Design, synthesis, and evaluation of 9-D-ribityl-1,3,7-trihydro-2,6,8-purinetrione, a potent inhibitor of riboflavin synthase and lumazine synthase

Article abstract of DOI:10.1021/jo010706r

Reduction of 5-nitro-6-D-ribitylaminouracil (9) afforded 5-amino-6-D-ribitylaminouracil (1), which reacted with ethyl chloroformate to yield 5-ethylcarbamoyl-6-D-ribitylaminouracil (12). The latter compound was cyclized to 9-D-ribityl-1,3,7-trihydropurine-2,6,8-trione (13), which was found to be a relatively potent inhibitor of both Escherichia coli riboflavin-synthase (K_i 0.61 μM) and Bacillus subtilis lumazine synthase (K_i 46 μM). Molecular modeling of the lumazine synthase-inhibitor complex indicated the possibility for hydrogen bonding between the Lys135 ε-amino group of the enzyme and both the 8-keto group and the 4′-hydroxyl group of the ligand. A bisubstrate analogue of the riboflavin synthase-catalyzed reaction, 1,4-bis[1-(9-D-ribityl-1,3,7-trihydropurine-2,6,8-trionyl)]-butane (18), was also synthesized using a similar route and was found to be inactive as an inhibitor of both riboflavin synthase and lumazine synthase.

Full text of DOI:10.1021/jo010706r

8320
J . Org. Chem. 2001, 66, 8320-8327
Design , Syn th esis, a n d Eva lu a tion of
9-D-Ribityl-1,3,7-tr ih yd r o-2,6,8-p u r in etr ion e, a P oten t In h ibitor of
Ribofla vin Syn th a se a n d Lu m a zin e Syn th a se
Mark Cushman,*,† Donglai Yang,† Klaus Kis,‡ and Adelbert Bacher‡
Department of Medicinal Chemistry and Molecular Pharmacology, School of Pharmacy and
Pharmacal Sciences, Purdue University, West Lafayette, Indiana 47907, and Lehrstuhl fu¨r
Organische Chemie und Biochemie, Technische Universita¨t Mu¨nchen, D-85747 Garching, Germany
cushman@pharmacy.purdue.edu
Received J uly 13, 2001
Reduction of 5-nitro-6-D-ribitylaminouracil (9) afforded 5-amino-6-D-ribitylaminouracil (1), which
reacted with ethyl chloroformate to yield 5-ethylcarbamoyl-6-D-ribitylaminouracil (12). The latter
compound was cyclized to 9-D-ribityl-1,3,7-trihydropurine-2,6,8-trione (13), which was found to be
a relatively potent inhibitor of both Escherichia coli riboflavin synthase (Ki 0.61 µM) and Bacillus
subtilis lumazine synthase (Ki 46 µM). Molecular modeling of the lumazine synthase-inhibitor
complex indicated the possibility for hydrogen bonding between the Lys135 ꢀ-amino group of the
enzyme and both the 8-keto group and the 4′-hydroxyl group of the ligand. A bisubstrate analogue
of the riboflavin synthase-catalyzed reaction, 1,4-bis[1-(9-D-ribityl-1,3,7-trihydropurine-2,6,8-trionyl)]-
butane (18), was also synthesized using a similar route and was found to be inactive as an inhibitor
of both riboflavin synthase and lumazine synthase.
Sch em e 1
In tr od u ction
Riboflavin synthase catalyzes an unusual dismutation
reaction in which a four-carbon unit is transferred from
one molecule of 6,7-dimethyl-8-D-ribityllumazine (3),
bound at the donor site of the enzyme, to a second
molecule of 3, bound at the acceptor site of the enzyme,
resulting in the formation of one molecule of riboflavin
(4) and one molecule of the pyrimidinedione 1 (Scheme
1).1-5 In an efficient recycling reaction, lumazine synthase
catalyzes the condensation of the pyrimidinedione 1 with
the four-carbon unit 2, resulting in the substrate 3 used
by riboflavin synthase.6
A plausible mechanistic pathway for the lumazine
synthase-catalyzed reaction involves an initial Schiff base
formation between the ketone 2 and the primary amine
1, resulting in the imine 5, which eliminates phosphate
to form the enol 6 (Scheme 2). Tautomerization of the
enol 6 to the ketone 7, followed by the addition of the
secondary amino group to the ketone, leads to the
formation of the covalently hydrated lumazine 8. Elimi-
nation of water yields the observed product 3.
The X-ray structure of a reconstituted, icosahedral â60
Bacillus subtilis lumazine synthase capsid complexed
* Ph: 765-494-1465. Fax: 765-494-6790.
† Purdue University.
‡ Technische Universita¨t Mu¨nchen.
(1) Plaut, G. W. E.; Smith, C. M.; Alworth, W. L. Annu. Rev.
Biochem. 1974, 43, 899-922.
(2) Plaut, G. W. E. In Comprehensive Biochemistry; Florkin, M.,
Stotz, E. H., Eds.; Elsevier: Amsterdam, 1971; Vol. 21, pp 11-45.
(3) Beach, R. L.; Plaut, G. W. E. J . Am. Chem. Soc. 1970, 92, 2913-
2916.
(4) Bacher, A.; Eberhardt, S.; Richter, G. In Escherichia coli and
Salmonella: Cellular and Molecular Biology, 2nd ed.; Neidhardt, F.
C., Ed.; ASM Press: Washington, DC, 1996; pp 657-664.
(5) Bacher, A.; Fischer, M.; Kis, K.; Kugelbrey, K.; Mo¨rtl, S.;
Scheuring, J .; Weinkauf, S.; Eberhardt, S.; Schmidt-Ba¨se, K.; Huber,
R.; Ritsert, K.; Cushman, M.; Ladenstein, R. Biochem. Soc. Trans.
1996, 24, 89-94.
with the substrate analogue 9 has allowed the construc-
tion of a hypothetical model of the phosphate 5 bound in
the active site.7 The model has also been substantiated
(6) Volk, R.; Bacher, A. J . Am. Chem. Soc. 1988, 110, 3651-3653.
10.1021/jo010706r CCC: $20.00 © 2001 American Chemical Society
Published on Web 11/09/2001

9-D-Ribityl-1,3,7-trihydro-2,6,8-purinetrione
J . Org. Chem., Vol. 66, No. 25, 2001 8321
Sch em e 2
shown that riboflavin-deficient mutants of both of these
bacteria require riboflavin concentrations well above 10
mg/L of culture medium. Similar concentrations of ribo-
flavin are also required to support the growth of riboflavin-
deficient mutants of the yeasts Candida guilliermondii
and S. cerevisiae.12-14 These observations support the idea
that the enzymes involved in riboflavin biosynthesis are
rational targets for the development of antibiotics to treat
Gram-negative bacterial and yeast infections.
A collaborative effort between our research groups has
resulted in the design and synthesis of a number of
riboflavin synthase and lumazine synthase inhibitors.8,15-20
These included 2,6-dioxo-(1H,3H)-9-N-ribitylpurine (10)
and 2,6-dioxo-(1H,3H)-8-aza-9-N-ribitylpurine (11).20 When
tested as an inhibitor of riboflavin synthase, the purine
10 displayed a Ki of 540 µM, and the azapurine 11
resulted in a Ki of 390 µM. On the other hand, versus
lumazine synthase, the purine 10 had a Ki of 470 µM and
the azapurine produced a Ki of 330 µM. In considering
the relative activities of 10 and 11 versus lumazine
synthase, it was proposed that the greater potency of the
azapurine 11 might possibly reflect an electrostatic and/
or hydrogen bonding interaction between the side chain
amino group of Lys135 of the enzyme and the N-8
nitrogen of 11, which would not be possible with 10. This
interpretation has led to the consideration of other
systems that might more effectively target Lys135,
including 9-D-ribityl-1,3,7-trihydropurine-2,6,8-trione (13).
On the basis of molecular modeling, the 8-keto group of
13 seemed to be ideally positioned to hydrogen bond with
the side chain amino group of Lys135.
by a crystal structure of a phosphonate analogue of 5
bound to Saccharomyces cerevisiae lumazine synthase.8,9
The placement of the phosphate group of 5 in the model
of the complex was accomplished by overlapping it with
an inorganic phosphate that had been incorporated from
the solvent during crystallization of the complex of 9 with
the enzyme for X-ray structure determination. The
geometry of the enzyme-bound imine in 5 is not estab-
lished, and both the E- and the Z-imines can be modeled
in the active site of lumazine synthase reasonably well,
with the phosphate and pyrimidine ring overlapping with
the pyrimidine ring an inorganic phosphate present in
the crystal structure of bound 9. However, a Z-imine is
required in intermediate 7 in order for cyclization to
occur. The details of the conformational change of the
side chain in the conversion of 5 (or its Z-isomer) to 7
have not been defined. Since the phosphate is not near
the amino group, an extensive conformational reorgani-
zation of the side chain must occur after phosphate
elimination from 5 in order to allow the ketone to be
brought into close proximity with the amine so the
cyclization can occur.
Resu lts a n d Discu ssion
The synthesis of the desired compound 13 is outlined
in Scheme 3. Catalytic hydrogenation of 5-nitro-6-ribi-
tylaminouracil (9)21 over palladium on charcoal afforded
(10) Bandrin, S. V.; Beburov, M. Y.; Rabinovich, P. M.; Stepanov,
A. I. Genetika 1979, 15, 2063-2065.
(11) Wang, A. I Chuan Hsueh Pao 1992, 19, 362-368.
(12) Oltmanns, O.; Lingens, F. Z. Naturforschung 1967, 22, 751-
754.
(13) Logvinenko, E. M.; Shavlovsky, G. M. Mikrobiologiya 1967, 41,
978-979.
(14) Roberts, J . C.; Gao, H.; Gopalsami, A.; Kongsjahju, A.; Patch,
R. J . Tetrahedron Lett. 1997, 38, 355-358.
Certain Gram-negative bacteria, including Escherichia
coli and Salmonella typhimurium, lack an efficient
riboflavin uptake system and are therefore dependent on
endogenous synthesis.10,11 More specifically, it has been
(15) Cushman, M.; Patrick, D. A.; Bacher, A.; Scheuring, J . J . Org.
Chem. 1991, 56, 4603-4608.
(16) Cushman, M.; Patel, H. H.; Scheuring, J .; Bacher, A. J . Org.
Chem. 1992, 57, 5630-5643.
(17) Cushman, M.; Patel, H. H.; Scheuring, J .; Bacher, A. J . Org.
Chem. 1993, 58, 4033-4042.
(7) Ritsert, K.; Huber, R.; Turk, D.; Ladenstein, R.; Schmidt-Ba¨se,
K.; Bacher, A. J . Mol. Biol. 1995, 253, 151-167.
(18) Cushman, M.; Mavandadi, F.; Yang, D.; Kugelbrey, K.; Kis, K.;
Bacher, A. J . Org. Chem. 1999, 64, 4635-4642.
(19) Cushman, M.; Mavandadi, F.; Kugelbrey, K.; Bacher, A. J . Org.
Chem. 1997, 62, 8944-8947.
(8) Cushman, M.; Mihalic, J . T.; Kis, K.; Bacher, A. J . Org. Chem.
1999, 64, 3838-3845.
(9) Meining, W.; Mo¨rtl, S.; Fischer, M.; Cushman, M.; Bacher, A.;
Ladenstein, R. J . Mol. Biol. 2000, 299, 181-197.
(20) Cushman, M.; Mavandadi, F.; Kugelbrey, K.; Bacher, A. Bioorg.
Med. Chem. 1998, 6, 409-415.

8322 J . Org. Chem., Vol. 66, No. 25, 2001
Cushman et al.
Sch em e 3a
four carbon atoms (structure 14, n ) 4). This suggests
that a similar bis(purinetrione) might also be an effective
inhibitor of riboflavin synthase. Accordingly, the desired
1,4-bis[1-(9-D-ribityl-1,3,7-trihydropurine-2,6,8-trionyl)]-
butane (18) was synthesized.
The synthesis of 18 proceeded through intermediates
15, 16, and 17, and was carried out using the methodol-
a
Reagents and conditions: (a) ClCOOEt, Et3N, aq CH3CN, 23
°C (10 h); (b) NaOEt, EtOH, reflux (24 h).
the unstable amino compound 1.15 Since intermediate 1
is extremely sensitive to oxidation, leading to a multitude
of products, it was treated in situ in the hydrogenation
reaction mixture with ethyl chloroformate. After the
reaction was complete, the catalyst was removed by
filtration and the desired 5-ethylcarbamoyl-6-ribitylami-
nouracil (12) was isolated from the filtrate. Cyclization
of 12 was then carried out with sodium ethoxide in
refluxing ethanol to afford the desired product 13.
The dismutation reaction catalyzed by riboflavin syn-
thase requires the presence of two substrate molecules,
or derivatives thereof, to be bound in the active site of
the enzyme at the same time.3,22,23 In an attempt to
prepare bisubstrate inhibitors of riboflavin synthase, a
series of bis(lumazines) of general structure 14 were
ogy outlined for the monomer 13 in Scheme 3. In contrast
to the riboflavin synthase inhibitory activity seen with
the bis(lumazine) 14 (n ) 4) (Ki ) 37 µM), the bis-
(purinetrione) 18 proved to be inactive as a riboflavin
therefore previously prepared and tested as inhibitors of
riboflavin synthase.18 Maximal activity was found in the
bis(lumazine) in which the linker chain connecting the
two substrate molecules was a straight chain containing
(21) Cresswell, R. M.; Wood, H. S. C. J . Chem. Soc. 1960, 4768-
4775.
(22) Otto, M.; Bacher, A. Eur. J . Biochem. 1981, 115, 511-517.
(23) Scheuring, J .; Fischer, M.; Cushman, M.; Lee, J .; Bacher, A.;
Oschkinat, H. Biochemistry 1996, 35, 9637-9646.

9-D-Ribityl-1,3,7-trihydro-2,6,8-purinetrione
J . Org. Chem., Vol. 66, No. 25, 2001 8323
synthase inhibitor, and it was also inactive versus
lumazine synthase (Table 1). Evidently, the bisubstrate
inhibitor approach requires lumazines very close in
structure to the natural substrate to be effective with
riboflavin synthase.
The purinetrione 13 was tested for inhibition of lu-
mazine synthase â60 capsids and recombinant riboflavin
synthase from E. coli. The resulting Lineweaver-Burk
plots are shown in Figure 1, and the inhibition constants
and types of inhibition displayed by 13 are listed in Table
1, together with the results from previously synthesized
riboflavin synthase and lumazine synthase inhibitors.8,15-20
lished, and it might be possible to explain the high
potency of 13 versus riboflavin synthase in structural
terms when the coordinates become available.24 A great
deal is also known about the structure of lumazine
synthase as a result of the availability of a 2.4 Å
resolution X-ray structure of the substrate analogue 9
complexed with B. subtilis lumazine synthase.7 In addi-
tion, the related 1.85 Å resolution X-ray structure of the
phosphonate inhibitor 31b8 (Table 1) (Ki 180 µM versus
B. subtilis lumazine synthase) bound to S. cerevisiae
lumazine synthase has also been published,9 and its
homologue 31c has been docked in the active site of B.
subtilis lumazine synthase by computer graphics molec-
ular modeling.8 The modeled structure of 31c in complex
with the enzyme was in close agreement with that of 31b,
which was later determined by X-ray crystallography.8,9
These structures might allow a possible explanation of
the high potency of 13 versus lumazine synthase. Ac-
cordingly, a molecular model was constructed by overlap-
ping the structure of 13 with the structure of 9 in the
active site of B. subtilis lumazine synthase. The structure
of 9 was then removed, and the energy of the complex
was minimized using the MMRR94 force field while
allowing the ligand and the hydrated protein structure
contained within a 6 Å sphere surrounding the ligand to
remain flexible with the remainder of the protein struc-
ture frozen. The 6 Å flexible sphere surrounding the
ligand was employed in order to allow for an induced fit
of the ligand to the protein. The resulting Figure 2 was
constructed by displaying the amino acid residues and
one water molecule calculated to be involved in hydrogen
bonding with the ligand 13.
Different kinetic models were considered, and the most
likely inhibition mechanisms for 13 were found to be
partial inhibition for lumazine synthase and strictly
competitive inhibition for riboflavin synthase. The Ki of
0.61 µM observed for 3-D-ribityl-2,6,8-purinetrione (13)
makes it the most potent E. coli riboflavin synthase
inhibitor ever reported. Its Ki of 46 µM versus the B.
subtilis lumazine synthase, along with that of 43 µM
presently determined for the synthetic intermediate 9,
rate these two compounds as the most potent lumazine
synthase inhibitors described to date.
According to the model in Figure 2, the 8-keto group
of 13 can hydrogen bond with the ꢀ-amino group of
Lys135, which positions the 4′-hydroxyl group of the
The crystal structure of E. coli riboflavin synthase
containing selenomethionine residues has been pub-
(24) Liao, D.-I.; Wawrzak, Z.; Calabrese, J . C.; Viitanen, P. V.;
J ordan, D. B. Structure 2001, 9, 399-408.

8324 J . Org. Chem., Vol. 66, No. 25, 2001
Cushman et al.
Ta ble 1. In h ibition Con sta n ts ver su s Lu m a zin e Syn th a se a n d Ribofla vin Syn th a se
Ki (µM)
compound
lumazine synthasea
riboflavin synthaseb
4
9
200
120
43.4 ( 4.1 (Ki, partial inhibition)
40 ( 2.8 (Ks)
250 ( 47 (Kis)c
79.2 ( 2 (kcat)
16 ( 33 (kcat′)
470c
10
11
12
13
540c
390c
>1000
330c
>1000
46 ( 5 (Ki, partial inhibition)
0.61 ( 0.05 (Ki, competitive inhibition)
9.6 ( 0.9 (Ks)
250 ( 42 (Kis)d
39.9 ( 0.7 (kcat)
14 (n ) 3)
14 (n ) 4)
840 ( 490 (Ki, partial inhibition)e
2200 ( 660 (Kis)e
320 ( 470 (Ki, partial inhibition)e
300 ( 30 (Kis)e
>1000e
37 ( 30 (Ki, mixed inhibition)f
57 ( 12 (Kis)e
14 (n ) 5)
15
15a
15b
>1000e
3700e
>1000e
3000e
280e
5000e
1500e
670e
17
18
>1000
>1000
>1000
>1000
19 (epimer A)
19 (epimer A)
120 (pH 7.4)g
38 (pH 6.8)g
20
21
22
23
24
25
26
27
28
29
30
31a
55g
75g
58h
70h
17h
17h
15
20h
200i
360i
430i
440 ( 200 (Ki, mixed inhibition)j
640 ( 300 (Kis)j
>1000j
31b
31c
180 ( 88 (Ki, mixed inhibition)j
350 ( 22 (Kis)j
>1000j
>1000j
130 ( 33 (Ki, mixed inhibition)j
140 ( 15 (Kis)j
32a
32b
290 ( 120 (Ki, competitive inhibition)j
690 ( 290 (Ki, mixed inhibition)j
1500 ( 640 (Kis)j
>1000j
>1000j
32c
290 ( 130 (Ki, mixed inhibition)j
580 ( 170 (Kis)j
Recombinant â60 capsids from B. subtilis. b Recombinant riboflavin synthase from E. coli. c Reference 20. Kis is the equilibrium constant
a
d
for the reaction EI + S h EIS. e Reference 18. f Strictly speaking, the data observed for bis(lumazine) 14 (n ) 4) cannot be interpreted in
terms of any simple inhibition type, but it is most consistent with mixed type inhibition. Reference 16. Reference 17. Reference 19.
Reference 8.
g
h
i
j
ligand to also accept a hydrogen bond from ꢀ-amino group
of Lys135. Similar contacts are not made when the nitro
compound 9 binds to the enzyme.7 The Kd for compound
9 versus the lumazine synthase of B. subtilis was
previously estimated to be 650 nM.25 When tested as an
inhibitor of B. subtilis lumazine synthase for direct
comparison with 13, compound 9 produced a Ki of 43 µM.
A comparison of the computer-generated hydrogen bond-
ing patterns calculated for the binding of 13 to lumazine
synthase with that resulting from the X-ray structure of
9 bound to lumazine synthase is shown in Figure 3.
Although the inhibitor 9 does not hydrogen bond to
Lys135, the hydrogen bonds between its 5′-hydroxyl
group and Phe113, as well as its 2′-hydroxyl group and
Phe57, are within the normal range for hydrogen bonding
(N to O distances 3.1 and 2.9 Å, respectively), as
compared to the corresponding distances calculated for
13 of 4.1 and 4.1 Å, which are outside of the normal
hydrogen bonding lengths. These factors seem to balance
out, resulting in similar enzyme inhibitory potencies.
Another difference between the two structures is that
Glu58 in 9 hydrogen bonds with the 3′-hydroxyl group
of the ligand, whereas in 13, it is calculated to hydrogen
bond with the 2′-hydroxyl group.
In contrast to the situation with lumazine synthase,
the purinetrione 13 (Ki 0.61 µM) is a much more potent
inhibitor of E. coli riboflavin synthase than the nitropy-
rimidine 9 (Ki 120 µM).20 These differences in riboflavin
synthase inhibitory activity can possibly be rationalized
in the future when the crystal coordinates of riboflavin
synthase are published.24
To place the activity of the purinetrione 13 in perspec-
tive, a complete listing of the inhibitory activities of all
of the known B. subtilis lumazine synthase inhibitors and
E. coli riboflavin synthase inhibitors has been provided
in Table 1.8,15-20 The Ki of 0.61 µM observed for 13 versus
(25) Bacher, A.; Ludwig, H. C. Eur. J . Biochem. 1982, 127, 539-
545.

9-D-Ribityl-1,3,7-trihydro-2,6,8-purinetrione
J . Org. Chem., Vol. 66, No. 25, 2001 8325
N-8 of 11.20 According to this line of reasoning, the
greater potency of 13 versus 11 might reflect more
favorable hydrogen bonding between the 8-keto group of
13 and the amino group of Lys135, which could also
position the Lys135 amino group to participate in a
second hydrogen bond with the oxygen of the 4′-hydroxyl
group of the ribityl side chain (Figure 2).
It seems logical to suspect that an inhibitor of both
lumazine synthase and riboflavin synthase would be a
good candidate for antibiotic drug development, since
resistance would be less likely to emerge against an agent
that is active against more than one enzyme in the
biosynthetic pathway.
Exp er im en ta l Section
5-Eth ylca r ba m oyl-6-D-r ibityla m in ou r a cil (12). A mix-
ture of palladium on charcoal (10%, 0.3 g) in water (50 mL)
was stirred in a hydrogen atmosphere for 5 min. 5-Nitro-6-D-
ribitylaminouracil (9)21 (1.0 g, 3.26 mmol) was added, and the
solution was hydrogenated at room temperature and atmo-
spheric pressure for 12 h to afford intermediate 1. The reaction
mixture was cooled to 0 °C, triethylamine (5 mL), acetonitrile
(25 mL), and ethyl chloroformate (3.0 mL, 31 mmol) were
added, and the mixture was stirred for 10 h. The catalyst was
removed by filtration, and the filtrate was concentrated under
reduced pressure. The residue was applied to a cation-
exchange column (Dowex 50 × 2-200, 30 g) and eluted with
water (200 mL). The fractions were then applied to an anion-
exchange column (Dowex 1 × 2-400) and eluted with water
and 10% aq HCOOH. After concentration of the combined
fractions, the residue was triturated with EtOAc to generate
a precipitate, which was then collected by filtration to yield
12 (0.8 g, 70%) as an amorphous solid: 1H NMR (300 MHz,
D2O) δ 4.08 (q, J ) 7 Hz, 2 H), 3.81 (m, 1 H), 3.69 (m, 2 H),
3.59-3.45 (m, 3 H), 3.39 (m, 1 H), 1.17 (t, J ) 7 Hz, 3 H).
Anal. Calcd for C12H20N4O8‚1.0(H2O)‚0.2(HCO2H): C, 39.41;
H, 5.97; N, 14.82. Found: C, 39.33; H, 5.84; N, 14.78.
9-D-Ribityl-1,3,7-tr ih ydr opu r in e-2,6,8-tr ion e (13). A mix-
ture of 5-ethylcarbamoyl-6-D-ribitylaminouracil (12) (0.20 g,
0.57 mmol) and NaOEt (0.10 g, 1.47 mmol) in ethanol (50 mL)
was heated at reflux for 24 h. After removal of the solvent
under reduced pressure, the residue was purified by anion-
exchange resin column chromatography (Dowex 1 × 2-400,
eluted with distilled water followed by 10% HCOOH) to afford
13 (120 mg, 70%) as an amorphous solid: [R]20D +3.7° (DMSO);
1H NMR (300 MHz, DMSO-d6) δ 11.50 (s, 1 H), 10.86 (s, 1 H),
10.79 (s, 1 H), 5.10-4.00 (br m, 4 H), 3.89-3.4 (m, 7 H). Anal.
Calcd for C10H14N4O7‚2.0(H2O)‚0.5(HCOOH): C, 34.91; H, 5.30;
N, 15.51. Found: C, 34.67; H, 5.05; N, 15.50.
1,4-Bis[3-(5-eth ylca r ba m oyl-6-D-r ibityla m in ou r a cilyl)]-
bu ta n e (17). A suspension of palladium on charcoal (10%, 0.3
g) in water (20 mL) was stirred in a hydrogen atmosphere for
5 min. 1,4-Bis[3-(5-nitro-6-ribitylaminouracilyl)butane (15)18
(0.8 g, 1.2 mmol) and triethylamine (1 mL) were added, and
the solution was hydrogenated at room temperature and
atmospheric pressure for 12 h to afford intermediate 16. The
reaction mixture was cooled to 0 °C, triethylamine (1 mL),
acetonitrile (5 mL), and ethyl chloroformate (1.0 mL, 11 mmol)
were added, and the mixture was stirred for 10 h at room
temperature. The catalyst was removed by filtration, and the
filtrate was concentrated under reduced pressure. The residue
was triturated with ethyl acetate to give a white precipitate,
which was collected by filtration. It was dissolved in H2O (5
mL), applied on a cation-exchange column (Dowex 50 × 2-200,
20 g), and eluted with water (100 mL). After concentration of
the combined fractions, the residue was triturated with H2O/
EtOH, and then collected by filtration to provide 17 (0.55 g,
61%) as an amorphous solid: 1H NMR (D2O) δ 4.08 (q, J ) 7
Hz, 4 H), 3.82 (m, 2 H), 3.7-3.5 (m, 14 H), 3.35 (m, 2 H), 1.48
(m, 4 H), 1.17 (t, J ) 7 Hz, 6 H).
F igu r e 1. Linweaver-Burk plots of the inhibition of lumazine
synthase and riboflavin synthase by compound 13. (A) Lu-
mazine synthase. The concentration of the substrate 2 ranges
from 17 to 166 µM. Inhibitor concentrations: 0, 862; (, 431;
b, 259; 9, 172; O, 86; 4, 0 µM. (B) Riboflavin synthase. The
concentration of the substrate 3 ranged from 17 to 166 µM.
Inhibitor concentrations: 0, 40; (, 20; b, 12; 9, 8; O, 4; 4, 0
µM. The initial velocities of product formation, v, were
determined at steady-state conditions with varying amounts
of substrate, s, and inhibitor. The kinetic data were fitted with
a nonlinear regression method using the program DynaFit
from P. Kuzmic.27 Different kinetic models were considered.
The most likely inhibition mechanisms found were partial
inhibition for lumazine synthase and strictly competitive
inhibition for riboflavin synthase.
riboflavin synthase is significantly lower than that
observed for the nearest-ranking compound, the D,L-
dithiothreitol adduct 26 (Ki 15 µM). Likewise, both 9
(Ki ) 43 µM) and 13 (Ki ) 46 µM) are significantly more
potent as lumazine synthase inhibitors than the next-
ranking compound, the phosphonate 31c (Ki ) 130 µM).
It is interesting to compare the activities of the
ribitylxanthine 10 (Ki 470 µM), the corresponding ribi-
tylazapurine 11 (Ki 330 µM), and the ribitylpurinetrione
13 (Ki 46 µM) versus lumazine synthase. We had previ-
ously proposed that the slightly more potent activity of
11 in comparison to 10 might reflect the possible elec-
trostatic interaction between the positively charged
Lys135 primary ammonium ion and the N-8 nitrogen of
11, as well as a possible hydrogen bond between the
Lys135 amino group and the nonbonded electron pair on
1,4-Bis[1-(9-D-r ibityl-1,3,7-tr ih ydr opu r in e-2,6,8-tr ion yl)]-
bu ta n e (18). Sodium (0.1 g, 4.3 mmol) was dissolved in

8326 J . Org. Chem., Vol. 66, No. 25, 2001
Cushman et al.
F igu r e 2. Hypothetical model for the binding of compound 13 to B. subtilis lumazine synthase. The figure is programmed for
walleyed stereoviewing.
Calcd for C24H34N8O14‚2.6(H2O): C, 40.86; H, 5.60; N, 15.88.
Found: C, 40.86; H, 5.32; N, 15.62.
Molecu la r Mod elin g. Using Sybyl (Tripos, Inc., version
6.5, 1998), the X-ray crystal structure of the complex of 5-nitro-
6-ribitylamino-2,4-(1H,3H)pyrimidinedione (9) and the heavy
riboflavin synthase of B. subtilis (1RVV)7 was clipped to
include information within a 15 Å sphere of one of the 60
equivalent ligand molecules. The residues that were clipped
in this cut complex were capped with either neutral amino or
carboxyl groups. The structure of the inhibitor 13 was over-
lapped with the structure of 5-nitro-6-ribitylamino-2,4-(1H,3H)-
pyrimidinedione (9), which was then deleted. Hydrogen atoms
were added to the complex. MMFF94 charges were loaded, and
the energy of the complex was minimized using the Powell
method to a termination gradient of 0.05 kcal/mol while
employing the MMFF94 force field. During the minimization
of the complex, inhibitor 13 and a 6 Å sphere surrounding it
were allowed to remain flexible, while the remaining portion
of the complex was held ridged using the aggregate function.
Figure 2 was constructed by displaying the amino acid residues
of the enzyme and one water molecule that are calculated to
be involved in hydrogen bonding with the inhibitor 13. The
maximum distance of the hydrogen bonds shown in Figure 2
was set at 3.4 Å.
Lu m a zin e Syn th a se Assa y.26 Reaction mixtures contained
100 mM potassium phosphate, pH 7.0, 5 mM EDTA, 5 mM
dithiothreitol, inhibitor (0-862 µM), 153 µM 5-amino-6-ribi-
tylamino-2,4(1H,3H)-pyrimidinedione (1), and recombinant B.
subtilis lumazine synthase capsids (30 µg, specific activity 12.5
µmol mg-1 h-1) in a total volume of 560 µL. The solution was
incubated at 37 °C, and the reaction was started by the
addition of a small volume (20 µL) of L-3,4-dihydroxy-2-
butanone 4-phosphate to a final concentration of 17-166 µM.
F igu r e 3. (Top) Hydrogen bonds and distances in the
calculated model of the inhibitor 13 bound in the active site
of B. subtilis lumazine synthase. (Bottom) The corresponding
diagram based on the X-ray structure of the complex of 9 with
The velocity/substrate data were fitted for all inhibitor con-
centrations with a nonlinear regression method using the
program DynaFit.27 Different inhibition models were consid-
ered for the calculation. Ki values ( standard deviations were
obtained from the fit under consideration of the most likely
the enzyme.
inhibition model.
ethanol (60 mL). 1,4-Bis[3-(5-ethylcarbamoyl-6-D-ribitylamino-
uracilyl)]butane 17 (0.10 g, 0.13 mmol) was added to the
Ribofla vin Syn th a se Assa y.28 Reaction mixtures con-
tained buffer (100 mM potassium phosphate, 10 mM EDTA,
10 mM sodium sulfite), inhibitor (0-40 µM), and E. coli
solution, and the reaction mixture was heated at reflux for 48
h. After cooling, the precipitate was collected by filtration and
then purified by anion-exchange resin column chromatography
(Dowex 1 × 2-400, eluted with distilled water followed by 10%
HCOOH) to afford 18 (50 mg, 58%) as an amorphous solid:
1H NMR (300 MHz, DMSO-d6) δ 11.8 (s, 2 H), 10.86 (s, 2 H),
5.5-4.2 (br m, 8 H), 3.85-3.0 (m, 18 H), 1.5 (m, 4 H). Anal.
(26) Kis, K.; Bacher, A. J . Biol. Chem. 1995, 270, 16788-16795.
(27) Kuzmic, P. Anal. Biochem. 1996, 237, 260-273.
(28) Eberhardt, S.; Richter, G.; Gimbel, W.; Werner, T.; Bacher, A.
Eur. J . Biochem. 1996, 242, 712-718.

9-D-Ribityl-1,3,7-trihydro-2,6,8-purinetrione
J . Org. Chem., Vol. 66, No. 25, 2001 8327
riboflavin synthase (10 µg, specific activity 50 µmol mg-1 h-1).
After preincubation, the reactions were started by the addition
of various amounts of 6,7-dimethyl-8-ribityllumazine (3) (16-
166 µM) to a total volume of 620 µL. The formation of riboflavin
(4) was measured online with a computer-controlled photom-
eter at 470 nm (ꢀriboflavin ) 9100 M-1 cm-1). The Ki evaluation
was performed in the same manner as descibed above.
Ack n ow led gm en t. This research was made possible
by NIH grant GM51469 and by support from the
Deutsche Forschungsgemeinschaft and Fonds der
Chemischen Industrie.
J O010706R