Chemical Modification of Rapamycin: The Discovery of SDZ RAD
`
`R. Sedrani, S. Cottens, J. Kallen, and W. Schuler
`
`THE immunosuppressive macrolide rapamycin (RAP) 1
`
`(Fig 1) has attracted interest in recent years because
`of its potential in the prevention of both allograft rejection1
`and the development of graft vessel disease (GVD).2 This
`complex natural product, with its remarkable biological
`properties, unfortunately exhibits unfavorable physico-
`chemical properties. As a consequence, the galenical for-
`mulation and oral administration of RAP has proven to be
`rather difficult. Until recently, the majority of published
`work demonstrating the immunosuppressive effect of RAP
`in vivo has dealt with parenteral administration of the
`compound (for references, see Granger et al3). We there-
`fore embarked on a program aimed at overcoming these
`difficulties by chemical derivation of RAP. The nature of
`the chemical modification had to be carefully chosen. It had
`to allow for the introduction of a variety of chemical groups
`and functionalities.
`
`Another important criterion to be considered was the
`metabolic stability; the targeted RAP derivatives had to
`constitute the active principle, ie, they should not behave as
`prodrugs. Therefore, it was of utmost importance to find a
`position in RAP that could be chemically modified without
`resulting in loss of immunosuppressive activity. Alkylation
`of the hydroxyl (O-alkylation) in either position 28 or 40
`(Fig 1) was envisaged as a type of modification that could be
`expected to correspond to our criteria. The data presented
`herein show that O-alkylation in position 40 can indeed lead
`to novel, potently immunosuppressive RAP-derivatives,
`provided that the newly introduced alkyl group is properly
`chosen.
`O-alkylation in position 28 (eg, 2, Fig 1), on the other
`hand, leads to loss of immunosuppressive activity, which
`can be explained on the basis of structural results obtained
`by X-ray crystallography. The efforts described herein ulti-
`mately resulted in the identification of a potent RAP
`derivative, SDZ RAD (40-O-(2-hydroxy)ethylrapamycin) 4
`(Fig 1), which is currently undergoing clinical trials.
`
`MATERIALS AND METHODS
`Rapamycin and Analogs
`RAP was obtained by fermentation of the actinomycete strain
`A91-259211. The analogs were prepared in our laboratories by
`chemical modification of RAP. The experimental details of the
`chemical syntheses will be reported elsewhere.
`
`In Vitro Assays
`FKBP12 Binding Assay. Binding to the FK 506 binding
`protein (FKBP12) was indirectly assessed by means of an
`ELISA-type competition assay. FK 506 was included in
`each individual experiment as a standard, and the inhibitory
`activity is expressed as relative IC50 compared to FK 506
`(rIC50 ⫽ IC50 test compound/IC50 FK 506). Details regard-
`ing this assay have been reported.4
`Mixed Lymphocyte Reaction (MLR). The immunosup-
`pressive activities of RAP and its derivatives were assessed
`in a two-way MLR, using spleen cells of BALB/c and CBA
`mice. RAP was included in each individual experiment as a
`standard, and the inhibitory activity is expressed as relative
`
`Fig 1. Chemical structures of RAP 1, 28-O-methylrapamycin 2
`and 40-O-alkylated derivatives, including SDZ RAD 4.
`
`From Novartis Pharma AG, Basel, Switzerland.
`Address reprint requests to Richard Sedrani, Novartis Pharma
`AG, Preclinical Research, CH-4002, Basel, Switzerland.
`
`0041-1345/98/$19.00
`PII S0041-1345(98)00587-9
`
`2192
`
`© 1998 by Elsevier Science Inc.
`655 Avenue of the Americas, New York, NY 10010
`
`Transplantation Proceedings, 30, 2192–2194 (1998)
`
`West-Ward Exhibit 1020
`Sedrani 1998
`Page 001
`
`

`

`RAPAMYCIN MODIFICATION
`
`2193
`
`Compound
`
`1 RAP
`2
`3
`4 SDZ RAD
`5
`6
`
`Table 1. In Vitro Activities of Rapamycin and O-Alkylated Derivatives
`
`Substituent R1
`
`Substituent R2
`
`FKBP12 binding (rIC50)*
`
`MLR (rIC50)†
`
`H
`H
`Me
`HO(CH2)2⫺
`HO(CH2)6⫺
`Ph⫺
`
`H
`Me
`H
`H
`H
`H
`
`0.6
`1.6
`1.1
`2.0
`0.8
`23
`
`1
`1300
`6.5
`2.1
`18
`⬎430
`
`*The ability of the compounds to compete with immobilized FK 506 for binding to biotinylated FKBP12 was determined in a competitive binding assay. FK 506 was
`included as standard in each individual experiment. Results are expressed as means of the relative IC50 values (ie, IC50 test compound/IC50 FK 506). The range of
`absolute IC50 values for FK 506 was 0.8 –1.2 nmol/L.
`†The inhibitory effect on a two-way MLR performed with spleen cells from BALB/c and CBA mice was tested. RAP was included as standard in each individual
`experiment. Results are expressed as means of the relative IC50 values (ie, IC50 test compound/IC50 RAP). The range of absolute IC50 values for RAP was 0.06 – 0.9
`nmol/L.
`See Fig 1.
`
`IC50 compared to RAP (rIC50 ⫽ IC50 test compound/IC50
`RAP). Details regarding this assay have been reported.4
`
`RESULTS AND DISCUSSION
`As can be seen from the data shown in Table 1, alkylation
`in either position 28 (compound 2) or position 40 (com-
`pounds 3–5) did not greatly affect binding to FKBP12. A
`significant loss in affinity was only observed when the rather
`bulky phenyl group was added to the C40-hydroxyl (com-
`pound 6). Methylation of the C28-hydroxyl (compound 2)
`resulted in a 1000-fold loss of activity in the MLR. When
`the same modification was carried out in position 40
`(compound 3), the immunosuppressive activity was only
`reduced by a factor of 6.5 as compared to RAP. This clearly
`indicated that, in contrast to the C28-hydroxyl, the C40-
`hydroxyl was amenable to chemical modifications without
`resulting in a prohibitive loss of potency, and that further
`optimization could be envisaged. Indeed, introduction of a
`2-hydroxyethyl group in that same position, leading to SDZ
`RAD 4, enhanced the potency with respect to the methyl-
`derivative 3; the immunosuppressive activity of SDZ RAD
`in vitro was found to be comparable to that of RAP. When
`the hydroxyalkyl chain was extended, as in derivative 5, the
`trend was reversed, and a reduction of activity was ob-
`served.
`The results reported herein can be explained on the basis
`of structural results obtained by X-ray crystallography. RAP
`exerts its immunosuppressive activity by first binding to
`FKBP12. This binding is necessary, but not sufficient, as was
`recognized earlier (see Brown et al5 and references cited
`therein), and as can be seen from our data in Table 1 (ie,
`compounds 2 and 6). In the case of compound 6 the
`chemical modification results in decreased affinity for
`FKBP12 and, as a consequence, in a significant loss of
`immunosuppressive potency. Compound 2, on the other
`hand, exhibits a 1300-fold loss in activity despite its high
`affinity for FKBP12. The active principle is actually rather
`the FKBP12/RAP complex, which binds to the target
`protein mTOR6 (also termed FRAP5 or RAFT7.
`As far as 2 is concerned, we have previously shown that in
`the complex with FKBP12, the newly introduced O-methyl
`group in position 28 causes a 120° rotation around the
`
`C35–C36 bond (Fig 1), considerably shifting the cyclohexyl
`subunit.8 The macrocyclic part of 2 remains virtually un-
`changed with respect to FKBP12-bound RAP and can be
`superimposed with the macrocyclic part of RAP in the
`X-ray crystal structure of the FKBP12-RAP-FRB complex9
`(FRB ⫽ FKBP12-rapamycin-binding domain of mTOR).
`This analysis shows that the cyclohexyl subunit of 2, in its
`new position, collides with a loop of the mTOR fragment.
`This actually prevents binding of 2 in complex with FKBP12
`to mTOR, thus explaining the lack of activity in the MLR.
`The X-ray crystal structure of the FKBP12/SDZ RAD 4
`complex, on the other hand, reveals a three-dimensional
`structure for bound SDZ RAD, which very closely resem-
`bles that of RAP (details concerning the structure will be
`published elsewhere). From a structural point of view, there
`is no impediment to complex formation between FKBP12/
`SDZ RAD and the FRB fragment of mTOR, thus explain-
`ing its high potency. The loss of activity observed when
`going from SDZ RAD 4 to 5 is probably due to the larger
`size of the modification at the C40-hydroxyl in the latter.
`This can conceivably result in steric hindrance with, and
`reduced affinity for, mTOR. The modification in 6, finally,
`clearly affects binding to FKBP12, a prerequisite for activ-
`ity, and probably also, because of its size, binding of the
`FKBP12/6 complex to mTOR.
`In conclusion, we have shown that O-alkylation of RAP
`can lead to highly potent derivatives, provided that a
`suitable alkyl group is introduced into the appropriate
`position and that the structure-activity relationship of O-
`alkylated RAP derivatives can be explained by X-ray struc-
`tural results. The 40-O-(2-hydroxyethyl) derivative SDZ
`RAD 4 is presently undergoing clinical trials for use in
`combination with cyclosporine A to prevent acute and
`chronic rejection after solid-organ allotransplantation.4,10
`
`REFERENCES
`1. Kahan BD: Transplant Proc 29:48, 1997
`2. Gregory CR, Huie P, Billingham ME, Morris RE: Transplan-
`tation 55:1409, 1993
`3. Granger DK, Cromwell JW, Chen SC, et al: Transplantation
`59:183, 1995
`4. Schuler W, Sedrani R, Cottens S, et al: Transplantation 64:36,
`1997
`
`West-Ward Exhibit 1020
`Sedrani 1998
`Page 002
`
`

`

`2194
`
`SEDRANI, COTTENS, KALLEN ET AL
`
`5. Brown EJ, Albers MW, Shin TB, et al: Nature 369:756, 1994
`6. Sabers CJ, Martin MM, Brunn GJ, et al: J Biol Chem 270:815,
`1995
`7. Sabatini DM, Erdjument-Bromage H, Lui M, et al: Cell 78:35,
`1994
`
`8. Kallen JA, Sedrani R, Cottens S: J Am Chem Soc 118:5857,
`1996
`9. Choi J, Chen J, Schreiber SL, et al: Science 273:239, 1996
`10. Schuurman HJ, Cottens S, Fuchs S, et al: Transplantation
`64:32, 1997
`
`West-Ward Exhibit 1020
`Sedrani 1998
`Page 003
`
`