Fragment-Based Discovery of an Apolipoprotein E4 (apoE4) Stabilizer

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE

Article

Fragment-based Discovery of an ApolipoproteinE4 (apoE4) Stabilizer Andrew M. Petros, Alla Korepanova, Clarissa G. Jakob, Wei Qui, Sanjay C. Panchal, Jie Wang, Justin Dietrich, Jason Brewer, Frauke Pohlki, Andreas Kling, Kyle Wilcox, Viktor Lakics, Lamiaa Bahnassawy, Peter Reinhardt, Sarathy Karunan Partha, Pierre M. Bodelle, Marc Lake, Erik Charych, Vincent S. Stoll, Chaohong C. Sun, and Eric Mohler J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00178 • Publication Date (Web): 01 Apr 2019 Downloaded from http://pubs.acs.org on April 1, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Fragment-based Discovery of an ApolipoproteinE4 (apoE4) Stabilizer Andrew M. Petros*, Alla Korepanova, Clarissa G. Jakob, Wei Qiu, Sanjay C. Panchal, Jie Wang, Justin D. Dietrich, Jason T. Brewer, Frauke Pohlki†, Andreas Kling†, Kyle Wilcox, Viktor Lakics†, Lamiaa Bahnassawy†, Peter Reinhardt†, Sarathy Karunan Partha, Pierre M. Bodelle, Marc Lake, Erik I. Charych‡, Vincent S. Stoll, Chaohong Sun, Eric G. Mohler* Research & Development, AbbVie, 1 North Waukegan Road, North Chicago, Illinois 60064 †AbbVie Deutschland GmbH & Co. KG, Neuroscience Research, Knollstrasse, 67061 Ludwigshafen, Germany

‡AbbVie Neuroscience Research, 200 Sydney Street, Cambridge, Massachusetts 02139

ABSTRACT Apolipoprotein E is a 299-residue lipid carrier protein produced in both the liver and the brain. The protein has three major isoforms denoted apoE2, apoE3, and apoE4 which differ at positions 112 and 158 and which occur at different frequencies in the human population. Genome-wide association studies indicate that the possession of two E4 alleles is a strong genetic risk factor for late-onset Alzheimer’s disease (LOAD). In an attempt to identify a small molecule stabilizer of apoE4 function that may have utility as a therapy for Alzheimer’s disease, we carried out an NMR-based fragment screen on the N-terminal domain of apoE4 and identified a benzyl amidine based fragment binder. In addition to NMR, binding was characterized using various other biophysical techniques and a crystal structure of the bound core was obtained.

ACS Paragon Plus Environment

Journal of Medicinal Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Core elaboration ultimately yielded a compound which showed activity in an IL-6 and IL-8 cytokine release assay.

2


Page 2 of 48

Page 3 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION Apolipoprotein E (apoE) is a lipid carrier protein which is manufactured both in the liver and brain and plays a major role in overall lipid homeostasis.1 In the CNS apoE is primarily synthesized by astrocytes, but there is accumulating evidence that apoE is expressed in microglia as well.2 In addition, there are reports that neurons may produce apoE under conditions of stress.3 The 299-residue protein exists in the human population as three major polymorphic alleles, denoted apoE2 (E2), apoE3 (E3), and apoE (E4) which occur with a frequency of 8.4%, 77.9%, and 13.7%, respectively.4 Genome-wide association studies (GWAS) have consistently shown the possession of two E4 alleles to be the strongest genetic risk factor for late-onset Alzheimer’s disease (LOAD)5 with an odds ratio of 10-14.4 However, it is not clear whether this is a result of some loss of function for the E4 protein compared to the E2 and E3 proteins, from some toxic gain of function, or from a combination of the two.6 The E4 isoform is primarily associated with increased amyloid beta production and accumulation, as well as impaired clearance.7 In mice carrying a human tau mutation associated with frontal-temporal dementia and the E4 isoform, increased neurodegeneration has been reported.8 Additional potential mechanisms of action for the effects of the E4 isoform include impaired cerebral blood flow9, impaired insulin signaling10, aberrant microglial activity11, and impaired lipidation.12 Some or all of these effects of the E4 isoform may contribute to the increased risk for developing Alzheimer’s disease observed in E4 carriers. The apoE protein is composed of two domains, an N-terminal domain which runs from residue 1 to 164 and a C-terminal domain which runs from residue 165 to 299. The three alleles vary at positions 112 and 158 of the N-terminal domain, with the E2 protein having a cysteine residue at both positions, the E3 protein having a cysteine at position 112 and an arginine at

3



position 158, and the E4 protein having an arginine at both positions. In addition, the N-terminal domain contains the binding site for the low-density lipoprotein, and other lipoprotein receptors.13 Various biochemical, biophysical, and structural studies have been carried out on the apoE proteins over the years. A crystal structure has been obtained for the N-terminal domain of all three variants while an NMR-derived structure has been obtained for a mutant form of the full-length E3 protein.14 The N-terminal domain of apoE4 is an attractive target for application of fragment-based drug discovery methods as it is relatively small, stable, and can be readily expressed in E. coli. Over the past two decades fragment-based methods have come to play an integral role in the drug discovery process especially for non-enzymatic targets such as those involving proteinprotein interactions. For many of these targets traditional biochemical screening methods fail to find tractable small-molecule starting points and therefore, application of biophysical screening methods has been critical. Although the precise mechanism by which apoE4 contributes to the development Alzheimer’s disease is unclear, there are several lines of research which suggest that the effects of apoE4 result from a loss-of-function and/or a toxic gain-of-function.15 One possible explanation for these effects is that apoE4 is less stable than the other apoE isoforms.16 This may lead to greater protease degradation (loss-of-function) as well as a propensity to aggregate, alone, or with amyloid beta (toxic gain-of-function). Stabilization of apoE4 may decrease degradation or aggregation to an extent that may impact disease pathology. Others have reported putative apoE4 binders discovered via biochemical screening.17 In this paper we describe biophysical, and specifically, fragment-based methods aimed at finding an apoE4 stabilizer which could potentially be developed into an apoE4-based therapy for Alzheimer’s disease.

4


Page 4 of 48

Page 5 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


RESULTS AND DISCUSSION We initially purified the N-terminal domain of apoE4 along with the full-length protein from E. coli expression, and likewise, for comparison, similar constructs for apoE3. For both full-length proteins we introduced five point mutations in the C-terminal domain (F257A, W264R, V269A, L279Q, V287E) as described by Zhang et al. in order to reduce the tendency of the full-length protein to aggregate.1b Subsequently, we employed a mammalian expression system to obtain full-length, wild type, apoE4 and apoE3. As part of our initial characterization of the recombinant apoE proteins, we carried out a thermal shift analysis (TSA) of the N-terminal domain of both apoE3 and apoE4 and the fulllength mutant proteins, as described in the Experimental Section. Melting curves for the Nterminal domains of both apoE3 and apoE4 are presented in Figure 1A and as the data shows, the apoE3 N-terminal domain melts about ten degrees higher than the apoE4 N-terminal domain. The same holds true for the full-length mutant proteins as shown in Figure 1B. This supports the observations of Hatters et al. who showed that apoE4 is more sensitive to denaturation by guanidine HCl than apoE3.18 We later prepared the N-terminal domain of apoE2 and found it to be slightly more stable than that of apoE3 (Table 1). The question then arose as to whether this reduced stability of apoE4 with respect to the two other isoforms, might, at least in part, be related to its aberrant function with respect to apoE2 and apoE3, and whether a small molecule binder would be able to rescue this effect, through stabilization of apoE4.

5



Figure 1. (A) Melting curves for the Nterminal domain of apoE4 (blue) and the N-terminal domain of apoE3 (red). (B) Melting curves for full-length, E. coli derived, apoE4 (blue) and full-length, E. coli derived, apoE3 (red). Tm values are listed in Table 1.

6


Page 6 of 48

Page 7 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 1. Melting Temperature for apoE4 Proteins Protein form

Tm (οC)

σ

apoE3 NT*

60.9

0.09

apoE4 NT

52.8

0.3

apoE2 NT

63.2

0.18

apoE3 FL#

66.6

0.6

apoE4 FL

55.3

0.8

*N-terminal

domain mutant protein

#Full-length,

7



While numerous biophysical and biochemical techniques have been used to find fragment binders to protein targets, we chose to employ protein-based, two-dimensional NMR spectroscopy as our primary method to find fragment binders to apoE4. NMR methods based on collection of either

15N-HSQC

or

13C-HSQC

spectra using isotopically-labeled protein have

proven to be extremely robust, as any chance of getting false positive binders is completely eliminated. In our hands, the full-length apoE4 protein, even with the introduction of the five point mutations, showed a tendency to aggregate and yielded a sub-optimal NMR spectrum. The N-terminal domain, on the other hand, yielded a reasonable NMR spectrum suitable for proteinbased fragment screening. The N-terminal domain of apoE4 is composed of a four-helix bundle, as shown in Figure 2A. While a close examination of one of the crystal structures from the Protein Data Bank (PDBID: 1GS9) does not reveal any obvious potential binding site for a fragment-sized molecule, in another structure, utilizing a shorter protein construct (PDBID: 1B68), a Χ1 rotation of the Trp34 sidechain (Figure 2B) exposes a small hydrophobic pocket as a potential fragment binding site. This is reminiscent of the conformational remodeling we observed for the EED protein whereby a compound-induced change in rotamer orientation for Tyr365 of the trimethyllysine pocket exposed a larger, more ligandable pocket than observed when bound to the histone peptide.19 Therefore, based on these observations we decided to pursue an NMRbased fragment screen. In order to enable this NMR-based fragment screen, the apoE4 N-terminal domain protein was expressed in E. coli and labeled with 13C at the methyl groups of isoleucine (δ only), valine, leucine, and methionine as described in the Experimental Section. Screening was carried

8


Page 8 of 48

Page 9 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


A)

N

C B)

Figure 2. (A) Ribbon representation of crystal structure of the N-terminal domain of apoE4 (PDB ID: 1GS9). Nand C-termini are labeled. Sidechain of Trp34 is shown in sticks. (B) Overlay of two crystal structures of the Nterminal domain of apoE4 highlighting the rotamer differences in the Trp34 sidechain. Trp34 of the 1GS9 structure is in orange while that of the1B68 structure is in green.

9



out by monitoring changes induced in a two-dimensional

13C-HSQC

Page 10 of 48

spectrum of the protein

upon addition of potential fragment binders (Figure 3). At this point the methyl resonances of the protein had not been sequence specifically assigned. A total of 4068 fragments were initially tested in mixtures of 12 for binding to the apoE4 protein. All fragments were “rule of three” compliant20 with a solubility of at least 500 µM in PBS as determined by NMR. Those mixtures which induced significant chemical shift perturbations were then deconvoluted in order to uncover the individual fragment hits. In order to confidently count a crosspeak as shifted we typically set a minimum change of ~ 0.05 ppm in the proton dimension or ~ 0.5 ppm in the carbon dimension. The overall hit rate from the screen was ~ 0.3%. Binding affinities for these fragment hits were then determined via NMR titration. Among the best of the NMR-derived fragment binders was compound 1, shown in Figure 3 (inset), which has an affinity of 900 μM for the apoE4 N-terminal domain. This translates to a Binding Efficiency Index (BEI) of 15, a Ligand Efficiency (LE) of 0.30, and a Lipophilic Ligand Efficiency (LLE) of 2.9. Definitions for BEI, LE, and LLE are provided in the Supporting Information document. We also measured the affinity of this fragment for apoE4 using surface plasmon resonance (Biacore) as shown in Figure 4. The Biacore-derived affinity of this fragment for the N-terminal domain of apoE4 and for full length apoE4, obtained from mammalian cell expression, was 205 μM and 233 μM, respectively. In addition, we used Biacore to measure the affinity of this fragment for full-length apoE3 and found it to be 890 μM (data not shown). Our next step was to determine what effect, if any, this fragment might have on the thermal

10


Page 11 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


14

N Cl

N

16 18 20 22 24 26

2.0

1.5

1.0

0.5

0.0

1

H

Figure 3. 13C-HSQC spectra (600 MHz) of the N-terminal domain of apoE4 labeled at the methyl groups of isoleucine (δ only), valine, leucine, and methionine. Black spectrum is protein alone at 30 µM while red spectrum is in the presence of 1 mM compound 1 (inset).

11


13

C


A) RU 30

25

20

Response

15

10

5

0

-5

-10 -10

0

10

20

30

40

50

Tim e

s

B) RU 30

25

20

15

Response

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 48

10

5

0

-5

-10 -10

0

10

20

30

Tim e

Figure 4. Steady state fit (black curve) to equilibrium responses (colored diamonds) of fragment binding to N-terminal domain of apoE4 (A) and for full length apoE4 (B). Inset shows the binding sensorgrams. Note fast on and fast off kinetics for fragment.

12


40

50 s

Page 13 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


stability of apoE4. Addition of this fragment to full-length, mutant, apoE4 at a concentration of 5 mM resulted in a ~4° stabilization of the protein (data not shown). In contrast, fragment addition to apoE3 resulted in a thermal stabilization of only ~0.5° (data not shown). To further assess the effect of fragment binding on apoE4 with respect to its physical properties, and potentially its function, we implemented the liposome breakdown assay first described by Weers and coworkers.21 These researchers reported that apoE proteins can break down large DMPG liposomes into smaller discoidal complexes and that this breakdown results in a dramatic reduction of light scattering intensity. While this assay is not intended to exactly mimic the lipid binding function of apoE4 in the body, it may, nonetheless, provide some insight as to how apoE4 stabilization via a small molecule may modulate endogenous lipid binding. For the N-terminal domain of the three isoforms the rate of liposome breakdown followed the order of E4>E3>E2.

Since the N-terminal domain most likely unfolds, at least somewhat, for

liposome engagement, Weers and coworkers hypothesize that the breakdown efficiency for the different isoforms is related to protein stability with apoE4 being the least stable and thus the most efficient at liposome breakdown. As shown in Figure 5A, our results are very similar to those obtained by Weers et al. We observed a T½ for liposome breakdown of 3, 15, and 30 minutes for the N-terminal domain of apoE4, apoE3, and apoE2, respectively. In subsequent experiments we set out to clarify the potential effect of fragment binding on the kinetics of liposome breakdown by apoE4. Since we had already shown that fragment binding stabilized apoE4 with respect to thermal denaturation,

13



Figure 5. Liposome breakdown assay. (A) Kinetics of liposome breakdown for the Nterminal domain of apoE2 (square), apoe3 (triangle), and apoE4 (circle). DMPG alone (control) is shown as diamonds. (B) Dose response for addition of compound 1 to apoE4 N-terminal domain. Compound concentration as follows: 15.6 μM (star), 62.5 μM (circle), 125 μM (triangle), 250 μM (square), 500 μM (diamond). (C) Dose response for addition of compound 1 to fulllength, wild-type apoE4 from mammalian cell expression. Compound concentration as follows: 15.6 μM (star), 31.25 μM (circle), 62.5 μM (triangle), 250 μM (square), 500 μM (diamond).

14


Page 14 of 48

Page 15 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


we expected fragment binding to slow the kinetics of apoE4 liposome breakdown. Figure 5B shows the effect of increasing fragment concentration on the breakdown of DMPG liposomes by the apoE4 N-terminal domain while Figure 5C shows the effect of increasing fragment concentration on the breakdown of DMPG liposomes by the full-length, wild-type protein derived from mammalian cell expression. At the highest fragment concentration tested, the kinetics of liposome breakdown by apoE4 approaches the same rate as observed for apoE3 and apoE2 alone. Having shown that fragment binding both stabilizes apoE4 and has an effect on the kinetics of liposome breakdown, we wanted to improve the potency of our fragment using structure-based drug design. The structure of compound 1 bound to the N-terminal domain of apoE4 (Figure 6A) was determined by soaking this fragment into a crystal of the apo protein as described in the Experimental Section. In this structure, the Trp34 sidechain is rotated up, similar to that observed in one of the crystal structures of the unliganded protein; and the compound binds into the pocket formed, at least partly, by this sidechain rotation. The phenyl ring of the ligand sits on a relatively flat surface presented by the sidechain of Leu30 while the appended chloro atom points into a hydrophobic sub-pocket formed mainly by sidechains from Trp34, Leu148, and Leu149. The cyclobutyl ring sits in another hydrophobic sub-pocket formed by Trp26, Leu30, and Ala152. Finally, the amidine moiety makes a direct hydrogen bond with the sidechain of Asp153. Based on the structure of compound 1 bound to the apoE4 N-terminal domain, we chose to elaborate off the central phenyl ring at the position meta to both the attached chloro and the attached cyclobutyl. Through the initial addition of an unsubstituted phenyl ring at this meta position we obtained compound 2 (Table 2), which exhibited an affinity of 230 µM (by NMR).

15



A)

W34 L149

D153 L30 W26

B)

D35

Figure 6. (A) Crystal structure of compound 1 bound to the N-terminal domain of apoE4. Hydrogen bonds are denoted with dashed magenta lines (PDB ID 6NCN). (B) Crystal structure of compound 3 bound to the N-terminal domain of apoE4 (PDB ID 6NCO).

16


Page 16 of 48

Page 17 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


This 4-fold gain over the parent likely results from favorable Van der Waals contact between the added phenyl ring and the protein. To further elaborate we then made various substitutions around this appended phenyl ring and found that substitution at the para position was the most productive. Addition of a hydroxyisopropyl moiety at the para position of the appended phenyl yielded compound 3 (Table 2) which exhibited an affinity of 30 µM, a 30-fold gain over the parent hit. A crystal structure of this compound bound to the apoE4 N-terminal domain (Figure 6B) shows that the appended hydroxyl group of the compound makes a hydrogen bond with the sidechain of Asp35, and this is the likely explanation for the gain in potency over compound 2. Based on the observed hydrogen bonding of compound 3 to Asp35, various other polar, hydrogen-bonding, substitutions were made at the para position as shown in Table 2. The most potent binders, which have an NMR-derived affinity of < 5 µM, are appended with either a primary amino group 7 or an N-(2-(dimethylamino)ethyl)amide group 8 at the para position of the phenyl ring. Compound 8 stabilized the full-length, mutant, apoE4 protein in TSA by about the same amount (~4° C) as the parent fragment, but at a 5-fold lower concentration (data not shown). Scheme 1 shows the preparation of biaryl cyclobutaneamidine 2 starting from commercially available 3-chloro-5-bromphenylacetonitrile 9. Reaction of acetonitrile 9 in the presence of potassium hydroxide and dibromopropane provided cyclobutane nitrile 10. Subsequent Suzuki coupling with phenylboronic acid provided biarylcyclobutane nitrile 11 and addition of Weinreb’s methylchloroaluminumamide.22 derived from trimethylaluminum and ammonium chloride produced desired amidine 2 in low (12%) but isolatable yield.23

17



Page 18 of 48

Table 2. Phenyl SAR HN

NH2

Example

R

NMR KD µM*

BEI#

LE

LLE

900

15

0.30

2.9

230

13

0.25

1.8

30

13

0.26

2.8

20

13

0.26

4.0

20

12

0.25

3.6

8

15

0.31

3.5

17

>0.34

>6.4

13

>0.27

>5.6

Cl

1

H

2 3

OH O

4

HN O O S HN

5 Cl

6 OH

7

NH2 O

8 *Estimated

HN

N

error is ± 10%, #BEI, LE, and LLE are defined in Supporting Information

18


Page 19 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Scheme 1. Synthesis of Compound 2a Cl Cl

a

N

N

b

Cl N

Br

Br 10

9

11

NH

Cl c

NH2

2

aReagents

and Conditions; (a) 1,3-dibromopropane, KOH, DMSO, RT, 87%; (b) Phenylboronic acid, Pd(dppf)Cl2, K2CO3, 6:2:1 toluene:EtOH:H2O, 90ºC, 67%. (c) NH4Cl, AlMe3, toluene, reflux, 5h, 12%.

Scheme 2: Synthesis of Analogues 3 and 4a Cl N

a

NH

Cl

NH

Cl

NH2

Br

Br

10

12

aReagents

b

NH2 Ar 3-8

and Conditions; (a) (i) 4N HCL/EtOH, RT, (ii) 4N NH3/MeOH, RT, 57%; (b)

Arylboronic acid, Pd2(dba)3, meCgPPh, K3PO4, dioxane/H2O, 100ºC, 23-37%

19



Exploring the SAR around the biaryl region required a more robust chemistry route that is described in Scheme 2.

In this route, the Pinner reaction was utilized to first convert

phenylcyclobutane nitrile 10 to the iminoethylester salt by acid catalyzed addition of ethanol to the nitrile. Subsequent reaction with 4N ammonia in methanol provided amidine 12 in modest yield (57%).

The synthesis of compounds 3-8 commenced with a Suzuki coupling of the

respective boronic acid or boronic ester. Optimal conditions for the Suzuki coupling relied on the employment of potassium phosphate as base, shorter reaction times (

Fragment-Based Discovery of an Apolipoprotein E4 (apoE4) Stabilizer

Recommend Documents