Full Papers
doi.org/10.1002/cmdc.202000793
ChemMedChem
3-Methylbutanal 14 was condensed with (S)-2-meth-
with TFA or HCl to obtain the methoxy as well as the
corresponding unsubstituted analogues thereof (Scheme 5).
All new compounds were tested in vitro for their ability to
inhibit bacterial (Escherichia coli) or eukaryotic (rabbit reticulo-
cyte) protein synthesis (in vitro transcription translation assay
(IVTT)).[17] The results are summarised in Table 1.
Novel xenocoumacin analogues were identified that inhibit
protein synthesis in vitro at concentrations between 10 nM and
>100 μM and showed very different potencies against eukary-
otic or bacterial ribosomes. Surprisingly, despite virtually
identical RNA sequence at the binding sites, the ribosome of
eukaryotic systems is much less susceptible to structural
changes of the inhibitors. While even small modifications
drastically affected the inhibition of the ribosome of E. coli, the
compounds still remained potent inhibitors of eukaryotic
protein synthesis.
Conservative modifications on the aromatic portion of 2,
such as presence or absence of hydroxy or methoxy group,
presence or absence of double bond does not have a profound
effect on the activity (cf. compounds 2, 19–23) in vitro. Bigger
losses of potency were observed for the R,R analogue 24 with a
~100-fold higher IC50 value in the micromolar range.
We chose the unsubstituted isochromenone (R=H) inter-
mediate 13 as suitable simplified aromatic moiety for most of
our investigations.
Modifications of the isobutyl group (intercalating moiety,
compounds 25–30) led to much bigger loss of activity in the E.
coli IVTT assay (1.27!100 μM), whereas in the reticulocyte assay
sub-micromolar activities were generally still measured.
Variations in the linker region affected activity less than
expected. In the X-ray structure of 1, the two hydroxy groups
and the amine are tightly bound in a H-bond network and were
therefore expected to be essential for inhibitory activity. The
two analogues with one inverted hydroxy group (31 and 32,
absolute configuration of diol not assigned) as compared to 23
still retained some activity and showed IC50’s in the double-digit
micromolar range (about 100-fold less potent) in the bacterial
assay and are about 15 to 150 times more active in the
reticulocyte IVTT assay. The same range of activity was observed
1
2
3
4
5
6
7
8
9
ylpropane-2-sulfinamide[21] and treated with the anion of
trimethylsilylacetylene (generated in situ by treatment with
iPrMgCl) to give in good yield and with good diastereoselectiv-
ity the required intermediate which after cleavage of the
trimethylsilyl group with tetrabutylammonium fluoride (TBAF)
led to acetylene 15. Copper catalysed coupling of 15 with 2-
iodobenzoic acid gave the desired enantiomer of isochrome-
none 16 in modest yields, along with some phthalide by-
product. Deprotection with HCl in MeOH yielded the desired
unsubstituted isochromenone 13.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
All our attempts to synthesise the analogous methoxy
derivative 11 via the same route failed, as cyclisation of the
Sonogashira coupling product only led to the 5-membered
phthalide derivative via formal 5-exo-dig process (data not
shown). This methoxylated analogue 11 had to be accessed via
a different route.[17] To obtain hydroxy analogue 12 (Scheme 3)
we had to modify the sequence and use acetal protected
salicylic acid derivative 17 for the Sonogashira coupling which,
after hydrolysis of the acetal, was cyclised under Pd catalysed
conditions.[22]
The employed sulfinimide-based chemistry proved versatile
for the exploration of the intercalating moiety (Scheme 4, top),
as well as for the variation of the aromatic moiety in general
(Scheme 4, bottom) when using the opposite enantiomer of the
chiral auxiliary. Yields of this unoptimised sequence were rather
low but led rapidly to the desired enantiopure building blocks.
A range of additional side-chain analogues of 8 were
synthesised varying the stereochemistry of the aminodiol part
as well as the terminal moiety.[17] With side chains and aromatic
moieties in hand, we proceeded to the final coupling followed
by deprotection with BBr3 to access hydroxychromanones or
Scheme 3. Synthesis of hydroxyisochromenone 12. a) CuI, Pd(dppf), NEt3,
overnight, 70%; b) LiOH, THF/H2O, RT, 2.5 h, 100%; c) PdCl2(MeCN)2, NEt3,
°
THF, 50 C, 1 h, 43%; d) HCl/MeOH, 30 min, 70%.
Scheme 5. Coupling and final deprotection. a) ((Benzotriazol-1-yloxy)tris
(dimethylamino)phosphonium hexafluorophosphate or HBTU, NEt3, DMF/
°
°
CH2Cl2, 0 C–RT, overnight, 56–93%; b) BBr3 in CH2Cl2, À 78 C, 1 h, 11–55%; c)
°
Scheme 4. Synthetic access to various aromatic and intercalating moieties.
aq. TFA, or HCl in MeOH, 50 C, overnight, 19–56%.
ChemMedChem 2020, 15, 1–8
3
© 2020 Wiley-VCH GmbH
��
These are not the final page numbers!