Selective inhibition of type II dehydroquinases

Full text of DOI:10.1021/jo990004q

2612
J . Org. Chem. 1999, 64, 2612-2613
acidic 2-pro-S hydrogen of the substrate, in an anti elimina-
Selective In h ibition of Typ e II
Deh yd r oqu in a ses
tion of water (Scheme 1).13 The intermediate 3 is character-
ized by the flattening of the carbocyclic ring and probably
involves stronger hydrogen bonding to the enolate oxygen
in the transition state. We have exploited these two features
separately in the design of the first inhibitors of type II
dehydroquinases.
Martyn Frederickson,† Emily J . Parker,†,‡
Alastair R. Hawkins,§ J ohn R. Coggins, and
Chris Abell*,†
Our first target compound was 2,3-anhydroquinic acid 10.
This is structurally similar to the substrate 1 but lacks the
carbonyl group necessary to form an imine with the type I
enzyme. In addition, the C2-C3 double bond mimics the
flattening of the ring in the enolate intermediate 3. To assess
the contribution that this subtle conformational restriction
makes to binding, the reduced compound 3-deoxyquinic acid
11 was also prepared.
Both 10 and 11 were synthesized from (-)-quinic acid 4
(Scheme 2). Treatment of quinic acid with benzaldehyde in
hot acidic toluene afforded an epimeric mixture of 3,4-O-
benzylidenequinides 5 (75%),14 which were readily converted
to the trans-bromobenzoate 6 (86%) and thence to the silyl
ether 7 (83%) using previously reported protocols.15 Fluoride-
induced desilylation of 7 yielded allylic alcohol 8 (88%),
which upon treatment with catalytic methoxide underwent
lactone methanolysis with concomitant debenzoylation to
afford ester 9 (96%). Saponification of ester 9 followed by
ion exchange afforded the desired acid 10 (99%), which upon
hydrogenation yielded 3-deoxyquinic acid 11 (97%).
Our second design strategy was to look for extra binding
affinity in the carbonyl binding pocket of the type II
dehydroquinases, where it is assumed that stabilization is
provided for the formation of the enolate intermediate 3. The
target compound was the simple oxime 19. As a control we
made the compound 13 with the exo-methylene group
(Scheme 3). Three-step deprotection of the known16 exocyclic
olefin 12 proceeded smoothly to afford the desired acid 13.
The oxime 19 was prepared from the known lactone 14.14
Methoxide-catalyzed lactone cleavage occurred with con-
comitant silyl migration to afford a chromatographically
separable mixture of ketoesters 15 (35%) and 16 (49%).
These were readily converted to their corresponding oximes
17 (84%) and 18 (92%) (single isomers in each case),
formulated as the E isomers both on steric grounds and on
the basis of downfield shifts of the equatorial hydrogens at
C-2 due to the neighboring oxygen atom of the oxime moiety.
Fluoride-induced desilylation of 18 followed by saponification
afforded the oxime 19 (isolated as its sodium salt).
The four acids 10, 11, 13, and 19 were assayed in the
presence of dehydroquinate for their inhibitory properties
against type I dehydroquinase (from Salmonella typhimu-
rium) and against type II dehydroquinases (from Aspergillus
nidulans, M. tuberculosis, and Streptomyces coelicolor).17 The
inhibition data are summarized in Table 1. All of the
inhibitors showed some level of competitive reversible
inhibition of both type I and II dehydroquinases. The
inhibitors 10 and 19 were clearly selective for type II
dehydroquinases and exhibited unexpected discrimination
between different type II enzymes.
University Chemical Laboratory, Lensfield Road,
Cambridge CB2 1EW, England, Department of Biochemistry
and Genetics, University of Newcastle upon Tyne,
Newcastle upon Tyne NE2 4HH, England, and Department of
Biochemistry and Molecular Biology, University of Glasgow,
Glasgow G12 8QQ, Scotland
Received J anuary 4, 1999
The syntheses of the first inhibitors of the type II
dehydroquinase (3-dehydroquinate dehydratase) are de-
scribed. Dehydroquinase catalyses the reversible conversion
of 3-dehydroquinic acid 1 to 3-dehydroshikimic acid 2 on the
shikimic acid pathway to aromatic amino acids.1 The same
reaction is also a step on the quinic acid pathway used by
fungi to metabolize quinic acid as a carbon source.2 There
are two distinct types of dehydroquinase.3 Type I dehydro-
quinases are typified by the E. coli enzyme, which is a dimer
of subunit Mr 27466,4 whereas type II dehydroquinases are
dodecameric proteins of subunit Mr 12000-18500.5,6
The presence of type II dehydroquinase as part of the
shikimate pathway in Mycobacterium tuberculosis,7 Helico-
bacter pylori,8 and Streptomyces9 make it an interesting and
novel antibiotic target. However, in designing inhibitors for
this enzyme the problem of discrimination from the type I
enzyme must be addressed as both types of dehydroquinase
are likely to have similar recognition elements in the active
site. However, consideration of the two enzyme mechanisms
suggests a rational basis for this discrimination. The type I
enzyme catalyzes an overall syn dehydration involving loss
of the 2-pro-R hydrogen.10 The mechanism proceeds through
a series of imine and enamine intermediates11 formed with
a conserved active site lysine.4 In contrast, the mechanism
of the type II enzyme does not involve any covalent attach-
ment of the substrate to the enzyme. The elimination has
been shown to proceed by an E1CB mechanism, probably
via an enolate intermediate 3,12 and involves loss of the more
* To whom correspondence should be addressed. Tel: 44 1223 336405.
Fax: 44 1223 336362. E-mail: ca26@cam.ac.uk.
† University Chemical Laboratory.
‡ Current address: Department of Chemistry, Massey University, Palm-
erston North, New Zealand.
§ University of Newcastle upon Tyne.
University of Glasgow.
(1) Bentley, R. CRC Crit. Rev. Biochem. 1990, 25, 307.
(2) Giles, N. H.; Case, M. E.; Baum, J . A.; Geever, R. F.; Huiet, L.; Patel,
V. B.; Tyler, B. M. Microbiol. Rev. 1985, 49, 338.
(3) Kleanthous, C.; Deka, R.; Davis, K.; Kelly, S. M.; Cooper, A.; Harding,
S. E.; Price, N. C.; Hawkins, A. R.; Coggins, J . R. Biochem. J . 1992, 282,
687.
(4) Chaudhuri, S.; Duncan, K.; Graham, L. D.; Coggins, J . R. Biochem.
J . 1991, 275, 1.
(5) Euverink, G. J . W.; Hessels, G. I.; Vrijbloed, J . W.; Coggins, J . R.;
Dijkhuizen, L. J . Gen. Microbiol. 1992, 138, 2449.
(6) Hawkins, A. R.; Reinert, W. R.; Giles, N. H. Biochem. J . 1982, 203,
769.
(7) Garbe, T.; Servos, S.; Hawkins, A.; Dimitriadis, G.; Young, D.; Dougan,
G.; Charles, I. Mol. Gen. Genet. 1991, 228, 385.
(8) Bottomley, J . R.; Clayton, C. L.; Chalk, P. A.; Kleanthous, C. Biochem.
J . 1996, 319, 559.
(9) White, P. J .; Young, J .; Hunter, I. S.; Nimmo, H. G.; Coggins, J . R.
Biochem. J . 1990, 265, 735.
(10) Smith, B. W.; Turner, M. J .; Haslam, E. J . Chem. Soc., Chem.
Commun. 1970, 842.
(11) Shneier, A.; Kleanthous, C.; Deka, R.; Coggins, J . R.; Abell, C. J .
Am. Chem. Soc. 1991, 113, 9416.
(13) Harris, J .; Kleanthous, C.; Coggins, J . R.; Hawkins, A. R.; Abell, C.
J . Chem. Soc., Chem. Commun. 1993, 13, 1080.
(14) Manthey, M. K.; Gonza´lez-Bello, C.; Abell, C. J . Chem. Soc., Perkin
Trans. 1 1997, 625.
(15) Bartlett, P. A.; Maitra, U.; Chouinard, P. M. J . Am. Chem. Soc. 1986,
108, 8068.
(16) Montchamp, J .-L.; Frost, J . W. J . Am. Chem. Soc. 1997, 119, 7645.
(17) All assays were performed at 25 °C in quartz cuvettes (total reaction
volume 1 mL) buffered at pH 7.0 (type I, 50 mM phosphate; type II, 50 mM
Tris-HCl) and were monitored spectrophotometrically at 234 nm for the
appearance of dehydroshikimate.
(12) Harris, J . M.; Gonza´lez-Bello, C.; Kleanthous, C.; Hawkins, A. R.;
Coggins, J . R.; Abell, C. Biochem. J . 1996, 319, 333.
10.1021/jo990004q CCC: $18.00 © 1999 American Chemical Society
Published on Web 03/26/1999

Communications
J . Org. Chem., Vol. 64, No. 8, 1999 2613
Ta ble 1. In h ibition Con sta n ts (µM) for In h ibitor s a ga in st Typ e I a n d Typ e II Deh yd r oqu in a ses
a
enzyme
Km
Ki for 10
Ki for 11
Ki for 13
Ki for 19
type I S. typhimurium
type II A. nidulans
type II M. tuberculosis
type II S. coelicolor
1819
1503
647
3000 ( 1 000
60 ( 10
200 ( 20
30 ( 10
4500 ( 500
1500 ( 200
1200 ( 200
600 ( 200
>25 000
>25 000
15 ( 2
2 200 ( 100
700 ( 200
2 500 ( 500
20 ( 2
6509
500 ( 200
a
The Km values (µM) were measured for the respective enzymes with dehydroquinic acid 1.
Sch em e 1
Sch em e 3a
a
Reagents and conditions: (i) TFA, H2O, 60 °C, 2 h; (ii) NaOH, H2O,
20 °C, 30 min; (iii) Amberlite IR-120 (H), H2O, 20 °C, 10 min; (iv)
NaOMe, MeOH, 20 °C, 1 h; (v) HONH3Cl, NaOAc, MeCN, H2O, 20 °C,
16 h; (vi) Bu4NF, THF, 20 °C, 16 h; (viii) NaOH, H2O, 20 °C, 10 min.
Sch em e 2a
and 11 for the compound with sp2 geometry at both C-2 and
C-3 is consistent with stabilization by the enzyme of a
transition state that has some degree of ring flattening.
Observations concerning similar geometric selectivity for the
enzyme 3-dehydroquinate synthase have been reported
recently.16 Likewise, the positioning of the oxime group into
the pocket on the enzyme thought to stabilize the developing
negative charge in the enolate intermediate led to signifi-
cantly higher affinity for 19 than for the exo-methylene
compound 13.
This is the first report of inhibitors of type II dehydro-
quinase. Furthermore, the inhibition constants for 19 against
the A. nidulans enzyme and for 10 against the S. coelicolor
enzyme are over 1 order of magnitude lower than the
corresponding Km values for the substrate dehydroquinate.
This is an encouragingly high affinity for the first generation
of inhibitors but is not as high as might be expected for true
transition-state mimics. The results suggest that compounds
combining the separate strategies of flattening the ring and
having a hydrogen-bonding capability at C-3 should be
interesting targets.
a
Reagents and conditions: (i) PhCHO, TsOH‚H2O, PhMe, 110 °C,
16 h; (ii) NBS, AIBN, C6H6, 80 °C, 1 h; (iii) TBDMSCl, DBU, MeCN,
80 °C, 16 h; (iv) Bu4F, THF, 20 °C, 30 min; (v) NaOMe, MeOH, 20 °C,
2 hg; (vi) NaOH, H2O, 20 °C, 30 min; (vii) Amberlite IR-120 (H), H2O,
20 °C, 10 min; (viii) H2, Pt, H2O, 20 °C, 16 h.
All the compounds were poor inhibitors against the type
I enzyme. The exo-methylene compound 13 and the oxime
19 were particularly poor, presumably due to steric interac-
tions with the C-3 substituent, possibly from the active-site
lysine involved in imine formation. 2,3-Anhydroquinic acid
10 was a reasonably good inhibitor against all three type II
enzymes, especially that from Streptomyces. Reduction of the
2,3-double bond led to a marked loss in affinity for 11. The
two inhibitors functionalized at C-3 showed contrasting
inhibition against the type II enzymes. The exocyclic meth-
ylene compound 13 had approximately millimolar affinity
for the type II dehydroquinases. In comparison, the oxime
19 was a particularly potent inhibitor of the A. nidulans and
M. tuberculosis enzymes, but was surprisingly much less
potent against the Streptomyces enzyme.18
Ack n ow led gm en t. We thank J ohn Green for as-
sistance in enzyme preparation, Dr. Finian Leeper for
proofreading the manuscript, and the BBSRC for funding.
Su p p or tin g In for m a tion Ava ila ble: Experimental data for
all new compounds and inhibition data.
J O990004Q
(18) Compound 19 is not a substrate for type II dehydroquinase, nor was
deuterium exchange into C-2 observed when it was incubated with the type
II enzyme from S. coelicolor or M. tuberculosis.
(19) Moore, J . D.; Hawkins, A. R.; Charles, I. G.; Deka, R.; Coggins, J .
R.; Cooper, A.; Kelly, S. M.; Price, N. C. Biochem. J . 1993, 295, 277.
These results broadly supported our initial assumptions.
The selectivity shown by the type II enzymes between 10