An Efficient Synthesis of a Doxorubicin-Peptide Conjugate

Abstract

An efficient synthesis of the sodium salt of the doxorubicin-peptide conjugate 1, useful for the treatment of prostate cancer is described. The EDC-mediated amide formation between the heptapeptide 4 and doxorubicin (2) as the key step has been extensively studied employing peptide-coupling additives, such as 1-hydroxybenzotriazole (HOBt), 1-hydroxy-7-azabenzotriazole (HOAt) and 2-hydroxypyridine N-oxide (HOPO), surprisingly, mixing the additives HOAt and HOPO furnished the best results.

Key words

amides - coupling reaction - isomerization - peptide conjugate - protecting group

References

• 1 Dorr RT. Von Hoff DD. Cancer Chemotherapy Handbook   2nd ed.:  Appleton and Lange; Norwalk: 1994. 
• 2 Raghavan D. Koczwara B. Javle M. Eur. J. Cancer  1997,  33:  566 
• 3 DeFeo-Jones D, Feng D.-M, Garsky VM, Jones RE, and Oliff AI. inventors; PCT Patent WO  97/12624. 
• 4 Khan SR. Denmeade SR. The Prostate  2000,  45:  80 
• 5 DeFeo-Jones D. Garsky VM. Wong BK. Feng D.-M. Bolyar T. Haskell K. Kiefer DM. Leander K. McAvoy E. Lumma P. Wai J. Sendark ET. Motzel SL. Keenan K. Zwieten MV. Lin JH. Freidinger R. Huff J. Oliff A. Jones RE. Nature Medicine  2000,  6:  1248 
• 6 Garsky VM. Lumma PK. Feng D.-M. Wai J. Ramjit HG. Sardana MK. Oliff A. Jones RE. DeFeo-Jones D. Freidinger RM. J. Med. Chem.  2001,  44:  4216 
• 7 DiPaolo RS. Ebbinghaus S. Nemunatitis J. McCullough J. Ciardella M. Adams N. Williams A. Garsky VM. Wong BK. DeFeo-Jones D. Jones RE. Schwartz MS. Winchell GA. Arena CD. Deutsch PJ. Yao S.-L. Proced. Am. Soc. for Clin. Oncol.  2000,  19:  1370 
• 8a

  For the solid phase synthesis of the heptapeptide 4 and the preparative HPLC purification of salt 6 (cf. ref. 6).

• 8b Ashwood MS. inventors; PCT Patent, WO  2001029065. For the solution-based process that has been developed and demonstrated in our Laboratories to provide multi-kilograms of the heptapeptide 4. See:
• 9 Kessler H. Siegmeier R. Tetrahedron Lett.  1983,  24:  281 
• 10a Bednarek MA. Bodanszky M. Int. J. Pept. Protein Res.  1983,  21:  196 
• 10b

  The identity of 5 was confirmed by comparing with the authentic sample (cf. ref. 6) on HPLC and LC/MS.

• 11 Carpino LA. El-Fahan A. Tetrahedron  1999,  55:  6813 
• 13 Ho G.-J. Emerson KM. Mathre DA. Shuman RF. Grabowski EJJ. J. Org. Chem.  1995,  60:  3569 
• 16 The autoxidation of -COCH₂OH to -COOH at C-9 position of Doxorubicin was reported previously. See: Gianni L. Vigano L. Lanzi C. Niggeler M. Malatesta V. J. Natl. Cancer Inst.  1988,  80:  1104 
• 17 The Amadori rearrangement was not possible when the tertiary amine was used. Although quinuclidine was very effective for the deprotection, its supply was limited for any scale-up operation. See: Maruoka H. Yamamoto H. Comp. Org. Synth.  1991,  6:  789 

12

For instance, when the HOAt-ester derived from the heptapeptide 4 was prepared and aged for 1.0 h at -15 °C, followed by addition of Dox 2, the d-Leu-epimer was observed at an 11% level.

14

A Typical Procedure for Preparation of 5: To a 3-litter, 4-neck round bottom flask equipped with a truebore stirrer, N₂ inlet/vacuum inlet, and a thermocouple were charged with Dox 2∞HCl salt (20.0 g, 98.3 wt%, 34.5 mmol), HOPO (4.14 g, 37.3 mmol, 1.08 equiv), HOAt (0.56 g, 4.14 mmol, 0.12 equiv), water (4.7 mL),²⁰ DMF (472 mL), and 2,4,6-collidine (13.63 mL, 103.0 mmol, 3.0 equiv) at 20 ºC. The resulting slurry was cooled to -5 ºC, and the peptide 4 (40.6 g, 91.4 wt%, 35.3 mmol, 1.02 equiv) was introduced. After the slurry was stirred at -5 ºC for 0.5 h, the first portion of EDC (5.30 g, 27.6 mmol, 0.8 equiv) was charged. Again, the resulting slurry was stirred at -5 ºC for 1.5 h followed by the second portion of EDC (3.96 g, 20.7 mmol, 0.6 equiv) charge. The cooling bath was removed and the reaction was allowed to warm to 20 ºC and aged for 12 h. Ethyl acetate (354 mL) was added to the resulting solution at 20 ºC. It was followed by a subsurface addition of 1.2 L phosphate buffer (pH 6.05, 2 liters buffer prepared from 10.9 g K₂HPO₄,
43.54 g KH₂PO₄) using a peristaltic pump over 1.0 h, while maintaining the temperature between 18 °C to 20 ºC. The resulting red precipitation was then filtered. The cake was washed with water (2.3 L) and dried overnight on the filter by a vacuum/N₂ purge at ambient temperature. The product 5 was obtained as a red solid (54.3 g, 2.2% of the d-leu-epimer, 89.9 wt%). [6] The actual yield after correction for purity was 90.0%.

15

Adduct 10 was formed in situ and no attempt was made to isolate this adduct. NMR experiments showed a mixture of EDC, HOPO and adduct 10 when an equal mole of EDC and HOPO was mixed in DMF-d ₇ at r.t. However, when additional HOPO (up to 2-3 equiv) was introduced to the aforementioned solution, NMR showed HOPO and adduct 10 with the disappearance of EDC. For adduct 10 (broad signals are denoted with br): ¹H NMR (600.13 MHz, DMF-d ₇): δ = 7.95 (dd, J = 7.2, 1.9 Hz, 1 H), 7.51 (ddd, J = 8.9, 6.4, 1.9 Hz, 1 H), 6.69 (dd, J = 8.9, 1.5 Hz, 1 H), 6.30 (obscured m, 1 H), 3.28 (br m, 4 H), 3.17 (br t, J = 6 Hz, 2 H), 2.81 (s, 6 H), 1.96 (br m, 2 H), 1.16 (br t, J = 7 Hz, 3 H). ¹³C NMR (150.88 MHz, DMF-d ₇): δ = 157.6, 152.0(br), 140.0, 138.9, 121.8, 104.3, 55.9, 42.4, 42.0 (br), 38.0 (br), 25.3, 15.9.

18

When the reaction was run at r.t., the impurity 12 derived from the Amadori rearrangement was observed. It occurred even when the isolated solid 6 was kept at r.t.

19

Previously, pure 1 as a free acid was obtained as an amorphous solid by preparative HPLC purification of the piperidium salt 6 followed by lyophilization (cf. ref. 6).

20

The solubility of Dox 2∞HCl in DMF was increased by addition of H₂O. Therefore, the reaction rate was improved.