C. Yoshida et al.
Bioorganic & Medicinal Chemistry Letters 37 (2021) 127837
aldehydes 6b, 9b, 11b, and 13 were prepared in a stepwise manner.
Wittig reactions with these aldehydes gave ethyl esters 3a–e, phenyl
ketones 3o and 3p, and benzyl esters 3q and 3r. Since it was difficult to
purify ethyl ester 3d as a single component, it was used as a mixture of
cis/trans isomers in growth inhibition assays.
Ethyl ester 3f was prepared from the commercially available 1-
methyl-1H-pyrrole-2-carboxaldehyde, 4c, via aldehyde 9c with
a
repeated combination of Wittig reaction, DIBAL reduction, and hydro-
lysis, and a final Horner-Wadsworth-Emmons (HWE) reaction with
triethyl 4-phosphonocrotonate.
2
1
22
Ethyl esters 3g–i and 3j–l were prepared from furfural 4e and
thiazole-2-carboxaldehyde 4d, respectively. Ethyl esters 3g and 3j were
derived from 4d and 4e, respectively, using HWE reactions. DIBAL re-
ductions of 4d and 4e, and the following Dess-Martin oxidations, gave
aldehydes 9d and 9e, respectively. Wittig reactions of 9d and 9e with
2
3
ethyl (triphenylphosphoranylidene)acetate gave esters 3h and 3k
,
respectively, and HWE reaction of 9d with triethyl 4-phosphonocroto-
nate gave 3i.
The growth-suppressing activity of all synthesized analogs was
evaluated in T-ALL-derived CCRF-CEM cells and Burkitt lymphoma-
derived Raji cells to obtain a half-maximal inhibitory concentration
Fig. 1. Molecular design of the analogs based on the chemical structures of the
seed compounds 1 and 2. The 3-chloropyrrole and 4-methoxy-2-pyrone rings of
(
IC50) value. Also, we defined selectivity as the difference in cell viability
between CCRF-CEM cells and Burkitt lymphoma-derived Raji cells upon
the addition of 10 M of compounds. The detailed results are shown in
1
and 2 correspond to the partial structures A and B, respectively.
μ
related compounds and evaluated their growth inhibitory activities in
CCRF-CEM, Raji, and solid tumor-derived cell lines. We investigated the
structure–activity relationships for the prepared analogs, along with
their effect on tumor cell viability. These analogs were designed
depending on three structural key points based on the seed compounds 1
and 2: The pyrrole ring was replaced by another heterocyclic ring; the
number of double bonds was changed from 1 to 5; the pyrone ring was
simplified (Fig. 1). We aimed to successfully obtain an octate-
traenylpyrrole analog that is a promising and novel anti-T-ALL com-
pound with a superior biological activity to the seed compounds and
provide an important direction for establishing a new anti-T-ALL drug.
In order to design analogs of the seed compounds 1 and 2, we divided
their structures into the three parts, as shown in Fig. 1. That is, 1 and 2
have 3-chloropyrrole and 4-methoxy-2-pyrone rings as the partial
structures A and B, and they are conjugated by an octatetraene linker. In
this study, we planned to modify the linker by varying the number of
double bonds between 1 and 5. For the partial structure A, we planned to
introduce pyrrole, N-methylpyrrole, thiazole, and furan rings, as similar
structures to the 3-chloropyrrole ring. In addition, we intended to
introduce ethyl ester, benzyl ester, and phenyl ketone moieties as
simpler structures than the 4-methoxy-2-pyrone ring of the partial
structure B. By following the modification guidelines described above,
we prepared 20 analogs (3a–t) to investigate the structure–activity
relationship.
Supplemental Information (Figs. S1–18) and summarized in Table 1.
Analogs 3a–e had a pyrrole ring and the ethyl ester moiety at A and B
regions, respectively, and the number of double bonds varied from 1 to 5
(
Table 1). The killing effect in CCRF-CEM cells and the selectivity for
CCRF-CEM vs Raji cells increased as the number of double bonds
increased. Especially, pentaene 3e exhibited an anti-T-ALL activity,
comparable to that of 1 and 2, and an increased CCRF-CEM vs Raji
selectivity. Therefore, the decapentaene structure of 3e, similar to the
octatetraene linker structures of 1 and 2, was essential for the T-ALL
growth inhibitory activity.
Ethyl esters 3f, 3g–i, and 3j–l had N-methylpyrrole, thiazole, and
furan rings, respectively, at the A region. The number of double bonds of
each analog was between 2 and 4 (Table 1). While octatetraene, 3d, had
moderate anti-T-ALL activity, the N-methylpyrrole analog, 3f, showed a
negligible activity. Thiazole analogs 3g and 3i had no cytotoxic activity
in both CCRF-CEM and Raji cells. Only triene 3h showed a weak cyto-
toxic activity in CCRF-CEM cells, but the CCRF-CEM vs Raji selectivity
was low. Furan analogs 3j–l showed no T-ALL growth inhibitory activity
even when the number of double bonds increased. These results indi-
cated that the pyrrole ring of the seed compounds was critically crucial
for potent inhibition of T-ALL proliferation. In addition, it was also
necessary to have the NH moiety in the pyrrole ring for T-ALL growth-
inhibiting activity.
Ethyl esters 3m, 3n, and 3s had the 3-chloropyrrole ring, similar to
the seed compounds 1 and 2. The number of double bonds varied from 2
to 4 (Table 1). The T-ALL growth-inhibiting activity increased with the
increased number of double bonds, which was observed in 3a–e with the
unsubstituted pyrrole ring as well. The IC50 value of octatetraene 3s was
approximately five times higher than that of 3d (Table 1, Fig. 3 left).
Thus, the chlorine atom at the C3 position of the pyrrole ring in the seed
compounds had an essential role in inhibiting T-ALL growth.
Designed analogs were synthesized following the procedure shown
1
5
in Fig. 2. As a starting compound for preparing ethyl esters 3m, 3n, 3s
,
and benzyl ester 3t15, aldehyde 4a was prepared by the photoreaction
of commercially available 4-chloropyridine N-oxide in an aqueous so-
lution containing copper sulfate with a low-pressure mercury lamp.
Nitrile 5a was synthesized by a Wittig reaction of 4a, followed by the
diisobutylaluminum hydride (DIBAL) reduction and hydrolysis, which
produced aldehyde 6a. Following a similar procedure, aldehyde 9a was
prepared from 6a via nitrile 7a, and aldehyde 11a was prepared from 9a
via nitrile 10a. Ethyl esters 3m, 3n and 3s were prepared from the
corresponding aldehydes 6a, 9a, and 11a, respectively, by Wittig re-
actions with ethyl (triphenylphosphoranylidene)acetate. Benzyl ester 3t
was prepared from 11a by a Wittig reaction with benzyl (triphenyl-
phosphoranylidene)acetate.
16
As another modification pattern, 3o and 3p, along with 3q and 3r
had phenyl ketone, and benzyl ester moieties, respectively, at the B re-
gion. They had the pyrrole ring at the A region with the linker, con-
taining 3 or 4 double bonds (Table 1). Similar to the observed tendency,
octatetraenes 3p and 3r had higher anti-T-ALL activities than the cor-
responding hexatrienes 3o and 3q. Among the series of octatetraene
analogs 3d, 3p, and 3r, the T-ALL growth-inhibiting activity increased
in the order of ethyl ester, phenyl ketone, and benzyl ester. The CCRF-
CEM vs Raji selectivity also improved in the order of 3d, 3p, and 3r.
This result implied that the phenyl group of 3p and 3r did not induce a
drastic decrease in the anti-T-ALL activity. An intermolecular
Ethyl esters 3a–e17–19, phenyl ketones 3o, 3p20 and benzyl esters 3q,
were prepared from commercially available pyrrole-2-
3
r
carboxaldehyde 4b as a starting compound. Similar to the production
of 6a from 4a via Wittig reaction, DIBAL reduction and hydrolysis,
2
Yoshida, Chihiro
