Solid-Phase Synthesis and Combinatorial Technologies
.pdf7.4 ENCODING METHODS FOR SOLID-PHASE POOL LIBRARIES 303
INDIVIDUAL CODES:
T1 corresponds to M1,1; T2 to M1,2; T3 to M1,3;
T4 to M1,4; T5 to M1,5; T6 to M1,6
6 tags = 6 encoded monomers
|
|
|
|
|
|
|
BINARY CODES: |
|
|
|
|
|
|
|
|
|
|
|||
|
T1 T2 T3 T4 T5 T6 |
|
T1 T2 T3 T4 T5 T6 |
|
T1 T2 T3 T4 T5 T6 |
|||||||||||||||
M1,1 |
1 |
0 |
0 |
0 |
0 |
0 |
M1,22 |
1 |
1 |
1 |
0 |
0 |
0 |
M1,43 |
1 |
1 |
1 |
0 |
1 |
0 |
M1,2 |
0 |
1 |
0 |
0 |
0 |
0 |
M1,23 |
1 |
1 |
0 |
1 |
0 |
0 |
M1,44 |
1 |
1 |
1 |
0 |
0 |
1 |
M1,3 |
0 |
0 |
1 |
0 |
0 |
0 |
M1,24 |
1 |
1 |
0 |
0 |
1 |
0 |
M1,45 |
1 |
1 |
0 |
1 |
1 |
0 |
M1,4 |
0 |
0 |
0 |
1 |
0 |
0 |
M1,25 |
1 |
1 |
0 |
0 |
0 |
1 |
M1,46 |
1 |
1 |
0 |
1 |
0 |
1 |
M1,5 |
0 |
0 |
0 |
0 |
1 |
0 |
M1,26 |
1 |
0 |
1 |
1 |
0 |
0 |
M1,47 |
1 |
1 |
0 |
0 |
1 |
1 |
M1,6 |
0 |
0 |
0 |
0 |
0 |
1 |
M1,27 |
1 |
0 |
1 |
0 |
1 |
0 |
M1,48 |
1 |
0 |
1 |
1 |
1 |
0 |
M1,7 |
1 |
1 |
0 |
0 |
0 |
0 |
M1,28 |
1 |
0 |
1 |
0 |
0 |
1 |
M1,49 |
1 |
0 |
1 |
1 |
0 |
1 |
M1,8 |
1 |
0 |
1 |
0 |
0 |
0 |
M1,29 |
1 |
0 |
0 |
1 |
1 |
0 |
M1,50 |
1 |
0 |
1 |
0 |
1 |
1 |
M1,9 |
1 |
0 |
0 |
1 |
0 |
0 |
M1,30 |
1 |
0 |
0 |
1 |
0 |
1 |
M1,51 |
1 |
0 |
0 |
1 |
1 |
1 |
M1,10 |
1 |
0 |
0 |
0 |
1 |
0 |
M1,31 |
1 |
0 |
0 |
0 |
1 |
1 |
M1,52 |
0 |
1 |
1 |
1 |
1 |
0 |
M1,11 |
1 |
0 |
0 |
0 |
0 |
1 |
M1,32 |
0 |
1 |
1 |
1 |
0 |
0 |
M1,53 |
0 |
1 |
1 |
1 |
0 |
1 |
M1,12 |
0 |
1 |
1 |
0 |
0 |
0 |
M1,33 |
0 |
1 |
1 |
0 |
1 |
0 |
M1,54 |
0 |
1 |
1 |
0 |
1 |
1 |
M1,13 |
0 |
1 |
0 |
1 |
0 |
0 |
M1,34 |
0 |
1 |
1 |
0 |
0 |
1 |
M1,55 |
0 |
1 |
0 |
1 |
1 |
1 |
M1,14 |
0 |
1 |
0 |
0 |
1 |
0 |
M1,35 |
0 |
1 |
0 |
1 |
1 |
0 |
M1,56 |
0 |
0 |
1 |
1 |
1 |
1 |
M1,15 |
0 |
1 |
0 |
0 |
0 |
1 |
M1,36 |
0 |
1 |
0 |
1 |
0 |
1 |
M1,57 |
1 |
1 |
1 |
1 |
1 |
0 |
M1,16 |
0 |
0 |
1 |
1 |
0 |
0 |
M1,37 |
0 |
1 |
0 |
0 |
1 |
1 |
M1,58 |
1 |
1 |
1 |
1 |
0 |
1 |
M1,17 |
0 |
0 |
1 |
0 |
1 |
0 |
M1,38 |
0 |
0 |
1 |
1 |
1 |
0 |
M1,59 |
1 |
1 |
1 |
0 |
1 |
1 |
M1,18 |
0 |
0 |
1 |
0 |
0 |
1 |
M1,39 |
0 |
0 |
1 |
1 |
0 |
1 |
M1,60 |
1 |
1 |
0 |
1 |
1 |
1 |
M1,19 |
0 |
0 |
0 |
1 |
1 |
0 |
M1,40 |
0 |
0 |
1 |
0 |
1 |
1 |
M1,61 |
1 |
0 |
1 |
1 |
1 |
1 |
M1,20 |
0 |
0 |
0 |
1 |
0 |
1 |
M1,41 |
0 |
0 |
0 |
1 |
1 |
1 |
M1,62 |
0 |
1 |
1 |
1 |
1 |
1 |
M1,21 |
0 |
0 |
0 |
0 |
1 |
1 |
M1,42 |
1 |
1 |
1 |
1 |
0 |
0 |
M1,63 |
1 |
1 |
1 |
1 |
1 |
1 |
6 tags = 63 encoded monomers
Figure 7.24 Chemical encoding: individual and binary encoding systems.
monomers, we would encode just six monomers (Fig. 7.24, top), while the binary code would accommodate a much higher number of 2n – 1 monomers (63 from six code structures, removing the null code, Fig. 7.24, bottom).
A family of chemical tags should ideally possess a number of features. It should be stable to all, or to the large majority, of the reaction conditions that may be used in an SP library synthesis; its synthesis should be easy and almost quantitative using commercially available precursors; its presence should not interfere with the library synthesis scheme; it should be inert in library screening when on-bead methods are used; its structure should be determined with a fast and reliable method. We will now present the various chemical encoding strategies reported in the literature and will comment on their fulfillment of the above criteria.
The first encoding approaches used peptides (177) or oligonucleotides (178) as tags. We will consider two reported examples of resin-bound constructs able to support peptide-encoded (7.37, Fig. 7.25) (177) and oligonucleotide-encoded (7.38, Fig. 7.25) (179) peptide or peptidomimetic library synthesis. They both contain a core moiety, derived from L-lysine (7.37) or L-serine (7.38), which simultaneously multiplies the
304 SYNTHETIC ORGANIC LIBRARIES: SOLID-PHASE POOL LIBRARIES
|
|
|
|
|
|
|
H |
|
|
|
|
|
|
|
N |
|
|
|
|
|
|
|
Peptide |
|
|
|
|
|
NH |
O |
|
|
|
|
|
|
|
|
|
|
|
|
1st release |
|
|
|
|
|
|
(screening in |
|
O |
|
|
|
|
|
|
pools) |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
O |
|
|
|
|
|
|
|
O N |
|
|
|
|
O |
O |
|
NH2 |
|
|
|
|
|
|
|
|
||
|
|
|
|
NH |
|
|
Peptide |
|
|
|
|
|
H |
|
|
|
|
|
|
|
N |
|
|
|
|
|
|
|
N |
|
|
|
|
NH |
O |
|
H |
|
|
|
|
|
|
|
|||
|
|
|
O |
O |
|
|
|
|
Peptide |
|
|
|
|
||
|
|
|
|
|
|
||
|
|
N |
O |
|
2nd release |
|
|
|
|
H |
|
|
|
||
|
|
|
7.37 |
|
(screening as |
|
|
tag sites, |
|
|
single beads) |
|
|
||
Edman sequencing, |
|
|
|
|
|
|
|
structure determination |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
library sites |
|
H |
O |
|
|
|
O |
|
|
|
O |
|
|
|
NHFmoc |
|
CPG |
N |
N |
|
N |
|
||
|
|
|
|
|
|||
|
O |
|
O |
7.38 |
H |
H |
|
|
|
|
O |
||||
|
|
|
|
||||
DMT
ON tag sites
Figure 7.25 Peptide and oligonucleotide encoding: structures 7.37 and 7.38.
loading of the resin bead and presents either orthogonally protected sites, which are used in turn to grow the library structure (NHfluorenylmethoxycarbonyl, NHFmoc) and to grow the tag structures (O-dimethoxytrityl, DMT), such as in 7.38, or multiple cleavage sites such as in 7.37, which allow the release and screening of library individuals with final sequencing of the positive beads and determination of positive structures via the remaining resin-bound peptide copies. Such tags are very convenient for the synthesis of on-bead screened oligomeric libraries. In fact, these libraries (usually made by peptides, peptoids, peptidomimetics, oligonucleotides) require standard reaction conditions, which allow the presence of peptide/ON codes without interfering in library, or tag, synthesis. The tags are made from commercial, cheap nucleotides or α-amino acids; the codes do not usually influence the biological activities of the library components, but when this does happen, this undesired effect can be easily spotted; the codes can be easily read with high sensitivity after the library screening by Edman degradation using an automated peptide sequencer for peptides
7.4 ENCODING METHODS FOR SOLID-PHASE POOL LIBRARIES 305
or by polymerase chain reaction (PCR) for oligonucleotides. Several examples of the successful determination of active oligomeric sequences from SP-encoded pool libraries have been reported, with either oligonucleotide (180–184) or peptide (185–188) tags.
The use of such tags for small organic molecule libraries is prevented by the sensitivity of peptides and, even more so, of oligonucleotides to some of the reaction conditions that are common in organic synthesis: strong bases (proton abstractions, racemizations, alkylations), strong reducing agents (amide reductions), acidic conditions (sugar dehydrations), and so on. More stable tags were thus needed, and a major achievement was reported by Ohlmeyer et al. (111) and Nestler et al. (189) with the introduction of the so-called electrophoric tags. Their optimized structure is reported in Fig. 7.26 (7.40), together with their synthesis (path a), the mechanism of their
|
|
|
HO |
O |
(n) |
O |
|
|
|
|
|
a,b Ar |
|
|
|
|
|
O |
(n) |
OH |
+ |
|
|
|
|
|
7.39 |
|
|
|
|
||||
Ar |
|
|
|
OMe |
|
COOH |
||
|
|
|
OMe |
COOMe |
|
|||
|
|
|
|
|
|
|
||
|
|
|
|
|
Cl |
|
|
|
|
O |
(n) |
O |
Cl |
|
Cl |
Cl |
Cl |
c,d |
Ar |
|
|
|
||||
|
|
Ar = |
|
|
|
|
||
|
7.40 |
|
|
|
|
|
|
|
|
OMe |
N2 |
|
Cl |
|
|
||
|
|
|
|
|
||||
|
n = 1-10 |
|
O |
Cl |
|
|
Cl |
|
|
|
|
|
|
|
|||
|
|
|
|
|
|
|
||
A B
A1-A10 path a B1-B10
O (n) O |
|
|
|
O |
|
(n) O |
|
|
|
|
|
|
Ar |
||
Ar |
|
|
e |
|
|
|
7.41 |
|
|
|
|
|
|
||
7.40 |
|
|
|
|
MeO |
|
|
OMe |
N2 |
|
|
|
|
||
|
|
|
O |
|
|
||
|
|
O |
path b |
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
O |
(n) O |
Ar |
|
|
|
|
|
|
|
|
|
|
|
|
|
MeO |
7.41 |
f,g |
O |
(n) O |
h |
STRUCTURE |
|
|
DETERMINATION |
|||||
|
|
Ar |
|
Si |
|
||
|
|
|
|
|
|
||
|
O |
|
path c |
|
7.42 |
|
|
|
|
|
|
|
|
|
|
a:PPh3, DEAD, rt; b: LiOH; c: (COCl)2; d: CH2N2; e: resin, [(TFA)2Rh]2; f: (NH4)2Ce(NO3)2;
g:Bis-TMSacetamide; h: EC-GC characterization.
Figure 7.26 Electrophoric tags: structure (7.40), synthesis, encoding (7.41), and decoding (7.42) protocols.
306 SYNTHETIC ORGANIC LIBRARIES: SOLID-PHASE POOL LIBRARIES
support onto the solid phase (path b), and their cleavage (path c). The synthesis of the tags is straightforward, so that large quantities of A1−10 and B1−10 can be easily obtained, and they are stable and can be stored for months at room temperature. They are added to the support as acylcarbenes, generated by the treatment of diazoacetates 7.40 with rhodium trifluoroacetate, with no need for orthogonal sites on the resin beads, as the carbenes react aspecifically with the support (the aromatic rings) or with the library individuals. Fortunately, the large excess of the former versus the latter and the low amount of tags used (around 1% of the resin loading) will produce negligible amounts of modified library compounds while ensuring a correct encoding with tags 7.41. Using the binary encoding system where tag mixtures encode for specific monomers, the acylcarbene addition is performed with equimolar amounts of different tags to follow the planned encoding scheme. Tag cleavage is performed, after library synthesis and screening, via oxidative cleavage with cerium ammonium nitrate (CAN) and silylation of the resulting alcohols to give 7.42. The silyl derivatives 7.42 are then characterized via electron capture gas chromatography (EC-GC), and the retention times of each peak produce the structural information needed to identify the library positive(s).
This method employs tags that are stable to the large majority of the reaction conditions used in an SP library synthesis, their synthesis is easy using commercially available precursors, and their presence does not interfere with the library synthesis scheme; they are inert in library screening when on-bead methods are used; and their structure is determined with a fast and reliable method, even though EC-GC is not a very common analytical technique and the cleavage/derivatization procedure is somewhat time-consuming. This encoding system has been widely used by the group that discovered these tags (190–193) as well as by other researchers (194–196) and was instrumental in decoding the structure of positive library components for a wide panel of biological targets, employing both off-bead (194–196) and on-bead (190–194) screening methods.
Another important chemical encoding method, based on secondary amine tags, has been reported (197). The structure, synthesis (path a), and coupling to the solid support (path b) of these tags are reported in Fig. 7.27. The acid tags 7.44 are easily made from N-protected iminodiacetic acid anhydride 7.43 and require an amino function on the resin to be coupled. The use of orthogonally protected resin sites is also necessary; usually 90% of the sites are functionalized with a photolinker, then protected with Fmoc (P1, Fig. 7.27), while the tag sites (≈10%) are protected with an orthogonal group (either Boc or, more recently, Alloc; P2, Fig. 7.27). The coupling of tag units or library units then takes advantage of orthogonal deprotections, as in path b, until the encoded library 7.45 is prepared and QCed. The binary coding system will require some monomers to be encoded by the coupling of equimolar tag mixtures.
After library cleavage, screening, and selection of positives (steps a–c, Fig. 7.28), the positive beads, such as 7.46, are decoded simply by strong aqueous acid hydrolysis, neutralization, and dansylation of the residual amines to give the dansyl derivatives 7.47 (steps d–g), which are all distinct in an HPLC spectrum. Several recent reports have provided optimized HPLC/fluorescence decoding protocols (198), which shortened the average decoding procedure from 1 h to around 6 min, and also alternative
7.4 |
ENCODING METHODS FOR SOLID-PHASE POOL LIBRARIES 307 |
|||
|
|
|
|
P2 |
COOH |
COOH |
b |
N |
|
|
a |
|
|
|
NH COOH |
N |
COOH |
O |
O O |
|
P2 |
|
||
|
O |
P2 |
|
7.43 |
|
|
|
||
c |
R1 N |
N |
COOH |
|
path a
R2 7.44
40 tags prepared and characterized
library sites
H |
|
H |
P1 |
|
H |
P1 |
|
|
N |
d,e,f |
N |
g,h |
N |
7.45 |
|
||
|
P1 |
M1 |
|
M1 |
|
|
||
N |
P2 |
NH |
|
|
NH |
P2 |
O |
|
|
|
|
|
|||||
H |
|
P2 |
path b |
|
||||
|
|
|
N |
|
R |
|||
|
|
|
|
O |
|
|
N |
2 |
|
|
|
|
|
|
|
||
tag sites |
|
|
|
|
|
R1 |
|
|
a: N-protection; b: (COCl2)3, TEA; c: coupling with R1R2NH; d: P1 deprotection; e: resin portioning (1 to n); f: coupling with M1(P1); g: P2 deprotection; h: coupling with 7.44.
Figure 7.27 Secondary amine tags: structure (7.44), synthesis, and encoding (7.45) protocols.
analytical decoding techniques involving capillary electrochromatography (CEC) (199, 200), so as to produce coding patterns that can be attributed to all the positive library individuals (step h, Fig. 7.28).
This method employs tags that are stable to many of the reaction conditions used in an SP library synthesis, even though they should be less stable in general than the aforementioned electrophoric tags; their synthesis is easy using commercially available precursors, and many tags can be prepared and characterized; their presence requires the use of ≈10% of the loading sites for tagging and a library synthesis scheme starting from a resin-bound amine function; and their structure is determined by fast and reliable methods. Researchers at Affymax have used this approach for several bead-based libraries (201–204) and have positively decoded the structures of active compounds.
Other approaches have been reported (127, 205–211), but their usefulness is, as of today, not comparable with the previously reported examples. The reader may refer to some recent reviews (212–217) to have more details on chemical encoding.
308 SYNTHETIC ORGANIC LIBRARIES: SOLID-PHASE POOL LIBRARIES
SELECTION OF
POSITIVES
H |
P1 |
|
|
|
|
|
|
c |
|
|
|
|
|
|
|
|
|
|
||
N |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|||
M1 |
|
|
|
|
|
|
|
a,b |
|
M1 |
+ |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
||
NH |
P |
|
O |
H2N |
|
|
|
|
NH |
P2 O |
|
|
||||||||
2 |
|
|
|
|
|
|
|
|
R2 |
|||||||||||
|
|
|
|
|
|
|
|
R |
|
|
|
|
|
|
|
N |
|
|||
O |
N |
|
|
N |
|
|
|
|
|
O |
|
|||||||||
|
|
|
|
|
|
|
||||||||||||||
|
|
|
|
|
2 |
|
|
|
|
|
|
|
|
N |
||||||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7.46 |
|
|
|
||||
7.45 |
|
|
R1 |
|
|
|
|
|
|
|
|
R1 |
||||||||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
d,e |
|
|
|
|
|
|
|
|
|
|
|
|
|
R1 |
|
Dansyl |
|
f,g R |
+ |
|
|
|
|
H |
||
|
|
|
|
|
|
|
|
|
N |
1 |
|
NH2 + |
|
HOOC |
|
N COOH |
||||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
||||
|
|
|
|
|
|
|
|
|
R2 |
|
|
|
R |
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2 |
|
|
|
|
|
|
7.47
h
STRUCTURE
DETERMINATION
a: P1 deprotection; b: cleavage of library components; c: screening; d: P2 deprotection; e: HCl 6N, reflux, 15 hrs; f: Li2CO3;g: dansyl chloride; h: HPLC/CEC decoding.
Figure 7.28 Secondary amine tags: decoding (7.46) protocols.
7.4.2 An Example: Synthesis and Structure Determination of Positives from Encoded Dihydrobenzopyran Libraries
Baldwin (194) reported the synthesis of four encoded libraries of dihydropyrans L6–L9, whose synthetic scheme is shown in Figs. 7.29–7.32. The monomer sets used are M1–M6, shown in Figs. 7.33 and 7.34. The chemical diversity was built using amines (seven representatives, monomer set M1, step b, Fig. 7.29), carboxyacetophenones (six representatives, two three-member subsets, monomer sets M2a,b, step g, Fig. 7.29) and ketones (two subsets: seven ketones, monomer set M3a, step i, to give the intermediates 7.50, and three cyclic N-protected aminoketones, monomer set M3b, step j, to give the intermediates 7.51, Fig. 7.29). The intermediates 7.51 were treated with acylating/alkylating agents (seven representatives, monomer set M4, step b, Fig. 7.30) to give 7.52. These intermediates were thoroughly mixed with resin 7.50 (step d), and the resin was split into 10 portions (step e, Fig. 7.30). Three of these portions were withdrawn (step a, Fig. 7.31): one was archived (L6, step b), one was reduced and archived (L7, steps c and b), and one was converted to a dithioketal and archived (L8, steps b and d, Fig. 7.31). The remaining seven portions were further diversified with amines (seven representatives, monomer set M5, steps a and b, Fig. 7.32) and, after encoding and mixing/splitting of the resin, with acylating agents (10 representatives, monomer set M6, to give L9, step f, Fig. 7.32).
310 SYNTHETIC ORGANIC LIBRARIES: SOLID-PHASE POOL LIBRARIES
|
|
NO2 |
|
|
|
|
|
|
|
|
R1 |
|
|
|
|
|
|
H |
N |
|
|
O |
|
|
|
O |
|
|
|
||
|
|
N |
|
|
|
|
|
|
|
|
O X |
|
|
|
|
|
|
O |
7.51 |
|
|
|
R5 |
|
|
|
|
|
|
||
|
|
126 intermediates |
|
|
O |
R6 NBoc |
|
|
|
7 identical pools, 126 |
|
R2 |
|
||
|
|
compounds/pool |
|
|
|
||
|
|
NO2 |
|
|
|
|
|
|
|
|
R1 |
|
|
|
|
|
H |
|
N |
|
O |
|
|
a-d |
|
O |
|
|
|
||
N |
O |
|
|
+ |
|||
|
|
X |
|
|
|||
|
O |
7.52 |
|
|
|
R5 |
|
|
|
|
|
|
|
||
|
|
882 intermediates |
|
O |
|
R7 |
|
|
|
single pool |
|
R |
N |
||
|
|
|
R2 |
|
6 |
|
|
|
|
|
|
|
H |
|
|
|
|
NO2 |
|
|
|
|
|
|
|
N |
R1 |
|
|
|
O |
|
H |
|
O |
|
|
||
|
|
O |
|
|
|
||
d,e |
N |
|
|
|
|
|
|
|
O |
X |
|
|
and |
||
|
O |
|
|
|
R3 |
|
R5 |
|
|
|
|
|
O |
||
|
|
|
|
|
|
|
|
|
|
|
|
O |
R4 |
|
R6 N R7 |
|
|
|
R2 |
|
|
H |
|
7.53
1,176 intermediates
10 identical pools
a: TFA; b: coupling with M4; c: encoding; d: mix in one pool; e: resin portioning (1 to 10).
Figure 7.30 SP synthetic strategy to the SP benzopyran encoded (electrophoric tags) pool libraries L6–L9: synthesis of SP intermediates 7.53.
The authors did not report the biological activity of the primary libraries L6–L9, which were tested as a source of biologically relevant compounds. The libraries’ quality was inferred from the parallel synthesis of six library individuals from L6–L9 (7.55–7.60, Fig. 7.35), which were characterized by NMR, MS, HPLC, and gravimetric yield after photorelease. The encoding/decoding procedure was defined in the paper as being “greatly facilitated by encoding with electrophoric tags” and indirectly “accelerating the early phases of drug discovery” (194).
7.4.3 Nonchemical Encoding
The use of tagging methods that do not require a covalent bond between the tag and the solid support is appealing because it reduces the chemical complexity while
