96
S. B. Christensen et al.
groups prevents conjugation to any peptide. In addition, the following problems must
be addressed: (1) The prodrug should be hydrophilic, preventing diffusion into cells
and thereby making interaction with the intracellular SERCA pump impossible. (2)
The prodrug must be a substrate for the protease, in this case, either PSA, KLK2
or PSMA. (3) The drug cleaved from the prodrug after interaction with the enzyme
must still be a toxin and be able to reach and neutralize the target (SERCA).
According to structure-activity relationships, a flexible acyl group attached to O-8
should enable potent interaction of the analog with SERCA. A long chain affords a
distance between the bulky guaianolide skeleton and the peptide binding, enabling the
amide group to come into the active site cleft, which in all three proteases are located
in a cavity inside the enzyme [205–208]. Consequently, the butanoyl group at O-8
was substituted with ω-aminoacyl groups to give thapsigargin analogs possessing an
amino group for conjugating with an appropriate peptide. Among the many analogs
prepared, the 8-(12-aminododecanoate) (Fig. 18, compound 73) showed the same
cytotoxicity as 1 [152]. The high cytotoxicity is explained mainly by the structural similarities with 1 and by the ability of the O-8 acyl group to penetrate the
transmembrane sections enabling the molecule to fit into the binding site [79, 94].
The high hydrophilicity of the peptide moieties of the molecules 74–76 was
hypothesized to prevent penetration of the cell membrane and thereby make it impossible for the molecules to reach the SERCA enzyme. Since advantage was taken of
the expression of PSA [160], KLK2 [176, 209], and PSMA [210] in prostate cancer
tumors, substrates for these enzymes were conjugated with the amino group. Based
on kinetic studies, compound 75 was found to be an excellent substrate for KLK2
[176], 74 for PSA [93], and 76 for PSMA [210–213]. Cleavage of the relevant peptide
bond in 75 with KLK2 and in 74 with PSA gives the lipophilic toxin 77, which penetrates the cell membrane (Fig. 19). In both cases, the C-terminal amino acid (Leu)
was not removed by the enzyme. However, a C-terminal Leu was essential for the
peptide to become a substrate [176].
In the case of PSMA, 76 was found to be an optimal substrate. The prodrug is
rapidly cleaved by PSMA to give 78. PSMA is expressed in prostate cancer cell
lines and in neovascular tissues in a number of tumors. Toxin 78 is somewhat slowly
converted to 79 by PSMA. A C-terminal Asp in the peptide is required for cleavage
by PSMA [94].
Preclinical studies were performed particularly on prodrugs 74 and 76. In vitro
studies of 74 revealed that the prodrug was cleaved efficiently by PSA-producing
LNCaP prostate cancer cells, to give the toxin 77, meaning that the C-terminal Leu
was not removed. Encouragingly, 77 efficiently killed PSA-producing cells [176,
214]. Non-PSA-producing cell lines like HCT116 were not affected by the prodrug
74. The compound was stable in human plasma. In vivo studies revealed that injection of 7 mg/kg in mice afforded a plasma concentration of 15 μM and a half-life of
2.8 hours. Only trace amounts of the toxin 74 was observed in the plasma. Continuous injection of the prodrug caused complete growth inhibition of PSA-producing
xenograft tumors and had no effect on non-PSA producing xenografts. No general
toxicity was seen [93].
S. B. Christensen et al.
groups prevents conjugation to any peptide. In addition, the following problems must
be addressed: (1) The prodrug should be hydrophilic, preventing diffusion into cells
and thereby making interaction with the intracellular SERCA pump impossible. (2)
The prodrug must be a substrate for the protease, in this case, either PSA, KLK2
or PSMA. (3) The drug cleaved from the prodrug after interaction with the enzyme
must still be a toxin and be able to reach and neutralize the target (SERCA).
According to structure-activity relationships, a flexible acyl group attached to O-8
should enable potent interaction of the analog with SERCA. A long chain affords a
distance between the bulky guaianolide skeleton and the peptide binding, enabling the
amide group to come into the active site cleft, which in all three proteases are located
in a cavity inside the enzyme [205–208]. Consequently, the butanoyl group at O-8
was substituted with ω-aminoacyl groups to give thapsigargin analogs possessing an
amino group for conjugating with an appropriate peptide. Among the many analogs
prepared, the 8-(12-aminododecanoate) (Fig. 18, compound 73) showed the same
cytotoxicity as 1 [152]. The high cytotoxicity is explained mainly by the structural similarities with 1 and by the ability of the O-8 acyl group to penetrate the
transmembrane sections enabling the molecule to fit into the binding site [79, 94].
The high hydrophilicity of the peptide moieties of the molecules 74–76 was
hypothesized to prevent penetration of the cell membrane and thereby make it impossible for the molecules to reach the SERCA enzyme. Since advantage was taken of
the expression of PSA [160], KLK2 [176, 209], and PSMA [210] in prostate cancer
tumors, substrates for these enzymes were conjugated with the amino group. Based
on kinetic studies, compound 75 was found to be an excellent substrate for KLK2
[176], 74 for PSA [93], and 76 for PSMA [210–213]. Cleavage of the relevant peptide
bond in 75 with KLK2 and in 74 with PSA gives the lipophilic toxin 77, which penetrates the cell membrane (Fig. 19). In both cases, the C-terminal amino acid (Leu)
was not removed by the enzyme. However, a C-terminal Leu was essential for the
peptide to become a substrate [176].
In the case of PSMA, 76 was found to be an optimal substrate. The prodrug is
rapidly cleaved by PSMA to give 78. PSMA is expressed in prostate cancer cell
lines and in neovascular tissues in a number of tumors. Toxin 78 is somewhat slowly
converted to 79 by PSMA. A C-terminal Asp in the peptide is required for cleavage
by PSMA [94].
Preclinical studies were performed particularly on prodrugs 74 and 76. In vitro
studies of 74 revealed that the prodrug was cleaved efficiently by PSA-producing
LNCaP prostate cancer cells, to give the toxin 77, meaning that the C-terminal Leu
was not removed. Encouragingly, 77 efficiently killed PSA-producing cells [176,
214]. Non-PSA-producing cell lines like HCT116 were not affected by the prodrug
74. The compound was stable in human plasma. In vivo studies revealed that injection of 7 mg/kg in mice afforded a plasma concentration of 15 μM and a half-life of
2.8 hours. Only trace amounts of the toxin 74 was observed in the plasma. Continuous injection of the prodrug caused complete growth inhibition of PSA-producing
xenograft tumors and had no effect on non-PSA producing xenografts. No general
toxicity was seen [93].