Tải bản đầy đủ - 0 (trang)
V. SOLID-PHASE PARALLEL SYNTHESIS AND NONPEPTIDE PHENYL PHOSPHATE LIBRARIES

V. SOLID-PHASE PARALLEL SYNTHESIS AND NONPEPTIDE PHENYL PHOSPHATE LIBRARIES

Tải bản đầy đủ - 0trang

pattern of the plate. For example, if different diversity elements are

added across the rows of the plate (one diversity element, repeated 8

times, per row), with the 12 columns housing the second set of diversity

elements, 96 discrete compounds (e.g., 8 A-group  12 B-group = 96

compounds; Fig. 7) will result. This format permits the rapid synthesis of

relatively large structurally biased libraries by systematically combining

sets of diversity groups.

To streamline plate synthesis, we developed with Cyberlab, Inc. [25] a

custom high throughput organic synthesizer designed to process the 96deepwell reaction blocks (Fig. 8). This instrument was constructed to

tolerate a wide range of chemistries; therefore, all liquid contacts (syringes,

needles, tubes, and valves) are made of glass, stainless steel, or Teflon.

Coaxial tip needles with N2 inlets (connected to a bubbler) allow inert

dispensing and withdrawal of liquid reagents from the closed vessels

without excessive negative or positive pressure buildup. The instrument

head, which can access all positions on the deck, is fitted with single-needle

and four-needle probes. The tandem use of both needle probes facilitates

the transfer (via 5 mL syringe pumps, not shown) of all reaction intermediates from the reagent vials (left side) to the resin-containing 96deepwell reaction blocks (right side).

The reaction block, which provides a fully enclosed reaction

environment (Teflon, polypropylene, and silicone rubber seals) is a

slightly modified version of a design first disclosed by Sphynx Pharmaceuticals [23]. Figure 9 shows the reaction block and the reagent

vials (100, 30, and 10 mL sizes) in their fully assembled and disassembled states. Holes at the bottom of the wells of the 96-deepwell

polypropylene plate (sealed in fully assembled reaction block) allow the

reaction solutions to be removed from the wells (via a separate vacuum

plenum) and the functionalized resin (retained by Teflon frits) to be

washed with solvents.

The use of the phenyl phosphate group as both a solid support

attachment site and a crucial binding element represents what has been

referred to as a ‘‘pharmacophore-linking’’ strategy [26]. We explored a

variety of phenyl phosphate tether functionalities to provide resins varying

in substitution pattern and in chemical flexibility (Scheme 1 and Table 4)

[22]. All phenyl phosphate resins were synthesized in batch quantities of

20 g or more. Resin synthesis began with the addition of either p-methoxybenzyl alcohol or benzyl alcohol to commercially available bis(diisopropylamino)chlorophosphine, followed by addition of the diversity phenol

[(R1)-OH, DIAT (diisopropylamino tetrazole)]. Displacement of the



Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.



Figure 8 High throughput organic synthesizer developed in collaboration with

Cyberlab, Inc. [25] and designed to process the 96-deepwell reaction blocks. The

instrument is capable of tolerating a wide range of chemistry (liquid contacts are

glass, stainless steel, or Teflon) and accomplishes the transfer of reagents with

coaxial tip (N2 inlets) single-needle and four-needle probes.



remaining diisopropylamino group from 1 with Wang resin and oxidation

with 4-methylmorpholine N-oxide (NMO) provided the protected phenyl

phosphate resins 2 and 3 in excellent yields, as shown in Table 4. Two types

of functionality, namely, protected carboxy and amino groups, differentiated the starting phenols. Reaction schemes demonstrating compound

synthesis using both phenol types are shown in Scheme 2 [22]. Mild Fmoc

deprotection (1% DBU/DMA) of resin 2a and amide formation using

standard coupling conditions [TBTU, DIEA, p-(CO2H)PhCH2NHFmoc)] resulted in attachment of the first diversity element to provide

resin product 4. A second deprotection followed by a double, one-pot



Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.



Figure 9 The 96-deepwell reaction block and the reagent vials (100, 30, and 10

mL sizes) used in the organic synthesizer in their fully assembled and

disassembled states.



Scheme 1



Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.



Table 4 Yields and Loading of Phenylphosphate Resins

Resin

2a

2b

2c

2d

2e

3f

3g

3h

3i



R1



Loading (mmol/g)



Yield (%)



p-(FmocNHCH2CH2)Ph

p-(FmocNHCH2)Ph

m-(FmocNHCH2)Ph

p-(Allyl-O2CCH2CH2)Ph

p-(Allyl-O2CCH2)Ph

m-(Allyl-O2CCH2)Ph

p-[ p-(FmocNHCH2)PhO]Ph

p-(Allyl-O2CCH=CH)Ph

m-(Allyl-O2CCH=CH)Ph



0.659

0.637

0.627

0.730

0.672

0.807

0.552

0.535

0.796



93

89

88

92

84

98

81

66

98



Source: Ref. 22.



Scheme 2 Abbreviations: DBU, 1,8-diazabicyclo[5.4.0]undec-7-ene; DMA,

N,N-dimethylacetamide; TBTU, O-benzotriazole-1-yl-N,N,NV,NV-tetramethyluronium tetrafluoroborate; DIEA, diisopropylethylamine; RA, reductive amination; TFA, trifluoroacetic acid; DCM, dichloromethane.



Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.



reductive amination [Na(OAc)3BH, 2-ethylbutyraldehyde] provided the

fully coupled compound with a branch point at the second diversity

site. Mild conditions (30% TFA/CH2Cl2) to cleave the final compound

from the solid support as well as to remove the p-methoxybenzyl protecting group resulted in the isolation of compound 5 in 86% HPLC purity,

following in vacuo concentration. For the synthesis of compound 7, the

allyl ester of resin 2d was deprotected under palladium-mediated conditions, followed by amide coupling [TBTU, DIEA, m-(NH2CH2) PhCO2allyl] to generate the functionalized phenyl phosphate resin 6. A second

deprotection and coupling [TBTU, DIEA, NH(Me)CH2Ph] provided the

bisamide resin-bound compound, which was cleaved and isolated as

described earlier to yield compound 7 in 66% HPLC purity.

The compound types synthesized by using the foregoing combinatorial approach are represented in Figure 10. Variations in functional

group connectivity (e.g., amides, olefins, sulfonamides) reflect the wide

range of chemistry that was pursued in the generation of these libraries.

Some bifunctional A-group and monofunctional B-group diversity elements used in the coupling reactions are shown in Figure 11. Alkyl

and aryl phosphate ester groups (R; see Fig. 10) were also explored to



Figure 10 Representations of some of the compound types synthesized in the

nonpeptide phenyl phosphate libraries.



Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.



investigate the binding consequences of reduced charge at the phosphate

group. Initial Src SH2 screening produced hits, which were resynthesized

and then retested in the binding assay. Some of the higher affinity

compounds are shown in Figure 12. Although more than 10,000 compounds were produced by this methodology, only marginal binding

affinities and no high-resolution x-ray or NMR structures were achieved;

the poor aqueous solubility and undesirable physical properties of the

molecules are likely to have hampered these efforts. At this point a decision

was made to pursue a much more structure-based approach. Compounds



Figure 11 List of some of the molecular diversity building blocks used in the

construction of the nonpeptide phenyl phosphate libraries.



Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.



Figure 12 Resynthesized library hits identified from the high throughput

fluorescence-polarization assay along with their Src SH2 binding data.



Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.



lending themselves to possible x-ray and/or NMR co structure determination were emphasized.



VI. STRUCTURE-BASED, SMALL-MOLECULE

LIBRARIES TO EXPLORE Src SH2 BINDING

Refocusing our drug-discovery strategy prompted us to revisit the initial

lead compound, pTyr-Glu-Glu-Ile. It was clear that to generate high

affinity, small-molecule compounds for Src, we would likely need to

maintain the key binding interactions of the pTyr-Glu-Glu-Ile motif, as

well as to explore molecules capable of mimicking or interacting with the

structural waters found in the Src SH2-phosphopeptide complexed x-ray

structure. A template would be required that allowed access to both

pockets (pY and pY+3), mimicking the ‘‘two-pronged’’ binding mode

of pTyr-Glu-Glu-Ile. Noteworthy in this regard, a novel, de novo designed

nonpeptide 8 was disclosed [14] with comparable binding to Ac-pTyrGlu-Glu-Ile-NH2 (phosphopeptide 9) (Fig. 13). Significant interactions

involving the benzamide functionality were revealed in an x-ray structure

of 8 bound to Src SH2 [14]. In addition to interacting with several key sites

of Src SH2 (e.g., the pY/pY+3 pockets and the CO of His204), this

compound displaces both structural water molecules and makes a direct

hydrogen bond contact with the backbone NH of Lys206 through its

benzamide CO moiety. The effect of this carboxamide group on Src SH2

binding is demonstrated by the related compounds 10 and 11 [14], in which

the desamide compound 11 binds with over 15-fold lower affinity than 10

(Fig. 13). ARIAD’s strategy was to utilize this high affinity benzamide

template to gain a better understanding of nonpeptide interactions with Src

SH2, and then to advance a database of structure–activity relationships

(SARs) to ultimately develop novel, proprietary Src SH2 inhibitors.

Subsequent to the disclosure of compound 8, a second-generation, higher

affinity compound, containing a methylated benzamide template in the

context of a pTyr group, was reported [27]. A literature procedure [27] was

used to synthesize this compound (12, AP21733) [16], and a 2.5 A˚ x-ray

crystal structure of Lck SH2 (S164C), a protein homologue of Src SH2,

complexed with AP21733, was obtained (M. H. Hatada, unpublished

results). The proposed S-configuration of the benzylic methyl stereocenter

of AP21733 was confirmed through independent asymmetric synthesis

[28]. The Lck SH2–nonpeptide structure reveals adherence to the historical

pTyr-Glu-Glu-Ile interactions in the pY pocket and shows carboxamide



Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.



Figure 13 Series of de novo designed nonpeptides containing a benzamide

template (exemplified by compound 12, AP21733) designed to interact favorably

with Src SH2 and specifically to displace structural waters found in complexed

Src SH2 structures [14,27]. The Src SH2 binding IC50 is shown for each

compound, as well as a comparative IC50 for Ac-pTyr-Glu-Glu-Ile-NH2

(compound 9).



contacts with Lys182 (206 in Src) and Ile193 (217 in Src). The phenyl ring

of the benzamide template also forms favorable stacking interactions with

Tyr181 (205 in Src). Although the cyclohexylmethyl group interacts with

the pY+3 pocket, the contacts are primarily surface type and do not

extend as deeply into the pocket as the Ile of pTyr-Glu-Glu-Ile. Consequently, SAR exploration of the pY+3 pocket, which had not been

rigorously studied with nonpeptide (peptidomimetic) small molecules

[13,14], became the first objective to be investigated.

Parallel synthesis provides the means of rapidly preparing discrete

analogues for both lead generation and lead optimization strategies,

which makes it an attractive option for developing compound databases

for therapeutic targets. Furthermore, the incorporation of structure-based

methods into the design and evaluation of parallel synthetic libraries has

proven to be a successful strategy for integrating the two drug discovery

technologies [29]. For the synthesis of the benzamide-containing compounds, we devised a hitherto unreported solid phase, parallel synthetic



Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

V. SOLID-PHASE PARALLEL SYNTHESIS AND NONPEPTIDE PHENYL PHOSPHATE LIBRARIES

Tải bản đầy đủ ngay(0 tr)

×