Appeal No. 2004-0245 Application 09/265,926 page 5, lines 1-17), which encompasses the purine derivatives of formula (A) of appealed claim 10. In the reaction scheme disclosed in Grinter, the purine derivative intermediates with a diethyl-carboxylate containing group in the “9” position encompassed by formula (I) are prepared according to sequence “(i)” wherein a purine derivative of formula (II) is coupled with, inter alia, a triethyl-carboxylate derivative of formula (V) to obtain a purine derivative intermediate with a triethyl-carboxylate containing group in the “9” position encompassed by formula (VI) which is then decarboxylated to obtain the diethyl-carboxylate containing derivative (e.g., page 5, lines 18-47, and page 6, lines 21-45 and 48-49).4 The sequence “(i)” in the reaction scheme thus disclosed by Grinter involves coupling and decarboxylation reactions falling within appealed claim 10, and the purine derivative intermediates of formulae (I), (II) and (VI) of Grinter encompass the 6-chloro-purine derivative intermediates of formulae (I), (II) and (VI) of appealed claim 10, while formula (V) of Grinter encompasses the same tri-carboxylate derivatives as formula (V) in claim 10. The sequence “(ii)” of Grinter converts the purine derivative intermediate of formula (I) to purine derivative products of formula (A) by, inter alia, the same reduction and acylation reactions (e.g., page 6, lines 51-53) specified in claim 10, wherein the acylation step of Grinter encompasses the acetylation step of claim 10. We note in this respect, that formula member R3 of formula (I) of Grinter can be methyl, and thus is the acetyloxy group (page 5, line 32; and page 7, line 7). The principal issue raised by the parties in this appeal, with respect to appealed process claim 10, and indeed also applies to appealed product claim 5, involves the step in the claimed scheme and that of Grinter where the chloro-substituent in the “6” position is removed from the purine derivative intermediates (see, e.g., answer, page 6; brief, page 6). In appealed claim 10, the 6-chloro substituent is removed by hydrogenolysis from the 6-chloro-purine derivative intermediate subsequent to the acetylation step, as the last step in sequence “(ii).” 4 No transesterification is disclosed in Grinter for this synthesis route, characterized by appellant as the “bromotriester route” of Grinter (brief, e.g., paragraph bridging pages 6-7). The only transesterification reaction disclosed in Grinter is in an alternative synthesis route wherein a purine derivative of formula (II) is reacted with a dioxane derivative of formula (III) to obtain the purine derivative intermediate of formula (IV) which is converted to the purine derivative intermediate of formula (I) by transesterification (e.g., page 5, line 34, to page 6, line 20, page 6, lines 47-48, and page 7, lines 35-46). - 4 -Page: Previous 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 NextLast modified: November 3, 2007