optimal for HDAC inhibition. This correlated with the lower efficiency of CHAP
6.11 (HC-toxin derivative), which does present a favorable LDLD configuration but
with the Pro residue in position 2 instead [68, 70]. Peptide T1.1, which was the most
promising compound from this SAR (cyl-1 derivative), and the chlamydocinderived compound 6.9 have been the subject of in vivo studies, which are discussed
in Sect. 4.1.
Chlamydocin analogues have also been subject of SAR studies at positions 2 and
3 of the macrocycle [45, 71]. In particular, the replacement of 2-aminoisobutyric
acid (Aib) at position 2 by more bulky and constrained cyclic amino acids leads to
significant improvements in the antiproliferase activity against cancer cells (MCF-7,
HeLa, and K562 cells) [71]. Moreover, a separate study on chlamydocin derivatives
was taken into account for the design of bicyclic inhibitors with an aliphatic linkage
between side chains 2 and 3. Increase in hydrophobicity and, presumably, change of
conformation are possible explanations for the improved antiproliferase activity of
these compounds against HL-60, K562, and U937 leukemia cells in culture. These
compounds presented improved in vitro selectivity toward HDAC1 and HDAC4
when compared to HDAC6, which resembles the behavior of chlamydocin [72, 73].
The cyclic tetrapeptide scaffold, characteristic of most macrocyclic HDAC inhibitors, consists of 12 atoms and is highly conformationally constrained. This is a
challenge for the synthesis of analogues, since the cyclization step from linear
peptide to 12-member ring is usually low yielding. In this regard, some cyclic
peptides have been designed with a larger ring size, which is more prone to
cyclization. Together with the original CHAP studies, an octapeptide was prepared
as a result of combining two linear precursors of compound 6.15. However, this
compound showed 26-fold loss in in vitro HDAC inhibition and no activity in cellbased assays [70]. Jose et al. designed a substrate-based cyclic hexapeptide inhibitor
selective toward HDAC6 [74]. Since acetylated α-tubulin had been reported as an
HDAC6 target, a macrocycle containing amino acids 38–43 of α-tubulin was
prepared, with K40 replaced by Asuha. The hypothesis was that the Zn
2+ -binding
group would help accommodate the rest of the peptide in a conformation similar to
the native substrate. However, the cyclic peptide adopted a different spatial projection of side chains and was inactive [74]. A more successful approach, carried out by
the group of M. Reza Ghadiri, was based on introduction of β-amino acid residues in
the structure of natural tetrapeptides in order to increase the size of the cycle. The
lead structure, apicidin A (6.3), was simplified and modified with one or two
β-amino acids (α 3 β or α 2 β 2 scaffolds, respectively) yielding 13- and 14-member
ring scaffolds. This resulted in improved synthetic yields but also in a single
conformation of the macrocycles as determined by NMR spectroscopy in DMSOd 6 , as opposed to the native compound, which adopts at least three conformations on
the NMR timescale. Thus, increasing the ring size by a single methylene group
relieved the strain sufficiently to improve the yields of ring closure as well as the
conformational flexibility. In terms of HDAC inhibitory activity, the analogues
bearing one β-amino acid in position 3 retained potency and isotype selectivity
compared to the parent compound. Then, when the native Zn
2+ -binding group was
changed to hydroxamic acid (Fig. 7, compounds 7.1 and 7.2), the overall HDAC
Hydroxamic Acid-Containing Peptides in the Study of Histone Deacetylases
41
6.11 (HC-toxin derivative), which does present a favorable LDLD configuration but
with the Pro residue in position 2 instead [68, 70]. Peptide T1.1, which was the most
promising compound from this SAR (cyl-1 derivative), and the chlamydocinderived compound 6.9 have been the subject of in vivo studies, which are discussed
in Sect. 4.1.
Chlamydocin analogues have also been subject of SAR studies at positions 2 and
3 of the macrocycle [45, 71]. In particular, the replacement of 2-aminoisobutyric
acid (Aib) at position 2 by more bulky and constrained cyclic amino acids leads to
significant improvements in the antiproliferase activity against cancer cells (MCF-7,
HeLa, and K562 cells) [71]. Moreover, a separate study on chlamydocin derivatives
was taken into account for the design of bicyclic inhibitors with an aliphatic linkage
between side chains 2 and 3. Increase in hydrophobicity and, presumably, change of
conformation are possible explanations for the improved antiproliferase activity of
these compounds against HL-60, K562, and U937 leukemia cells in culture. These
compounds presented improved in vitro selectivity toward HDAC1 and HDAC4
when compared to HDAC6, which resembles the behavior of chlamydocin [72, 73].
The cyclic tetrapeptide scaffold, characteristic of most macrocyclic HDAC inhibitors, consists of 12 atoms and is highly conformationally constrained. This is a
challenge for the synthesis of analogues, since the cyclization step from linear
peptide to 12-member ring is usually low yielding. In this regard, some cyclic
peptides have been designed with a larger ring size, which is more prone to
cyclization. Together with the original CHAP studies, an octapeptide was prepared
as a result of combining two linear precursors of compound 6.15. However, this
compound showed 26-fold loss in in vitro HDAC inhibition and no activity in cellbased assays [70]. Jose et al. designed a substrate-based cyclic hexapeptide inhibitor
selective toward HDAC6 [74]. Since acetylated α-tubulin had been reported as an
HDAC6 target, a macrocycle containing amino acids 38–43 of α-tubulin was
prepared, with K40 replaced by Asuha. The hypothesis was that the Zn
2+ -binding
group would help accommodate the rest of the peptide in a conformation similar to
the native substrate. However, the cyclic peptide adopted a different spatial projection of side chains and was inactive [74]. A more successful approach, carried out by
the group of M. Reza Ghadiri, was based on introduction of β-amino acid residues in
the structure of natural tetrapeptides in order to increase the size of the cycle. The
lead structure, apicidin A (6.3), was simplified and modified with one or two
β-amino acids (α 3 β or α 2 β 2 scaffolds, respectively) yielding 13- and 14-member
ring scaffolds. This resulted in improved synthetic yields but also in a single
conformation of the macrocycles as determined by NMR spectroscopy in DMSOd 6 , as opposed to the native compound, which adopts at least three conformations on
the NMR timescale. Thus, increasing the ring size by a single methylene group
relieved the strain sufficiently to improve the yields of ring closure as well as the
conformational flexibility. In terms of HDAC inhibitory activity, the analogues
bearing one β-amino acid in position 3 retained potency and isotype selectivity
compared to the parent compound. Then, when the native Zn
2+ -binding group was
changed to hydroxamic acid (Fig. 7, compounds 7.1 and 7.2), the overall HDAC
Hydroxamic Acid-Containing Peptides in the Study of Histone Deacetylases
41