The substrate-based inhibitors were also designed by incorporating modifications
into the peptide scaffold. The thioacetyl-lysine containing tri-, tetra-, and pentapeptides based on α-tubulin (SDK(thioAc)TI) and p53 (HKK(thioAc)LM) were developed, and they inhibited SIRT1 at the nanomolar scale and SIRT2 at a low
micromolar scale [58]. It was also observed that the p53 sequence displayed a better
inhibition than the α-tubulin sequence. Interestingly, SIRT1 inhibition was not very
sensitive to the length of peptide, whereas the pentapeptide (HKK(thioAc)AM) was
more potent toward SIRT2 with an IC 50 value of 3.8 μM. Some of these peptides
such as H 2 N-KK(thioAc)L-OH showed clear selectivity for SIRT1 over SIRT2
[58]. Subsequently, it has been reported that the simple peptide Ac-Ala-Lys
(thioAc)-Ala could achieve SIRT1 and SIRT2 inhibition similar to the longer and
more complicated peptides [56]. A set of p53 and α-tubulin mimetic peptides was
also screened against SIRT6. The most potent compounds exerted more than 60%
inhibition activity at the 200 μM concentration, and subsequently they displayed
IC 50 values below 100 μM [59]. The exploitation of peptides as therapeutic agents
has disadvantages due to their delivery, poor bioavailability, and short half-life, and
thus, other types of inhibitors have been developed.
The macrocyclic peptide inhibitors had improved potency, cell permeability, and
resistance toward proteolytic enzymes compared to the analog linear peptide. However, most of them were pan-SIRT1/2/3 inhibitors [60]. Simple cyclic peptides
containing N
ε -thioacetyl-lysine showed inhibitory properties in the low nanomolar
range for SIRT1, SIRT2, and SIRT3. Interestingly, macrocyclic peptides containing
trifluoroacetyl-lysine (e.g., compound 14 in Fig. 3) were able to act as catalytic
mechanism-based inhibitors for sirtuins. They have a greater inhibitory potency
toward SIRT2 (IC 50 ¼ 3–4 nM) than toward SIRT1 (IC 50 ¼ 5–6 nM) and SIRT3
(IC 50 ¼ 480 nM). However, their linear counterparts were equally potent SIRT2
inhibitors with IC 50 values of 5–6 nM. The short and linear RIK(Trifluoroacetyl)RY
(compound 15 in Fig. 3) was less potent toward SIRT2 (IC 50 ¼ 31 nM) than the
cyclic compound, and the inhibition was even weaker for SIRT1 (IC 50 ¼ 280 nM)
and SIRT3 (IC 50 ¼ 1 μM) [61]. Investigation of these macrocyclic peptides with
SIRT2 revealed that the trifluoroacetyl-lysine moiety was orientated into the catalytic tunnel of the binding site and the macrocyclic structure was stabilized via
multiple intramolecular hydrogen bonds [62].
Non-peptide N-thioacetyl-lysine analogs represented another attempt to improve
the inhibitory activity of the substrate-based inhibitors. The first non-peptide
N-thioacetyl-lysine analogs were developed by Asaba et al. [63] and Suzuki et al.
[64]. The most potent inhibitor of these compounds had a thioacetyl-lysine scaffold
with an aniline group attached to the carbonyl terminal and a benzyloxycarbonyl
group to the amino terminal (compound 16 in Fig. 3). The compound exhibited low
micromolar inhibition for SIRT1 (IC 50 ¼ 3 μM) and SIRT2 (IC 50 ¼ 23 μM), but it
was selective over SIRT3 (IC 50 > 100 μM) [64]. Jing and coworkers developed
Cbz-Lys(ThioAc)-NH-Phe that showed IC 50 values of 10 μM for SIRT1, 6 μM for
SIRT2, 28 μM for SIRT3, and >200 μM for SIRT5–7 [65]. This inhibitor exhibited
also a dose-dependent increase in p53 acetylation in human colon cancer cells [64].
62
M. Rahnasto-Rilla et al.
into the peptide scaffold. The thioacetyl-lysine containing tri-, tetra-, and pentapeptides based on α-tubulin (SDK(thioAc)TI) and p53 (HKK(thioAc)LM) were developed, and they inhibited SIRT1 at the nanomolar scale and SIRT2 at a low
micromolar scale [58]. It was also observed that the p53 sequence displayed a better
inhibition than the α-tubulin sequence. Interestingly, SIRT1 inhibition was not very
sensitive to the length of peptide, whereas the pentapeptide (HKK(thioAc)AM) was
more potent toward SIRT2 with an IC 50 value of 3.8 μM. Some of these peptides
such as H 2 N-KK(thioAc)L-OH showed clear selectivity for SIRT1 over SIRT2
[58]. Subsequently, it has been reported that the simple peptide Ac-Ala-Lys
(thioAc)-Ala could achieve SIRT1 and SIRT2 inhibition similar to the longer and
more complicated peptides [56]. A set of p53 and α-tubulin mimetic peptides was
also screened against SIRT6. The most potent compounds exerted more than 60%
inhibition activity at the 200 μM concentration, and subsequently they displayed
IC 50 values below 100 μM [59]. The exploitation of peptides as therapeutic agents
has disadvantages due to their delivery, poor bioavailability, and short half-life, and
thus, other types of inhibitors have been developed.
The macrocyclic peptide inhibitors had improved potency, cell permeability, and
resistance toward proteolytic enzymes compared to the analog linear peptide. However, most of them were pan-SIRT1/2/3 inhibitors [60]. Simple cyclic peptides
containing N
ε -thioacetyl-lysine showed inhibitory properties in the low nanomolar
range for SIRT1, SIRT2, and SIRT3. Interestingly, macrocyclic peptides containing
trifluoroacetyl-lysine (e.g., compound 14 in Fig. 3) were able to act as catalytic
mechanism-based inhibitors for sirtuins. They have a greater inhibitory potency
toward SIRT2 (IC 50 ¼ 3–4 nM) than toward SIRT1 (IC 50 ¼ 5–6 nM) and SIRT3
(IC 50 ¼ 480 nM). However, their linear counterparts were equally potent SIRT2
inhibitors with IC 50 values of 5–6 nM. The short and linear RIK(Trifluoroacetyl)RY
(compound 15 in Fig. 3) was less potent toward SIRT2 (IC 50 ¼ 31 nM) than the
cyclic compound, and the inhibition was even weaker for SIRT1 (IC 50 ¼ 280 nM)
and SIRT3 (IC 50 ¼ 1 μM) [61]. Investigation of these macrocyclic peptides with
SIRT2 revealed that the trifluoroacetyl-lysine moiety was orientated into the catalytic tunnel of the binding site and the macrocyclic structure was stabilized via
multiple intramolecular hydrogen bonds [62].
Non-peptide N-thioacetyl-lysine analogs represented another attempt to improve
the inhibitory activity of the substrate-based inhibitors. The first non-peptide
N-thioacetyl-lysine analogs were developed by Asaba et al. [63] and Suzuki et al.
[64]. The most potent inhibitor of these compounds had a thioacetyl-lysine scaffold
with an aniline group attached to the carbonyl terminal and a benzyloxycarbonyl
group to the amino terminal (compound 16 in Fig. 3). The compound exhibited low
micromolar inhibition for SIRT1 (IC 50 ¼ 3 μM) and SIRT2 (IC 50 ¼ 23 μM), but it
was selective over SIRT3 (IC 50 > 100 μM) [64]. Jing and coworkers developed
Cbz-Lys(ThioAc)-NH-Phe that showed IC 50 values of 10 μM for SIRT1, 6 μM for
SIRT2, 28 μM for SIRT3, and >200 μM for SIRT5–7 [65]. This inhibitor exhibited
also a dose-dependent increase in p53 acetylation in human colon cancer cells [64].
62
M. Rahnasto-Rilla et al.