Cephalostatin analogues--synthesis and biological activity.

Starting off in the early 90's the field of cephalostatin analogues has continually expanded over the last 10 years. First syntheses prepared symmetric analogues like 14b (119) and 26 (65), which were subsequently desymmetrized to provide analogues like beta-hydroxy ketone 31 (19). Importantly the straightforward approach provided already compounds with mu-molar potency and the same pattern of activity as cephalostatin 1 (1) (see Chapter 2.1). Chemically more demanding, two new methods for the directed synthesis of (bissteroidal) pyrazines were devised and subsequently applied to a wide variety of differently functionalized coupling partners. These new methods allowed for the synthesis of various analogues (Chapter 2.2.; and, last but not least, for the totals synthesis of several cephalostatin natural products; Chapter 1.). Functionalization and derivatization of the 12-position was performed (Chapter 2.1 and 3) and synthetic approaches to establish the D-ring double bond were successfully investigated (Chapter 3). [figure: see text] Dealing synthetically with the spiroketal moiety, novel oxidative opening procedures on monomeric delta 14, 15-steroids were devised as well as intensive studies regarding spiroketal synthesis and spiroketal rearrangements were conducted (Chapter 3.2. and 4.). Last but not least direct chemical modification of ritterazines and cephalostatins were studied, which provided a limited number of ritterazine analogues (Chapter 4.). All these synthetic activities towards analogues are summarized in Fig. 18. During this period of time the growing number of cephalostatins and ritterazines on the one hand and of analogues on the other hand provided several SAR trends, which can guide future analogue synthesis. The combined SAR findings are displayed in Fig. 19. So far it is apparent that: Additional methoxylations or hydroxylations in the steroidal A ring core structure (1-position) are slightly decreasing activity (compare cephalostatin 1 1 to cephalostatins 18, 19, 10, and 11). Not investigated by preparation of analogues. Additional hydroxylations in the B-ring (7- and 9-position) do not have a strong effect. They appear to decrease slightly the activity in the case of 9-position (compare cephalostatin 1 1 to cephalostatin 4) and are neutral in the case of the 7-position (compare ritterazines J and K). Analogue synthesis confirmed this: 7-ring-hydroxylation has little impact on activity, e.g. 109a (Table 6). C'-ring aryl compounds with a 12,17 connected spiroketal area are much less active (cephalostatins 5 and 6), meaning South 6 moiety reduces activity [figure: see text] Confirmed by analogue synthesis, e.g. 190a and 190b (Table 9). Regarding 12-functionalization it is apparent, that all cephalostatins/ritterazines possess either a free hydroxy or a keto function at this position (exemption: cephalostatins 5 and 6--very low activity). However, it is not apparent whether a 12,12'-diol or a 12-keto-12'-ol is favored. In the cephalostatin series the most potent compounds possess a 12-keto-12'-ol function, while in the ritterazine series the direct comparison of ritterazine B and ritterazine H clearly favors the 12,12'-diol setting. Synthesis of simple analogues like 31 showed a "cephalostatin trend" for favoring the 12-keto, 12'-alcohol functionalization. Synthesis of a cephalostatin 1-12'-alcohol 1a supported that trend (2 fold drop in activity). Synthesis of acylated ritterazine B derivatives proved that free hydroxy groups in 12-position are necessary for high activity. At least one 14,15-double bond is part of all highly active cephalostatins/ritterazines. All ritterazines lacking this feature display only low potency (but most of them possess the unfavorable North A moiety or have unfavorable combinations of moieties; vide infra). However, the 14,15-double bond may be necessary "only" for stereochemical reasons creating a specific "curvature" of the molecule by "bending" the D-ring down (for an in depth discussion on this topic: see Chapter 3). In line with this are the observations that 14,15-alpha-epoxides do substantially decrease activity (cephalostatins 14 and 15) while a 14,15-beta-epoxide does not decrease activity (cephalostatin 4). Also in line with the "curvature theory" is the fact that ritterazine B (14-beta-hydrogen) is even more potent than ritterazine G (14,15-double bond). Therefore it is not clear if--at least one--14,15-double bond is essential for high activity. The synthesis and biological evaluation of completely 14-beta-saturated analogues (like 14'-beta-hydrogen ritterazine B) could answer this question. Synthesis of the partially saturated analogues 14' alpha-cephalostatin 1 1c and 7-deoxy-14' alpha-ritterazine B 2a showed that the stronger the divergence of conformation implied by the saturation is, the higher is the loss of activity, thus underlining the "curvature hypothesis". Synthesis showed, that analogues possessing the 14,15-double bond(s) are substantially better soluble, e.g. 26. Furthermore, the D-Ring area turned out to be sensitive for modifications, since substantially differing analogues, like 162, 163, and 164 were completely inactive. At least one 17-hydroxy group is part of all highly active cephalostatins/ritterazines. Loss of one out of two 17-hydroxy groups does not decrease activity (compare ritterazine K and L) but of the second 17-hydroxy groups (along with the 7-hydroxy group) as seen in the ritterazine series (compare ritterazines A/T and B/Y) leads to a significant decrease in activity. Increased activity of 17-ether analogues 178 and 179 points into the same direction All highly active cephalostatins and ritterazines are substantially asymmetric. Cephalostatins and ritterazines that are symmetric--either consisting of two polar units (cephalostatin 12 and ritterazine K) or two unpolar units (ritterazine N and ritterazine R)--or almost symmetric (cephalostatin 13 and ritterazine J, L, M, O, S) show substantially diminished potency. However, one has to keep in mind, that even some of the symmetrical compounds (e.g. ritterazine K--96 nM in the NCI panel) still show strong cytostatic properties. Same trend was identified with simple analogues, e.g. compare 26 to 31. In addition to the basic requirement of overall substantial asymmetry for high activity there appears to be the necessity for a "polarity match" between both steroidal units (33)--as one has to be substantially more polar (high hydroxylation grade) than the other. (e.g. cephalostatin 1 (1): North 1--high hydroxylation grade--and South 1--low hydroxylation grade; or: ritterazine B (2): South 7--medium hydroxylation grade--and North G--very low hydroxylation grade). Not directly confirmed by Analogue Synthesis--some "polarity matched analogues" did not show appropriate activity, e.g. 198 and 197. 4 core moieties are privileged, meaning all highly active ritterazines/cephalostatins (see table 1) are constructed out of them. Namely these are North 1, South 1, South 7 and North G. Numerous analogues were prepared to probe questions regarding the mechanism of action of the cephalostatins, e.g. close cephalostatin analogues like 197 and 198 (70) with increased energy content in the spiroketal. However, so far the mechanism and mode of action of the cephalostatins remains unknown. In the absence of any structural information of the biological target(s), the understanding about the structural necessities for high cytostatic activity is still limited and thus the rational design of more simple, yet highly active analogues seems at the current stage elusive. Additionally, there are many open questions, e.g. how the "monomeric" OSW-1 (3) relates to the "dimeric" cephalostatins. It remains the hope that forthcoming studies will bring light into this so far nebulous area--enabling chemists in the long run to provide highly active analogues in substantial amounts for advanced pharmacological studies. In conclusion one can state that the first decade after the extraordinarily complex cephalostatin 1 (1) entered the scene was necessary for the chemists to explore novel ways towards cephalostatins and cephalostatin analogues. They have provided methods to prepare basically every thinkable cephalostatin analogue, have delivered simple analogues (< 10 steps) with substantial activity and shaped first SAR trends in the class of cephalostatins. Now the time has come for chemists to harvest the fruits of their long and enduring synthetic ventures by aiming towards highly active, yet still not too complex analogues, which could be available in substantial amounts for advanced pharmacological studies. And for pharmacologists to explore the therapeutic potential of the cephalostatins along with elucidation of the unknown mechanism. Clearly, there is much more to expect of the cephalostatins in the coming years.