Stereochemistry and conformational analysis of hemirubin

Article abstract of DOI:10.1016/S0040-4020(00)00087-9

Intramolecularly hydrogen-bonded analogs of the natural bile pigment, bilirubin, are very scarce. Nuclear Overhauser effect NMR studies of the newest analog, a hemirubin (1), confirms intramolecular hydrogen bonding and a ridge-tile-shaped conformation. <

Full text of DOI:10.1016/S0040-4020(00)00087-9

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 1797–1810
Stereochemistry and Conformational Analysis of Hemirubin
*
Michael T. Huggins and David A. Lightner
Department of Chemistry, University of Nevada, Reno, NV 89557, USA
Received 14 December 1999; revised 25 January 2000; accepted 27 January 2000
Abstract—Intramolecularly hydrogen-bonded analogs of the natural bile pigment, bilirubin, are very scarce. Nuclear Overhauser effect
1
NMR studies of the newest analog, a hemirubin (1), confirms intramolecular hydrogen bonding and a ridge-tile-shaped conformation. H
NMR studies of 1 at sufficiently low temperatures detect conformational enantiomerization, for which an activation barrier of DG‡ϳ16 kcal/
mol at 25ЊC in CDCl3 has been estimated. Evidence for the presence of dimeric association at low temperatures or high concentrations of 1 is
found by the appearance of new sets of resonances, data from which may be used to calculate: DGЊ298 K ϩ3.4 kcal/mol, DHЊ Ϫ5.6 kcal/mol
and DSЊ Ϫ30.3 cal/deg/mol. ᭧ 2000 Elsevier Science Ltd. All rights reserved.
Introduction
Various dipyrrinone models for bilirubin have been synthe-
sized and found useful in understanding the spectroscopic
properties and photochemical reactions of the natural
pigment.3,9,10 However, these analogs typically lacked the
intramolecular hydrogen bonding so essential to bilirubin—
until the very recent syntheses of two analogs with (i) three
pyrrole rings (3)11 and (ii) only two pyrrole rings (hemi-
rubin, 2).12 Tripyrrole 3 was found to engage in intra-
molecular hydrogen bonding in the crystal and in solution
in nonpolar solvents. Hemirubin 2 is believed to engage in
intramolecular hydrogen bonding, but its solubility in
nonpolar solvents such as CDCl3 and CD2Cl2 was too
limited to permit a detailed analyses of conformation by
NMR techniques. To overcome this obstacle, we synthe-
sized a new hemirubin (1) and its 10-oxo analog (4), both
of which have excellent solubility in non-polar organic
Bilirubin (Fig. 1), the yellow–orange end-product of heme
metabolism and the colorful herald of hepatobiliary
disease,1,2 consists of twin dipyrrinones conjoined to a
–CH2– group. Dipyrrinones are bright yellow chromo-
phores and avid participants in hydrogen bonding.3,4 In bili-
rubin they may rotate independently about the –CH2–, but
only one conformation, shaped like a half-opened book, lies
at a global energy minimum.5,6 And this conformation (Fig.
1B) is further stabilized by intramolecular hydrogen bonds
that link each dipyrrinone with an opposing propionic
acid.7,8 Taken collectively, the matrix of six intramolecular
hydrogen bonds dominates the stereochemistry of bilirubin
and profoundly controls its solution, spectroscopic and
metabolic properties.
Figure 1. (A) Bilirubin in a high energy linear conformation with angles of rotation designated as f1 and f2 about the C(9)–C(10) and C(10)–C(11) bonds.
(B) Preferred bilirubin conformation shaped like a ridge-tile (f1ˆf2ϳ60Њ) with an interplanar angle uϳ100Њ. This conformation achieves considerable
stabilization from intramolecular hydrogen bonds (hatched lines). (C) Dipyrrinone propionic acid analogs of bilirubin.
Keywords: pyrroles; hydrogen bonding; stereoisomerism.
* Corresponding author. Tel.: ϩ1-775-784-4980; fax: ϩ1-775-784-6804; e-mail: lightner@unr.edu
0040–4020/00/$ - see front matter ᭧ 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00087-9

1798
M. T. Huggins, D. A. Lightner / Tetrahedron 56 (2000) 1797–1810
solvents. In the following we describe (i) their syntheses,
and (ii) their conformational analysis by molecular
dynamics calculations and by systematic analyses of nuclear
Results and Discussion
Synthesis
1
Overhauser effects (NOEs) and variable temperature H
NMR (VT-NMR) spectroscopic analysis. The study of 1
provides insights into how the conformation and solution
properties of bilirubin are controlled by intramolecular
hydrogen bonding. The 10-oxo-hemirubins, e.g. 4, serve
as models for 10-oxo-bilirubin, which is thought to be impli-
cated in alternate (oxidative) pathways for bilirubin elimi-
nation and mammalian metabolism.13,14
The target hemirubin (1) was prepared following SnCl4-
catalyzed Friedel–Crafts acylation of dipyrrinone 9 with
acid chloride 7 to afford a 59% yield of 10-oxo-hemirubin
4 as its methyl ester (4M) (see Scheme 1). Saponification of
4M gave an 85% yield of 4, reduction of which by NaBH4 in
refluxing 2-propanol afforded the target (1) in 93% yield.
Although acid chloride 7 was known from earlier work,12
the dipyrrinone coupling partner (9) was not. It was
prepared as outlined in Scheme 1 from two monopyrroles:
11 and 12 following decarboxylation of the coupled product
10. Pyrrolinone 11 was known from earlier work;9e alde-
hyde 12 was prepared by oxidation of the 5-methyl group
of pyrrole ester 1311 using ceric ammonium nitrate (CAN).
Ester 13 was prepared from simple, inexpensive starting
materials: reaction of 2-pentanone and propionic anhydride
in the presence of BF3·Et2O15 gave condensation product 14
in 38% yield; reaction of 14 with base and Fischer–Knorr
type condensation with the oxime of diethyl malonate using
Zn and acetic acid16 gave 13. Simplified hemirubins 5 and 6,
which are incapable of intramolecular hydrogen bonding,
served as comparison components and were prepared
smoothly from dipyrrinone 9 and o-toluoyl chloride (8), as
outlined in Scheme 1. The procedures parallel those used in
the syntheses of 1 and 4.
Constitutional structure
The structures of hemirubins 1 and 4, and their methyl esters
1M and 4M, follow from the structures of the component
Scheme 1.
a
NaBH4=2-propanol; refl:;
b
CH3OH=H2SO4;
c
NaOH=aq: THF;
d
SnCl4=CH2Cl2;
e
NaOAc–KOAc=D;
f
KOH=CH3OH;
g CeꢀNH3†2ꢀNO3†4=AcOH–H2O; h NaOH; i HON ˆ CꢀCO2Et†2=Zn–AcOH; j BF3·Et2O.

M. T. Huggins, D. A. Lightner / Tetrahedron 56 (2000) 1797–1810
Table 1. Comparison of 13C NMR chemical shifts of hemirubin analogs
1799
a
Chemical shift in CDCl3
Carbon position
9H-Dipyrrinone
1
1M
5
4
4M
6
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
174.5
130.4
142.6
129.9
101.4
126.0
123.8
124.5
120.5
–
–
–
–
–
–
–
–
–
174.7
128.8
142.0
129.5
101.1
123.0
123.9
134.6
122.8
27.2
138.4
138.4
130.4
128.8
126.6
126.3
24.6
173.5
129.0
142.1
131.5
100.6
123.1
123.7
132.3
122.0
30.0
138.2
137.7
129.4
128.7
126.6
126.6
27.9
173.9
128.8
142.0
131.6
100.7
123.1
123.7
132.3
122.0
30.3
135.8
138.0
129.8
128.3
126.0
126.0
19.6
–
175.7
137.5
141.6
128.3
99.1
127.3
129.6
135.3
129.3
188.1
141.4
139.7
129.2
127.3
125.4
130.1
26.0
175.7
137.5
141.7
130.1
96.0
129.9
129.9
135.4
128.7
187.8
140.1
138.1
129.6
126.1
127.1
130.0
28.3
174.0
135.9
141.9
130.4
96.6
130.3
130.4
134.7
128.6
187.0
140.3
136.9
125.4
127.0
129.6
130.6
19.4
–
a
b
32.0
34.6
34.4
35.3
CO2R
2-CH3
3-CH3
7-CH2
7-CH3
8-CH2
8-CH3
–
10.2
8.5
18.5
17.9
16.8
15.1
179.1
8.1
9.7
17.9
17.3
173.8
8.3
9.9
17.8
17.3
–
8.1
9.8
17.8
17.3
17.0
15.7
177.9
8.2
9.7
18.2
17.2
173.7
8.6
9.9
18.0
17.3
–
8.6
9.9
17.8
17.3
16.3
15.9
17.1
16.8
17.0
15.6
16.3
15.6
16.2
15.7
a Chemical shifts in d (ppm) downfield from the residual solvent resonances. Solutions were ϳ5×10Ϫ3 M.
dipyrrinone (9) and benzene derivative (7) used in their
syntheses, and they are confirmed by 13C NMR and
1H{1H}-homonuclear Overhauser effect (NOE) spectro-
scopy (Table 1). The carbon chemical shift assignments of
Table 1 were made by a combination of HMQC and HMBC
techniques. The condensation product (4) of 5 and 6 shows
the expected ten new signals from the aromatic ring com-
ponent 6, including the new C(10) ketone carbonyl at
ϳ188 ppm. As expected, addition of the benzoyl group to
dipyrrinone 9 caused the greatest 13C NMR deshieldings in
the carbons of the pyrrole ring of 4M (or 4). In addition,
C(5) became more shielded, particularly in 4M and much
less so in 4; and, at a remote site, C(2) became strongly
deshielded. Conversion of 4M to 1, of CvO to CH2, led
to the expected strong shielding at C(10) in 1 and to other
changes in 13C NMR chemical shifts in 1 relative to 4,
especially in the pyrrole ring, but also at C(5) and C(2).
Noticeable differences between certain signals in the free
acids and their methyl esters (C(1), C(2), C(4), C(7), C(10)
in 1 and 1M; C(1), C(4), C(5), C(6) in 4 and 4M) suggest
differences in conformation in CDCl3 solvent.
Comparisons (Table 1) with 5 and 6, which are incapable
of intramolecular hydrogen bonding, indicate a closer
Figure 2. Structurally significant NOEs found in 1, 1M and 4 are shown by solid, double-headed, curved arrows. Weak NOEs are shown by dashed, double-
headed arrows.

1800
M. T. Huggins, D. A. Lightner / Tetrahedron 56 (2000) 1797–1810
Figure 3. (Left) Intramolecularly hydrogen-bonded structures of 1 and 4. (Right) Intermolecularly hydrogen-bonded dimeric structure of ester 1M.
similarity in dipyrrinone chemical shifts of 6 with 4M than
with 4, and of 5 with 1M than with 1. These comparisons are
consistent with the notion that 1 and 4 engage in strong
intramolecular hydrogen bonding in CDCl3, and 1M and
4M might engage in weaker intramolecular hydrogen
bonding, or in intermolecular hydrogen bonding.
NOEs are also found between the COOH and the lactam
NHs of 1 and 4, suggesting a close proximity between these
groups. In 1M, an NOE is found between the C(21)CH3 and
the C(10) CH2, suggesting a close proximity between these
groups. Such an NOE is not found in 1 and 4. The data are
consistent with intramolecular hydrogen bonding between
the COOH and lactam groups in 1 and 4, and with an inter-
molecularly hydrogen-bonded dimeric structure (Fig. 3) in
1M.
Nuclear Overhauser effects and stereochemistry
Consistent with the constitutional structures of 1 and 4,
NOEs showed that the dipyrrinone unit retained the syn-Z
configuration (Fig. 2). Thus, strong NOEs are seen between
the pyrrole and lactam NHs and between the C(5)H and
C(31)CH3 and C(71)CH3. Interestingly and significantly,
1H NMR and hydrogen bonding
Supporting evidence for intramolecular hydrogen bonding
in 1 and 4 may be found from an examination of H NMR
1
chemical shifts of the lactam and pyrrole NHs (Table 2). In
(CD3)2SO these NH chemical shifts of 1, 1M and 5 are
essentially identical and rather close to those of the parent
9H-dipyrrinone, as expected from previous studies, for
monomeric structures solvated (and hydrogen-bonded) to
DMSO. The NH chemical shifts of 4 and 4M are nearly
identical to those of 6. In CDCl3, however, where hydrogen
bonding is promoted—either intramolecular or inter-
molecular between two dipyrrinones—the chemical shifts
differ in a significant way. While the lactam NH becomes
deshielded in 1, 1M and 5, relative to its chemical shift in
(CD3)2SO, the pyrrole NH becomes strongly shielded in 1
but only slightly shielded in 1M and 5. This behavior, noted
earlier for various bilirubins and their dimethyl esters10,16,17
and found in numerous dipyrrinone esters10,17 and in
Table 2. Comparison of hemirubin NH and COOH 1H NMR chemical
shifts (all spectra measured in 10Ϫ3 M solutions at 500 MHz and reported
relative to residual solvent resonances)
Compoundsa
d (ppm) in CDCl3
d (ppm) in DMSO-d6
Lactam Pyrrole Acid Lactam Pyrrole Acid
9H-Dipyrrinone 11.05
10.41
8.58
10.22
10.08
8.14
–
9.72
9.75
9.74
9.76
10.48
10.33
10.35
10.34
11.05
11.04
11.01
–
1
1M
5
4
4M
6
10.58
10.83
10.58
10.70
9.06
13.47
–
–
12.37 10.59
–
–
12.12
–
–
12.11
–
–
8.43
9.46
10.58
10.58
9.66
a See Table 1 for structures.
Figure 4. Ball and stick drawings for the global energy minimum conformational enantiomers of 1. Hydrogen bonds are shown by hatched lines. See Table 6
for torsion angles, dihedral angles and hydrogen bond distances.

M. T. Huggins, D. A. Lightner / Tetrahedron 56 (2000) 1797–1810
1801
Figure 5. Variable low temperature 1H NMR spectra of hemirubin 1 in CD2Cl2, showing (a) the region for the C(10) CH2 resonance and (b) the region for the
NH and COOH resonances.
recently reported tripyrrinone acids and esters,11 is in accord
with a predominantly intramolecularly hydrogen-bonded
conformation of 1 and the presence of intermolecularly
hydrogen-bonded planar dimer in 1M and in 5. In 1, the
intramolecularly hydrogen-bonded conformation (Fig. 3)
does not have the phenyl-ring coplanar with the dipyrrinone
but twisted out of plane at nearly right angles into either of
two mirror image ridge-tile-shaped conformations (Fig. 4).
In the ridge-tile conformation, the pyrrole NH lies above (or
below) the benzene ring, which deshields the NH resonance.
as in 6, are considerably deshielded relative to those of 1 and
1M. In CDCl3, the pyrrole NHs of 4 and 4M exhibit strongly
shielded (ϳ8 ppm) chemical shifts, similar to that of 1 (but
not 1M). In fact, the pyrrole NH chemical shifts in CDCl3 of
4 and 4M are also much more deshielded than that of 6,
consistent with differing ridge-tile conformations. The
lactam NH of 4 exhibits about the same shielding as that
found in 1 and 1M but differs considerably from that of 4M
and of 5.
Molecular weights in solution from vapor pressure
osmometry
The chemical shifts of 4 and 4M are influenced somewhat
by the C(10) carbonyl. Thus, in (CD3)2SO, the lactam and
pyrrole NHs, while nearly identical and essentially the same
There are only a few vapor pressure osmometry (VPO)

1802
M. T. Huggins, D. A. Lightner / Tetrahedron 56 (2000) 1797–1810
measurements of bilirubin pigments in nonpolar solvents—
largely due to limited solubility of the pigment.3 No VPO
measurements of bilirubin acids have been performed, but
bilirubin dimethyl ester (mol. wt. 616) was shown to exhibit
a molecular weight of 850^20 in CDCl3, 1100^30 in THF
and 595^30 in CH3OH.3 These results indicate a dimer in
THF, a monomer in CH3OH, and a mixture of monomer and
dimer in CHCl3.
The molecular weights of 1, 1M, 4, 4M, 5 and 6 in CHCl3
are shown in Table 3. Acids 1 and 4 are clearly monomeric,
but the methyl analogs (5 and 6) are mainly dimeric. Esters
1M and 4M differ: surprisingly, 4M is monomeric, while
1M appears to be a mixture of monomer and dimer.
1
Variable low temperature H NMR spectroscopic
analysis
Figure 6. Partial 1H NMR spectra in CD2Cl2 showing the region of NH
resonance: (A) hemirubin 1 at ϩ30Њ, (B) 1 at Ϫ90ЊC, and (C) hemirubin
methyl ester 1M at Ϫ90ЊC.
We prefer CD2Cl2 solvent to CDCl3 for low temperature
measurements, as it has a lower freezing point. The H
1
NMR spectrum of 1 at 20ЊC is similar to that in CDCl3
and shows the expected sharp singlet for the C(10) CH2
protons near 4.05 ppm (Fig. 5a) in addition to signals for
the remaining hydrogens with essentially no change in
chemical shift. We focused on the C(10) CH2 signal; it is
well-isolated from other proton resonances. It could be seen
to broaden and shift upfield as the temperature is lowered.
At approximately Ϫ50ЊC a new, broad signal appears near
4.5 ppm. Below Ϫ50ЊC, these two signals begin to sharpen
and resolve into doublets, while shifting upfield slightly. At
about Ϫ60ЊC a second set of broad signals appears and
sharpens gradually into two doublets while moving slightly
upfield as the temperature is lowered further to Ϫ90ЊC. We
attribute these interesting changes to two distinct events: (i)
a slowing down of conformational inversion (Fig. 4)
between mirror image intramolecularly hydrogen-bonded
conformers, and (ii) a monomer–dimer equilibrium.
Evidence for the first may be found (Fig. 5a) when the
conformational enantiomerism of Fig. 4 becomes slow on
the NMR timescale and the C(10) hydrogens give clear
evidence of their diastereotopicity. Note that the singlet
near 4.05 ppm at 20ЊC broadens then splits into two doublets
at ϳϪ60ЊC, doublets (one near 4.42–4.52 ppm, the other
near 3.92–4.00 ppm) which sharpen and remain of nearly
equal intensity from Ϫ60–Ϫ90Њ. Evidence for the second, a
monomer–dimer equilibrium may also be found in Fig. 5a
Figure 7. Partial 1H NMR spectra of hemirubin 1 in CD2Cl2 at Ϫ90ЊC, and
the influence of changing concentration on the lactam and pyrrole NH and
COOH.
Table 3. Molecular weights of 1, 1M, 4, 4M, 5 and 6 in CHCl3 from vapor pressure osmometry (solutions were 2.1×10Ϫ3–9.5×10Ϫ3 M. VPO measurements
were carried out at 45ЊC and calibrated vs. benzil (MWˆ210) in CHCl3 (obs MW 220^15))
a
Calculated molecular weight of the monomer.

M. T. Huggins, D. A. Lightner / Tetrahedron 56 (2000) 1797–1810
Table 4. Monomer–dimer composition and Keq for hemirubin 1 at low temperatures
1803
1
Temperature (ЊC)
Monomer/dimer composition by integrationa of 1H NMR signals
Acid COOH Lactam NH Pyrrole NH
Keq based on H NMR signals from
Monomer
Dimer
Monomer
Dimer
Monomer
Dimer
Acid
Lactam
Pyrrole
Ϫ50
Ϫ60
Ϫ70
Ϫ80
0.55
0.81
1.21
1.45
0.50
0.50
0.50
0.41
0.67
1.00
1.46
1.82
0.51
0.55
0.55
0.45
0.63
1.05
1.49
1.80
0.56
0.55
0.55
0.44
1.65
0.76
0.34
0.195
1.14
0.55
0.26
0.14
1.41
0.58
0.25
0.14
a Integration performed using the NUTS NMR data processing program.
with the emergence of a second set of C(10) hydrogen
signals at Ϫ60ЊC, signals (one near 4.25, the other near
3.78 ppm) which also appear as doublets.
sharpen upon further cooling. The new set of signals
continues to grow, relative to the old set, and at ϳ80Њ yet
another new set of very weak signals (ϳ15, ϳ10.9 and
ϳ9 ppm) appears.
The dimerization may be recognized more clearly by
examining the NH and COOH chemical shifts of 1 in
CD2Cl2 over the same temperature range. Figure 5b shows
the anticipated three sharp singlets for each of lactams and
pyrrole NH and COOH hydrogens, with chemical shifts
nicely correlated with an intramolecularly hydrogen-bonded
conformation, as discussed earlier for 1 in CDCl3. Upon
cooling, the signals begin to broaden near 0ЊC, and a new
set of signals begins to emerge at approx. Ϫ40ЊC. Upon
further cooling, the signals sharpen and one clearly sees
(at Ϫ60ЊC) a new COOH signal near 13 ppm and new NH
signals near 11.4 and 10.6 ppm. The chemical shifts of the
original set of signals remains unchanged over the tempera-
ture range studied, and both sets of signals continue to
The lactam and pyrrole NH signals attributed to the dimer of
1 and seen in CD2Cl2 at Ϫ60–Ϫ90ЊC correlate nicely with
those of its ester, 1M in CD2Cl2 at Ϫ90ЊC (Fig. 6), which is
thought to exist mainly as an intermolecularly hydrogen
bonded dimer (as in Fig. 3). In further support of a monomer
Y dimer equilibrium of 1 in CD2Cl2 we observed a concen-
tration dependence of the NH and COOH resonances (Fig.
7). At Ϫ90Њ and 0.002 M concentration, these resonances
from the dimer are noticeably weaker than at 0.01 M, and at
a 0.02 M concentration of 1, they are at least as intense as
the monomer. In addition, one may detect very weak signals
near 15, 11 and 9 ppm. We assume that this weak set of new
signals (seen at Ϫ80 and Ϫ90ЊC) might be associated with a
different sort of dimer, one akin to the stacked dimer
reported earlier for dipyrrinones such as xanthobilirubic
acid with alkanoic acid b-substituents.18
Table 5. Thermodynamic data for the monomer–dimer equilibrium of
hermirubin 1 (from van’t Hoff plots (ln Keq vs. 1/T) the data of Table 4,
with fitting parameters, R-values Ͼ0.998)
Energya
Acid
Lactam
Pyrrole
Average
Similar VT-NMR experiments on the 10-oxo analog 4 show
no temperature dependence of the NH signals. However, as
with 1 (Fig. 5a), we do observe variations in the CH2
resonances of the propionic acid group. At ϩ30ЊC one
observes a broad 2H resonance for the a-protons of the
–C(b)H2–C(a)H2–CO2H segment, and a triplet for the b.
DGЊ (kcal/mol)
DHЊ (kcal/mol)
DSЊ (cal/mol)
3.2
Ϫ5.6
Ϫ29.4
3.4
Ϫ5.6
Ϫ30.2
3.6
Ϫ5.7
Ϫ31.2
3.4
Ϫ5.6
Ϫ30.3
a DGЊ and DHЊ ^0.5 kcal/mol; DSЊ^1.0 cal/mol.
Figure 8. Ball and stick drawings for the global energy minimum conformational enantiomers of 4. Hydrogen bonds are shown by hatched lines. See Table 6
for torsion angles, dihedral angles and hydrogen bond distances.

1804
M. T. Huggins, D. A. Lightner / Tetrahedron 56 (2000) 1797–1810
˚
Table 6. Comparison of torsion and interplanar angles (Њ) and hydrogen bonding distances (A) in global minimum conformations of hemirubins 1 and 4 with
bilirubin (BR) and a 10-oxo-analog
Torsion (f, c) and interplanar
(u) angles (Њ) and hydrogen
˚
bonding distances (d, A)
1 (XˆH2)
4 (XˆO)
BR
64
64
Ϫ17
Ϫ17
96Њ
10-oxo-MBR
f1 (N-9-10-11)
f2 (9-10-11-16) or (9-10-11-N)
c1 (4-5-6-N)
c2 (N-14-15-16)
u
COOH···OvC (lactam) d
HO–CvO···HN (lactam) d
HO–CvO···NH (pyrrole) d
64
66
Ϫ18
–
27
60
29
59
60
Ϫ0.9
Ϫ3.4
88Њ
1.55
1.58
1.58
–
97
76Њ
1.54
1.55
1.89
1.55
1.53
1.58
1.52
1.57
1.59
As the temperature is lowered, these signals broaden.
One of the a-Hs coalescesces into the b-H2 resonance,
then both sharpen somewhat. The isolated a-H reso-
nance is not sharp and thus not completely resolved at
Ϫ90ЊC, and the other a-H (and two b-Hs) lie over-
lapped with the C(7)-ethyl CH2 quartet, making detailed
analysis very difficult.
DG‡298 K are smaller in the more polar CD2Cl2 solvent.
Previously,
DG‡326 K ϳ18 kcal/mol in CDCl3 was
a
determined for conformational enantiomerism in bili-
rubin,19 corresponding to an inversion rate constant of
k ϳ7.2 sϪ1. Since conformational inversion in bilirubin is
thought to involve breaking at least three of six hydrogen
bonds and conformational inversion in 1 involves breaking a
minimum of one of three hydrogen bonds, the similarity in
DG‡ values in CDCl3 for these two pigments seems reason-
able.
Using VT NMR, we were able to sort out some thermo-
dynamic parameters for the: (i) monomer Y dimer equi-
librium and (ii) M Y P conformational enantiomerism of
1. In (i) the exchange rate between monomer and dimer at
low temperature is slow on the NMR timescale; thus, the
equilibrium constant (Keq) may be calculated at various low
temperatures from the relative intensities of the pyrrole NH,
the lactam NH or the acid proton signals (Table 4). The
pyrrole and lactam data gave similar Keq values;
whereas the acid data gave a somewhat larger Keq.
Plots (van’t Hoff) of ln Keq vs. 1/T gave excellent linear
fits and reasonably consistent data for DHЊ (ϳϪ5.6
kcal/mol), DSЊ (ϳϪ30.3 e.u.) and DGЊ (ϳ3.4 kcal/mol)
(Table 5).
Conformational analysis by molecular dynamics
computations
In order to gain further insight into the shape of 1 and 4,
molecular dynamics calculations (Sybyl)20 gave two
enantiomeric global minimum conformations for both.
Those for 1 are shown in Fig. 4; those for 4 are very similar
(Fig. 8) but show a smaller torsion angle about the C(10)–
C(11) bond. The global minima are intramolecularly
hydrogen-bonded, lying some 13 kcal/mol below the
non-hydrogen-bonded global minima. The intramolecularly
hydrogen-bonded global minimum conformation of 1 has
various skeletal torsion angles and an interplanar dihedral
angle very similar to that found in bilirubin (Table 6). Those
for the oxo-analog 4 differ somewhat, especially at f2,
which leads to a smaller u —apparently a reflection of the
change in geometry at C(10) from sp3 to sp2. This change in
geometry also translates into a longer hydrogen bond
between the carboxyl carbonyl oxygen and pyrrole NH in
4 than in 1, thus predicting less effective hydrogen bonding
in the former. The parameters for 4 are similar to those
computed for 10-oxo-mesobilirubin-XIIIa.
For the activation barrier to conformational inversion in 1
(Fig. 4), we used a line-shape analysis of the C(10) CH2
resonance (Fig. 5a). From those data we could determine
k‡ over a wide temperature range in both CD2Cl2 and CDCl3
solvents, and from the k‡ values, using an Eyring plot
(ln k‡/T vs. 1/f ), we determined: DG2‡98 K 9.94 kcal/mol, DH‡
Ϫ0.518 kcal/mol and DS‡ Ϫ35.1 cal/mol/deg for 1 in
CD2Cl2; and DG‡298 K 16.4 kcal/mol, DH‡ Ϫ9.08 kcal/mol
and DS‡ Ϫ85.5 cal/mol/deg for 1 in CDCl3. The nine points
of the two Eyring plots graphed rather well to straight lines,
with an excellent fit (R-values of Ͼ0.998). The computed
DH‡ values are oddly solvent dependent; both DH‡ and
Conformational energy maps5 (Fig. 9) for enantiomerism in

M. T. Huggins, D. A. Lightner / Tetrahedron 56 (2000) 1797–1810
1805
Figure 9. Potential energy surface (left) and contour map (right) for conformations of 1 (upper) and 4 (lower) generated by rotating their dipyrrinone and
phenyl groups independently about the C(9)–C(10) and C(10)–C(11) bonds (f1 and f2, respectively). The energy scale is in kcal/mol. Isoenergetic global
minima for 1 (set to 0 kcal/mol) are found near (f1, f2)ˆ(60, 60Њ) (P-chirality) and near (f1, f2)ˆ(Ϫ 70, Ϫ60Њ), (Ϫ70, 300Њ), (290, Ϫ60Њ), (290, 300Њ)
(M-chirality).Local minima for 1 (ϳ6 kcal/mol above the global minima) are found near the (60, 60Њ) global minimum at (f1, f2)ˆ(110, 130Њ), (250, 110Њ),
(Ϫ60, 140Њ), (Ϫ110, 230Њ) and (250, Ϫ230Њ). Isoenergetic global minima (set to 0 kcal/mol) 4 are found near (f1, f2)ˆ(30, 60Њ) (P-chirality) and near (f1,
f2)ˆ(330, Ϫ60Њ), (330, 350Њ), (Ϫ30, Ϫ60Њ) and (Ϫ30, 300Њ) (M-chirality). Local minima for 4 (ϳ9 kcal/mol above the global minimum at (30, 60Њ)) are found
at (f1, f2)ˆ(150, 130Њ), (230, 210Њ), (230, Ϫ150Њ), (Ϫ130, Ϫ150Њ) and (Ϫ130, 210Њ).
1 and 4 (Figs. 4 and 8) show small, deep valleys ringed by
ridges and peaks, for rotations about the C(9)–C(10) and
C(10)–C(11) bonds, corresponding to f1 and f2, respec-
tively. The centrally-located valley corresponds to the
M-helical global energy minimum (Figs. 4 and 8); the
four surrounding valleys correspond to the P-helical global
energy minima. Interconversion pathways between the M
and P enantiomers may be found from such maps. For 1
(Fig. 9, top), we find the lowest energy path connects the
global minimum at f1, f2ˆ(ϩ60, ϩ60Њ) to its enantiomer
at (Ϫ60, ϩ300Њ) via conformers at (ϩ60, ϩ60Њ) Y (ϩ100,
ϩ120Њ) Y (ϩ60, ϩ180Њ) Y (ϩ40, ϩ260Њ) Y (Ϫ10,
ϩ300Њ) Y (Ϫ60, ϩ300Њ). The saddle point at (ϩ60,
ϩ180Њ) represents the highest energy conformer along the
path, some 10 kcal/mol above the M and P-helical enantio-
mers. Two somewhat higher barriers may be found along
the different pathways (ϩ60, ϩ60Њ) Y (ϩ180, ϩ60Њ) Y
(ϩ220, ϩ40Њ) Y (ϩ270, Ϫ10Њ) Y (ϩ300, Ϫ60Њ), and

1806
M. T. Huggins, D. A. Lightner / Tetrahedron 56 (2000) 1797–1810
Figure 10. Ball and stick representations of the transition state structures lying on the lowest energy inter-conversion pathways for the M Y P conformational
enantiomerism of Figs. 4 and 8. Hemirubin 1 at (A) f1; f2 ϳ ϩ40; ϩ270Њ; (B) f1; f2 ϳ ϩ220; ϩ40Њ; and (C) f1; f2 ϳ Ϫ90; ϩ50Њ: 10-oxo-hemirubin 4 at
(D) f1, f2 ϳ ϩ150, ϩ100Њ; (E) f1, f2 ϳ ϩ190; ϩ40Њ; and (F) f1, f2 ϳ Ϫ70, ϩ50Њ:
Table 7. UV-visible spectral data for hemirubin analogs and the parent 9H-dipyrrinone
l
max (emax)a for
5
Solvent
C6H6
9H-Di-pyrrinone
394(29,400)
389(25,400)
398(22,000)
384(29,700)
395(18,600)
1
1M
4
4M
6
430(28,500)
427(29,800)
414(32,000)
416(27,600)
412(33,900)
410(27,700)
407(29,600)
415(21,300)
402(29,600)
413(22,700)
410(27,500)
407(29,400)
413(39,600)
402(33,600)
413(19,300)
399(23,000)
420(19,100)
401(17,900)
420(14,800)
402(34,500)
423(30,200)
392(22,200)
415(16,100)
405(33,600)
428(32,700)
396(34,000)
415(28,800)
400(27,200)
418(23,400)
403(23,600)
423(20,800)
394(37,200)
414(28,000)
405(30,800)
428(23,400)
405(23,100)
429(21,600)
402(28,800)
425(22,200)
402(16,300)
424(15,000)
394(17,500)
414(13,700)
404(14,800)
428(13,600)
CHCl3
CH3OH
CH3CN
(CH3)SO
a lmax in nm, e in L·molϪ1
.
(ϩ60, ϩ60Њ) Y (Ϫ40, ϩ80Њ) Y (Ϫ90, ϩ50Њ) Y (Ϫ90,
Ϫ30Њ) Y (Ϫ60, Ϫ60Њ), both with barriers of ϳ15 kcal/
mol at transition points (ϩ220, ϩ40Њ) and (Ϫ90, ϩ50Њ),
respectively. For 4, we find three low energy intercon-
version pathways on its conformational energy maps (Fig.
9 bottom). The lowest energy connects the global minimum
at f1, f2ˆ(ϩ60, ϩ60Њ) to its enantiomer at (Ϫ60, ϩ300Њ)
via conformers at (ϩ60, ϩ60Њ) Y (ϩ130, ϩ90Њ) Y (ϩ150,
ϩ130Њ) Y (ϩ120, ϩ170Њ) Y (ϩ30, ϩ240Њ) Y (ϩ60,
ϩ300Њ). The saddle point at (ϩ150, ϩ130Њ) corresponds
to the highest energy conformer along the path, ϳ11 kcal/
mol above the M and P-helical global minima. Two other
pathways, higher energy by 2–5 kcal/mol, are found: (i)
(ϩ60, ϩ60Њ) Y (ϩ120, ϩ70Њ) Y (ϩ190, ϩ40Њ) Y
(ϩ210, ϩ20Њ) Y (ϩ240, 0Њ) Y (300, Ϫ60Њ) and (ii)
(ϩ60, ϩ60Њ) Y (Ϫ20, ϩ70Њ) Y (Ϫ70, ϩ50Њ) Y (Ϫ70,
Ϫ20Њ) Y (Ϫ60, Ϫ60Њ). Route (i) crosses a barrier of
ϳ13 kcal/mol at (ϩ190, ϩ40Њ); route (ii) crosses a barrier
of ϳ15 kcal/mol at (Ϫ70, ϩ50Њ). The structures of the
transition state conformers at the saddle points are shown
in Fig. 10.
Optical spectra
The UV-visible spectral data for hemirubins 1 and 1M and
their 10-oxo-analogs (4 are 4M) in solvents with a wide
range of polarity are shown in Table 7. In polar solvents
capable of engaging in hydrogen bonding, such as CH3OH
or (CH3)2SO, the long wavelength absorption of 1 has nearly
the same lmax and emax as its methyl ester (1M) and very
similar to its o-xylyl analog. In nonpolar solvents, however,
the spectra of 1 and 1M are noticeably different, with 1
showing a strong bathochromic shift of lmax; whereas,
l
max of 1M and the o-xylyl analog are relatively unchanged.
Although the spectral shifts do not unambiguously confirm
an intramolecularly hydrogen bonded structure for 1,

M. T. Huggins, D. A. Lightner / Tetrahedron 56 (2000) 1797–1810
1807
they are consistent with the ability of 1 to adopt a unique
conformational structure in nonpolar solvents. The UV-
visible spectral data for 4 and 4M differ from those of 1
and 1M due to the presence of the carbonyl group at C(10)
and show the major long wavelength absorption near
400 nm, with an inflection or shoulder near 420–430 nm
over the range of solvents studied. The spectra of 4 do not
show the same sensitivity to change in solvent polarity, as
seen in 1. It may be noted that in both 4 and 4M, the main
absorption is bathochromically shifted upon changing from
nonpolar solvents to those capable of hydrogen bonding;
whereas in 1 it is hypsochromically shifted. The spectral
shift seen in comparing the major absorption bonds of 1 to
1M in nonpolar solvents is not evident in the spectra of 4
and 4M.
using a Chromatotron (Harrison Research, Inc., Palo Alto,
CA). Spectral data were obtained in spectral grade solvents
(Aldrich or Fischer). Ceric ammonium nitrate was from
Aldrich, and stannic chloride was from Baker. Dichloro-
methane, methanol, DMSO, acetic acid, tetrahydrofuran,
hexane, and 2-propanol were from Fischer, and 2-pentanone
and propionic anhydride were from Acros.
o-(Methoxycarbonylethyl)benzoyl chloride (7) was
prepared from b-naphthol as described previously.12
o-Toluoyl chloride (8) was prepared by reaction of o-toluic
acid with SOCl2.
9-[2-(Carboxyethyl)benzyl]-3,4-dimethyl-7,8-diethyl-di-
pyrrinone (1). Dipyrrinone (4M) (200 mg, 0.46 mmols)
and 200 mL of 2-propanol were placed in a 100 mL
round-bottom flask equipped for magnetic stirring. Sodium
borohydride (100 mg, 1.7 mmols) was added, and the
reaction mixture was heated at reflux for 2 h. The hot reac-
tion mixture was poured into 100 mL of ice water and the
solution was acidified with 10% aq. HCl. The suspension
was extracted with dichloromethane (3×50 mL), and the
combined organic extracts were washed with water
(3×100 mL) and dried over Na2SO4 (anhydr.). The solvent
was removed (Rotovap), and the crude product was purified
by radial chromatography (97:3 by vol. CH2Cl2:MeOH) and
recrystallized from CH2Cl2–hexane to give 0.182 g of 1,
93% yield. It had mp 209–10ЊC; IR (KBr) n 3440, 3351,
2964, 2928, 2869, 1707, 1664, 1636, 1459, 1373, 1247,
Concluding Comments
The new hemirubin 1 with improved solubility in organic
solvents is found to adopt preferentially either of two enan-
tiomeric, intramolecularly hydrogen-bonded, ridge-tile-like
conformation in nonpolar solvents. The interconversion
barrier to conformational enantiomerism has been deter-
mined by variable low temperature 1H NMR measurements
to be DG‡298 K ϳ16 kcal/mol in CDCl3 and ϳ10 kcal/mol in
CD2Cl2. Variable low temperature 1H NMR studies of 1 also
reveal a monomer Y dimer equilibrium, evident at low
temperatures, with DG‡298 K (equiv.) ϳ3.4 kcal/mol.
Spectroscopic (1H NMR) and molecular modelling analyses
suggest that the 10-oxo-analog (4) also adopts an intra-
1
1181, 942, 735, 694 cmϪ1; H NMR (CDCl3, 500 MHz) d
1.18 (t, Jˆ7.5 Hz, 3 H), 1.24 (t, Jˆ7.5 Hz, 3 H), 1.84 (s, 3
H), 2.05 (s, 3 H), 2.58 (q, Jˆ7.5 Hz, 2 H), 2.61 (q,
Jˆ7.5 Hz, 2 H), 2.97 (t, Jˆ7.5 Hz, 2 H), 3.07 (t,
Jˆ7.5 Hz, 2 H), 4.10 (s, 2 H), 6.07 (s, 1 H), 7.02 (d,
Jˆ7.5 Hz, 1 H), 7.05 (t, Jˆ7.5 Hz, 1 H), 7.14 (t,
Jˆ7.5 Hz, 1 H), 7.23 (d, Jˆ7.5 Hz, 2 H), 8.85 (bs, 1 H),
10.58 (bs, 1 H), 13.47 (bs, 1 H) ppm; 1H NMR (DMSO-d6,
500 MHz) d 0.84 (t, Jˆ7.5 Hz, 3 H), 1.08 (t, Jˆ7.5 Hz, 3
H), 1.76 (s, 3 H), 2.06 (s, 3 H), 2.24 (q, Jˆ7.5 Hz, 2 H), 2.46
(t, Jˆ7.5 Hz, 2 H), 2.51 (q, Jˆ7.5 Hz, 2 H), 2.91 (t,
Jˆ7.5 Hz, 2 H), 3.95 (s, 2 H), 5.95 (s, 1 H), 6.78 (d,
Jˆ7.5 Hz, 1 H), 7.11 (m, 3 H), 9.75 (bs, 1 H), 10.33 (bs,
1 H), 12.12 (bs, 1 H) ppm; with 13C NMR data in Table 1;
and UV-vis data in Table 7. Anal. Calcd for C25H30N2O3
(406.2): C, 73.85; H, 7.44; N, 6.89. Found: C, 73.57; H,
7.53; N, 6.87.
molecularly hydrogen-bonded conformation.
VPO
measurements indicate that while 1, 4 and 4M are clearly
monomeric in CHCl3, 1M shows the presence of some
dimer at 45ЊC.
Experimental
General procedures
All ultraviolet-visible spectra were recorded on a Perkin–
Elmer l-12 spectrophotometer. Nuclear magnetic reso-
nance (NMR) spectra were obtained on GE QE-300 or GE
GN-300 spectrometers operating at 300 MHz, or on a
Varian Unity Plus 500 MHz spectrometer in CDCl3 solvent
(unless otherwise specified). Chemical shifts were reported
in d ppm referenced to the residual CHCl3 1H signal at
7.26 ppm and 13C signal at 77.0 ppm. Heteronuclear Multi-
ple Quantum Coherence (HMQC) and Heteronuclear Multi-
ple Bond Correlation (HMBC) spectra were used to assign
13C NMR spectra. Vapor Pressure Osmometry (VPO)
measurements were performed using an Osmomat 070
(Gonotec, Berlin, Germany) in CHCl3 at 45ЊC with benzil
used for calibration. Melting points were taken on a
MelTemp capillary apparatus and are uncorrected. Com-
bustion analyses were carried out by Desert Analytics,
Tucson, AZ. Analytical thin layer chromatography was
carried out on J. T. Baker silica gel IB-F plates (125 m
layers). Flash column chromatography was carried out
using Woelm silica gel F, thin layer chromatography
grade. Radial chromatography was carried out on Merck
Silica Gel PF254 with gypsum preparative layer grade,
9-[2-(Methoxycarbonylethyl)benzyl]-3,4-dimethyl-7,8-di-
ethyl-dipyrrinone (1M). Dipyrrinone (1) (27 mg, 0.067
mmols) and 50 mL of methanol were added to a 100 mL
round-bottom flask equipped for magnetic stirring. Ten
milliliters of 10% aq. H2SO4 were added to the solution
dropwise over 5 min, and the reaction mixture was heated
at reflux for 1 h. The reaction mixture was cooled to room
temperature, taken up in dichloromethane (50 mL), and
washed with water (2×100 mL) and saturated aq. sodium
bicarbonate solution (2×100 mL). The organic extract was
dried over Na2SO4 (anhydr.) and the solvent removed (Roto-
vap). The residue was purified by radial chromatography
(97:3 by vol. CH2Cl2:MeOH) and recrystallized from
CH2Cl2–hexane) to give the desired product (27 mg) in
95% yield. It had mp 197–199ЊC; IR (KBr) n 3360, 2964,
2929, 1742, 1665, 1631, 1462, 1369, 1277, 1178, 943, 756,

1808
M. T. Huggins, D. A. Lightner / Tetrahedron 56 (2000) 1797–1810
1
692 cmϪ1; 1H NMR (CDCl3, 500 MHz) d 1.00 (t, Jˆ7.5 Hz,
3 H), 1.28 (t, Jˆ7.5 Hz, 3 H), 1.78 (s, 3 H), 2.20 (s, 3 H),
2.41 (q, Jˆ7.5 Hz, 2 H), 2.61 (q, Jˆ8.0 Hz, 2 H), 2.68 (q,
Jˆ7.5 Hz, 2 H), 3.16 (t, Jˆ8.0 Hz, 2 H), 3.77 (s, 3 H), 4.30
(s, 2 H), 6.14 (s, 1 H), 7.15–7.25 (m, 4 H), 10.22 (bs, 1 H),
10.83 (bs, 1 H) ppm; 1H NMR (DMSO-d6, 500 MHz) d 0.80
(t, Jˆ7.5 Hz, 3 H), 1.07 (t, Jˆ7.5 Hz, 3 H), 1.77 (s, 3 H),
2.06 (s, 3 H), 2.22 (q, Jˆ7.5 Hz, 2 H), 2.46 (m, 4 H), 2.93 (t,
Jˆ8.0 Hz, 2 H), 3.55 (s, 3 H), 3.95 (s, 2 H), 5.94 (s, 1 H),
6.84 (d, Jˆ7.5 Hz, 1 H), 7.14 (m, 3 H), 9.74 (bs, 1 H), 10.33
(bs, 1 H) ppm; with 13C NMR data in Table 1; and UV-vis
data in Table 7. Anal. Calcd for C26H32N2O3 (420.2): C,
74.24; H, 7.67; N, 6.66. Found: C, 73.98; H, 7.49; N, 6.54.
1282, 1168, 928, 758, 690, 668 cmϪ1; H NMR (CDCl3,
500 MHz) d 0.96 (t, Jˆ7.5 Hz, 3 H), 1.15 (t, Jˆ7.5 Hz, 3
H), 1.91 (s, 3 H), 2.10 (s, 3 H), 2.37 (q, Jˆ7.5 Hz, 2 H), 2.51
(q, Jˆ7.5 Hz, 2 H), 2.67 (t, Jˆ7.5 Hz, 2 H), 3.00 (t,
Jˆ7.5 Hz, 2 H), 3.62 (s, 3 H), 6.07 (s, 1 H), 7.26–7.40
1
(m, 4 H), 8.43 (bs, 1 H), 9.06 (bs, 1 H) ppm; H NMR
(DMSO-d6, 500 MHz) d 0.77 (t, Jˆ7.5 Hz, 3 H), 1.05 (t,
Jˆ7.5 Hz, 3 H), 1.80 (s, 3 H), 2.06 (q, Jˆ7.5 Hz, 4 H), 2.09
(s, 3 H), 2.60 (t, Jˆ7.0 Hz, 2 H), 2.83 (t, Jˆ7.0 Hz, 2 H),
3.55 (s, 3 H), 5.93 (s, 1 H), 7.31–7.45 (m, 4 H), 10.58 (bs, 1
H), 11.04 (bs, 1 H) ppm; with 13C NMR data in Table 1; and
UV-vis data in Table 7. Anal. Calcd for C26H30N2O4
(434.2): C, 71.85; H, 6.96; N, 6.45. Found: C, 71.45; H,
6.77; N, 6.33.
9-[2-(Carboxyethyl)benzoyl]-3,4-dimethyl-7,8-diethyl-di-
pyrrinone (4). Dipyrrinone 4M (73 mg, 0.16 mmols) and
50 mL of THF were placed in a 100 mL round-bottom flask
equipped for magnetic stirring. Ten milliliters of 2 M NaOH
(aq.) were added; then the mixture was heated at reflux for
3 h, quenched by pouring into ice–water and acidified with
10% aq. HCl. The suspension was extracted into dichloro-
methane (3×70 mL) and washed with water (3×100 mL).
The combined organic extracts were dried over anhydr.
sodium sulfate and evaporated (Rotovap). The residue was
purified by radial chromatography (97:3 by vol.
CH2Cl2:MeOH) and recrystallized from CH2Cl2–hexane).
Pure acid 4 (70 mg) was obtained in 85% yield. It had mp
237–238ЊC; IR (KBr) n 3365, 3194, 2966, 2921, 1688,
9-(2-Methylbenzyl)-3,4-dimethyl-7,8-diethyl-dipyrrinone
(5). Dipyrrinone 6 (97 mg, 0.27 mmols) and 200 mL of
2-propanol were placed in a 100 mL round-bottom flask
equipped for magnetic stirring. Sodium borohydride
(100 mg, 1.7 mmols) was added, and the reaction mixture
was heated at reflux for 2 h. The hot reaction mixture was
poured into 100 mL of ice–water and the solution was acidi-
fied with 10% aq. HCl. The suspension was extracted with
dichloromethane (3×50 mL), and the combined organic
extracts were washed with water (3×100 mL) and dried
over Na2SO4 (anhydr.). The solvent was removed (Rotovap),
and the crude product was purified by radial chroma-
tography (97:3 by vol. CH2Cl2:MeOH) and recrystallized
from CH2Cl2–hexane to give 93 mg of 5, 92%. It had mp
258–259ЊC; IR (KBr) n 3469, 3348, 2967, 2915, 1666,
1
1602, 1429, 1398, 1290, 1158, 1024, 928, 758 cmϪ1; H
NMR (CDCl3, 500 MHz) d 1.18 (t, Jˆ7.5 Hz, 3 H), 1.27
(t, Jˆ7.5 Hz, 3 H), 1.85 (s, 3 H), 2.08 (s, 3 H), 2.55 (q,
Jˆ7.5 Hz, 2 H), 2.84 (bs, 2 H), 2.90 (bs, 2 H), 3.07 (bs, 2
H), 6.05 (s, 1 H), 7.10–7.40 (m, 4 H), 8.14 (bs, 1 H), 10.70
(bs, 1 H), 12.37 (bs, 1 H) ppm; 1H NMR (DMSO-d6,
500 MHz) d 0.77 (t, Jˆ7.5 Hz, 3 H), 1.05 (t, Jˆ7.5 Hz, 3
H), 1.80 (s, 3 H), 2.06 (m, 4 H), 2.09 (s, 3 H), 2.52 (t,
Jˆ7.5 Hz, 2 H), 2.79 (t, Jˆ7.5 Hz, 2 H), 5.93 (s, 1 H),
7.30–7.50 (m, 4 H), 10.59 (bs, 1 H), 11.05 (bs, 1 H),
12.11 (bs, 1 H) ppm; with 13C NMR data in Table 1; and
UV-vis data in Table 7. Anal. Calcd for C25H28N2O4
(420.2): C, 71.39; H, 6.76; N, 6.66. Found: C, 71.12; H,
6.64; N, 6.52.
1
1636, 1457, 1369, 1274, 1179, 739, 694 cmϪ1; H NMR
(CDCl3, 500 MHz) d 0.91 (t, Jˆ7.5 Hz, 3 H), 1.16 (t,
Jˆ7.5 Hz, 3 H), 1.62 (s, 3 H), 2.06 (s, 3 H). 2.33 (s, 3 H),
2.57 (q, Jˆ7.5 Hz, 2 H), 2.71 (q, Jˆ7.5 Hz, 2 H), 4.10 (s, 2
H), 6.11 (s, 1 H), 6.90–7.25 (m, 4 H), 10.08 (bs, 1 H), 10.85
1
(bs, 1 H), ppm; H NMR (DMSO-d6, 500 MHz) d 0.86 (t,
Jˆ7.5 Hz, 3 H), 1.08 (t, Jˆ7.5 Hz, 3 H), 1.76 (s, 3 H), 2.06
(s, 3 H), 2.24 (q, Jˆ7.5 Hz, 2 H), 2.33 (s, 3 H), 2.52 (q,
Jˆ7.5 Hz, 2 H), 3.87 (s, 2 H), 5.95 (s, 1 H), 6.78–7.23 (m, 4
H), 9.76 (bs, 1 H), 10.34 (bs, 1 H), ppm; with 13C NMR data
in Table 1; and UV-vis data in Table 6. Anal. Calcd for
C23H28N2O (348.2): C, 79.26; H, 8.10; N, 8.04. Found: C,
79.27; H, 8.19; N, 8.26.
9-[2-(Methoxycarbonylethyl)benzoyl]-3,4-dimethyl-7,8-
diethyl-dipyrrinone
(4M).
3,4-Dimethyl-7,8-diethyl
9-(2-Methylbenzoyl)-3,4-dimethyl-7,8-diethyl-dipyrrinone
(6). 3,4-Dimethyl-7,8-diethyl dipyrrinone (9) (0.30 g, 1.22
mmols) and 100 mL of dichloromethane were added to a
500 mL round-bottom flask equipped for magnetic stirring.
The solution was cooled in an ice bath for 20 min with
stirring; then a solution of acid chloride (8) (0.6 g,
3.9 mmols) and SnCl4 (2.4 g, 9.2 mmols) in 100 mL of
dichloromethane was added in one portion. The reaction
mixture was stirred for 1.5 h at room temperature then
poured into a mixture of conc. HCl (200 mL) and 100 g of
ice and stirred for 2 h. The organic layer was separated, and
the aqueous layer extracted with dichloro-methane
(2×100 mL). The combined organic extracts were washed
with sat. aq. NaHCO3 (2×200 mL) then with water
(400 mL), and dried over anhydr. Na2SO4. The solvent
was removed (Rotovap), and the crude product was purified
by radial chromatography (97:3 by vol. CH2Cl2:MeOH) and
recrystallized from CH2Cl2–hexane to afford 0.223 g, 50%
of 6. It had mp 237ЊC; IR (KBr) n 3460, 3351, 2967, 2930,
dipyrrinone (9) (0.300 g, 1.22 mmols) and 100 mL of
dichloromethane were added to a 500 mL round-bottom
flask equipped for magnetic stirring. The solution was
cooled in an ice bath for 20 min with stirring; then a solution
of acid chloride 7 (1.0 g, 4.4 mmols) and SnCl4 (4.5 g,
17.3 mmols) in 100 mL of dichloromethane was added in
one portion. The reaction mixture was stirred for 1.5 h at
room temperature then poured into a mixture of conc. HCl
(200 mL) and 100 g of ice and stirred for 2 h. The organic
layer was separated, and the aqueous layer extracted with
dichloromethane (2×100 mL). The combined organic
extracts were washed with sat. aq. NaHCO3 (2×200 mL)
then with water (400 mL), and dried over anhydr. Na2SO4.
The solvent was removed (Rotovap), and the crude product
was purified by radial chromatography (97:3 by vol.
CH2Cl2:MeOH) and recrystallized from CH2Cl2–hexane
to afford 0.313 g, 59% of 4M. It had mp 167–168ЊC; IR
(KBr) n 3344, 2966, 2872, 1735, 1660, 1607, 1431, 1405,

M. T. Huggins, D. A. Lightner / Tetrahedron 56 (2000) 1797–1810
1809
1659, 1609, 1431, 1403, 1275, 1166, 926, 762, 728,
682 cmϪ1 1H NMR (CDCl3, 500 MHz) d 0.82 (t,
(2×300 mL), and dried over sodium sulfate (anhydr.). The
solvent was removed (Rotovap) to yield an oily residue
which was crystallized from methanol–water to give 6.0 g
of solid 12 (80%). It had mp 50–51ЊC (lit.5 mp 53ЊC); and
1H NMR (CDCl3, 300 MHz) 1.15 (t, Jˆ7.5 Hz, 3 H), 1.23 (t,
Jˆ7.5 Hz, 3 H), 1.37 (t, Jˆ7.5 Hz, 3 H), 2.73 (q, Jˆ7.5 Hz,
4 H), 4.36 (q, Jˆ7.5 Hz, 2 H), 9.52 (bs, 1 H), 9.78 (bs, 1 H)
ppm.
;
Jˆ7.5 Hz, 3 H). 1.13 (t, Jˆ7.5 Hz, 3 H), 1.94 (s, 3 H),
2.10 (s, 3 H), 2.13 (q, Jˆ7.5 Hz, 2 H), 2.32 (s, 3 H), 2.50
(q, Jˆ7.5 Hz, 2 H), 5.96 (s, 1 H), 7.26–7.40 (m, 4 H), 9.46
(bs, 1 H), 9.66 (bs, 1 H) ppm; 1H NMR (DMSO-d6,
500 MHz) d 0.77 (t, Jˆ7.5 Hz, 3 H), 1.05 (t, Jˆ7.5 Hz, 3
H), 1.80 (s, 3 H), 2.08 (s, 3 H), 2.10 (q, Jˆ7.5 Hz, 2 H), 2.24
(s, 3 H), 2.49 (q, Jˆ7.5 Hz, 2 H), 5.93 (s, 1 H), 7.25–7.50
(m, 4 H), 10.58 (bs, 1 H), 11.01 (bs, 1 H) ppm; with 13C
NMR data in Table 1; and UV-vis data in Table 7. Anal.
Calcd for C23H26N2O2 (362.2): C, 76.20; H, 7.23; N, 7.73.
Found: C, 76.05; H, 7.12; N, 7.63.
5-Methyl-3,4-diethyl-2-carboethoxy-1H-pyrrole (13). To
a 250 mL round-bottom flask equipped for magnetic stirring
was added 3-ethyl-2,4-hexanedione-O2, O4-difluoroboro-
nate (9.0 g, 47 mmols) and 25 mL of methanol. A solution
of 50% aqueous NaOH was added to adjust the pH to ϳ9.0.
The reaction mixture was heated at reflux for 30 min
followed by removal of the methanol (Rotovap). The resi-
due was taken up in 75 mL of glacial acetic acid, and diethyl
oximinomalonate (17.5 g, 95 mmols) was added in one
portion. Zinc dust (12.5 g) was added slowly, such that
the reaction temperature was maintained between 85–
95ЊC. After the addition was complete, the reaction mixture
was heated at reflux for 1 h then decanted into 2 L of ice-
water. The solid product was collected by filtration and
recrystallized from methanol–water to yield 6.1 g pure
product (62% yield). It had mp 73–75ЊC (lit.9 mp 74–
76ЊC); GC-MS, m/z (rel. intens): 209, 194, 162, 148
(100%), 120, 91, 77, 53 amu; and 1H NMR (CDCl3,
300 MHz) 1.08 (t, Jˆ7.5 Hz, 3 H), 1.15 (t, Jˆ7.5 Hz, 3
H), 1.35 (t, Jˆ7.5 Hz, 3 H), 2.21 (s, 3 H), 2.39 (q,
Jˆ7.5 Hz, 2 H), 2.72 (q, Jˆ7.5 Hz, 2 H), 4.30 (q,
Jˆ7.5 Hz, 2 H), 8.80 (s, 1 H) ppm.
3-Ethyl-2,4-Hexanedione-O2,O4-difluoroboronate (14).
Boron trifluoride etherate (59 mL, 0.48 mol) and 2-penta-
none (50 mL, 0.47 mol) were added to a 1 L round-bottom
flask fitted with a drying tube (anhydr. CaCl2) and equipped
for magnetic stirring. The solution was stirred for 2 h at
room temperature, then propionic anhydride (60 mL,
0.47 mol) was added, and the reaction mixture was heated
at reflux for 2 h. The solution was then poured into 2 L of
ice-water, and the solid complex was collected by filtration
(vacuum). The crude product was recrystallized from
methanol–water to give the desired product, 32 g, in 38%
yield. It had mp 76–77ЊC (lit.15 mp 75–77ЊC); and 1H NMR
(CDCl3, 300 MHz) 1.10 (t, Jˆ7.5 Hz, 3 H), 1.25 (t,
Jˆ7.5 Hz, 3 H), 2.34 (s, 3 H), 2.36 (q, Jˆ7.5 Hz, 2 H),
2.63 (q, Jˆ7.5 Hz, 2 H) ppm.
2,3,-Dimethyl-7,8-diethyl-9H-dipyrrinone (9). 3,4-Diethyl-
5-formyl-2-carboethoxy-1H-pyrrole (12) (6.0 g, 27 mmols),
3,4-dimethyl-1H-pyrrolin-2-one (11) (3.0 g, 27 mmols) and
50 mL of methanol were placed in a 500 mL round-bottom
flask equipped for magnetic stirring. 200 milliliters of 4 M
KOH (aq.) were added, and the reaction mixture was heated
at reflux for 4 h then chilled in an ice bath for 30 min
followed by acidification with HCl (conc) to ϳpH 3. The
resultant solid product (10) was collected by filtration
(vacuum), washed with water (100 mL), and dried in
vacuo. The resulting powder was placed in a 500 mL
round-bottom flask and mixed with 3.0 g of potassium
acetate and 3.0 g of sodium acetate trihydrate which had
been ground with a mortar and pestle until intimately
mixed. The solid mixture was heated to ϳ160Њ at which
time it melted with the evolution of CO2. The temperature
was maintained at ϳ160Њ until the evolution of CO2 ceased,
ϳ10 min The reaction mixture was cooled to room tempera-
ture and 400 mL of water was added followed by vigorous
stirring for 1 h. The product was collected by filtration
(vacuum) and triturated with cold acetone (50 mL) to give
pure dipyrrinone 9, 2.1 g, 56% yield. It had mp 203–204ЊC;
IR (KBr) n 3397, 3193, 3150, 2953, 2910, 1648, 1507,
1398, 1267, 1174, 1114, 939, 793, 750, 690 cmϪ1 1H
;
NMR (CDCl3, 500 MHz) d 1.13 (t, Jˆ7.5 Hz, 3 H), 1.21
(t, Jˆ7.5 Hz, 3 H), 1.94 (s, 3 H), 2.13 (s, 3 H), 2.52 (q,
Jˆ7.5 Hz, 2 H), 2.82 (q, Jˆ7.5 Hz, 2 H), 6.20 (s, 1 H),
6.86 (d, Jˆ2.5 Hz, 2 H), 10.41 (bs, 1 H) 11.05 (bs, 1 H)
ppm; 1H NMR (DMSO-d6, 500 MHz) d 1.04 (t, Jˆ7.5 Hz, 3
H), 1.13 (t, Jˆ7.5 Hz, 3 H), 1.77 (s, 3 H), 2.05 (s, 3 H), 2.37
(q, 7.5 Hz, 2 H), 2.47 (q, Jˆ7.5 Hz, 2 H), 5.94 (s, 1 H), 6.72
(d, Jˆ2.5 Hz, 2 H), 9.72 (bs, 1 H) 10.48 (bs, 1 H) ppm; and
13C NMR (CDCl3, 125 MHz) d 8.52, 10.18, 13.10, 16.78,
17.94, 18.47, 101.36, 120.54, 123.83, 124.34, 125.94,
129.90, 130.41, 142.63, 174.46 ppm; The UV-vis data are
in Table 7. Anal. Calcd for C215H20N2O (244.2): C, 73.72;
H, 8.26; N, 11.47. Found: C, 73.93; H, 8.19; N, 11.19.
Acknowledgements
We thank the US National Institutes of Health (HD 17779)
for generous support of this work. We also thank Prof. T. W.
Bell of our department for generous use of the vapor
pressure osmometry apparatus. Michael T. Huggins is an
R. C. Fuson Graduate Fellow.
3,4-Diethyl-5-formyl-2-carboethoxy-1H-pyrrole (12). To
a 1 L round-bottom flask equipped for magnetic stirring was
added 5-methyl-3,4-diethyl-2-carboethoxy-1H-pyrrole (13)
(7.0 g, 33 mmols), 120 mL of tetrahydrofuran, 240 mL of
acetic acid, and 200 mL of water. The reaction mixture was
cooled in a ice bath for 30 min; then ceric ammonium nitrate
(73.0 g, 132 mmols) was added in one portion, and the
mixture was stirred at room temperature for 2 h. The
reaction mixture was extracted with dichloromethane
(3×150 mL) and the combined organic extracts were
washed with water (3×400 mL), sat. aq. sodium bicarbonate
References
1. Chowdury, J. R.; Wolkoff, A. W.; Chowdury, N. R.; Arias, I. M.
Hereditary Jaundice and Disorders of Bilirubin Metabolism. In The
Metabolic and Molecular Bases of Inherited Disease, Scriver,

1810
M. T. Huggins, D. A. Lightner / Tetrahedron 56 (2000) 1797–1810
C. R.; Beaudet, A. L.; Sly, W. S.; Valle, D., Eds.; McGraw-Hill:
New York, 1979; II, pp 2161–2208.
6033. (f) McDonagh, A. F.; Lightner, D. A.; Agati, G. Montash.
Chem. 1998, 129, 649–660.
2. McDonagh, A. F. In The Porphyrins, Dolphin, D., Ed.;
Academic Press: New York, 1979; 6, pp 293–491.
3. For leading references, see Falk, H. The Chemistry of
Linear Oligopyrroles and Bile Pigments; Springer: Vienna,
1989.
4. Nogales, D. F.; Ma, J.-S.; Lightner, D. A. Tetrahedron 1993, 49,
2361–2372.
5. Person, R. V.; Peterson, B. R.; Lightner, D. A. J. Am. Chem.
Soc. 1994, 116, 42–59.
6. (a) Falk, H. Molecular Structure of Bile Pigments. In Bilirubin;
Heirwegh, K. P. M.; Brown, S. B., Eds.; CRC Press: Boca Raton,
FL, 1982; Vol. 1, pp 7–29. (b) Shelver, W. L.; Rosenberg, H.;
Shelver, W. H. Intl. J. Quantum Chem. 1992, 44, 141–163. (c)
Shelver, W. L.; Rosenberg, H.; Shelver, W. H. J. Mol. Struct.
1994, 312, 1–9.
10. Trull, F. R.; Ma, J.-S.; Landen, G. L.; Lightner, D. A. Israel J.
Chem. 1983, 23, 211–218.
11. Tipton, A. K.; Lightner, D. A. Montash. Chem. 1999, 130,
425–440.
12. Chen, Q.; Lightner, D. A. J. Org. Chem. 1998, 63, 2665–2675.
13. (a) Galliani, G.; Monti, D.; Speranza, G.; Manitto, P.
Tetrahedron Lett. 1984, 25, 6037–6040. (b) Ostrow, J. D. Semin.
Hematol. 1972, 9, 113–125.
14. Chen, Q. Q.; Huggins, M. T.; Lightner, D. A.; Norona, W.;
McDonagh, A. F. J. Am. Chem. Soc. 1999, 40, 9253–9264.
15. Wang, C.-B.; Chang, C. K. Synthesis 1979, 548–549.
16. (a) Xie, M.; Lightner, D. A. Tetrahedron 1993, 49, 2185–
2200. (b) Kar, A.; Lightner, D. A. Tetrahedron 1998, 54,
12671–12690. (c) Boiadjiev, S.; Lightner, D. A. J. Org. Chem
1998, 63, 6220–6228.
7. (a) Nogales, D.; Lightner, D. A. J. Biol. Chem. 1995, 270, 73–
17. (a) Kaplan, D.; Navon, G. Israel J. Chem. 1983, 23, 177–186.
(b) Kaplan, D.; Navon, G. Biochem. J. 1982, 201, 605–613. (c)
Navon, G.; Frank, S.; Kaplan, D. J. Chem. Soc., Perkin Trans. 2
1984, 1145–1149.
18. (a) Boiadjiev, S. E.; Anstine, D. T.; Lightner, D. A. J. Am.
Chem. Soc. 1995, 117, 8727–8736. (b) Boiadjiev, S. E.; Anstine,
D. T.; Maverick, E.; Lightner, D. A. Tetrahedron: Asymmetry
1995, 6, 2253–2270.
¨
77. (b) Dorner, T.; Knipp, B.; Lightner, D. A. Tetrahedron 1997,
53, 2697–2716.
8. (a) Bonnett, R.; Davies, J. E.; Hursthouse, M. B.; Sheldrick,
G. M. Proc. R. Soc. London, Ser. B 1978, 202, 249–268. (b)
LeBas, G.; Allegret, A.; Mauguen, Y.; DeRango, C.; Bailly, M.
Acta Crystallogr., Sect. B 1980, B36, 3007–3011. (c) Becker, W.;
Sheldrick, W. S. Acta Crystallogr., Sect. B, 1978, B34, 1298–1304.
(d) Mugnoli, A.; Manitto, P.; Monti, D. Acta Crystallogr., Sect. C
1983, 38, 1287–1291.
19. Manitto, P.; Monti, D. J. Chem. Soc. Chem. Commun. 1976,
122–123.
9. (a) Lightner, D. A. Photochem. Photobiol. 1977, 26, 427–436.
(b) Lightner, D. A.; Park, Y.-T. Tetrahedron 1979, 35, 463–471.
(c) Lamola, A. A.; Braslavsky, S. E.; Schaffner, K.; Lightner, D. A.
Photochem. Photobiol. 1983, 37, 263–270. (d) Landen, G. L.;
Park, Y.-T.; Lightner, D. A. Tetrahedron 1983, 39, 1893–1907.
(e) Byun, Y. S.; Lightner, D. A. J. Org. Chem. 1991, 56, 6027–
20. Molecular dynamics calculations were carried out on an SGI
Octane workstation using version 6.5.1 of SYBAL (Tripos Assoc.,
St. Louis, MO) as described in Ref. [5]. The ball and stick drawings
were created from the atomic coordinates of the molecular
dynamics structures using Mu¨ller and Falk’s ‘Ball and Stick’
program for the Macintosh.