Discovery of GDC-0853: A Potent, Selective, and Noncovalent Bruton's Tyrosine Kinase Inhibitor in Early Clinical Development

Article abstract of DOI:10.1021/acs.jmedchem.7b01712

Bruton's tyrosine kinase (Btk) is a nonreceptor cytoplasmic tyrosine kinase involved in B-cell and myeloid cell activation, downstream of B-cell and Fcγ receptors, respectively. Preclinical studies have indicated that inhibition of Btk activity might offer a potential therapy in autoimmune diseases such as rheumatoid arthritis and systemic lupus erythematosus. Here we disclose the discovery and preclinical characterization of a potent, selective, and noncovalent Btk inhibitor currently in clinical development. GDC-0853 (29) suppresses B cell- and myeloid cell-mediated components of disease and demonstrates dose-dependent activity in an in vivo rat model of inflammatory arthritis. It demonstrates highly favorable safety, pharmacokinetic (PK), and pharmacodynamic (PD) profiles in preclinical and Phase 2 studies ongoing in patients with rheumatoid arthritis, lupus, and chronic spontaneous urticaria. On the basis of its potency, selectivity, long target residence time, and noncovalent mode of inhibition, 29 has the potential to be a best-in-class Btk inhibitor for a wide range of immunological indications.

Full text of DOI:10.1021/acs.jmedchem.7b01712

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE
Drug Annotation
Discovery of GDC-0853, a Potent, Selective, and Non-Covalent
Bruton’s Tyrosine Kinase Inhibitor in Early Clinical Development
James J. Crawford, Adam R. Johnson, Dinah L. Misner, Lisa D. Belmont, Georgette M. Castanedo, Regina
Choy, Melis Coraggio, Liming Dong, Charles Eigenbrot, Rebecca Erickson, Nico Ghilardi, Jonathan Hau,
Arna Katewa, Pawan Bir Kohli, Wendy Lee, Joseph W. Lubach, Brent S. McKenzie, Daniel Fred Ortwine,
Leah Schutt, Suzanne Tay, Binqing Wei, Karin Reif, Lichuan Liu, Harvey Wong, and Wendy B. Young
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 19 Feb 2018
Downloaded from http://pubs.acs.org on February 20, 2018
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.
Journal of Medicinal Chemistry 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 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of GDCꢀ0853, a Potent, Selective, and
NonꢀCovalent Bruton’s Tyrosine Kinase Inhibitor in
Early Clinical Development.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
*
James J. Crawford , Adam R. Johnson, Dinah L. Misner, Lisa D. Belmont, Georgette Castanedo,
$
Regina Choy, Melis Coraggio, Liming Dong, Charles Eigenbrot, Rebecca Erickson, Nico
Ghilardi, Jonathan Hau, Arna Katewa, Pawan Bir Kohli, Wendy Lee, Joseph W. Lubach, Brent
S. McKenzie, Daniel F. Ortwine, Leah Schutt, Suzanne Tay, BinQing Wei, Karin Reif, Lichuan
*
Liu, Harvey Wong, and Wendy B. Young
Genentech, Inc., 1 DNA Way, South San Francisco, California 94080, USA
$
formerly of ChemPartner, No. 1 Building, 998 Halei Road, Zhangjiang HiꢀTech Park, Pudong
New Area, Shanghai, China 201203
KEYWORDS Kinase inhibitor, Bruton's tyrosine kinase, Btk, Rheumatoid arthritis, Lupus,
kinase mutants, C481S, GDCꢀ0853.
ABSTRACT Btk is a nonꢀreceptor cytoplasmic tyrosine kinase involved in Bꢀcell and myeloid
cell activation, downstream of Bꢀcell and Fcγ receptors, respectively. Preꢀclinical studies have
indicated that inhibition of Btk activity might offer a potential therapy in autoimmune diseases
such as rheumatoid arthritis and systemic lupus erythematosus. Here we disclose the discovery
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and preꢀclinical characterization of a potent, selective, and nonꢀcovalent Btk inhibitor currently
in clinical development. GDCꢀ0853 (29) suppresses B cellꢀ and myeloid cellꢀmediated
components of disease, and demonstrates doseꢀdependent activity in an in vivo rat model of
inflammatory arthritis. It demonstrates highly favorable safety, pharmacokinetic (PK), and
pharmacodynamic (PD) profiles in preꢀclinical and Phase 2 studies are ongoing in patients with
rheumatoid arthritis, lupus and chronic idiopathic urticaria. Based on its potency, selectivity,
long target residence time, and nonꢀcovalent mode of inhibition, 29 has the potential to be a bestꢀ
inꢀclass Btk inhibitor for a wide range of immunological indications.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Introduction
Bruton’s tyrosine kinase (Btk) is a nonꢀreceptor Tec family tyrosine kinase that is broadly
1
,2
expressed in hematopoietic cells, with the exception of T cells. Btk plays a crucial role in
signaling through the Bꢀcell antigen receptor (BCR) and the Fcγ receptor (FcγR) in B cells and
1
,3ꢀ5
myeloid cells, respectively.
Spontaneous mutations in the pleckstrin homology domain of the
6
Btk gene found in xid mice, and targeted mutations in Btk knockout mice, both result in Bꢀcell
2,7
development and proliferation defects. These deficiencies are manifested by reduced numbers
of mature circulating B cells, greatly reduced serum titers of certain immunoglobulin (Ig)
isotypes—particularly IgM and IgG3—and poor Tꢀindependent antibody responses. Xꢀlinked
agammaglobulinemia (XLA) is a human congenital immunodeficiency condition that occurs in
2
,7
males, which is associated with low to undetectable btk mRNA levels and Btk protein
8
expression and, as a consequence, low to no Btk kinase activity. In contrast to the phenotype
observed in xid mice, Bꢀcell development in XLA patients is almost completely blocked at the
proꢀBꢀcell to preꢀBꢀcell transition leading to extremely low to undetectable levels of mature B
ACS Paragon Plus Environment
2

Page 3 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
,7
cells and, in turn, immunoglobulins. Left untreated, XLA patients succumb to recurring
9
ꢀ11
infections, but can be treated with Ig replacement therapy.
Taken together, these factors
indicate that inhibition of Btk represents an attractive potential therapeutic approach for the
treatment of immunological disorders in which B cells and myeloid cells induce or sustain an
excessive autoimmune response such as rheumatoid arthritis (RA) or systemic lupus
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
,5,7,12ꢀ14
erythematosus (SLE).
For over a decade, there have been significant efforts expended
towards the goal of identifying Btk inhibitors for clinical application in chronic inflammatory
1
5ꢀ23
diseases as well as hematologic malignancies.
While the irreversible Btk inhibitor ibrutinib
1
5,24
(PCIꢀ32765, Imbruvica®)
has been successful in treating Bꢀcell malignancies and is
25,26
approved for chronic lymphocytic leukemia (CLL),
relapsed or refractory mantleꢀcell
2
7,28
29
lymphoma (MCL),
and Waldenström's macroglobulinemia (WM), there is currently no
approved Btkꢀtargeted therapy for chronic autoimmune indications. A number of Btk inhibitors
have shown efficacy in several preclinical models of inflammatory or autoimmune diseases that
5
,14,24,30ꢀ37
are driven by pathogenic B cells or myeloid cells,
but thus far none of these has
progressed into advanced clinical trials for the treatment of autoimmune diseases. This slow
progress in the autoimmune/inflammatory arena may, at least in part, be due to the stringent
safety requirements these indications demand.
There continues to be significant interest in the pursuit of Btk inhibitors and the leading
31
38
inhibitors to date are covalent: ibrutinib (1), spebrutinib (2) (CCꢀ292, AVLꢀ292), tirabrutinib
39
(
3) (GSꢀ4059, ONOꢀ4059) and acalabrutinib (4) (ACPꢀ196) (Figure 1). All of these compounds
react covalently and irreversibly with cysteineꢀ481 in the ATP binding site of Btk. This
dependence on irreversible inhibition poses a selectivity risk – since ten other human kinases
40
possess an equivalent cysteine residue in their active site – and raises the spectre of unwanted
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
offꢀtarget inhibition of these kinases. Indeed, the limited kinase selectivity data published so far
for these covalent inhibitors align with this expectation, as off targets include some or all of these
2
4,31
1
0 kinases that contain an analogous cysteine residue.
Furthermore, due to their ATPꢀlike
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
binding mode, these covalent molecules can also bind nonꢀcovalently to and inhibit even more
kinases, further increasing safety concerns due to selectivity. Finally, the reactive groups on
covalent compounds may lead to haptenization of serum proteins, potentially leading in rare
cases to serum sicknessꢀlike reactions in patients. Together, these offꢀtarget effects could
theoretically erode the safety profiles of these molecules. Because autoimmune indications such
as RA and SLE require exquisitely safe drugs, our strategy was to develop nonꢀcovalent
inhibitors that interact with Btk in a different manner than existing ATPꢀmimetic covalent
inhibitors.
Figure 1. Selected examples of clinical stage Btk inhibitors.
ACS Paragon Plus Environment
4

Page 5 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Results
We previously reported the discovery and structureꢀactivity relationships (SAR) of our first
2
0,21
clinicalꢀstage inhibitor, GDCꢀ0834, 5,
which suffered from rapid amide bond hydrolysis in
2
1
humans , but which we evolved into more advanced compounds including Gꢀ278, 6 (Figure
22
2
). With a Btk enzyme IC of 4 nM and a BCRꢀinduced CD69 IC of 35 nM in human whole
5
0
50
blood, 6 had a desirable profile. This, in addition to good kinase selectivity and an acceptable
preꢀclinical pharmacokinetic profile, allowed us to evaluate 6 in rat and dog tolerability studies
with the intent of advancing this compound into the clinic. Unfortunately, 6 induced doseꢀ
limiting toxicities with unacceptably low safety margins in both species. In SpragueꢀDawley
rats, an unusual pancreatic toxicity was also observed with this as well as a number of other Btk
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
1
inhibitors, and this was the subject of previously reported investigations (vide infra), whereas
in dogs liver toxicity was observed with 6 (as well as related analogues). The observed
hepatotoxicity in dogs was characterized by increases in ALT and AST liver enzymes, as well as
other findings in the liver such as perivascular mixed cell infiltrates, Kupffer cell hypertrophy,
and hepatocellular degeneration. In addition, increased circulating neutrophils were observed at
higher dose levels (Supporting Information). As a result, hepatotoxicity was the primary issue
that we sought to eliminate in next generation analogues.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Progression of Genentech’s published Btk inhibitors from GDCꢀ0834 to Gꢀ278.
In an attempt to better understand the hepatotoxicity that was primarily observed in dogs, a cell
viability (cytotoxicity) assay using cryopreserved human hepatocytes was used to identify highꢀ
4
2,43
risk compounds (Supporting Information).
Compounds with an IC < 50 ꢀM in this assay
5
0
were classified as having potentially high risk for cytotoxicity while compounds with IC >100
5
0
ꢀ
M were considered low risk. Once a sufficient data set was obtained, we set out to develop an
in silico model for the prediction of hepatocyte cytotoxicity to be applied prior to compound
synthesis and to flag any possible causative factors (Figure Sꢀ1, Supporting Information). A
ACS Paragon Plus Environment
6

Page 7 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
training data set consisting of 70 Btk inhibitors was used, with their cytotoxicity classified into
three toxicity risk bins: high (IC < 50 ꢀM), medium (IC = 50ꢀ100 ꢀM), or low (IC > 100
5
0
50
50
ꢀM). We explored a wide range of commonly used structural and physicochemical descriptors
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
and a variety of modeling algorithms. A decision tree model was chosen based on its
performance in 5ꢀfold crossꢀvalidation as well as its easy interpretability. It was subsequently
tested using the data for 121 molecules outside the training set, including some compounds that
originated from other internal discovery projects. The model performed reasonably well on this
more structurally diverse set of molecules (see details in Supporting Information), which
supported its application on this programme.
The model relied on only three calculated properties (pK of the most basic nitrogen, plasma
a
4
4
protein binding (PPB), and logP) and assigned a predicted hepatocyte IC range of a given
50
molecule into one of the three defined toxicity risk categories. While we had previously
hypothesized basicity (i.e., high pK ) to be a major risk factor for cytotoxicity, the decision tree
a
model identified logP and PPB as two additional predictive factors and defined specific
thresholds for each of them. Although often related—logP and PPB within this set of compounds
2
were linked enough to give a modest correlation (R =0.55)—removing either parameter
significantly reduced the performance of the model. Interestingly, inclusion of structure
descriptors did not improve the accuracy of predictions. This suggested that, within this series of
Btk inhibitors, physicochemical properties rather than particular substructures were the likely
drivers of hepatocyte cytotoxicity (for more details, see the Supporting Information). Given the
influence of basicity and logP, moving forward we focused our efforts on avoiding the
incorporation of strongly basic groups, and lowering logD (i.e., taking into account the combined
impact of pK and logP), while aiming to maintain or improve whole blood potency and
a
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
pharmacokinetic (PK) properties.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
SAR: Increasing Potency, Improving Metabolic Stability, and Minimizing Toxicity
One region of this Btk inhibitor scaffold that we sought to optimize was the hydroxymethylꢀ
substituted aryl ring that links the pyridone hingeꢀbinding motif to the selectivity pocket (or
2
2,45
‘H3’) motif. While hydroxymethyl and F substitution had been shown to improve potency,
changes to the ring itself had not.
We sought to lower logD by replacing the aryl ring with a heterocycle such as pyridine. The
three isomeric pyridyl analogues were prepared and tested (7ꢀ9, Table 1). In all cases,
incorporation of the nitrogen reduced the calculated and measured logD by at least one unit and
resulted in improved metabolic stability in human liver microsomes as compared to the benzene
analogue 6. Examples 7 and 8 maintained whole blood potency within twoꢀfold of the benzene
analogue 6, however, the isomer with the pyridyl nitrogen ortho to the H3 motif (7) retained the
most intrinsic potency with a Btk K of 1.8 nM and thus was regarded as the preferred isomer.
i
Table 1. Initial SAR around the core aryl ring.
d
cmpd
Core
Btk Ki
WB CD69 IC50
HLM CLhep
clogD / logD
ACS Paragon Plus Environment
8

Page 9 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
b
c
e
(nM)
(nM)
(mL/min/kg)
(pH 7.4)
6
7
36.5±10.2
n=7
0
.69±0.44
12.3
2.6 / 3.1
1.6 / 2.0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
54.9±19.4
n=3
1.80±1.20
8.1
8
9
64.4±6.1
n=2
6
.30±2.80
7.9
8.0
1.3 / 1.9
1.2 / 1.8
387.0±101.8
n=2
6.90±2.20
a
Inhibitor K values were determined in at least triplicate in an assay that monitors Btkꢀcatalyzed phosphorylation of
a synthetic peptide substrate. See Supporting Information for assay details. CD69 expression on CD19 B cells in
human whole blood was measured after stimulation with antiꢀIgM (Supporting Information). Human liver
microsomeꢀpredicted hepatic clearance. Calculated logD (logD was calculated using v2 of the MoKa program,
available from Molecular Discovery Ltd. (www.moldiscovery.com)). LogD measured in house using a high
throughput microscale shake flask with liquid chromatography tandem mass spectrometry quantitation.
i
b
+
c
d
e
4
6
Over time, many additional pyridyl compounds were prepared and the same trends toward
improved metabolic stability and LLE were apparent as illustrated in matched pair plot analyses
4
7,48
(
Figure 3).
Figure 3. Matched pair plots for Btk inhibitors containing an ArꢀF (l) or pyridyl (r) linker. Lines
4
7,48
join matched pairs. (a) HLM stability (mL/min/kg);
(b)
LLE = pK – clogD
i
(a) HLM (mL/min/kg)
(b) LLE = pK ꢀ clogD
i
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
HO
HO
N
N
As previously noted, much of the potency and selectivity of our chemical scaffold results from
the ability of these compounds to stabilize and interact with binding site in Btk that is generated
5
,22,50
via a conformational rearrangement of the activation loop of the kinase domain.
It is
instructive to note that, given our inducedꢀfit binding mode, we used our whole blood assay as
the main driver of SAR in this program since it could encompass the effects of a combination of
therapeutically relevant parameters, including Btk affinity, cell permeability, PPB, and favorable
targetꢀbinding kinetics. In this respect, potency in whole human blood was determined as
1
4
previously described and monitored via either Btk Y223 phosphorylation or antiꢀIgMꢀinduced
CD69 expression on B cells.
The generally lipophilic groups at the tricyclic H3 motif termini present hydrophobic surfaces
that interact with Y551 of the activation loop of Btk (among other residues), ultimately
occupying a lipophilic specificity/selectivity pocket. We decided to synthesize a set of inhibitors
ACS Paragon Plus Environment
10

Page 11 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
with a variety of H3 groups in combination with the pyridyl linker to see if we could take
advantage of improved LLE, allowing us to incorporate groups that might afford improved
whole blood potency. In the case of thiopheneꢀcontaining examples 10 and 11, the resulting
increase in lipophilicity relative to pyrrole analogue 7 led to increased logD and a drop in
metabolic stability in human liver microsomes (Table 2). The dimethyl substituted cyclopentyl
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
11) offered improved whole blood potency relative to the corresponding cyclohexyl analogue
10), so the next compound prepared was the corresponding pyrroleꢀcontaining derivative, 12.
(
This molecule displayed improved whole blood potency (IC = 15.4 nM), and maintained the
50
metabolic stability and physicochemical property profile of 6 (kinetic solubility = 79.8 ꢀM for 12
vs. 12 ꢀM for 6). Small changes such as unsaturation of the lactam ring or fluoro substitution of
4
9
the pyrrole ring—albeit not in tandem, as was subsequently reported to be beneficial —did not
afford any appreciable whole blood potency benefits in our series. Lastly, although the tertꢀ
butylꢀsubstituted bicyclic thiophene lactam 15 was a potent inhibitor, it had an unfavorable high
measured logD (>3) like the other thiopheneꢀcontaining examples. From this exploration,
compound 12 offered the best balance of properties and was not cytotoxic in primary human
hepatocytes (IC50 >100 ꢀM). However, when dosed in a cassette rat PK study, 12 exhibited high
plasma clearance (CL = 165 mL/min/kg) and a relatively short halfꢀlife (0.4 hours), so SAR
p
efforts continued in order to improve the metabolic stability of this scaffold.
a
Table 2. Selectivity pocket / H3 SAR.
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
HLM
CLhep
(mL/min/kg)
WB
CD69 IC50
nM)
a
clogD /
logD (pH 7.4)
cmpd
7
H3 group
Btk K (nM)
i
(
54.9±19.4
n=3
1
.80±1.20
.97±0.32
8.1
1.6 / 2.0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
0
121.1±24.5
n=2
5
13.0
2.1 / 3.4
1
1
1
2
39.0±20.4
n=3
0.58±0.01
1.80±0.56
4.20±0.51
13.5
7.2
2.3 / 3.3
1.8 / 1.9
0.8 / 2.5
1.7 / 2.5
2.1 / 3.6
15.4±8.2
n=5
13
77.8±52.9
n=3
9.2
1
1
4
5
42.4
n=1
2.50±1.20
8.1
61.9±11.3
n=3
1.00±0.22
10.0
a
See footnotes from Table 1 for experimental details.
We previously reported that a pyridopiperazine moiety, such as 12 (Table 3), that spans a flat,
lipophilic channel leading to a partially solvent exposed region of the protein (“H2 pocket”),
optionally as the Nꢀoxetanyl derivative, displayed improved kinetic solubility and
2
2
pharmacokinetic properties. With a preference for compounds with a lower pK , we next set
a
ACS Paragon Plus Environment
12

Page 13 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
out to examine the SAR around the H2 region of our inhibitors. Moving the oxetanylpiperazine
ortho to the pyridyl nitrogen (example 16) was tolerated with respect to Btk inhibition (Ki
maintained relative to 12), but resulted in a significant drop in whole blood potency and
metabolic stability (Table 3). Structural modeling suggested that substitution at the other ring
positions would not be tolerated, so our subsequent efforts focused on 2,5ꢀdisubstituted pyridyl
inhibitors. Replacement of the oxetane with an acyl group, either as the piperazine (17) or the
bisꢀazetidine isostere thereof (18), was tolerated, but offered no advantage. Similarly, while the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
introduction of other substituents that allowed for modulation of the piperazine nitrogen pK (19ꢀ
a
22) did not result in a significant loss in potency (with the exception of 21, with a whole blood
potency of 116 nM), each of these suffered from decreased metabolic stability in human liver
microsomes (CLhep of 12ꢀ16 mL/min/kg) relative to 12. A closeꢀin analogue of 12 was prepared
in the form of oxetanylazetidine 23. While 23 maintained much of the Btk affinity and whole
blood potency relative to 12, it suffered from poor metabolic stability and a higher measured
logD value and, therefore, was not advanced.
a
Table 3. Piperazine replacements.
ACS Paragon Plus Environment
13

HLM
CLhep
(mL/mi
n/kg)
Journal of Medicinal Chemistry
clogD /
logD
Page 14 of 61
BTK K
WB CD69
IC (nM)
MDCK
i
cmpd
R1
R2
b
(nM)
AB:ER
5
0
(pH 7.4)
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
2
6
15.4±8.2
n=5
15.7 /
1.8
H
1.80±0.56
1.80±0.08
7.2
1.8 / 1.9
240.9
n=1
H
17
2.1 / 3.5
ꢀ / ꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
17
66.5
n=1
11.1 /
1.2
H
H
1.80±0.44
3.30±0.55
8.8
1.7 / 2.9
1.5 / 2.5
1
1
8
9
33.4
n=1
10.0
ꢀ / ꢀ
36.6±22.2
n=2
12.5 /
1.0
H
H
1.20±0.28
1.00±0.27
12.0
12.0
1.9 / 3.2
2
0
32.2
n=1
9 / 1.3
2.3 / 3.3
3.4 / 4.4
2
2
1
2
116.1
n=1
14.2 /
0.6
H
H
4.00±1.60
0.45±0.08
0.43±0.15
16.0
14.0
16.0
31.6
n=1
ꢀ / ꢀ
2.3 / 3.0
23
23.5
n=1
48.5 /
0.4
H
1.6 / 3.0
a
b
See footnotes from Table 1. Permeability in MadinꢀDarby Canine Kidney (MDCKI) cells, ApicalꢀtoꢀBasolateral
AB); Efflux Ratio (ER).
(
ACS Paragon Plus Environment
14

Page 15 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
One aspect of the piperazine SAR that remained relatively unexplored was substitution of the
piperazine ring itself. Modeling suggested that we might be able to increase van der Waals
contacts with the edge of the protein and/or perhaps influence the conformation of the ring itself
using small alkyl substituents, e.g., methyl (Figure 4A). We subsequently synthesized and tested
a set of such compounds, with a methyl group in the four possible positions on the ring, using
available intermediates containing the Fꢀaryl linker motif (Table 4).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Solvent exposed surface of Btk around the oxetanylpiperazine motifs: (A) Model of
2
2
oxetanylpiperazine. (B) From the Xꢀray coꢀcrystal structure of 29 bound to Btk (see Figure 5).
(A)
(B)
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In each case whole blood potency improved without any negative impact on other properties.
Most significantly, in the case of (S)ꢀ2,4ꢀdimethylꢀ1ꢀ(pyridinꢀ3ꢀyl)piperazine 26 (as in an Xꢀray
coꢀcrystal structure of 29, Figure 4B), we observed a 5ꢀfold improvement in CD69 whole blood
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
potency relative to 24. Since this particular example had a higher logD and pK than desired (per
a
our in silico model for prediction of human hepatotoxicity), we applied this change to compound
1
2.
a
Table 4. SAR for piperazine ring substitution.
BTK Ki
WB CD69 IC50
(nM)
HLM CLhep
(mL/min/kg)
clogD / logD
(pH 7.4)
cmpd Piperazine
a
(nM)
2
4
27.6±19.9
n=7
0.35±0.11
15.0
13.0
14.0
2.5 / 3.1
2.9 / 3.1
2.9 / 3.2
25
16.9
n=1
0
0
.37±0.13
.65±0.28
2
6
4.6±3.5
n=4
ACS Paragon Plus Environment
16

Page 17 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
7
8
27.5±0.6
n=2
0.32±0.01
13.0
13.0
2.9 / 2.3
2.9 / 3.3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
12.7±4.5
n=4
0
.43±0.23
a
See footnotes from Table 1.
We chose to transfer the (S)ꢀmethyl group of 26 onto 12, which afforded a compound with an
overall excellent profile. This resulting compound, GDCꢀ0853, 29 (Table 5), had a ~2ꢀfold
improvement in whole blood B cell CD69 potency (IC = 8.4 nM) relative to 12 (IC =15.4 nM)
50
50
and greater than 4ꢀfold relative to 6 (IC = 36.5 nM). The corresponding enantiomer of 29 (30)
5
0
suffered from poor metabolic stability, and had whole blood potency of only 40 nM. Increasing
the size of the substituent to an ethyl group (31) resulted in an expected logD increase and
eroded stability in human liver microsomes and hepatocytes. Adding an additional methyl to the
ring as in 32 gave a similar result, albeit with excellent whole blood potency. Finally,
incorporation of nitrogen atoms as in hinge motif pyridazinone 33 and H2 pyrazine 34 yielded no
further benefit. Compound 29 exhibited the best overall profile of any compound prepared, and
as such was further characterized. When tested for cytotoxicity in primary human hepatocytes,
2
9 had a clean profile with an IC > 300 ꢀM, in contrast to 6, which had an IC of 39 ꢀM.
5
0
50
a
Table 5. Combination of core aryl ring, H3, piperazine, and late stage SAR.
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
WB
CD69
IC50
HLM / HH
CLhep
clogD /
logD
(mL/min/kg) (pH 7.4)
Btk K
MDCK
AB:ER
i
Cmpd
R1
R2
X
Y
a
(nM)
(nM)
29,
GDC-
8.4±5.6
n=5
9.4 / 7.6
2.2 / 1.6
2.2 / 2.8
OH
CH
CH 0.91±0.34
CH 1.80±0.30
CH 0.96±0.21
15.1 / 0.9
7.8 / 0.8
0
853
3
0
40.0±15.5
n=4
OH
CH
15.0 / 14.0
3
3
3
1
2
3
15.9
n=1
OH
OH
OH
OH
CH
CH
N
12.0 / 16.0
12.0 / 12.0
11.0 / 13.0
12.0 / 10.0
2.7/ 3.6
2.7 / 3.5
2.3 / 3.7
9.5 / 0.9
7.2 / 0.9
17.3 /0.9
10.2±1.5
n=1
CH
CH
N
1.30±0.65
16.1±0.7
n=1
ꢀ
0.8
51.5
n=1
15.6
/ 1.2
3
4
CH
1/4 / ꢀ
n=1
a
See footnotes from Tables 1 & 3.
We obtained a 1.59ꢀÅ resolution Xꢀray coꢀcrystal structure of GDCꢀ0853 (29) bound to Btk
(Figure 5). Within the Btk kinase domain, 29 shows the expected hinge binding by the central
ACS Paragon Plus Environment
18

Page 19 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
pyridone, with hydrogen bonding between the pyridone carbonyl and pendant aniline NH
hydrogen and the protein backbone atoms of Met477. Consistent with other previously solved
crystal structures with earlier molecules, Tyr551 in the activation loop has moved 18Å from its
position in apoꢀBtk to pack against the hydrophobic right hand side of the tricyclic ring system,
forming part of the H3 site that confers substantial potency and kinase selectivity. Additionally,
the gemꢀdimethyl cyclopentane presents nonplanar structure that complements the shape of the
pocket in this area, in a similar vein to tertꢀbutyl analogues. The hydoxymethyl projects into an
interior waterꢀfilled cavity, forming hydrogen bonds with Lys430 and Asp539. The pyridyl ring
occupies part of the active site cleft and packs against Gly480 and Leu408 while its nitrogen
atom accepts a nonꢀtraditional Hꢀbond from an aromatic hydrogen on the adjacent pyridone ring,
contributing to the molecule’s rigidity. The 2ꢀMe substituent on the piperidine ring imparts a
restricted rotation relative to the unsubstituted analogue, favoring a more outꢀofꢀplane orientation
of the piperidine ring. Energetics from a quantum torsion scan of the piperazineꢀtoꢀpyridyl bond
support this observation (Supporting Information). The unsubstituted piperazine is calculated to
have relatively free rotation about this bond, while the 2ꢀMe substituent on the piperazine ring
imparts a more restricted rotation. The conformation observed in the structure of 29 lies
essentially at the calculated global minimum. Thus, in addition to being conformationally rigid,
it presents a more nonꢀplanar motif at each end of its structure, which aids in conferring
favorable physiochemical, cellular potency, and ultimately pharmacokinetic properties. Detailed
physicochemical properties of 29 are included in the Supplementary Information. The most
stable known crystalline form has a very high melting point of 274 °C, which contributes to the
relatively low water solubility of 0.003 mg/mL. When protonated in 0.1 N HCl, aqueous
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
solubility jumps to 35.9 mg/mL, affording excellent solubility and dissolution in the, typically
acidic, mammalian stomach.
Figure 5. Xꢀray coꢀcrystal structure of the 29/Btk kinase domain complex (PDB 5VFI) with key
residues labeled. The molecular surface of 29 is in blue, and proteinꢀligand Hꢀbonds are shown
as dashed lines.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
20

Page 21 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Pre-clinical PK Profile
We assessed the in vivo pharmacokinetic properties of 29 (Table 6) in cassette studies in both
rats (0.2 mg/kg IV and 1.0 mg/kg PO) and dogs (0.2 mg/kg IV and 0.5 mg/kg PO). In rats, 29
had moderate clearance of 27.4 mL/min/kg and excellent bioavailability (F=65%) relative to the
desꢀmethyl analogue 12. The plasma clearance was reduced 6ꢀfold, from 165 to 27.4 mL/min/kg,
while the volume of distribution (V ) and oral bioavailability were relatively unchanged, which
d
resulted in a 5.5ꢀfold improvement in plasma halfꢀlife from 0.4 to 2.2 hours. Because efficacy in
the collagenꢀinduced arthritis model in rat likely requires chronic inhibition of Btk throughout
3
0
the study, this improvement enabled adequate exposures following oral dosing such that preꢀ
clinical efficacy could be assessed. Compound 29 also demonstrated favorable PK properties in
dogs. The 3.8ꢀhour halfꢀlife (Cl 10.9 mL/min/kg, V 2.96 L/kg) and high oral bioavailability
p
d
(85%) also enabled attainment of sufficient exposures in dog toxicology studies. Overall, these
enhancements in the in vivo pharmacokinetics provided adequate exposures to enable rat and dog
preꢀclinical toxicology studies.
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 6: Rat pharmacokinetic parameters estimated from singleꢀdose intravenous (0.2 mg/kg)
and oral (1 mg/kg) cassette pharmacokinetic study.
29
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
(GDCꢀ0853)
CL (mL/min/kg)
165
4.97
0.4
27.4
V (L/kg)
d
5.42
t1/2 (h)
F (%)
2.2
65.8
65.1
a
Table 7: In Vitro Safety Profile of 29 (GDCꢀ0853).
Assay
Result
Receptor Binding Profile (41 targets)
hERG Channel Inhibition (Functional)
No binding > 50% at 10 ꢀM
IC > 30 ꢀM (>1,500ꢀfold window)
5
0
hNa 1.5 Channel Inhibition (Functional)
0.6% at 1 ꢀM, 1.9% at 10 ꢀM
v
hCa 1.2 Channel Inhibition (Functional)
1.5% at 1 ꢀM, 3.8% at 10 ꢀM
Negative
v
Ames
MNT (HPBLs)
Negative
Cytotoxicity in Human Hepatocytes
IC > 300 ꢀM
5
0
CYP 3A4, 1A2, 2C19, 2D6, 2D9 Inhibition
IC s > 5 ꢀM
5
0
a
See Supporting Information for assay methods and details.
ACS Paragon Plus Environment
22

Page 23 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
9 was tested for offꢀtarget activity against a panel of in vitro safety assays and was found to
have a clean offꢀtarget profile (Table 7), was negative in the in vitro Ames and MNT assays,
indicating a low risk for genotoxicity, and has low potential cardiovascular risk liability as
indicated in the in vitro hERG assay (IC > 30 ꢀM).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
Like other structurally related molecules that bind in the H3 region of Btk, 29 is an extremely
selective Btk inhibitor. When compared with six other clinical stage Btk inhibitors, each tested at
1
µM against a broad panel of human kinase biochemical assays, 29 was found to be the most
Btkꢀselective molecule and inhibited only 3 of 286 offꢀtarget kinases (Figure 6a). Based on the
determined IC values, the selectivity for Btk was >100ꢀfold against each of these 3 offꢀtargets:
5
0
Bmx (153ꢀfold), Fgr (168ꢀfold), and Src (131ꢀfold) (Figure 6b and Supporting Information). In
2
3,51
contrast, BMSꢀ986142
showed >50% inhibition of 20 of 219 offꢀtarget kinases tested and
ibrutinib (1) inhibited 31 of 221 offꢀtarget kinases by >50% (Figure 6b and Supporting
Information). The IC values (Figure 6b and Supporting Information) demonstrate that ibrutinib
5
0
is rather nonꢀselective, being <100ꢀfold selective against 18 other kinases, while BMSꢀ986142 is
<100ꢀfold selective for 6 kinases. Although spebrutinib (2), tirabrutinib (3), and acalabrutinib (4)
inhibit fewer total offꢀtargets than ibrutinib or BMSꢀ986142 at the same 1ꢀꢀM test concentration,
for the additional targets they do inhibit, the IC50 values are quite close in magnitude to the
activity they display against Btk [spebrutinib: Bmx (1ꢀfold selectivity ratio), Tec (1ꢀfold), Txk
(5ꢀfold), JAK3 (6ꢀfold), and STK16 (7ꢀfold); tirabrutinib: Bmx (2ꢀfold) and Tec (12ꢀfold);
acalabrutinib: ErbB4 (4ꢀfold), Bmx (22ꢀfold), and RIPK2 (47ꢀfold)] (Figure 6b and Supporting
5
2,53
Information). Evobrutinib
inhibited 3 of 221 offꢀtargets (Bmx, Tec, and ErbB4), but it was
also the weakest Btk inhibitor tested (IC = 37.9 nM) (Supporting Information). As noted above,
5
0
ten other human kinases have a Cys residue in the analogous location as Cysꢀ481 of Btk, and as
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
0
such may be particularly sensitive to inhibition by covalent inhibitors. When these clinical
stage Btk inhibitors were tested against ten of the Cysꢀcontaining kinases, 29 was clearly more
selective for Btk than any of the covalent Btk inhibitors, or even the nonꢀcovalent inhibitor,
BMSꢀ986142 (Figure 6c). Thus, based on these in vitro kinase profiles, 29 is overall a more
selective Btk inhibitor than any of these other molecules. The complete kinase selectivity profiles
of 29 and the six other clinical stage Btk inhibitors (tested at the 1ꢀꢀM test concentration and in
IC titrations) are provided in the Supporting Information.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
Figure 6. (a) Kinase selectivity of Btk inhibitors ibrutinib 1, spebrutinib 2, tirabrutinib 3,
acalabrutinib 4, evobrutinib, BMSꢀ986142, and GDCꢀ0853 29. In vitro kinase selectivity for Btk
inhibitors was determined in the SelectScreen panel (ThermoFisher, Madison, WI) comprised of
activity assays and ATP site competitive binding assays. Percent (%) inhibition was assessed at a
1ꢀµM test concentration: red, ≥80% inhibition; yellow, 50ꢀ80% inhibition; green, ≤50%
inhibition or not tested. Compound 29 was tested in 287 total kinase assays, whereas the other 6
inhibitors were tested against a subset of approximately 220 of these kinases, and only those
kinases that were tested against all 7 inhibitors are represented (see Supporting Information for
all of the % Inhibition data against all kinases). (b) Potency of Btk inhibitors against offꢀtarget
kinases. IC50 values were determined as described in the Methods against kinases that were
inhibited by >50% at 1 µM. The IC values for 29 and six other clinical stage Btk inhibitors are
50
provided in the Supporting Information. Due to its low potency toward inhibition of Btk (IC =
5
0
3
7.9 nM), no offꢀtarget IC values were collected for evobrutinib. (c) Inhibition of kinases that
50
have a cysteine that is analogous to Btk Cysꢀ481. Inhibitors were tested at 1 µM as described
above for Figure 6a and the % Inhibition is shown.
ACS Paragon Plus Environment
24

Page 25 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(a)
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(b)
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Ibrutinib
Spebrutinib
BMSꢀ986142
Btk
Src
Btk
Src
Btk
Src
Bmx
Fgr
Bmx
Fgr
Bmx
Fgr
Blk
Brk
Blk
Brk
Blk
Brk
CSK
EGFR
ErbB2
ErbB4
EphA1
EphA7
EphB1
Flt3
CSK
EGFR
ErbB2
ErbB4
EphA1
EphA7
EphB1
Flt3
CSK
EGFR
ErbB2
ErbB4
EphA1
EphA7
EphB1
Flt3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Frk
Frk
Frk
Hck
Hck
Hck
IGF1R
ITK
IGF1R
ITK
IGF1R
ITK
JAK3
Lck
JAK3
Lck
JAK3
Lck
Lyn
Lyn
Lyn
MuSK
Ret
MuSK
Ret
MuSK
Ret
RIPK2
Ros
RIPK2
Ros
RIPK2
Ros
Srm
Srm
Srm
STK16
TEC
TNK2
TXK
TrkA
TrkB
Yes
STK16
TEC
TNK2
TXK
TrkA
TrkB
Yes
STK16
TEC
TNK2
TXK
TrkA
TrkB
Yes
1.0
10.0 100.0 1000.0
1.0
10.0 100.0 1000.0
1
.0
10.0 100.0 1000.0
IC50 (nM)
IC50 (nM)
IC50 (nM)
Tirabrutinib
Acalabrutinib
GDCꢀ0853
Btk
Src
Btk
Src
Btk
Src
Bmx
Fgr
Bmx
Fgr
Bmx
Fgr
Blk
Blk
Blk
Brk
Brk
Brk
CSK
EGFR
ErbB2
ErbB4
EphA1
EphA7
EphB1
Flt3
CSK
EGFR
ErbB2
ErbB4
EphA1
EphA7
EphB1
Flt3
CSK
EGFR
ErbB2
ErbB4
EphA1
EphA7
EphB1
Flt3
Frk
Frk
Frk
Hck
Hck
Hck
IGF1R
ITK
IGF1R
ITK
IGF1R
ITK
JAK3
Lck
JAK3
Lck
JAK3
Lck
Lyn
Lyn
Lyn
MuSK
Ret
MuSK
Ret
MuSK
Ret
RIPK2
Ros
RIPK2
Ros
RIPK2
Ros
Srm
Srm
Srm
STK16
TEC
TNK2
TXK
TrkA
TrkB
Yes
STK16
TEC
TNK2
TXK
TrkA
TrkB
Yes
STK16
TEC
TNK2
TXK
TrkA
TrkB
Yes
1.0
10.0 100.0 1000.0
1.0
10.0 100.0 1000.0
1.0
10.0 100.0 1000.0
IC50 (nM)
IC50 (nM)
IC50 (nM)
(c)
ACS Paragon Plus Environment
26

Page 27 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
% Inhibition of Cys-Containing Kinase by 1 ꢀM Inhibitor
Inhibitor
Btk
Bmx
Tec
Txk
ErbB4
Blk
Jak3
ErbB2
EGFR
Itk
1
, ibrutinib
101
100
98
99
99
97
77
89
80
56
100
92
86
71
80
98
23
93
95
66
31
30
81
4
93
76
28
93
57
38
1
102
44
47
16
42
82
6
80
89
-7
-3
0
91
6
89
34
-2
-1
1
81
19
-5
2
2, spebrutinib
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
, tirabrutinib
7
4, acalabrutinib
evobrutinib
100
90
36
1
-2
47
-1
BMS-986142
91
-2
5
1
7
29, GDC-0853
99
-1
-7
GDC-0853 inhibits human BCR and FcγR signaling
Since Btk plays a critical role in signal transduction of BCR and FcγR, we evaluated the cellular
1
2,13
potency of 29 in isolated primary human B cells and monocytes.
Specifically, we evaluated
inhibition of antiꢀIgM induced Btk phosphorylation, as well as antiꢀIgM or CD40L stimulated
14
proliferation in B cells, and FcγRꢀdependent TNFα production in human monocytes.
In isolated primary human B cells, 29 potently inhibited antiꢀIgMꢀinduced Btk Y223 tyrosine
phosphorylation (IC = 3.1 nM) and antiꢀIgM and CD40Lꢀinduced Bꢀcell proliferation (IC =
50
50
1.2 nM and 1.4 nM, respectively) (Table 8). Consistent with the previously reported role of Btk
5
in Fc receptor function, Btk inhibition by 29 prevented FcγRIIIꢀtriggered TNFα production in
human monocytes (IC50 = 1.3 nM). As previously mentioned, in human blood 29 potently
inhibited upꢀregulation of the early activation marker CD69 on B cells after overnight antiꢀIgM
stimulation (IC = 8 nM). Furthermore, GDCꢀ0853 also suppressed antiꢀIgM induced Btk Y223
5
0
autophosphorylation in human whole blood (IC50 = 11 nM). Taken together, these data
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
demonstrate that 29 is a highly selective Btk inhibitor that blocks both Bꢀcell BCR and monocyte
FcγR signaling.
Table 8. Potency of 29 in human whole blood and isolated primary cell assays.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Blood
B cells
Monocytes
Stimulus
Readout
Donors
AntiꢀIgM
AntiꢀIgM
rhCD40L
AntiꢀFCγRIII
TNFα
pBtk
CD69
(n=6)
8.0
pBtk
(n=4)
3.1
Proliferation Proliferation
(n=6)
11
(n=3)
1.2
(n=3)
1.4
(n=3)
IC ±SD
1.3
50
(nM)
± 4.2
± 6.0
± 0.7
± 0.9
± 0.6
± 0.8
In the course of our SAR efforts, we wondered if our H3ꢀpocket binding compounds might show
timeꢀdependent inhibition of Btk. We found that some inhibitors did show an increase in potency
after preꢀincubation with the enzyme. As just one example, 29 showed an 8ꢀfold increase in
potency in the enzyme assay after a 2ꢀhour preꢀincubation with Btk. We were interested to
determine whether this timeꢀdependent inhibition would be accompanied by slow dissociation of
29 from Btk and an extended target residence time, as this feature could lead to an extended
54,55
duration of pharmacological action in vivo.
In order to quantify the kinetics of dissociation of GDCꢀ0853 from Btk and to determine the Btk
residence time for 29, we adapted our standard in vitro biochemical Btk enzyme assay
50
conditions to a 'jump dilution' experiment (Supporting Information). Btk that had been
incubated with DMSO vehicle before being extensively diluted into the assay mixture rapidly
ACS Paragon Plus Environment
28

Page 29 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
catalyzed the phosphorylation of the peptide substrate to form phosphoꢀpeptide product in a
nearly linear fashion over time (Figure 7, blue circles). On the other hand, Btk that had been preꢀ
incubated with 29 and then diluted into the assay mixture showed a lag in product formation that
extended for several hours (Figure 7, red circles). This delayed recovery of activity suggests that
29 dissociates extremely slowly from Btk. The product formation time course data were fitted to
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
6
an equation that describes the firstꢀorder dissociation of E•I into free enzyme and inhibitor with
54
an observed rate constant, k (Supporting Information). The residence time is defined as 1/k .
off
off
In three experiments 29 displayed an average residence time with Btk of 18.3 ± 2.8 hours (k =
off
ꢀ1
1.52e–5 s ).
Figure 7. An in vitro biochemical jump dilution experiment illustrates the long Btk residence
time of 29.
2
1
0
5
DMSO
10
5
GDCꢀ0853
0
0
1
2
3
4
5
6
Time, h
ACS Paragon Plus Environment
29