Synthesis of N(3),N′(3)-polymethylene-bis-hydantoins and their macrocyclic derivatives

Article abstract of DOI:10.1021/jo060370r

An efficient and straightforward two-step approach toward N(3),N′(3)-polymethylene-bis-hydantoins was developed. As a first step, a pyroglutamate is reacted with a diisocyanate to produce a bis-carbamoyllactam. The second step is a double-ring transformat

Full text of DOI:10.1021/jo060370r

Synthesis of N(3),N′(3)-Polymethylene-bis-hydantoins and Their
Macrocyclic Derivatives
Nicolai Dieltiens, Diederica D. Claeys,† and Christian V. Stevens*
Research group Synbioc, Department of Organic Chemistry, Faculty of Bioscience Engineering,
Ghent UniVersity, Coupure links 653, B-9000 Ghent, Belgium
chris.steVens@ugent.be
ReceiVed February 22, 2006
An efficient and straightforward two-step approach toward N(3),N′(3)-polymethylene-bis-hydantoins was
developed. As a first step, a pyroglutamate is reacted with a diisocyanate to produce a bis-carbamoyllactam.
The second step is a double-ring transformation by treatment of this bis-carbamoyllactam with KOtBu
in ethanol. In this fashion N(3),N′(3)-polymethylene-bis-hydantoins are produced in two quantitative
steps and under very mild conditions. When properly derivatized, these compounds can be converted to
their macrocyclic derivatives upon treatment with 5 mol % of second-generation Grubbs’ catalyst. These
macrocyclic derivatives are so far not described in the literature. It was proven that exclusively (E)-
isomers are formed.
Introduction
The exploration of privileged structures in drug discovery
has gained significant popularity in pharmaceutical chemistry
over the years. Hydantoins (or imidazolidine-2,4-diones, 1) have
been extensively studied and are reported to possess a wide
range of biological activities (Figure 1). Phenetoin (5,5-
diphenylhydantoin, 2), for example, was already synthesized
in 19081 and is now still the drug of choice for the treatment of
certain types of epileptic seizures.2 They have proven useful
not only in human medicine (antiarrhythmic,3 anticonvulsant,4
antitumor,5 antidiabetic,6 and antimuscarinic activity,7 etc.) but
also in the agrochemical sector (herbicidal and fungicidal
FIGURE 1. Hydantoin nucleus 1 and phenetoin 2.
activity).8 In recent years, many new synthetic approaches have
been developed toward this interesting heterocycle.9-15
Recently, we described the pyroglutamate-hydantoin rear-
rangement.16 It was found that treating pyroglutamates 3 with
NaH in the presence of isocyanates or isothiocyanates in THF
† D.D.C. is aspirant of the Fund for Scientific Research Flanders (FWO
Vlaanderen).
(1) Biltz, H. Chem. Ber. 1908, 1379-1393.
(2) Back, P.; Maurois, P.; Dupont, C.; Pages, N.; Stables, J. P.; Gressens,
P.; Evrard, P. J. Neurosci. 1998, 18, 4363-4373.
(3) Knabe, J.; Baldauf, J.; Ahlhelm, A. Pharmazie 1997, 52, 912-919.
(4) (a) Luer, M. S. Neurol. Res. 1998, 20, 178-182. (b) Sholl, S.; Koch,
A.; Henning, D.; Kempter, G.; Kleinpeter, E. Struct. Chem. 1999, 10, 355-
366.
(5) Rodgers, T. R.; LaMontagne, M. P.; Markovac, A.; Ash, A. B. J.
Med. Chem. 1977, 20, 591-594.
(6) (a) Sarges, R.; Oates, P. J. Prog. Drug. Res. 1993, 40, 99-159. (b)
Somsak, L.; Kovacs, L.; Toth, M.; O¨ sz, E.; Szilagyi, L.; Gyo¨rgydeak, Z.;
Dinya, Z.; Docsa, T.; Toth, B.; Gergely, P. J. Med. Chem. 2001, 44, 2843-
2848.
(8) Mappes, C. J.; Pommer, E.-H.; Rentzea, C.; Zeeh, B. (BASF A.-G.,
Fed. Rep. Ger.). US Patent 4,198,423, 1980; Chem. Abstr. 1980, 93, 71784.
(9) Charton, J.; Delarue, S.; Vendeville, S.; Debreu-Fontaine, M.-A.;
Girault-Mizzi, S. Sergheraert. C. Tetrahedron Lett. 2001, 42, 7559-7561.
(10) Zhang, W.; Lu, Y. Org. Lett. 2003, 5, 2555-2558.
(11) Muccioli, G. G.; Poupaert, J. H.; Wouters, J. Norberg, B.; Poppitz,
W.; Scriba, G. K. E.; Lambert, D. M. Tetrahedron 2003, 59, 1301-1307.
(12) Meusel, M.; Ambrozak, A.; Hecker, T. K.; Gu¨tschow, M. J. Org.
Chem. 2003, 68, 4684-4692.
(13) Charton, J.; Gassiot, A. C.; Melnyk, P.; Girault-Mizzi, S. Sergheraert,
C. Tetrahedron Lett. 2004, 45, 7081-7085.
(14) Volonterio, A.; Ramirez de Arellano, C.; Zanda, M. J. Org. Chem.
2005, 70, 2161-2170.
(15) Manku, S.; Curran, D. P. J. Org. Chem. 2005, 70, 4470-4473.
(16) Dieltiens, N.; Claeys, D. D.; Zhdankin, V. V.; Nemykin, V. N.;
Allaert, B.; Verpoort, F.; Stevens, C. V. Eur. J. Org. Chem. 2006, in press.
(7) Bazil, C. W.; Pedley, T. A. Annu. ReV. Med. 1998, 49, 135-162.
10.1021/jo060370r CCC: $33.50 © 2006 American Chemical Society
Published on Web 04/22/2006
J. Org. Chem. 2006, 71, 3863-3868
3863

Dieltiens et al.
SCHEME 1
SCHEME 2
produces hydantoins 4 in high yield via a ring-closing-ring-
opening sequence (Scheme 1).
the desired compound 10b, the dicarbamoyllactam 9b was
observed in the 1H NMR of the crude reaction mixture
accompanied by a variety of compounds probably resulting from
the attack of the formed alkoxide anion on the second isocyanate
after reaction with the first isocyanate functionality.
In this paper, we disclose the application of this ring
transformation for the synthesis of N(3),N′(3)-polymethylene-
bis-hydantoins. Several of these bis-hydantoins 5 have been
evaluated as hexamethylenebis(acetamide) analogues (HMBA,
6) (Figure 2). HMBA is an agent that induces differentiation of
certain types of tumor cells to nonmalignant phenotypes. This
implies that it works not by killing the cancer cells but by
inducing them to differentiate and to express characteristics of
the normal nontransformed counterpart. This is a promising
approach to cancer therapy, potentially without many of the
disadvantages of cytotoxic agents. HMBA has even had some
modest success in clinical trials. The doses required to achieve
sufficient blood levels in human patients, however, led to some
undesirable side effects. It was found that compounds 5a and
5d were 10 times more potent than HMBA itself, but the activity
of 5b however was low, probably due to its insolubility.17-19
Compound 5c was found to show antiepileptic activity in the
maximal electroshock seizure test.20
To avoid these undesired reactions, it seemed necessary to
make a short synthetic detour. Previous observations had shown
that the formed sodium salt of the carbamoylated lactam
precipitates from diethyl ether, effectively preventing it from
reacting further.25 Thus, it would be possible to isolate the
carbamoylated intermediate and attempt the ring transformation
in a separate step. As expected, the double sodium salts 11
precipitates from solution after treating a mixture of 7 and 8
with NaH in ether and provided derivatives 9 in quantitative
yield and good purity after acidic workup (Scheme 3).
The double-ring transformation from 9 to 10 did not proceed
as smoothly as anticipated. Treatment of 9 with strong bases
such as NaH and KO-t-Bu in THF at room temperature or at
reflux did not result in the desired conversion. Switching to
ethanol as a solvent in combination with NaOH or KO-t-Bu as
a base however, resulted in clean and complete double
transformation to the bis-hydantoins 10 within 1 h at room
temperature (Scheme 4). Derivatives 10 are obtained as a viscous
oil after acidic workup but can be crystallized from THF to
afford white powders.
The synthesis of these compounds usually involves the
reaction of a hydantoin with a ω-dihaloalkane under basic
conditions. The problem with this reaction is the need for N(3)-
selective alkylation and the yield which is usually quite low.
Starting from pyroglutamates would avoid the problem of N(3)-
selectivity and also allows the synthesis of highly functionalized
derivatives since pyroglutamates can easily be derivatized.16,21-24
Special attention is required when distinguishing between
1-carbamoyl-2-pyrrolidines 9 and hydantoins 10. Indeed, from
their structures it is obvious that straightforward 1H NMR and
Results and Discussion
(17) Haces, A.; Breitman, T. R. Driscoll, J. S. J. Med. Chem. 1987, 30,
405-409.
Following our protocol for the ring transformation of pyro-
glutamates, a mixture of ethylpyroglutamate 7 and diisocyanate
8b in THF was treated with NaH (Scheme 2). It was found,
however, that a mixture of compounds was obtained. Besides
(18) Breslow, R. Jursic, B.; Yan, Z. F.; Friedman, E.; Leng, L.; Ngo,
L.; Rifkind, R. A.; Marks, P. A. Proc. Natl. Acad. Sci. U.S.A. 1991, 88,
5542-5546.
(19) Breslow, R.; Belvedere, S.; Gershell, L. HelV. Chim. Acta 2000,
83, 1685-1692.
(20) Poupaert, J. H.; Mergen, F. Lerot, T. Bull. Soc. Chim. Belg. 1988,
97, 469-470.
(21) Bran˜a, M. F.; Garranzo, M.; Pe´rez-Castells, J. Tetrahedron Lett.
1998, 39, 6569-6572.
(22) Najera, C.; Yus, M. Tetrahedron: Asymmetry 1999, 10, 2245-
2303.
(23) Stevens, C. V.; Rammeloo, T.; De Kimpe, N. Synlett. 2001, 10,
1519-1522.
(24) Dieltiens, N.; Stevens, C. V.; Masschelein, K. G. R.; Rammeloo,
T. Tetrahedron 2005, 61, 6749-6756.
(25) Stevens, C. V.; Dieltiens, N.; Claeys, D. D. Org. Lett. 2005, 7,
1117-1119.
FIGURE 2. Bis-hydantoins 5 and HMBA 6.
3864 J. Org. Chem., Vol. 71, No. 10, 2006

N(3),N′(3)-Polymethylene-bis-hydantoins
FIGURE 3. Most important COSY and HMBC couplings in 9a and 10a.
SCHEME 3
SCHEME 4
SCHEME 5
13C NMR measurements are not sufficient to distinguish between
the two. However, having both compounds in hand, two-
dimensional spectroscopy clearly allowed the unambiguous
determination (Figure 3). In the case of 9, there is a coupling
in the COSY spectrum between the NH protons and the CH2
next to the nitrogen. On the other hand, the NH protons of 10
show a coupling to the proton next to the carbonyl and not to
the CH2, proving that this methylene is connected to a tertiary
nitrogen. Furthermore, the protons of the CH2 next to nitrogen
of 9 only couple to the urea carbonyl in the HMBC (hetero-
nuclear multiple bound correlation) spectrum, whereas in the
case of 10 they couple to both the urea and the lactam carbonyl.
Another distinctive feature is the shift of the NH-protons.
Intramolecular hydrogen bridge formation in 9 to the lactam
carbonyl causes a downfield shift, resulting in a typical value
of 8.3 ppm, whereas the value in the case of 10 is typically
around 6.3 ppm.
As we reported before, the hydantoins can easily be further
functionalized by refluxing in acetone with a proper electrophile
in the presence of finely ground K2CO3.16 The same conditions
proved to work perfectly on the dimers, and in this fashion
several derivatives could be prepared in near-quantitative yields
(Scheme 5). The only losses occur during purification by flash
chromatography, which is only necessary to remove the excess
of electrophile in the case of 12e and 12f. For all other
derivatives, it is sufficient to filter off the solids and evaporate
the volatiles in vacuo.
One of the challenges of medicinal chemistry is the enhance-
ment of the affinity of a given ligand for its target by decreasing
its degrees of freedom, thereby reducing the cost in entropy.
To prevent free rotation around the central polymethylene axis,
we wanted to evaluate the possibility to use ring-closing
metathesis (RCM) for the macrocyclization of derivatives
J. Org. Chem, Vol. 71, No. 10, 2006 3865

Dieltiens et al.
SCHEME 6
SCHEME 7
12a,b,d,g.26,27 Fu¨rstner and co-workers have proven that sub-
strates devoid of any conformational constraints can be ef-
ficiently cyclized by RCM to macrocyclic products and that
neither the conformational predisposition nor the ring size
formed is a major concern. The decisive parameters are the
presence of polar relay substituents, their proper distance to the
alkene groups, and the low steric hindrance close to the double
bonds.28 Since all of these requirements are met in these
substrates, macrocyclization was expected to pose no special
problems. Indeed, it was found that treating derivatives 12a,b,d,g
with a catalytic amount of the second-generation Grubbs’
catalyst 17 (5 mol %) resulted in quantitative conversion to
macrocyclic derivatives 13 after 16 h of reflux in CH2Cl2
(Scheme 6). A significant drop in yield, however, was observed
during purification by flash chromatography, probably caused
by the high polarity of these compounds.
particular case, both 12 and 13 have two racemic centers and,
thus, the two observed signals could belong to another diaste-
3
reoisomer. Moreover, JHH coupling constants, which are
normally used to distinguish trans and cis isomers, are absent
since the macrocycles are symmetrical entities and as a
consequence the two alkene protons do not couple with each
other. They appear as small triplets. It was, however, possible
to separate the two forms of compound 13c by flash chroma-
tography. Now it was possible to prove whether form A and
form B were diastereoisomers or (E,Z)-isomers. First, it was
observed that treatment with base of one form of 13c resulted
in equilibration to the mixture of both forms (Scheme 7).
Second, hydrogenation of form A and form B gave two different
compounds. This can only be the case if the two compounds
were diastereoisomers since the (E,Z)-isomers should have
produced the same hydrogenation product.
It is known that when using RCM as a macrocyclization
method, usually (E,Z) mixtures of cyclized compounds are
formed. In this case, two separate signals (1:1 ratio) were
observed for the alkene protons of derivatives 13. This, however,
does not mean that an (E,Z) mixture is formed, since in this
The question rose whether the trans- or cis-fused cycles were
produced. The presence of only one isomer around the double
bound can be explained by secondary metathesis reactions. In
a kinetic study, Grubbs has proven that macrocycles isomerize
to produce a thermodynamically controlled (E/Z) ratio regardless
the stereochemistry of the initial alkene.29 Although usually a
mixture of both isomers is isolated, Fu¨rstner has already reported
that only the (E)-isomer is formed in some cases.30 A stereo-
selective approach to macrocyclic (E)- and (Z)-isomers is
possible using ring-closing alkyne metathesis (RCAM).31 In the
first case, the macrocyclic alkyne is converted to the (E)-isomer
(26) For our previous work on using RCM for the synthesis of aza-
heterocycles, see: (a) Dieltiens, N.; Stevens, C. V.; De Vos, D.; Allaert,
B.; Drozdzak, R.; Verpoort, F. Tetrahedron Lett., 2004, 45, 8995-8998.
(b) Dieltiens, N.; Stevens, C. V.; Allaert, B.; Verpoort, F. ArkiVoc 2005, i,
92-97.
(27) For recent reviews on RCM, see: Fu¨rstner, A. Angew. Chem., Int.
Ed. 2000, 39, 3012-3043. Grubbs, R. H. Tetrahedron 2004, 60, 7117-
7140. Dragutan, I.; Dragutan, V.; Filip, P. ArkiVoc 2005, x, 105-129.
Nicolaou, K. C.; Bulger, P. G.; Sarlah, D. Angew. Chem., Int. Ed. 2005,
44, 4490-4527. For a recent review on the use of RCM for the synthesis
of azaheterocycles, see: Deiters, A.; Martin, S. F. Chem. ReV. 2004, 104,
2199-2238.
(28) (a) Fu¨rstner, A.; Langeman, K. J. Org. Chem. 1996, 61, 3942-
3943. (b) Fu¨rstner, A. Top. Catal. 1997, 4, 285-299. (c) Fu¨rstner; A.;
Mu¨ller, T. J. Org. Chem. 1998, 63, 424-425 and references cited therein.
(29) (a) Lee, C. W.; Grubbs, R. H. Org. Lett. 2000, 2, 2145-2147. For
a computational study, see: (b) Vyboishchikov, S. F.; Thiel, W. Chem.
Eur. J. 2005, 11, 3921-3935.
(30) Fu¨rstner, A.; Thiel, O. R.; Kindler, N.; Bartkowska, B. J. Org. Chem.
2000, 65, 7990-7995.
(31) For a review on RCAM, see: Fu¨rstner, A. Angew. Chem., Int. Ed.
2000, 39, 3012-3043.
3866 J. Org. Chem., Vol. 71, No. 10, 2006

N(3),N′(3)-Polymethylene-bis-hydantoins
3
FIGURE 5. JH,H coupling of the 13C satellites of both isotopomers of 13c.
SCHEME 8
by trans-selective hydrosilylation and subsequent protodesily-
lation.32 In the second case, the macrocyclic alkyne is converted
to the (Z)-isomer by Lindlar reduction.33 In this perspective,
derivative 12c was prepared and was treated with a mixture of
Mo(CO)6 and p-chlorophenol as an “instant” catalyst system
under an Ar atmosphere (Scheme 8).34 Unfortunately, even after
prolonged reaction times only starting material could be
recovered.
In a further attempt to selectively synthesize the trans- and
the cis-isomer, derivative 10c was refluxed in two separate
experiments with trans-1,4-dichloro-2-butene and cis-1,4-
dichloro-2-butene. Only in the second case was some reaction
observed, resulting in a mixture that was tentatively identified
as a mixture of starting material, monoalkylated material, and
cyclized material (<10%). The last two components proved
impossible to separate, but a 13C-spectrum of this mixture proved
it to be different from compound 13c, suggesting that the
macrocycles produced by RCM are in the trans-geometry. The
fact that these derivatives are so difficult to cyclize again proves
the importance of RCM as the macrocyclization method and of
the functional group as an essential relay in order to accomplish
this.35 Probably chelation complexes of type 16 are formed
during metathesis, effectively bringing the two alkene moieties
in the correct position (Figure 4).
FIGURE 4. Possible chelation complex formed during metathesis.
Conclusions
A short and high-yield approach was developed to N(3),N′-
(3)-polymethylene-bis-hydantoins. This two-step approach con-
sists of carbamoylation of a pyroglutamate and subsequent
double-ring transformation. After proper functionalization, these
dimers can be cyclized by ring-closing metathesis using the
second-generation Grubbs’ catalyst. It was proven that only the
trans-isomers are isolated.
Experimental Section
1. Synthesis of Derivatives 9. As a representative example, the
synthesis of ethyl 1-({[4-({[2-(ethoxycarbonyl)-5-oxopyrrolidin-1-
yl]carbonyl}amino)butyl]amino}carbonyl)-5-oxopyrrolidine-2-car-
boxylate 9a is described. Ethyl pyroglutamate 7 (5 g, 31.8 mmol)
was dissolved in diethyl ether (100 mL, freshly distilled from Na
metal) and kept under a positive N2-pressure. To this solution was
added 1,4-diisocyanatobutane (2.23 g, 15.9 mmol) followed by NaH
(0.92 g, 38.2 mmol, washed with hexanes). The reaction was
allowed to stir under nitrogen atmosphere at room temperature for
1 h. The reaction was quenched by the addition of saturated aqueous
NH4Cl until the pH was neutral. The mixture was extracted with
EtOAc, and the organics were dried (MgSO4) and filtered. The
solvent was removed in vacuo. Compound 9a was obtained as a
viscous oil in quantitative yield.
A more decisive confirmation of the trans-geometry was
obtained by looking at the 13C satellites of the alkene protons
of one pair of enantiomers of 13c (Figure 5).36 When one of
the two alkene carbon atoms is a 13C-isotope, the symmetry is
lost due to the presence of a magnetically active 13C in place
of an inactive 12C. In this case, the two isotopomers give two
separate signals (br d, 5.14 ppm, 14.3 Hz and br d, 5.12 ppm,
14.9 Hz). They are not each others mirror image because of
the two other chiral centers. A 3JH,H coupling of 14.3 and 14.9
Hz is observed, pointing to the trans-geometry.37
(32) Lacombe, F.; Radkowski, K.; Seidel, G.; Fu¨rstner, A Tetrahedron
2004, 60, 7315-7324.
Ethyl
1-({[4-({[2-(ethoxycarbonyl)-5-oxopyrrolidin-1-yl]-
carbonyl}amino)butyl]amino}carbonyl)-5-oxopyrrolidine-2-car-
boxylate 9a: 1H NMR (300 MHz, CDCl3) δ 1.29 (6H, t, J ) 7.2
Hz), 1.60 (4H, t, J ) 3.0 Hz), 2.00-2.10 (2H, m), 2.25-2.45 (2H,
m), 2.57 (2H, ddd, J ) 17.6 Hz, J ) 3.4 Hz, J ) 9.5 Hz), 2.74
(2H, ddd, J ) 17.6 Hz, J ) 9.9 Hz, J ) 9.9 Hz), 3.23-3.40 (4H,
(33) (a) Fu¨rstner, A.; Guth, O.; Rumbo, A.; Seidel, G. J. Am. Chem.
Soc. 1999, 121, 11108-11113. (b) Fu¨rstner, A.; Mathes, C.; Lehmann, C.
W. Chem. Eur. J. 2001, 7, 5299-5317.
(34) Mortreux, A.; Blanchard, M. J. Chem. Soc., Chem. Commun. 1974,
786-787.
(35) (a) Fu¨rstner, A.; Langemann, K. Synthesis 1997, 792-803. (b)
Fu¨rstner, A.; Thiel, O. R.; Lehmann, C. W. Organometallics 2002, 21, 331-
335.
3
(37) For articles concerning the recovery of JH,H, see: (a) Mucci, A.;
(36) For clarity the diastereoisomer with the two ester groups in a cis
position is depicted.
Parenti, F.; Schenetti, L. Eur. J. Org. Chem. 2002, 938-940. (b) Janssen,
C. E.; Krause, N. Eur. J. Org. Chem. 2005, 2322-2329.
J. Org. Chem, Vol. 71, No. 10, 2006 3867

Dieltiens et al.
m), 4.24 (4H, q, J ) 7.2 Hz), 4.77 (2H, dd, J ) 9.5 Hz, J ) 2.6
Hz), 8.31 (2H, t, J ) 5.5 Hz); 13C NMR (75 MHz, CDCl3) δ 14.2,
21.3, 27.0, 31.9, 39.5, 58.2, 61.7, 152.3, 171.5, 176.5; IR (NaCl,
cm-1) νmax 1694 (CdO), 1721 (br CdO), 3317 (NH); MS (70 eV)
m/z 455.7 (M+ + 1, 100); HRMS calcd for C20H30N4O8 (M + H+)
455.2136, found 455.2148.
172.4, 172.6; IR (NaCl, cm-1) νmax 1645 (CdC), 1709 (br CdO),
1769 (CdO); MS (70 eV) m/z 535.7 (M+ + 1, 100); HRMS calcd
for C26H38N4O8 (M + H+) 535.2762, found 535.2769.
4. Synthesis of Derivatives 13. As a representative example,
the synthesis of 3-[18-(2-ethoxycarbonylethyl)-8,17,19,20-tetraoxo-
1,6,9,16-tetraazatricyclo[14.2.1.16,9]icos-3-en-7-yl]propionic acid
ethyl ester 13b is described. Compound 12b (0.2 g, 0.36 mmol)
was dissolved in CH2Cl2 (20 mL, freshly distilled from CaH2), and
the second-generation Grubbs’ catalyst (0.015 g, 0.018 mmol) was
added. The reaction was allowed to reflux for 16 h under a N2
atmosphere. The product was coated on silica gel by removal of
the solvent in vacuo and purified by flash chromatography (silica
gel; hexane/EtOAc). Compound 13a is obtained as a 1:1 mixture
of the two diastereoisomers (combined yield 46%) which can be
separated and obtained as viscous oils.
2. Synthesis of Derivatives 10. As a representative example,
the synthesis of 3-(1-{4-[4-(2-ethoxycarbonylethyl)-2,5-dioxoimi-
dazolidin-1-yl]butyl}-2,5-dioxoimidazolidin-4-yl)propionic acid eth-
yl ester 10a is described. Compound 9a (7.23 g, 15.9 mmol) was
dissolved in absolute ethanol (75 mL) and kept under a positive
N2 pressure. To this solution was added KO-t-Bu (3.92 g, 35 mmol).
The reaction was allowed to stir under nitrogen atmosphere at room
temperature for 1 h. The reaction was quenched by the addition of
saturated aqueous oxalic acid until the pH was acidic. The ethanol
was removed in vacuo, and the residue was extracted with EtOAc,
the organics were dried (MgSO4) and filtered. The solvent was
removed in vacuo. For the crystallization, the product was dissolved
in a minimal amount of boiling THF and hexane was added until
the solution remained cloudy. The flask was put in the freezer and
afterward the solids were filtered. This procedure was repeated until
no more precipitate was formed. Compound 10a was obtained in
quantitative yield.
3-(1-{4-[4-(2-Ethoxycarbonylethyl)-2,5-dioxoimidazolidin-1-
yl]butyl}-2,5-dioxoimidazolidin-4-yl)propionic acid ethyl ester
10a: white powder; mp 124-127 °C; 1H NMR (300 MHz, CDCl3)
δ 1.26 (6H, t, J ) 7.2 Hz), 1.63 (4H, br s), 2.03 (2H, ddt, J ) 13.9
Hz, J ) 7.2 Hz, J ) 7.3 Hz), 2.20 (2H, ddt, J ) 13.9 Hz, J ) 7.3
Hz, J ) 6.3 Hz), 2.47 (2H, t, J ) 7.3 Hz), 3.52 (4H, br s), 4.14
(4H, q, J ) 7.2 Hz), 4.10 (2H, dd, J ) 7.2 Hz, J ) 6.3 Hz), 6.29
(2H, s); 13C NMR (75 MHz, CDCl3) δ 14.3, 25.2, 26.9, 29.9, 38.0,
56.4, 61.1, 157.4, 172.9, 173.7; IR (KBr, cm-1) νmax 1713 (br Cd
O), 1764 (CdO), 3249 (NH); MS (70 eV) m/z 455.7 (M+ + 1,
100); HRMS calcd for C20H30N4O8 (M + H+) 455.2136, found
455.2148.
3. Synthesis of Derivatives 12. As a representative example,
the synthesis of 3-(3-allyl-1-{4-[3-allyl-4-(2-ethoxycarbonylethyl)-
2,5-dioxoimidazolidin-1-yl]butyl}-2,5-dioxoimidazolidin-4-yl)pro-
pionic acid ethyl ester 12a is described. Compound 10a (7.23 g,
15.9 mmol) was dissolved in acetone (75 mL). To this solution
were added K2CO3 (11 g, 79.5 mmol) and allyl bromide (7.70 g,
63.6 mmol). This mixture was refluxed until TLC showed complete
consumption of the starting material. The mixture was filtered, and
the solvent was removed in vacuo. Compound 12a was obtained
as a viscous oil in quantitative yield.
3-(3-Allyl-1-{4-[3-allyl-4-(2-ethoxycarbonylethyl)-2,5-diox-
oimidazolidin-1-yl]butyl}-2,5-dioxoimidazolidin-4-yl)propion-
ic acid ethyl ester 12a: 1H NMR (300 MHz, CDCl3) δ 1.26 (6H,
t, J ) 7.2 Hz), 1.64 (4H, br s), 2.00-2.12 (2H, m), 2.18-2.42
(6H, m), 3.53 (4H, br s), 3.62 (2H, dd, J ) 15.7 Hz, J ) 7.4 Hz),
4.01 (2H, dd, J ) 3.0 Hz, J ) 6.6 Hz), 4.13 (4H, q, J ) 7.2 Hz),
4.34 (2H, dd, J ) 15.7 Hz, J ) 5.0 Hz), 5.24 (2H, d, J ) 4.7 Hz),
5.28 (2H, s), 5.70-5.83 (2H, m); 13C NMR (75 MHz, CDCl3) δ
14.3, 23.8, 25.3, 28.2, 38.3, 43.5, 57.8, 60.9, 119.5, 131.8, 156.3,
3-[18-(2-Ethoxycarbonylethyl)-8,17,19,20-tetraoxo-1,6,9,16-
tetraazatricyclo[14.2.1.16,9]icos-3-en-7-yl]propionic acid ethyl
ester 13b, diastereoisomer 1: TLC Rf 0.42 (petroleum ether/ethyl
1
acetate, 2:8); H NMR (300 MHz, CDCl3) δ 1.15-1.30 (4H, m),
1.24 (6H, t, J ) 7.2 Hz), 1.62-1.74 (4H, m), 2.00-2.16 (2H, m),
2.17-2.52 (6H, m), 3.43-3.49 (4H, m), 3.57 (2H, ddd, J ) 13.8
Hz, J ) 6.3 Hz, J ) 3.9 Hz), 4.04 (2H, dd, J ) 2.9 Hz, J ) 6.5
Hz), 4.11 (4H, q, J ) 7.2 Hz), 4.35 (2H, d, J ) 16.0 Hz), 5.35
(2H, t, J ) 2.2 Hz); 13C NMR (75 MHz, CDCl3) δ 14.2, 23.6,
27.5, 27.8, 28.5, 39.6, 41.8, 57.7, 60.9, 126.3, 157.0, 172.6, 172.7;
IR (NaCl, cm-1) νmax: 1709 (CdO), 1769 (CdO); MS (70 eV)
m/z 535.2 (M+ + 1, 100); HRMS calcd for C26H38N4O8 (M + H+)
535.2762, found 535.2772.
Diastereoisomer 2: TLC Rf 0.38 (petroleum ether/ethyl acetate,
2:8); 1H NMR (300 MHz, CDCl3) δ 1.16-1.35 (4H, m), 1.26 (6H,
t, J ) 7.2 Hz), 1.58-1.75 (4H, m), 1.97-2.09 (2H, m), 2.13-2.55
(6H, m), 3.43-3.49 (4H, m), 3.48 (2H, ddd, J ) 13.3 Hz, J ) 6.4
Hz, J ) 3.8 Hz), 3.58 (2H, ddd, J ) 13.3 Hz, J ) 3.9 Hz, J ) 3.8
Hz), 3.68 (2H, dt, J ) 16.2 Hz, J ) 2.4 Hz), 4.04 (2H, dd, J ) 3.6
Hz, J ) 7.2 Hz), 4.09 (2H, d, J ) 16.2 Hz), 4.14 (4H, q, J ) 7.2
Hz), 5.63 (2H, t, J ) 2.4 Hz); 13C NMR (75 MHz, CDCl3) δ 14.3,
24.3, 26.8, 27.6, 28.6, 38.9, 42.4, 58.1, 60.9, 128.2, 156.9, 172.7,
173.0; IR (NaCl, cm-1) νmax 1709 (CdO), 1768 (CdO); MS (70
eV) m/z 535.2 (M+ + 1, 100); HRMS calcd for C26H38N4O8 (M +
H+) 535.2762, found 535.2779.
Acknowledgment. Financial support of this research by the
BOF (Bijzonder Onderzoeksfonds Universiteit Gent, Research
Fund Ghent University) and by the FWO (Fonds voor Weten-
schappelijk Onderzoek - Vlaanderen, Fund for Scientific
Research - Flanders) is gratefully acknowledged.
Supporting Information Available: General information and
spectroscopic data of all compounds synthesized with complete
peak assignment. Copies of the 13C NMR spectra of all compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
JO060370R
3868 J. Org. Chem., Vol. 71, No. 10, 2006