to the active site. Notably, 1 completely inhibits AGAO at stoichiometric concentrations and is not a substrate, but is an excellent substrate of PSAO and inhibition is observed only at very high concentrations. Structural models of 1 in Schiff base linkage to the TPQ cofactor in AGAO and PSAO (for which crystal structures are available) reveal substantial differences in the degree of interaction of bound 1 with side-chain residues, consistent with the widely divergent activities. Collectively, these results suggest that the development of highly selective amine oxidase inhibitors is feasible.     

Cyanide as a copper-directed inhibitor of amine oxidases: implications for the mechanism of amine oxidation

 The interactions of five copper-containing amine oxidases with substrates and substrate analogues in the presence of the copper ligands cyanide, azide, chloride, and 1,10-phenanthroline have been investigated. While cyanide inhibits, to varying degrees, the reaction of phenylhydrazine with porcine kidney amine oxidase (PKAO), porcine plasma amine oxidase (PPAO), bovine plasma amine oxidase (BPAO), and pea seedling amine oxidase (PSAO), it enhances the reaction of Arthrobacter P1 amine oxidase (APAO) with this substrate analogue. This indicates that cyanide exerts an indirect effect on topa quinone (TPQ) reactivity via coordination to Cu(II) rather than through cyanohydrin formation at the TPQ organic cofactor. Moreover, cyanide binding to the mechanistically relevant TPQ^• semiquinone form of substrate-reduced APAO and PSAO was not observable by EPR or resonance Raman spectroscopy. Hence, cyanide most likely inhibits enzyme reoxidation by binding to Cu(I) and trapping the Cu(I)-TPQ^• form of amine oxidases, and thus preventing the reaction of O₂ with Cu(I). In contrast, ligands such as azide, chloride, and 1,10-phenanthroline, which preferentially bind to Cu(II), inhibit by stabilizing the aminoquinol Cu(II)-TPQ_red redox state, which is in equilibrium with Cu(I)-TPQ^•. Received: 12 December 1996 / Accepted: 20 March 1997     

An important lysine residue in copper/quinone-containing amine oxidases

The interaction of xenon with copper/6-hydroxydopa (2,4,5-trihydroxyphenethylamine) quinone (TPQ) amine oxidases from the plant pulses lentil (Lens esculenta) and pea (Pisum sativum) (seedlings), the perennial Mediterranean shrub Euphorbia characias (latex), and the mammals cattle (serum) and pigs (kidney), were investigated by NMR and optical spectroscopy of the aqueous solutions of the enzymes. (129)Xe chemical shift provided evidence of xenon binding to one or more cavities of all these enzymes, and optical spectroscopy showed that under 10 atm of xenon gas, and in the absence of a substrate, the plant enzyme cofactor (TPQ), is converted into its reduced semiquinolamine radical. The kinetic parameters of the analyzed plant amine oxidases showed that the k(c) value of the xenon-treated enzymes was reduced by 40%. Moreover, whereas the measured K(m) value for oxygen and for the aromatic monoamine benzylamine was shown to be unchanged, the K(m) value for the diamine putrescine increased remarkably after the addition of xenon. Under the same experimental conditions, the TPQ of bovine serum amine oxidase maintained its oxidized form, whereas in pig kidney, the reduced aminoquinol species was formed without the radical species. Moreover the k(c) value of the xenon-treated pig enzyme in the presence of both benzylamine and cadaverine was shown to be dramatically reduced. It is proposed that the lysine residue at the active site of amine oxidase could be involved both in the formation of the reduced TPQ and in controlling catalytic activity.     

Gene organization and molecular modeling of copper amine oxidase from Aspergillus niger: re-evaluation of the cofactor structure

Amine oxidase AO-I from Aspergillus niger AKU 3302 has been reported to contain topa quinone (TPQ) as a cofactor; however, analysis of the p-nitrophenylhydrazine-derivatized enzyme and purified active site peptides showed the presence of a carboxylate ester linkage of TPQ to a glutamate. The catalytic functionality of such a cross-linked cofactor has recently been shown unlikely by spectroscopic and voltammetric studies on synthesized model compounds. We have obtained resonance Raman spectra of native and substrate-reduced AO-I demonstrating that the catalytically active cofactor is unmodified TPQ. The primary structure of the enzyme (GenBank acc. no. U31869) has been reviewed and updated by repeated isolation and sequencing of AO-I cDNA. This allowed rectification of several errors that account for previously reported low homology to other amine oxidases in the regions around copper binding histididyl residues. The results were confirmed by cloning the ao-1 structural gene (GenBank acc. no. AF362473). Analysis of the gene 5'-upstream region of the gene revealed potential binding sites for an analog of NIT2, the nitrogen metabolism regulatory protein found in Neurospora crassa and other fungi. The molecular structure of AO-I was modeled by a comparative method using published crystal structures of amine oxidases as templates.     

Comparison of the three primary structures of deoxyribonuclease isolated from bovine, ovine, and porcine pancreas. Derivation of the amino acid sequence of ovine DNase and revision of the previously published amino acid sequence of bovine DNase

Based on the published bovine DNase sequence (Liao, T.-H., Salnikow, J., Moore, S., and Stein, W. H. (1973) J. Biol. Chem. 248, 1489-1495), the ovine DNase sequence is derived from the amino acid compositions of isolated short peptides covering all regions of the intact polypeptide. The sequence is substantiated by results of automated Edman degradation of the intact polypeptide and of the two middle CNBr fragments, and by elucidation of the complete sequence of the COOH-terminal CNBr peptide. The 12 changes from bovine to ovine DNase are at residues 22 (Ala to Ser), 29 (Val to Leu), 35 (Val to Ala), 54 (Tyr to Asp), 62 (Thr to Ser), 83 (Leu to Val), 121 (His to Pro), 127 (Glu to Ala), 132 (Ala to Pro), 159 (His to Asp), 163 (Val to Ile), and 231 (Ala to Val). A minor genetic variant form of ovine DNase has Val at residue 163. The data from automated Edman degradation of the largest CNBr peptide of bovine DNase show that the published bovine DNase sequence is in error and that an Ile-Val-Arg tripeptide must be inserted between Arg-27 and Arg-28. The corrected sequence is substantiated by two peptides covering this region each with three amino acids more than the published sequence. Comparison of the bovine, ovine, and porcine DNase sequences reveals the following: with the revised bovine sequence, all three DNase sequences can be aligned without a gap; all three DNases have a carbohydrate side chain at Asn-18, but only porcine DNase has carbohydrate at Asn-106; there are 12 changes between bovine and ovine DNases, 56 between bovine and porcine, and 50 between ovine and porcine; there are six highly variable regions and four invariable ones; bovine and ovine DNases have the same length while porcine DNase is longer by 2 amino acid residues at the COOH terminus; the residues around the nucleotide-binding site, the four pairs of salt bridges, and the essential His-134 groups are not changed.     

Inhibition of six copper-containing amine oxidases by the antidepressant drug tranylcypromine

Potential inhibitory effects of the clinically utilized monoamine oxidase inhibitor tranylcypromine (TCP) on mammalian, plant, bacterial, and fungal copper-containing amine oxidases have been examined. The following enzymes have been investigated: human kidney diamine oxidase (HKAO), bovine plasma amine oxidase (BPAO), equine plasma amine oxidase (EPAO), pea seedling amine oxidase (PSAO), Arthrobacter globiformis amine oxidase (AGAO), and Pichia pastoris lysyl oxidase (PPLO). Only BPAO, EPAO, and AGAO were found to lose significant levels of activity when incubated with varying amounts of TCP. Inhibition of BPAO was completely reversible, with dialysis restoring full activity. TCP inhibition of AGAO was also found to be ultimately reversible; however, dialysis did not remove all bound compounds. Chemical displacement with either substrate or a substrate analogue successfully removed all bound TCP, indicating that this compound has a high affinity for the active site of AGAO. The notable lack of TCP inhibition on HKAO argues against the inhibition of diamine oxidase as a potential source for some of the deleterious side effects occurring in patients treated with this antidepressant. The marked differences observed in behavior among these enzymes speaks to the importance of intrinsic structural differences between the active sites of copper amine oxidases (CAO) which affect reactivity with a given inhibitor.     

Copper-containing amine oxidases. Biogenesis and catalysis; a structural perspective

This review will focus on how X-ray crystallographic studies of copper-containing amine oxidases have complemented the solution, kinetic, and spectroscopic research on this ubiquitous class of enzymes. These enzymes not only contain a copper ion at the active site, but also a unique organic cofactor, 2,4,5-trihydroxyphenylalanine quinone (TPQ), which is absolutely required for catalysis. Structural data have not only shed light on the catalytic mechanism of the enzyme, which converts primary amines, using molecular oxygen, to aldehydes, ammonia, and hydrogen peroxide, but also on biogenesis of the cofactor. The cofactor is derived from a tyrosine in the enzyme amino acid sequence and requires only the addition of copper(II) and molecular oxygen in a self-processing event.     

Purification and characterization of methylamine oxidase induced in Aspergillus niger AKU 3302

Crude extract of Aspergillus niger AKU 3302 mycelia incubated with methylamine showed a single amine oxidase activity band in a developed polyacrylamide gel that weakly cross-reacted with the antibody against a copper/topa quinone-containing amine oxidase (AO-II) from the same strain induced by n-butylamine. Since the organism cannot grow on methylamine and the already known quinoprotein amine oxidases of the organism cannot catalyze oxidation of methylamine, the organism was forced to produce another enzyme that could oxidize methylamine when the mycelia were incubated with methylamine. The enzyme was separated and purified from the already known two quinoprotein amine oxidases formed in the same mycelia. The purified enzyme showed a sharp symmetric sedimentation peak in analytical ultracentrifugation showing S20,w0 of 6.5s. The molecular mass of 133 kDa estimated by gel chromatography and 66.6 kDa found by SDS-PAGE confirmed the dimeric structure of the enzyme. The purified enzyme was pink in color with an absorption maximum at 494 nm. The enzyme readily oxidized methylamine, n-hexylamine, and n-butylamine, but not benzylamine, histamine, or tyramine, favorite substrates for the already known two quinoprotein amine oxidases. Inactivation by carbonyl reagents and copper chelators suggested the presence of a copper/topa quinone cofactor. Spectrophotometric titration by p-nitrophenylhydrazine showed one reactive carbonyl group per subunit and redox-cyclic quinone staining confirmed the presence of a quinone cofactor. pH-dependent shift of the absorption spectrum of the enzyme-p-nitrophenylhydrazone (469 nm at neutral to 577 nm at alkaline pH) supported the identity of the cofactor with topaquinone. Nothern blot analysis indicated that the methylamine oxidase encoding gene is largely different from the already known amine oxidase in the organism.     

Electroactive centers in Euphorbia latex and lentil seedling amine oxidases

The electrochemical behavior of redox centers in the active site of amine oxidases from lentil seedlings and Euphorbia characias latex was investigated using a mercury film electrode. Tyrosine-derived 6-hydroxydopa quinone (TPQ) and copper ions in the active site are redox centers of these amine oxidases. The enzymes undergo two reduction processes at negative potentials related to the reduction of the TPQ cofactor to the corresponding hydroquinones and the reduction of copper ions, (Cu(II)-->Cu(I)). Copper depleted enzymes, prepared by reduction with dithionite followed by dialysis against cyanide, undergo only one reduction process. Nyquist diagrams, recorded at potentials corresponding to the reduction of cofactors as dc-offset, represent charge transfer impedance followed by a Warburg-type line at low frequencies, indicating the occurrence of a diffusion controlled process in the rate-limiting step of the reduction process.     

Stereochemistry of copper amine oxidase reactions

The stereochemical course of the oxidation of stereospecifically deuterated dopamine and tyramine, catalyzed by porcine plasma amine oxidase, has been investigated using 1H NMR spectroscopy. The oxidation proceeds with loss of the pro-R hydrogen at C-1. This stereochemistry is in contrast to that observed with the analogous copper containing oxidases isolated from pea seedlings (pro-S) and bovine plasma (nonstereospecific). There is no precedent for these three distinct stereochemical reaction courses to be followed by enzymes in the same class. Mechanistic differences among the three enzymes are evident from the profiles of solvent exchange into reaction products; however, these differences cannot account for the overall differential stereochemical courses observed.     

Irreversible inhibition of pig kidney copper-containing amine oxidase by sodium and lithium ions.

Copper amine oxidase was found to be inhibited in a complex way by small alkali metal ions. Classic enzyme kinetic studies showed that Li+ and Na+ were weak noncompetitive inhibitors, whereas the larger alkali metals K+, Rb+ and Cs+ were not inhibitors. However, freezing in the presence of Na+ or Li+ surprisingly resulted in complete and irreversible inactivation. In the case of Li+, it was possible to show that one ion per subunit was retained permanently in the inactivated enzyme, suggesting a structural rearrangement. The mechanism of inhibition was studied using a wide range of spectroscopic and analytic techniques. Only minor changes in the protein structure could be detected, except for a significant change in the geometry of the copper site. The unique topaquinone cofactor was apparently functional and able to proceed through the reductive half of the catalytic cycle, but the enzyme no longer reacted with oxygen. The effect of Na+ and Li+ was source-specific for pig kidney and bovine kidney amine oxidases, while the enzymes from bovine serum or plants were not inactivated, consistent with a mechanism dependent on small structural differences. A model for irreversible inactivation is proposed in which the cofactor is co-ordinated directly to copper, in analogy with the inactivation reported for Escherichia coli amine oxidase under crystal growth conditions.     

Stereochemical course of tyramine oxidation by semicarbazide-sensitive amine oxidase.

Two semicarbazide-sensitive amine oxidases (SSAO's) from bovine and porcine aortic tissue were partially purified and characterized, and the stereochemical course of amine oxidation was evaluated. The porcine and bovine SSAO's were membrane bound glycoproteins, with Km values for benzylamine of 8 and 16 microM, respectively. The reactivity of SSAO with semicarbazide and phenylhydrazine suggests that the cofactor is a carbonyl type molecule. The stereochemical course of the bovine and porcine aortic semicarbazide-sensitive amine oxidase reaction was investigated using chiral tyramines, deuterated at C-1 and C-2, and 1H-NMR spectroscopy to establish the loss or retention of deuterium in product p-hydroxyphenethyl alcohols. The preferred mode of tyramine oxidation was found to occur with the loss of pro-S proton at C-1, coupled with solvent exchange into C-2, a pattern which has not been observed for any copper amine oxidase examined to date. The solvent exchange reaction also occurred stereospecifically, with loss from and reprotonation to the pro-R position, suggesting that these two processes occur from the same face of the enamine double bond.