Model studies support pyrrolylation of the topaquinone cofactor to explain inactivation of bovine plasma amine oxidase by 3-pyrrolines. Unusual processing of a secondary amine

Full text of DOI:10.1021/ja9543210

+
+
J. Am. Chem. Soc. 1996, 118, 7241-7242
Scheme 1
7241
Model Studies Support Pyrrolylation of the
Topaquinone Cofactor To Explain Inactivation of
Bovine Plasma Amine Oxidase by 3-Pyrrolines.
Unusual Processing of a Secondary Amine
Younghee Lee, He Huang, and Lawrence M. Sayre*
Department of Chemistry
Case Western ReserVe UniVersity
CleVeland, Ohio 44106
ReceiVed December 29, 1995
Since the report in 1990 that the active carbonyl cofactor used
by plasma and other copper amine oxidases is the quinone form
(topaquinone, TPQ) of an active site 2,4,5-trihydroxyphenyl-
alanine residue,1 there has been renewed interest in mechanistic
and directed model studies.2-7 A pyridoxal-like transamination
mechanism appears to be in force for the enzyme (Scheme 1,
path A), since anaerobic single turnover results in release of
aldehyde product, whereas NH3 is released only upon O2-
dependent reoxidation of the reduced cofactor.8 In contrast, an
addition-elimination mechanism (Scheme 1, path B), found to
be a competing pathway for benzylamine deamination using
pyrroloquinoline quinone as a model,9 would predict simulta-
neous release of aldehyde and NH3 after anaerobic single
turnover. The fact that secondary amines are not substrates for
the enzyme10 might be rationalized on the basis that transami-
nation in this case would require tautomerization between two
iminium intermediates (see Scheme 2). However, we here
communicate model study results demonstrating the occurrence
of transamination for a secondary amine that led us to detect
enzymatic processing in the form of a mechanism-based
inactivation event.
Scheme 2
We recently reported a pivalamidoethyl-based model 1a
which is active in the catalytic aerobic deamination of benzyl-
amine in buffered aqueous acetonitrile,4 whereas Mure and
Klinman investigated the ability of various topaquinone models
to catalyze aerobic deamination in anhydrous acetonitrile.5 These
initial model studies revealed several features of relevance to
the enzymes, but the observation of catalytic turnover did not
permit a distinction between transamination and addition-
elimination mechanisms. In fact, although the spectrophoto-
metric behavior of a TPQ model reaction with benzylamine
under pseudo-first-order conditions was shown to be consistent
with a transamination mechanism,6 results for PhCD2NH2
compared to those for PhCH2NH2 revealed that the spectro-
photometric data could not provide an unambiguous mechanistic
interpretation due to overlapping of “substrate” (quinone) imine
and “product” (aldehyde) imine chromophores.7
We hoped that a mechanistic distinction could be made simply
by determining whether the product of anaerobic single turnover
was the aminoresorcinol (from transamination) or benzenetriol
(from addition-elimination) (see Scheme 1). However, the
occurrence of “redox interchange” reactions, which scrambled
the nature of the initial product of quinone reduction as soon
as it was formed, prevented a mechanistic conclusion based on
product analysis, even upon continuous NMR monitoring of
the reaction.4 A solution to this dilemma was found (i) by
manipulating NMR reaction conditions to favor the unimolecular
throughput of the initial quinone-amine adduct at the expense
of bimolecular redox interchange events and (ii) by analyzing
products of a diagnostic substrate that was devoid of redox
interchange complications.7 In this way we were able to
demonstrate unambiguously a preferred transamination mech-
anism.7
* Correspondence: phone, (216) 368-3704; FAX, (216) 368-3006; e-mail,
lms3@po.cwru.edu.
(1) Janes, S. M.; Mu, S.; Wemmer, D.; Smith, A. J.; Kaur, S.; Maltby,
D.; Burlingame, A. L.; Klinman, J. P. Science 1990, 248, 981.
(2) Mure, M.; Klinman, J. P. J. Am. Chem. Soc. 1993, 115, 7117.
(3) Wang, F.; Bae, J. Y.; Jacobson, A. R.; Lee, Y.; Sayre, L. M. J. Org.
Chem. 1994, 59, 2409.
(4) Lee, Y.; Sayre, L. M. J. Am. Chem. Soc. 1995, 117, 3096.
(5) Mure, M.; Klinman, J. P. J. Am. Chem. Soc. 1995, 117, 8698.
(6) Mure, M.; Klinman, J. P. J. Am. Chem. Soc. 1995, 117, 8707.
(7) Lee, Y.; Sayre, L. M. J. Am. Chem. Soc. 1995, 117, 11823.
(8) Janes, S. M.; Klinman, J. P. Biochemistry 1991, 30, 4599.
(9) (a) Ohshiro, Y.; Itoh, S.; Bioorg. Chem. 1991, 19, 169. (b) Rodriguez,
E. J.; Bruice, T. C. J. Am. Chem. Soc. 1989, 111, 7947.
(10) A recent report (Houen, G.; Bock, K.; Jensen, A. L. Acta Chem.
Scand. 1994, 48, 52) suggests that the polyamines spermine and spermidine
are preferentially metabolized by copper amine oxidases at the secondary
amine positions. This conclusion was based on (i) the authors’ inability to
identify the (amino)dialdehydes expected from primary amine deamination
and (ii) the identification instead of 3-aminopropanal, the product expected
from secondary amine metabolism. However, our reading of the Houen et
al. paper suggests their inability to detect the (amino)dialdehydes reflects
their failure to recognize the earlier reported11 need to achieve rapid
enzymatic reaction so that the initial (amino)dialdehydes can be reductively
stabilized (BH4-) prior to their decomposition. This decomposition is a
retro Michael reaction producing acrolein, which we have found can
combine, under the reaction conditions, with the NH3 generated from
terminal deamination, thereby rationalizing the 3-aminopropanal detected.
Thus, as will be described elsewhere, we support the earlier interpretation11
that polyamines are metabolized at primary and not secondary amine
positions.
At this point, we believed that a secondary amine such as
the sterically “tied back” pyrrolidine might recruit an addition-
elimination pathway (Scheme 2, path B) in our model system.
However, our finding that reactions of pyrrolidine with both
1a and the tert-butyl-based model 1b5 in CD3CN led to the
(11) Tabor, C. W.; Tabor, H.; Bachrach, U. J. Biol. Chem. 1964, 239,
2194.
S0002-7863(95)04321-6 CCC: $12 00 © 1996 American Chemical Society

+
+
7242 J. Am. Chem. Soc., Vol. 118, No. 30, 1996
Communications to the Editor
Scheme 3
processing that occurs nonetheless constitutes what appears to
be the first precedent10 for enzymatic processing of a secondary
amine. We suggest that the usual nonsubstrate behavior of
typical secondary amines toward the copper amine oxidases is
active site steric exclusion rather than their inability to undergo
transamination. In this regard, it is clear that mechanism-based
inactivation serves as a much more sensitive test of enzymatic
processing than does substrate activity.
In an effort to improve the inhibitory efficiency of 3-pyrroline,
we considered adding an aryl ring “recognition element” to give
3-phenyl-3-pyrroline. This compound16 displayed a more potent
concentration- and time-dependent pseudo-first-order loss of
BPAO activity (50% inhibition was observed in 30 min at 0.1
mM) that was reduced in the presence of substrate benzylamine.
A Kitz and Wilson plot of the inactivation rate data in this case
comes too close to the O-O origin to permit an accurate
estimate of the intercepts. We suspect that the apparent lack
of saturation (see ref 17) merely reflects a more efficient
turnover (inactivation) at the expense of an insufficiently
improved binding. The inactivation is thus instead described
as a second-order rate constant, 230 M-1 min-1, which can be
compared to kinact/Ki ) 6 M-1 min-1 for 3-pyrroline itself.
Inhibition by the 3-pyrrolines was found to be irreversible
as indicated by reassay of activity following Sephadex G-25
chromatography and was accompanied by loss of the 480 nm
chromophore ascribed to the TPQ cofactor. The latter finding,
along with our inability to see the normal development at A450
following titration with phenylhydrazine, is consistent with our
proposed cofactor pyrrolylation mechanism (Scheme 2). It will
be of obvious interest to confirm for the enzyme that the cofactor
has actually been pyrrolylated in the manner we see in the model
reaction.
Further analog development may lead to even more efficient
inactivators in the 3-pyrroline class, and it will be important to
determine if our results here generalize to other enzymes in the
TPQ-dependent family of copper amine oxidases. Presuming
that inactivation reflects cofactor pyrrolylation, 3-pyrrolines
represent examples of transamination-specific inactivators; that
is, for enzymes such as the flavin-dependent mitochondrial
amine oxidase which utilize nontransamination mechanisms for
amine oxidation, the 3-pyrrolines should simply be turned over
as normal substrates to pyrroles. We hope to exploit this
potential mechanistic distinction in the development of selective
inhibitors of the TPQ-containing enzymes.
generation of the redox interchange product, (pyrrolidino)-
resorcinol 6, as the major product,12 could only be understood
in terms of generation of the corresponding quinoneiminium
intermediate 3, which was being reduced by benzenetriol 2 or
by (alkylamino)resorcinol 5. The existence of 3 implied that
transamination (Scheme 2, path A) might still be the predomi-
nant reaction pathway, and we were in fact able to detect (1H
NMR, D2O) the subsequent iminium intermediate 4 following
aqueous HCl quenching of the reaction of pyrrolidine with 1a.13
We then recognized that upon replacing pyrrolidine with the
activated (doubly allylic) analog 3-pyrroline, transamination and
addition-elimination pathways would be distinguished by
whether the observed product was pyrrole or the pyrrolated
cofactor 9 (Scheme 3). Redox interchange complications in
this case would presumably be thwarted on account of rapid
irreversible tautomeric aromatizations. In fact, reaction of 1b
with 3-pyrroline in CD3CN was found to give exclusively 9b.14
This result suggested to us that 3-pyrroline might act as a
cofactor-directed irreversible inactivator of TPQ-dependent
amine oxidases with little if any productive turnover. In fact,
no O2 uptake could be detected when bovine plasma amine
oxidase (BPAO, from Sigma) was incubated with up to 10 mM
pyrrolidine or 3-pyrroline,15 consistent with the alleged non-
substrate behavior of secondary amines. However, preincuba-
tion with 3-pyrroline resulted in a concentration- and time-
dependent pseudo-first-order loss of BPAO activity.15 Kitz and
Wilson analysis of the first-order inactivation rate constants at
30 °C, pH 7.2 yielded values of Ki ) 50 mM and kinact ) 0.3
min-1. The rather high Ki for 3-pyrroline is consistent with
very weak binding for this simple aliphatic amine (the preferred
substrates for BPAO are arylalkylamines), but the clear
Acknowledgment. We are grateful to NIH for support of this work
through Grant GM-48812.
(12) The reaction of 50-60 mM 1a or 1b and 150-180 mM pyrrolidine
in degassed CD3CN was monitored by 1H NMR spectroscopy. After 2
days (no further reaction), the major product was 6a or 6b, identified by
their independent “redox cycling” synthesis3 from triol 2a or 2b and
pyrrolidine. 6a: 1H NMR (CD3CN, 300 MHz) δ 1.11 (s, 9H), 1.89 (m,
4H), 2.62 (t, 2H, J ) 7.08 Hz), 2.92 (m, 4H), 3.23 (m, 2H), 6.35 (s, 1H),
6.67 (br s, NH, 1H), 6.82 (s, 1H). 13C NMR (CD3CN, 75 MHz) δ 25.0
(2C), 27.7 (3C), 30.7, 39.1, 41.0, 53.7 (2C), 102.7, 117.3, 123.2, 130.5,
152.1, 153.2, 180.0. 6b: 1H NMR (CD3CN, 300 MHz) δ 1.31 (s, 9H),
1.90 (m, 4H), 2.92 (m, 4H), 6.28 (s, 1H), 7.00 (s, 1H). 13C NMR (CD3-
CN, 75 MHz) δ 25.03 (2C), 30.08 (3C), 34.90, 53.91 (2C), 103.28, 120.03,
128.09, 129.67, 151.50, 153.30. HRMS (EI) calcd for C14H21NO2 m/z
235.1573; found 235.1579 (56.7%). The major side product in the case of
1b was identified, following isolation, as that arising from electrophilic
substitution of 1-pyrroline at C-2 of 6b, 2-(pyrrolidin-2-yl)-4-(pyrrolidin-
1-yl)-6-tert-butylresorcinol.
Supporting Information Available: Text giving experimental
details for the syntheses, chemical model reactions, and enzymologic
studies and kinetic plots of inactivation (6 pages). See any current
masthead page for ordering and Internet access instructions.
JA9543210
(15) Bovine plasma amine oxidase (Sigma, 40-80 units/gram of protein)
was incubated with various concentrations of inhibitors (pH 7.2, 30 °C),
and aliquots were assayed for activity by determining the rate of oxidation
of benzylamine (10 mM) to benzaldehyde at 250 nm. O2 uptake was
monitored using a Yellow Springs Instruments 5300 biological oxygen meter
at 25 °C, pH 7.2 using 0.7 µM BPAO in a total volume of 2 mL. Under
these conditions, complete consumption of O2 occurred in 20 min using 10
mM benzylamine.
(16) 3-Phenyl-3-pyrroline was prepared by addition of PhMgBr to
N-(carboethoxy)-3-pyrrolidone, N-deprotection of the resulting phenyl
carbinol by refluxing overnight in 1:1 n-propanol aqueous KOH (10 N),
and finally dehydration by refluxing in concd HCl for 1 h. The compound
has also been prepared from 2-phenyl-2-vinylaziridine: Hortmann, A. G.;
Koo, J.-y. J. Org. Chem. 1974, 39, 3781.
(13) Signals corresponding to 4a seen during 1H NMR spectral monitor-
ing of the reaction of pyrrolidine with 1a were identified through
independent generation of 4a by aqueous HCl treatment of the diethylacetal
of the precursor 4-anilinobutanal, in turn prepared by redox cycling3 reaction
of 2a with 4,4-diethoxybutanamine.
(14) The reaction of 160 mM 1b and 250 mM 3-pyrroline in degassed
CD3CN was monitored by 1H NMR spectroscopy. Compound 9 was formed
quantitatively in 2 h: 1H NMR (CD3CN, 300 MHz) δ 1.34 (s, 9H), 4.83
(bs, OH, 2H), 6.18 (t, 2H, J ) 2.14 Hz), 6.43 (s, 1H), 6.85 (t, 2H, J ) 2.14
Hz), 6.98 (s, 1H). 13C NMR (CD3CN, 75 MHz) δ 29.88 (3C), 34.61,
105.61, 109.13 (2C), 122.95 (2C), 125.49, 128.43, 128.84, 150.25, 155.87.
HRMS (EI) calcd for C14H17NO2 m/z 231.1260; found 231.1260 (88.7%).
(17) Silverman, R. B. Mechanism-Based Enzyme InactiVation: Chemistry
and Enzymology; CRC Press, Inc.: Boca Raton, FL, 1988; pp 3-30.