SAR optimization studies on modified salicylamides as a potential treatment for acute myeloid leukemia through inhibition of the CREB pathway

Article abstract of DOI:10.1016/j.bmcl.2019.06.023

Disruption of cyclic adenosine monophosphate response element binding protein (CREB) provides a potential new strategy to address acute leukemia, a disease associated with poor prognosis, and for which conventional treatment options often carry a significant risk of morbidity and mortality. We describe the structure-activity relationships (SAR) for a series of XX-650-23 derived from naphthol AS-E phosphate that disrupts binding and activation of CREB by the CREB-binding protein (CBP). Through the development of this series, we identified several salicylamides that are potent inhibitors of acute leukemia cell viability through inhibition of CREB-CBP interaction. Among them, a biphenyl salicylamide, compound 71, was identified as a potent inhibitor of CREB-CBP interaction with improved physicochemical properties relative to previously described derivatives of naphthol AS-E phosphate.

Full text of DOI:10.1016/j.bmcl.2019.06.023

Accepted Manuscript
SAR optimization studies on modified salicylamides as a potential treatment for
acute myeloid leukemia through inhibition of the CREB pathway
Hee-Don Chae, Nick Cox, Samanta Capolicchio, Jae Wook Lee, Naoki
Horikoshi, Sharon Kam, Andrew Ng, Jeffrey Edwards, Tae-León Butler, Justin
Chan, Yvonne Lee, Garrett Potter, Mark C. Capece, Corey W. Liu, Soichi
Wakatsuki, Mark Smith, Kathleen M. Sakamoto
PII:
S0960-894X(19)30403-2
https://doi.org/10.1016/j.bmcl.2019.06.023
BMCL 26501
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
23 April 2019
12 June 2019
17 June 2019
Please cite this article as: Chae, H-D., Cox, N., Capolicchio, S., Wook Lee, J., Horikoshi, N., Kam, S., Ng, A.,
Edwards, J., Butler, T-L., Chan, J., Lee, Y., Potter, G., Capece, M.C., Liu, C.W., Wakatsuki, S., Smith, M., Sakamoto,
K.M., SAR optimization studies on modified salicylamides as a potential treatment for acute myeloid leukemia
through inhibition of the CREB pathway, Bioorganic & Medicinal Chemistry Letters (2019), doi: https://doi.org/
10.1016/j.bmcl.2019.06.023
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

SAR optimization studies on modified salicylamides as a potential treatment for acute myeloid
leukemia through inhibition of the CREB pathway
Hee-Don Chaea, Nick Coxb, Samanta Capolicchiob, Jae Wook Leea, Naoki Horikoshic,d, Sharon Kama,
Andrew Ngb, Jeffrey Edwardsa, Tae-León Butlera, Justin Chana, Yvonne Leea, Garrett Potterb, Mark C.
Capecee, Corey W. Liuf, Soichi Wakatsukic,g, Mark Smithb*, Kathleen M. Sakamotoa*
aDepartment of Pediatrics, Stanford University School of Medicine, Stanford, CA, USA
bMedicinal Chemistry Knowledge Center, Stanford ChEM-H, Stanford, CA, USA
cDepartment of Structural Biology, Stanford University School of Medicine, Stanford, CA, USA
dLife Science Center for Survival Dynamics, Tsukuba Advanced Research Alliance (TARA), University
of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8575, Japan
eDepartment of Chemistry, Stanford University, Stanford, CA, USA
fMacromolecular Structure Knowledge Center, Stanford ChEM-H, Stanford, CA, USA
gBioSciences Division, SLAC National Accelerator Laboratory, Menlo Park, CA, USA
*Correspondence to:
Kathleen M. Sakamoto, M.D., Ph.D.
Division of Hematology/Oncology
Department of Pediatrics
Stanford University
269 Campus Drive, CCSR 1215C
Stanford, California 94305-5162
kmsakamo@stanford.edu
Telephone: 650-725-7126
Fax: 650-723-6700
Mark Smith, Ph.D.
Medicinal Chemistry Knowledge Center
Stanford ChEM-H
Stanford University
CCSR 3110
269 Campus Drive,
Stanford, CA 94304
mxsmith@stanford.edu
Telephone: 415-706-4498
Keywords: Salicylamide, Acute Myeloid Leukemia, small molecule, CREB, CBP

Abstract
Disruption of cyclic adenosine monophosphate response element binding protein (CREB) provides a
potential new strategy to address acute leukemia, a disease associated with poor prognosis, and for which
conventional treatment options often carry a significant risk of morbidity and mortality. We describe the
structure-activity relationships (SAR) for a series of XX-650-23 derived from naphthol AS-E phosphate
that disrupts binding and activation of CREB by the CREB-binding protein (CBP). Through the
development of this series, we identified several salicylamides that are potent inhibitors of acute leukemia
cell viability through inhibition of CREB-CBP interaction. Among them, a biphenyl salicylamide,
compound 71, was identified as a potent inhibitor of CREB-CBP interaction with improved
physicochemical properties relative to previously described derivatives of naphthol AS-E phosphate.

Acute leukemia is a rapidly progressing hematological malignancy that begins in either lymphoid
(acute lymphoblastic leukemia (ALL)) or myeloid (acute myeloid leukemia (AML)) cells. Acute
leukemias are characterized by the rapid clonal accumulation of immature lymphoid or myeloid
progenitors from the accumulation of several oncogenic mutations with resultant multilineage cytopenia.1,
2 AML is the most common cause of leukemia death, while ALL is the most common type of cancer in
childhood.3, 4 Even with intensive chemotherapy and hematopoietic stem cell transplantation, the overall
5-year survival for AML is less than 50% in younger patients and less than 10% in patients over 65-years
old, and for ALL is less than 80% in younger patients and less than 20% in patients over 65-years old.5 In
addition, treatment of acute leukemia is associated with a variety of long-term complications and a high
risk of mortality.6-8 Therefore, the discovery of novel therapies that are more effective and less toxic
would be of great clinical benefit.
Transcription factors control the expression of essential genes to maintain normal hematopoiesis.
Mutations or dysregulated expression of these transcription factors play important roles in
leukemogenesis.9 Transcription factors have been considered as difficult drug targets, however, recent
studies have shown that transcription factors can be targeted for cancer therapy by inhibiting their
association with interacting cofactors.10, 11 We have studied the cAMP response element binding protein
(CREB), a transcription factor that promotes AML cell proliferation and survival, as a potential
therapeutic target for acute leukemia. Overexpression of CREB protein in AML cells is associated with a
significantly worse prognosis.12, 13 CREB overexpression in AML cells augments their growth rate and
confers resistance to apoptosis.12 Conversely, CREB knockdown in AML cells decreases cell proliferation
and induces apoptosis without any effects on long-term engraftment of hematopoietic stem cells
(HSCs).14, 15 Upon activation by phosphorylation at Ser133 of the Kinase Inducible Domain (KID), CREB
recruits co-activator CREB binding protein (CBP) through the interaction of phosphorylated KID with the
KID Interacting (KIX) domain. The CREB/CBP complex turns on the expression of CREB-driven genes
that regulate cell proliferation and survival.16-18

Using an NMR-based screening approach to identify small molecules that bind to KIX,
Montminy et al. reported that KG-501 1 disrupted CREB-CBP interaction with an IC50 of around 90 μM,
but attenuated the transcriptional activity of CREB when added to cells with a much higher potency
(Figure 1).19 Subsequent work by Xiao and coworkers showed that KG-501 1 is neither stable in tissue
culture media nor cell permeable and that the dephosphorylated product, naphthol AS-E 2, is significantly
more potent in cells, presumably due to improved permeability (Figure 1).20 Subsequent structure-activity
relationship (SAR) studies of 2 by Xiao showed that a small, electron-withdrawing group at the para
position in the phenyl ring was favored for inhibiting KIX-KID interaction and cancer cell viability
through the downregulation of CREB-dependent gene expression.21 To demonstrate the feasibility of
targeting CREB as a treatment for acute leukemia, we recently described a small molecule inhibitor of
CREB, XX-650-23 (N-(4-cyanophenyl)-3-hydroxy-2-naphthamide) 3, a compound originally based on 2
in collaboration with Xiao’s group.22 Compound 3 changes the expression of CREB-target genes, leading
to cell-cycle arrest and apoptosis of AML cells and increased survival of AML xenograft mice with no
toxicity to normal cells or animals. Despite significant interest in the development of clinical candidates
based on 3, little is known about the SAR for this scaffold. In this work, we highlight our recent efforts to
develop a comprehensive understanding of SAR for 3 with the goal of identifying a lead candidate for
acute leukemia therapy with improved potency and physicochemical properties.
Figure 1. Structures of the CREB-CBP binding inhibitor KG-501 (1) and CREB-CBP pathway inhibitors
naphthol AS-E (2) and XX-650-23 (3).

Compound 3 binds specifically to the CREB KIX domain and was found to suppress the
proliferation of AML cell lines and primary human AML cells with low micromolar IC50. 3 inhibits
histone H3 K27 acetylation in CREB-bound genes and CREB KIX-CBP KID domain interaction,
resulting in suppression of CREB-dependent transcriptional activity to induce apoptosis and G1/S arrest
in AML cells. Furthermore, 3 prolonged the median survival and reduced disease burden in AML
xenograft mice.22 However, due to its moderate potency and relatively poor pharmacokinetic properties, 3
is not suitable for clinical application. In order to develop a more potent and metabolically stable CREB-
CBP inhibitor, we decided to develop a better understanding of the effect of each structural element on
the potency and metabolic stability of 3 following a strategy outlined in Figure 2. Using this approach,
we planned to explore the SAR of various salicylamides in four key areas: (a) the evaluation of naphthyl
mimetics, (b) substitutes for the inter-connecting amide group, and (c) the effect of different substituents
and substitution patterns on the salicyl and anilide ring systems. The inhibitory potency of these
compounds was then assessed on cellular viability of acute leukemia cell lines and KIX-KID interaction.
Potent lead compounds were further investigated for the CREB KIX-CBP KIX domain interaction
inhibition potency, in vitro toxicity in normal bone marrow cells and metabolic stability.
Figure 2. Strategy to probe structure-activity relationships (SAR) of XX-650-23 (3).
Our initial efforts focused on the role of the naphthalen-2-ol group of 3. As shown in Table 1,
removal (4) or methylation (5) of the naphthyl-OH group completely abolished the cytotoxic activity

against AML cell lines, HL-60 and KG-1. Interestingly, the acetylated derivative, 6, showed similar
potency in inhibiting the viability of AML cell lines. It is possible that this ester is hydrolyzed to release
compound 1 in vitro, which could explain the similar IC50 values in both AML cell lines tested. It is worth
noting that this could allow for the exploration of a pro-drug strategy in order to modulate the
physicochemical properties of 3. In an effort to explore other hydrogen-bond donors, anilines (7 and 8),
indole (9) and 2-pyridone (10) analogs were tested but all were observed to have little biological activity
against AML cells. This suggests that the presence of the naphthalen-2-ol is essential for the biological
activity of 3.
Table 1. SAR of naphthalen-2-ol modifications.
Cell IC50 (µM)a
Entry
3
R =
HL-60
KG-1
1.58
1.23
4
5
> 10
> 10
> 10
> 10
6
7
1.69
> 10
1.70
> 10
8
>10
> 10
9
> 10
>10
> 10
> 10
10

aCell viability measured via CellTiter-Glo kit; average of triplicates reported.
To determine the importance of the inter-connecting amide moiety, we first explored the
constitutional isomer of 3 in which the carbonyl and amide nitrogen are transposed (11), and found it to
be inactive against AML cells. Substitution with a ketone (12) or N-methylation (13) also displayed no
activity in AML cells, suggesting that the amide hydrogen plays a key role as a H-bond donor or perhaps
stabilizes the desired conformation. Similarly, replacement of the amide group with a sulfonamide 14
abolished activity. The salicylamide isosteres, 15 and 16, were also prepared and tested. The substitution
of salicylanilide motif with benzimidazole or aminoindazole groups did not produce active compounds
against AML cell lines, suggesting that the salicylanilide group may be uniquely active.
Table 2. SAR of amide group modifications.
Cell IC50 (µM)a
Entry
Structure
HL-60
KG-1
11
> 10
> 10
12
13
14
15
16
> 10
>10
> 10
>10
>10
> 10
> 10
> 10
> 10
> 10
aCell viability measured via CellTiter-Glo kit; average of triplicates reported.

Previous exploration of SAR for this motif focused on substitutions of the anilide ring while
much less was known about SAR of the salicyl portion of 3. We prepared the 1-hydroxyl-2-naphthamide
17 and 2-hydroxy-1-naphthamide 18 analogs (Table 3). These modifications slightly decreased the
potency in both AML cell lines, suggesting that substituents in positions 4 and 5 of the salicyl ring are
preferred.
Table 3. Antiproliferative effects of naphthamide analogs on AML cells.
Cell IC50 (µM)a
Entry
Structure
HL-60
KG-1
17
3.41
2.72
18
4.43
3.76
aCell viability measured via CellTiter-Glo kit; average of triplicates reported.
We further explored the effect of substituents in the salicyl ring. The synthetic scheme for
preparing a series focusing on the salicyl ring is depicted at the top of Table 4: generally, salicylic acid
derivatives were first converted to the corresponding acid chlorides using SOCl2, which was followed by
the amide coupling step using excess aniline (See Supplemental Information for experimental details).
Replacing the fused naphthyl ring with a biaryl system resulted in a slight gain of potency when a phenyl
group occupied position 4 (19), but a slight loss of potency with the phenyl group in position 5 (20).
Replacement of the phenyl ring with 4-pyridyl (21 and 22) and 3-pyridyl (23 and 24) rings led to a 5- to
10-fold drop in potency. Given these results, our next efforts were focused on understanding the SAR of
salicylamides containing a phenyl (25) rather than a naphthyl ring. Removal of substituents on the phenyl
ring decreases activity suggesting that functional groups meta and para to the carbonyl group are
necessary for potency (25). In general, it was found that the presence of a halogen in position 5 (R3) of the
salicyl ring, such compounds 29 and 30, correlated with improved potency relative to the unsubstituted

analog 25. A moderate increase in potency was observed with a benzyl group 26, whereas a minor boost
in potency was seen with trifluoromethyl 37 and trifluoromethyl 40. The only significant decrease was
seen with the para-methoxy substituent 44. A similar SAR pattern was observed for substituents at
position 4 (R2) and very little, if any, improvement could be made with substitution at positions 3 and 6
relative to the N-(4-cyanophenyl)salicylamide 25. The nature of the halogen also plays a role: the fluoro-
derivatives were, in all cases, slightly less potent than the bromo- or chloro- analogs. However, it is not
clear that this is an electronic effect or simply related to slight differences in lipophilicity affecting cell
permeability.
Next, we explored the effect of combined substituents on the salicyl ring in order to understand
whether double substitutions have a synergistic effect or lead to a loss of potency (46 and 47). We
observed the highest potency when a phenyl group occupied position 4 (19), or the introduction of
bromine (29) or chlorine (30) at position 5 of the salicyl ring. Compounds that contained both a phenyl
group at position 5 and a halogen (46 and 47) at position 4 showed no further significant improvement in
potency.
Table 4. SAR of 2-Hydroxy-phenyl modifications.
Cell IC50 (µM)a
Entry
R1
R2
R3
R4
HL-60
0.70
1.35
7.09
5.41
6.66
7.96
4.62
1.74
2.91
5.39
KG-1
0.41
1.30
8.84
7.99
7.94
8.33
5.37
1.93
3.26
9.69
19
20
21
22
23
24
25
26
27
28
H
H
H
H
H
H
H
H
Br
F
Ph
H
H
H
H
H
H
H
H
H
H
H
H
Ph
4-Pyridyl
H
H
4-Pyridyl
3-Pyridyl
H
H
H
H
H
H
3-Pyridyl
H
Bn
H
H

29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
H
H
H
H
Br
Cl
H
H
H
0.62
0.70
1.38
1.42
1.21
1.30
> 10
1.03
3.12
> 10
2.42
3.92
6.86
> 10
5.47
> 10
> 10
0.52
0.70
1.29
1.86
3.15
2.67
1.45
1.92
> 10
1.44
4.81
8.98
2.56
3.37
7.94
> 10
5.43
> 10
> 10
0.48
0.85
H
F
H
H
H
F
H
H
Br
Cl
H
H
H
H
H
H
H
H
F
H
CF3
H
H
H
H
CF3
H
H
Me
H
H
H
Me
H
H
H
H
Me
H
H
H
H
Me
H
MeO
H
H
H
MeO
H
H
H
H
MeO
H
H
H
H
MeO
H
H
Ph
Ph
Cl
Br
H
H
aCell viability measured via CellTiter-Glo kit; average of triplicates reported.
Concurrent with an exploration of SAR of the salicyl ring, a series focusing on the anilide moiety
was generated. As derivatives with halogens in either position 4 or 5 or a non-substituted phenyl in
position 4 of the salicyl ring were the most potent against AML cell lines (Table 4), we designed several
salicylamide derivatives bearing different substituents on the anilide ring and explored their activity
against leukemia cell lines (Table 5). Derivatives with a cyano group in position 4 (48) or without a
cyano group in the position 2 (49) were significantly less potent than compounds with the same group in
position 4 of the anilide group (32). In general, it was found that anilides bearing electron-withdrawing
groups were significantly more potent. When trifluoromethyl, trifluoromethoxy, and nitro groups
occupied the position 3 (R4) or 4 (R5) of the anilide ring, we observed increased potency in AML and
ALL cell lines (Table 5 entries 50–63), while derivatives containing fluoro, chloro, acetyl, ester, phenyl,
methoxy, and sulfonyl groups were significantly less potent (data not shown). Based on our observations
that the most potent compounds contained anilides bearing electron withdrawing groups, several analogs
containing two substituents on the anilide ring were prepared. We observed that compounds with two

trifluoromethyl groups (Table 5, entries 64–73) were, in general, more potent in AML and ALL cell lines
than compounds with single trifluoromethyl group on the anilide ring. While single halogen substituents
on the anilide ring decreased potency, we observed a boost in potency when two halogens were
introduced (74–78). The combination of a chloro group with either a trifluoromethoxy or a
trifluoromethyl substituent did not further enhance the potency against leukemic cells (79–84). Similarly,
derivatives with both a chloro and a cyano group did not significantly increase the potency compared to
derivatives bearing two chloride groups (85–87).
Table 5. SAR for anilide ring.
#
R1
R2
R3
R4
R5
R6
Cell IC50 (µM)a
HL-60
KG-1
Jurkat
Nalm6
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
H
H
H
F
H
H
Cl
H
Br
CF3
Ph
Ph
F
H
H
H
H
H
Cl
H
Br
Ph
Ph
Ph
F
F
F
F
H
F
Cl
H
Br
H
H
Cl
Br
Br
F
CN
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CF3
CF3
CF3
CF3
CF3
CF3
CF3
CF3
CF3
CF3
Cl
H
H
CF3
CF3
H
CF3
CF3
CF3
CF3
CF3
CF3
CF3
CF3
OCF3
H
NO2
H
H
H
H
H
H
H
H
H
H
H
H
CF3
H
H
H
H
H
H
H
H
H
OCF3
H
CF3
CF3
CF3
CF3
CF3
CF3
CF3
CF3
CF3
CF3
Cl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
7.75
> 10
0.49
0.68
0.97
0.51
0.63
0.58
0.36
0.38
0.49
0.54
0.45
0.85
0.73
0.52
0.32
0.16
0.20
0.29
0.27
1.14
0.79
0.43
0.43
0.25
0.60
0.89
9.10
> 10
1.70
0.74
1.01
0.74
0.58
0.61
0.34
0.78
0.77
0.40
0.82
0.78
0.83
1.10
0.47
0.50
0.62
0.30
0.83
0.86
0.43
0.30
0.47
0.51
0.78
0.85
n.d.b
n.d.
n.d.
n.d.
0.58
0.80
0.84
0.71
0.67
0.62
0.50
0.82
1.16
0.74
0.88
0.81
0.70
0.94
0.50
0.48
0.57
0.35
0.93
1.24
0.54
0.68
0.73
0.70
0.71
0.88
2.09
1.53
1.80
1.21
1.18
1.09
0.69
1.12
0.80
1.00
1.30
1.21
1.37
2.08
0.62
0.63
0.85
0.59
1.20
1.58
0.56
0.84
2.13
0.95
1.44
1.62
F
F
F
Cl
H
Br
H
H
Cl
Br
Cl
Br
F
H
H
H
Cl
F
H
H
F
H
H

76
77
78
79
80
81
82
83
84
85
86
87
88
89
H
H
H
H
H
H
CF3
H
H
Br
H
F
F
Cl
Cl
F
Cl
H
Cl
F
H
Cl
H
F
Cl
Cl
Br
Cl
Cl
H
Cl
Cl
CF3
Cl
H
Cl
Br
Cl
OCF3
OCF3
OCF3
OCF3
CF3
Cl
CN
CN
CN
NO2
NO2
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
H
H
H
H
0.66
0.60
0.78
0.45
0.55
0.65
0.38
0.46
0.60
0.44
0.50
0.73
0.28
0.30
0.76
0.67
0.68
0.46
0.37
0.62
0.64
0.45
0.54
0.96
0.86
0.79
0.49
0.36
0.67
0.86
1.35
0.84
0.70
0.88
0.79
0.69
0.53
0.75
0.62
0.75
0.53
0.40
1.37
1.25
3.07
1.05
1.00
1.35
0.95
1.22
0.97
0.92
0.73
1.86
0.91
0.62
Cl
Cl
Cl
Cl
Br
H
H
H
H
H
Cl
aCell viability measured via CellTiter-Glo kit; average of triplicates reported.
bn.d.: not determined.
We found compound 88 to be particularly intriguing, as it bears significant resemblance to the
widely-used anthelmintic Niclosamide 89,23-25 having the salicylanilide core, an electron withdrawing
halogen para to the naphthyl hydroxyl group of the salicyl ring, and an electron-withdrawing nitro group
para to the anilide. Niclosamide has been reported to have anti-tumor activity in several cancers by
targeting several oncogenic signaling pathways including Wnt/β-catenin, mTORC1, STAT3, NF-κB, and
Notch.25-30 Indeed, when Niclosamide was tested in our hands, it was found to inhibit cellular viability of
AML cells by inducing apoptosis and cell cycle arrest, with minimal toxicity in normal bone marrow
cells.31 Additionally, Niclosamide was observed to inhibit KIX- KID binding interaction and CREB-
dependent transcription in a dose-dependent manner. Primary xenograft models with AML patient
samples demonstrated that Niclosamide inhibits AML disease progression in vivo and prolongs survival.
Niclosamide also potentiates the cytotoxic effect of other chemotherapy drugs including Daunorubicin,
Vincristine, and Mitoxantrone.31
Finally, we decided to investigate the potency of 8-hydroxyquinolines derivatives. Replacement
of the naphthyl ring with an 8-hydroxyquinoline derivative (90) displayed impressive potency against
AML and ALL cell lines except KG1. We further investigated the activity of derivatives of 90 against
leukemia cell lines (Table 6). Interestingly, these derivatives with structural modification of the anilide

moiety slightly decreased the potency of all tested derivatives (92–100), while the introduction of
bromine in the quinoline moiety resulted in similar potency (91).
Table 6. 8-Hydroxyquinoline modifications.
Cell IC50 (µM)a
#
Structure
HL-60 KG-1 Jurkat Nalm6
90
0.37
0.33
2.32
n.d.b
0.18
0.23
0.13
91
0.21
92
93
0.68
0.51
3.51
3.34
0.55
0.43
0.38
0.31
94
95
0.61
0.63
0.89
0.73
0.91
0.71
0.79
3.56
3.30
3.77
3.26
3.67
3.24
3.54
0.43
0.51
0.27
0.42
0.55
0.27
0.17
0.45
0.38
0.35
0.37
0.42
0.23
0.28
96
97
98
99
100
aCell viability measured via CellTiter-Glo kit; average of triplicates reported.
bn.d.: not determined.

To confirm whether our lead compounds can inhibit the CREB-CBP pathway, we performed the
split Renilla luciferase (RLuc) complementation assay in which the RLuc activity can be reconstituted
only when N-terminal RLuc (RLucN) fused with KID and C-terminal RLuc (RLucC) fused with KIX are
brought to a close by KIX-KID interactions.20 Without interaction between KIX and KID domains, the
RLuc fragments cannot exhibit emission of light in the presence of substrate coelenterazine. Therefore,
the emitted light from the reconstituted Renilla luciferase represents the binding activity of KIX and KID,
allowing us to access the KIX-KID interaction inhibition potency of lead compounds in live cells. There
was a positive correlation between KIX-KID interaction inhibitory activity and the potency in leukemia
cells. Consistent with their loss of potency in leukemia cells (Tables 1-4), removal of naphthyl-OH (4),
modification in the amide group (11) or methoxy substituent in the salicyl ring (44 and 45) showed
significantly less CBP KIX-CREB KID interaction inhibitory activity (Figure 3). As the introduction of a
bromo (29) or chloro (30) in position 5 of the salicyl ring modestly improved potency in AML cell lines,
their KIX-KID inhibitory potency was marginally increased (Figure 3). The presence of two
trifluoromethyl groups in anilide ring of compounds with a halogen in position 5 (66 and 67) or a non-
substituted phenyl in position 4 with a halogen in position 5 (71) of the salicyl ring led to the significant
enhanced KIX-KID inhibitory activity (Figure 3), consistent with their improved potency in leukemia
cells. Furthermore, we have assessed the direct inhibitory effect of lead compounds (66, 67 and 71) on
KID-KIX interaction with a fluorescence polarization (FP) assay. The assay measures a BODIPY-labeled
phosphorylated KID domain of CREB and KIX domain of CBP using FP to provide the degree of KIX-
KID interaction. In the presence of an inhibitor, BODIPY-labeled phosphorylated KID domain could not
bind to the KID domain, which decreases the FP as a result of the increase in the mobility of the
BODIPY-labeled KID domain. Consistent with the split RLuc complementation assay data, these
compounds (66, 67 and 71) led to a decrease in fluorescence polarization (Supplementary Figure 1).
Compounds 66, 67 and 71 showed a similar potency in leukemia cells and KIX-KID inhibition with 88
and Niclosamide 89 (Figure 3), suggesting our lead compounds are targeting CREB-CBP for anti-
leukemic activity similar to Niclosamide 89.

Figure 3. Split Renilla luciferase assay for CBP KIX-CREB KID interaction. RLucC-KIX and KID-
RLucN expressing vectors were transfected into 293 cells. Transfected cells were treated with compounds
half hour before forskolin treatment. Cells were incubated with the indicated amounts of compounds for
additional 90 min. If compounds block the association of KIX and KID domain, light emission by Renila
luciferase activity will be diminished. Renilla luciferase activity was measured using coelenterazine. Data
are presented as mean ± SEM (n=3).
Though their enhanced potency in leukemia cells, 8-hydroxyquinoline derivatives did not inhibit
KIX-KID interaction (Figure 4), suggesting the naphthyl group is critical to provide specificity for the
inhibitory effect on KIX-KID interaction.

Figure 4. Split Renilla luciferase assay for assessing the effects of 8-Hydroxyquinoline derivatives
on CBP KIX-CREB KID interaction. HEK 293 cells transfected with RLucC-KIX and KID-RLucN
expressing vectors were treated with 8-Hydroxyquinoline derivatives, and then Renilla luciferase activity
was measured using coelenterazine. Data are presented as mean ± SEM (n=3).
Compound 3 and Niclosamide 89 suppress leukemia cell viability in vitro and leukemia disease
progression in vivo by inhibiting CREB-dependent gene transcription and KIX-KID interaction with little
toxicity on normal human hematopoietic cells or mice.22, 31 As CREB knockdown using shRNA inhibited
AML cell growth without affecting the proliferation of normal hematopoietic cells,14, 15 we expected that
targeting CREB would be associated with less toxicity than conventional chemotherapy. Thus, the
potential toxicity of lead compounds was assessed in normal bone marrow (BM) progenitor cells. In vitro
colony formation activity of normal BM hematopoietic progenitor cells was not significantly inhibited up
to 10 M of 67 and 71, consistent with 3 and Niclosamide 89 (Table 7). Though 90 inhibited colony
formation activity of normal bone marrow cells (IC50: 3.07 μM), its 10-fold therapeutic window is still
enough for clinical application.
Since 66, 67, 71, and 90 showed low toxicity in normal BM cells as well as excellent potency in
leukemia cells, these compounds were evaluated for in vitro metabolic stability. However, compound 3
and Niclosamide 89 is metabolically unstable, limiting its development as a drug for the treatment of
AML.32 Therefore, we evaluated in vitro metabolic stability of our promising lead compounds to

determine what should be further profiled in pharmacokinetic (PK) studies. The microsomal metabolic
stabilities of these compounds were determined with human (HLM) and mouse liver microsomes (MLM)
(Table 7). 3 and Niclosamide 89 showed poor in vitro metabolic stability in both HLM (T1/2: 44.9 min
and 21 min) and MLM (T1/2: 5.5 min and 6.6 min) (Table 7). 90 showed improved in vitro stability in
both HLM and MLM (T1/2: 124.1 min and 156.7 min). Compound 67 showed an excellent potency in
leukemia cells and KID-KIX inhibitory activity, however, it was unstable in HLM and MLM assays. N-
[3,5-Bis(trifluoromethyl)phenyl]-4-chlorosalicylamide 66 observed a significant improvement in stability
when incubated with human microsomes (T1/2: 159 min). However, this modification did not improve
metabolic stability in mouse liver microsomes (T1/2: 31.8 min). Liver microsomes from mouse and human
have the specific enzyme activities of cytochrome P450 with different cytochrome P450 isoforms 33,
which can account for the discrepancy we observed for compound 66. However, compound 71, bearing a
phenyl group in position 4 and bromine at position 5 was stable in both human and mouse liver
microsomes.
Table 7. Toxicity and in vitro metabolic stability data.
Structure
Normal BM cells
Metabolic Stability (T1/2)
#
(IC50 (M))
HLM (min)
MLM (min)
3
> 10
> 10
44.9
5.5
67
52
29.1
71
89
> 10
> 10
159
21
159
6.6

90
3.07
124.1
156.7
Due to their excellent in vitro potency and improved in vitro metabolic stability, 71 and 90 were
further evaluated by profiling pharmacokinetic (PK) properties in NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG)
mice after single intravenous (IV) administration (2 mg/kg). The time-concentration profiles of
Niclosamide 89, 90, and 71 in mouse plasma were shown in Figure 5, and the pharmacokinetic
parameters were listed in Table 8. Compared with Niclosamide 89, 71 displayed a higher plasma
concentration (0.12 ng/ml vs. 19.0 ng/ml) and longer elimination phase half-life (0.56 h vs. 2.05 h),
suggesting that introduction of a phenyl group in the salicyl ring of 71 led to better pharmacokinetic
properties than Niclosamide 89. Interestingly, compound 90 also showed enhanced pharmacokinetic
performance (T1/2: 2.05 h).
Figure 5. The plasma concentrations of 71, 90, and Niclosamide 89 after a single intravenous
administration to mice. The compounds were dissolved in DMSO/Kolliphor EL (Sigma)/water (3/5/82),
and then administered into NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) mice intravenously at a dose of
2mg/kg. The levels of compounds were determined in the plasma at various times after the intravenous
injection.

Table 8. Pharmacokinetic profile of 71, 90, and Niclosamide 89.
Dose
(mg/kg)
AUC 0-t
(ng/ml*h) (ng/ml*h)
AUC 0-inf
Vz_obs
Cl-obs
Vss_obs
(mg/(ng/ml))
#
Route T1/2 (h)
MRT 0-inf (h)
(mg/(ng/ml)) ((mg)/(ng/ml)/h
89
2
IV
0.56
536
536
0.404
0.0025
0.0037
0.0015
71
90
2
2
IV
IV
2.05
1.39
2249
4070
2334
4089
1.016
0.458
0.0030
0.00086
0.00049
0.00087
0.00022
0.00098
Compound 1 and Niclosamide 89 induce apoptosis and G1/S arrest.22, 31 Though 90 showed
excellent cellular potency, in vitro microsomal stability and PK properties, compound 90 did not inhibit
KIX-KID interaction (Figure 4). To identify the mechanism of anti-leukemic cell cytotoxicity of 90 and
71, apoptosis and cell cycle analysis studies of compound-treated cells were performed. Apoptosis was
assessed using Annexin-V/DAPI double staining and anti-cleaved PARP (poly(ADP-ribose)polymerase)
(cPARP) antibody staining. While both 90 and 71 elicited apoptosis at early (Annexin-V+/DAPI-) and late
(Annexin-V+/DAPI+) stages after 1day treatment, significantly fewer cells entered apoptotic stage by
compound 90 compared with compound 71 (Annexin-V+: 9.3% vs. 39.7%) (Figure 6A). We further
investigated whether these compounds induced apoptosis through the activation of Caspase-3 by
measuring cPARP-positive population 34. cPARP-positive population was increased 1d after treatment of
90 and 71 (cPARP+: 17.3% vs. 58.1%) (Figure 6B).

Figure 6. Treatment of compound 90 and 71 induces apoptosis. HL60 cells were treated with compounds
(1 μM) for 1d. (A) Apoptotic cells were assessed by Annexin-V/DAPI double staining. Compound 71
induced a higher percentage of early apoptotic (Annexin-V+/DAPI−) and late apoptotic (Annexin-
V+/DAPI+) populations than compound 90. Percentages of early and late apoptotic cells are shown as
meanꢀ± SEM (n=3). (B) Cells were fixed, and then stained with anti-cleaved PARP (cPARP) antibody
and DAPI. Percentages of cPARP+ apoptotic cells are indicated as meanꢀ± SEM (n=3).
Since CREB knockdown using shRNA or CREB inhibition by treatment with compound 3 and
Niclosamide causes aberrant cell-cycle progression at the G1/S transition and G1 phase arrest,22, 31 we
further investigated the effects of 90 and 71 on the cell cycle progression. Consistent with 3 and
Niclosamide 89 as a CREB pathway inhibitor, 71 induced G1 cell cycle arrest with increased cell
population at G1 phase (control vs. 71-treated cells, 48.37 ± 0.15% vs. 69.77 ± 0.58%, mean ± SEM (n =
3), p < .001), but decreased cell populations at S, G2, and M phases (control vs. 71-treated cells, S: 27.50

± 0.57% vs. 14.27 ± 0.15%; G2: 21.40 ± 0.42% vs. 14.37 ± 0.43%; M: 2.64 ± 0.06% vs. 1.23 ± 0.08%,
mean ± SEM (n = 3), p < .001) (Figure 7A & B). Conversely, 90 showed defective S/G2 transition with
enhanced percentage of S phase cells (control vs. 90-treated cells, 27.50 ± 0.57% vs. 33.70 ± 0.26%,
mean ± SEM (n = 3), p < .001, Figure 7A & B) and suppressed Cyclin B expression in G2 phase
population (Cyclin B MFI, control vs. 90-treated cells, 1196.33 ± 40.48 vs. 875.33 ± 41.90, mean ± SEM
(n = 3), p < .01, Figure 7C). Next, we investigated the effects of 90 and 71 on the expression of CREB
target genes, RFC3 and POLD2. RFC3 and POLD2, critical players in DNA synthesis and repair,35 have
been shown to be downregulated by CREB knockdown and CREB inhibitors including Niclosamide 89.
22, 36, 37 Our quantitative RT-PCR results showed that mRNA expression levels of RFC3 and POLD2 are
suppressed by compound 71 (Figure 8). This data indicates that 71 exerts its inh0ibitory effect on
leukemia cells by inducing cell cycle arrest at G1 as well as apoptosis through inhibiting CREB-
dependent pathway consistent with compound 3 and Niclosamide 89.

Figure 7. Effect of compound 71 and 90 on cell cycle progression. HL60 cells were treated with
compounds (1 μM) for 1d. Cells were stained for assessing DNA content (DAPI) and levels of p-H3,
Cyclin A and Cyclin B. (A) Cell cycle profiles of DMSO or compound-treated HL-60 cells were shown
as the bivariate distribution of DNA content versus the level of phosphorylated Histone H3. (B)
Percentages of cell populations residing at each cell cycle stage determined by DNA content (DAPI) and
levels of p-H3, Cyclin A and Cyclin B using FlowJo software were plotted as mean ± SEM (n = 3). (C)
Protein levels of the Cyclin A and Cyclin B in G2 phase cells were assessed by flow cytometry analysis.
Median Fluorescence Intensities (MFI) of Cyclin A and Cyclin B were graphed as mean ± SEM (n = 3).
**, p< .01.
Figure 8. Downregulation of CREB-target genes (RFC3 and POLD2) following treatment of compound
71, not compound 90. HL60 cells were treated with compound 71 or 90 (1 μM) for 1d. RFC3 and POLD2
mRNAs were assessed by RT-qPCR and normalized against 7SL lncRNA expression level. The data
represent means of three independent experiments ± SEM. *, p < .05; **, p < .01.

We have conducted a thorough investigation of the SAR around the lead salicylamide, XX-650-
23 3, and have discovered numerous molecules that have improved potency in acute leukemia cells, CBP
KIX-CREB KID interaction inhibitory activity, and physicochemical properties, whilst maintaining low
toxicity in normal human hematopoietic cells. Our study showed that adding trifluoromethyl,
trifluoromethoxy or nitro group to the position 3 or 4 of the anilide ring significantly increased the
potency of XX-650-23 3 in acute leukemia cells and the inhibitory KIX-KID interaction activity.
Furthermore, the addition of a phenyl group in position 4 and a halogen in position 5 of the salicyl ring
improved the physicochemical properties of compounds bearing two trifluoromethyl groups in the anilide
ring while maintaining potency in leukemia cells and KIX-KID interaction inhibition with low toxicity to
normal bone marrow cells.
Acknowledgments
We would like to acknowledge Vivian Zhang, Grace Lam, and Dr. Ludmila Alexandrova for technical
assistance. The authors are thankful to Nanosyn, Inc. for supply of specific salicylamides (see
Supplemental Information). This research was supported by SPARK Pilot Grant Program, Pediatric
Cancer Research Foundation, Bear Necessities Foundation, Leukemia & Lymphoma Society Screen to
Lead Program, Almalechs and Stanford Translation & Clinical Innovation Award, and Maternal & Child
Health Research Institute at Stanford (K.M.S.).
References
1.
Saygin C, Carraway HE. Emerging therapies for acute myeloid leukemia. J Hematol Oncol.
2017;10(1): 93.
2.
Phelan KW, Advani AS. Novel Therapies in Acute Lymphoblastic Leukemia. Curr Hematol
Malig Rep. 2018;13(4): 289-299.

3.
Dores GM, Devesa SS, Curtis RE, Linet MS, Morton LM. Acute leukemia incidence and patient
survival among children and adults in the United States, 2001-2007. Blood. 2012;119(1): 34-43.
4.
5.
Siegel RL, Miller KD, Jemal A. Cancer statistics, 2018. CA Cancer J Clin. 2018;68(1): 7-30.
Noone AM HN, Krapcho M, Miller D, Brest A, Yu M, Ruhl J, Tatalovich Z, Mariotto A, Lewis
DR, Chen HS, Feuer EJ, Cronin KA (eds). SEER Cancer Statistics Review, 1975-2015. National Cancer
Institute; 2018.
6.
Kanellopoulos A, Hamre HM, Dahl AA, Fossa SD, Ruud E. Factors associated with poor quality
of life in survivors of childhood acute lymphoblastic leukemia and lymphoma. Pediatr Blood Cancer.
2013;60(5): 849-855.
7.
Oeffinger KC, Hudson MM. Long-term complications following childhood and adolescent
cancer: foundations for providing risk-based health care for survivors. CA Cancer J Clin. 2004;54(4):
208-236.
8.
Schultz KA, Chen L, Chen Z, et al. Health conditions and quality of life in survivors of childhood
acute myeloid leukemia comparing post remission chemotherapy to BMT: a report from the children's
oncology group. Pediatr Blood Cancer. 2014;61(4): 729-736.
9.
Takei H, Kobayashi SS. Targeting transcription factors in acute myeloid leukemia. Int J Hematol.
2019;109(1): 28-34.
10.
Illendula A, Pulikkan JA, Zong H, et al. Chemical biology. A small-molecule inhibitor of the
aberrant transcription factor CBFbeta-SMMHC delays leukemia in mice. Science. 2015;347(6223): 779-
784.
11.
Uttarkar S, Dukare S, Bopp B, Goblirsch M, Jose J, Klempnauer KH. Naphthol AS-E Phosphate
Inhibits the Activity of the Transcription Factor Myb by Blocking the Interaction with the KIX Domain of
the Coactivator p300. Mol Cancer Ther. 2015;14(6): 1276-1285.
12.
Shankar DB, Cheng JC, Kinjo K, et al. The role of CREB as a proto-oncogene in hematopoiesis
and in acute myeloid leukemia. Cancer cell. 2005;7(4): 351-362.

13.
Kornblau S, Wu WH, Qiu YH, et al. CREB/ATF Family Protein Expression States in AML:
Active CREB1, but not ATF is an Adverse Prognostic Factor. American Society of Hematology Annual
Meeting. San Francisco, CA.; 2014:Abstract #2344.
14.
leukemogenesis. Blood. 2008;111(3): 1182-1192.
15. Pellegrini M, Cheng JC, Voutila J, et al. Expression profile of CREB knockdown in myeloid
leukemia cells. BMC cancer. 2008;8: 264.
16. Chrivia JC, Kwok RP, Lamb N, Hagiwara M, Montminy MR, Goodman RH. Phosphorylated
CREB binds specifically to the nuclear protein CBP. Nature. 1993;365(6449): 855-859.
17. Kwok RP, Lundblad JR, Chrivia JC, et al. Nuclear protein CBP is a coactivator for the
transcription factor CREB. Nature. 1994;370(6486): 223-226.
18. Haus-Seuffert P, Meisterernst M. Mechanisms of transcriptional activation of cAMP-responsive
element-binding protein CREB. Mol Cell Biochem. 2000;212(1-2): 5-9.
19. Best JL, Amezcua CA, Mayr B, et al. Identification of small-molecule antagonists that inhibit an
Cheng JC, Kinjo K, Judelson DR, et al. CREB is a critical regulator of normal hematopoiesis and
activator: coactivator interaction. Proceedings of the National Academy of Sciences of the United States of
America. 2004;101(51): 17622-17627.
20.
Chembiochem : a European journal of chemical biology. 2009;10(17): 2721-2724.
21. Li BX, Yamanaka K, Xiao X. Structure-activity relationship studies of naphthol AS-E and its
Li BX, Xiao X. Discovery of a small-molecule inhibitor of the KIX-KID interaction.
derivatives as anticancer agents by inhibiting CREB-mediated gene transcription. Bioorg Med Chem.
2012;20(23): 6811-6820.
22.
protein in human acute myeloid leukemia cells. Leukemia. 2016;30(12): 2302-2311.
23. Andrews P, Thyssen J, Lorke D. The biology and toxicology of molluscicides, Bayluscide.
Pharmacol Ther. 1982;19(2): 245-295.
Mitton B, Chae HD, Hsu K, et al. Small molecule inhibition of cAMP response element binding

24.
Pearson RD, Hewlett EL. Niclosamide therapy for tapeworm infections. Ann Intern Med.
1985;102(4): 550-551.
25.
Chen W, Mook RA, Jr., Premont RT, Wang J. Niclosamide: Beyond an antihelminthic drug. Cell
Signal. 2017.
26.
Li Y, Li PK, Roberts MJ, Arend RC, Samant RS, Buchsbaum DJ. Multi-targeted therapy of
cancer by niclosamide: A new application for an old drug. Cancer Lett. 2014;349(1): 8-14.
27.
niclosamide antitumor activity]. Biomed Khim. 2015;61(6): 680-693.
28. Jin Y, Lu Z, Ding K, et al. Antineoplastic mechanisms of niclosamide in acute myelogenous
Moskaleva EY, Perevozchikova VG, Zhirnik AS, Severin SE. [Molecular mechanisms of
leukemia stem cells: inactivation of the NF-kappaB pathway and generation of reactive oxygen species.
Cancer Res. 2010;70(6): 2516-2527.
29.
the STAT3 Signaling Pathway. ACS Med Chem Lett. 2010;1(9): 454-459.
30. Fonseca BD, Diering GH, Bidinosti MA, et al. Structure-activity analysis of niclosamide reveals
Ren X, Duan L, He Q, et al. Identification of Niclosamide as a New Small-Molecule Inhibitor of
potential role for cytoplasmic pH in control of mammalian target of rapamycin complex 1 (mTORC1)
signaling. J Biol Chem. 2012;287(21): 17530-17545.
31.
Chae HD, Cox N, Dahl GV, et al. Niclosamide suppresses acute myeloid leukemia cell
proliferation through inhibition of CREB-dependent signaling pathways. Oncotarget. 2018;9(4): 4301-
4317.
32.
Al-Hadiya BM. Niclosamide: comprehensive profile. Profiles Drug Subst Excip Relat Methodol.
2005;32: 67-96.
33.
Singh JK SA, Shirsath VS. Comparative in-vitro Intrinsic Clearance of Imipramine in Multiple
Species Liver Microsomes: Human, Rat, Mouse and Dog. J Drug Metab Toxicol. 2012;3(4): 126.
34.
Lazebnik YA, Kaufmann SH, Desnoyers S, Poirier GG, Earnshaw WC. Cleavage of poly(ADP-
ribose) polymerase by a proteinase with properties like ICE. Nature. 1994;371(6495): 346-347.

35.
implication in cancer and other pathologies. Gene. 2016;590(1): 128-141.
36. Chae HD, Mitton B, Lacayo NJ, Sakamoto KM. Replication factor C3 is a CREB target gene that
Nicolas E, Golemis EA, Arora S. POLD1: Central mediator of DNA replication and repair, and
regulates cell cycle progression through the modulation of chromatin loading of PCNA. Leukemia.
2015;29(6): 1379-1389.
37.
Mitton B, Hsu K, Dutta R, et al. Small molecule screen for inhibitors of expression from
canonical CREB response element-containing promoters. Oncotarget. 2016;7(8): 8653-8662.