Discovery of novel 3-hydroxypicolinamides as selective inhibitors of HIV-1 integrase-LEDGF/p75 interaction

Article abstract of DOI:10.1016/j.ejmech.2016.10.045

Currently, three HIV-1 integrase (IN) active site-directed inhibitors are in clinical use for the treatment of HIV infection. However, emergence of drug resistance mutations have limited the promise of a long-term cure. As an alternative, allosteric inhibition of IN activity has drawn great attention and several of such inhibitors are under early stage clinical development. Specifically, inhibitors of IN and the cellular cofactor LEDGF/p75 remarkably diminish proviral integration in cells and deliver a potent reduction in viral replicative capacity. Distinct from the extensively studied 2-(quinolin-3-yl) acetic acid or 1H-indol-3-yl-2-hydroxy-4-oxobut-2-enoic acid chemotypes, this study discloses a new class of selective IN-LEDGF/p75 inhibitors without the carboxylic acid functionality. More significantly, 3-hydroxypicolinamides also show low micromolar inhibition against IN dimerization, providing novel dual IN inhibitors with in vitro therapeutically selective antiviral effect for further development. Finally, our shape-based ROCS pharmacophore model of the 3-hydroxypicolinamide class of compounds provides a new insight into the binding mode of these novel IN-LEDGF/p75 inhibitors.

Full text of DOI:10.1016/j.ejmech.2016.10.045

Accepted Manuscript
Discovery of novel 3-hydroxypicolinamides as selective inhibitors of HIV-1 integrase-
LEDGF/p75 interaction
Feng-Hua Zhang, Bikash Debnath, Zhong-Liang Xu, Liu-Meng Yang, Li-Rui Song,
Yong-Tang Zheng, Nouri Neamati, Ya-Qiu Long
PII:
S0223-5234(16)30907-2
10.1016/j.ejmech.2016.10.045
EJMECH 9010
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 8 December 2015
Revised Date: 27 September 2016
Accepted Date: 19 October 2016
Please cite this article as: F.-H. Zhang, B. Debnath, Z.-L. Xu, L.-M. Yang, L.-R. Song, Y.-T. Zheng,
N. Neamati, Y.-Q. Long, Discovery of novel 3-hydroxypicolinamides as selective inhibitors of HIV-1
integrase-LEDGF/p75 interaction, European Journal of Medicinal Chemistry (2016), doi: 10.1016/
j.ejmech.2016.10.045.
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.

ACCEPTED MANUSCRIPT
Graphic Abstract
Discovery of novel 3-hydroxypicolinamides as selective inhibitors of HIV-1
integrase-LEDGF/p75 interaction
Feng-Hua Zhang, Bikash Debnath, Zhong-Liang Xu, Liu-Meng Yang, Yong-Tang
Zheng, Nouri Neamati*, and Ya-Qiu Long*
A novel class of HIV-1 integrase (IN) allosteric inhibitors targeting both IN-LEDGF/p75
interaction and IN dimerization were discovered bearing 3-hydroxypicolinamide scaffold, with
low micromolar IC50 values and therapeutically selective antiviral effect.

ACCEPTED MANUSCRIPT
(Eur. J. Med. Chem.)
Discovery of novel 3-hydroxypicolinamides as selective inhibitors of HIV-1
integrase-LEDGF/p75 interaction
Feng-Hua Zhang,a Bikash Debnath,b Zhong-Liang Xu,a Liu-Meng Yang,c Li-Rui Song,a
Yong-Tang Zheng,c Nouri Neamati,b,* and Ya-Qiu Longa,*
aCAS Key Laboratory of Receptor Research, Shanghai Institute of Materia Medica, Chinese
b
Academy of Sciences, 555 Zuchongzhi Road, Shanghai 201203, China; Department of
Medicinal Chemistry, College of Pharmacy, University of Michigan, Ann Arbor, MI
c
48109-2800, USA; Laboratory of Molecular Immunopharmacology, Key Laboratory of
Animal Models and Human Disease Mechanisms, Kunming Institute of Zoology, Chinese
Academy of Sciences, Kunming 650223, China
To whom correspondence should be addressed. Ya-Qiu Long, Phone: 86-21-50806876; Fax:
+86-21-50806876;
Email:
yqlong@simm.ac.cn and
Nouri
Neamati,
Phone:
+01-734-647-2732; E-mail: neamati@umich.edu
1

ACCEPTED MANUSCRIPT
Abstract:
Currently, three HIV-1 integrase (IN) active site-directed inhibitors are in clinical use for the
treatment of HIV infection. However, emergence of drug resistance mutations have limited
the promise of a long-term cure. As an alternative, allosteric inhibition of IN activity has
drawn great attention and several of such inhibitors are under early stage clinical
development. Specifically, inhibitors of IN and the cellular cofactor LEDGF/p75 remarkably
diminish proviral integration in cells and deliver a potent reduction in viral replicative
capacity. Distinct from the extensively studied 2-(quinolin-3-yl) acetic acid or
1H-indol-3-yl-2-hydroxy-4-oxobut-2-enoic acid chemotypes, this study discloses a new class
of selective IN-LEDGF/p75 inhibitors without the carboxylic acid functionality. More
significantly, 3-hydroxypicolinamides also show low micromolar inhibition against IN
dimerization, providing novel dual IN inhibitors with in vitro therapeutically selective
antiviral effect for further development. Finally, our shape-based ROCS pharmacophore
model of the 3-hydroxypicolinamide class of compounds provides a new insight into the
binding mode of these novel IN-LEDGF/p75 inhibitors.
Key words: HIV-1 integrase, allosteric inhibitor, LEDGF/p75, dimerization,
3-hydroxypicolinamide
2

ACCEPTED MANUSCRIPT
Introduction
Major efforts to control and ultimately eradicate the global AIDS pandemic have
focused on antiretroviral therapy, vaccines, microbicides, and educational awareness that
could decrease HIV-1 exposure. Currently approved treatment options targeting the HIV-1
life cycle mainly utilize cocktails of HIV-1 protease, reverse transcriptase, and integrase
inhibitors [1]. These comprise highly active anti-retroviral therapy (HAART), which has
proven effective in prolonging and improving the lives of those infected with the virus [2].
Despite the exceptional success of HAART, new HIV-1 infections continue to spread due to
the emergence of drug-resistant viral strains [3], thereby emphasizing the need for novel
classes of inhibitors that target currently unexploited aspects of the HIV-1 lifecycle [4, 5].
The virally encoded HIV-1 integrase (IN) is an essential enzyme for the viral replication
and has no mammalian counterparts [6]. The protein catalyzes the insertion of the
reverse-transcribed proviral cDNA into the host cell genome through a multistep process that
includes two catalytic reactions: 3’ endonucleolytic processing of the proviral DNA ends
(termed 3’-processing) and the integration of the 3’-processed viral DNA into the cellular
DNA (referred to as strand transfer). So far, three IN inhibitors, i.e. Raltegravir [7],
Elvitegravir [8] and Dolutegravir [9] (Figure 1) have been approved by the FDA for the
clinical treatment of HIV infection. These compounds are IN catalytic site inhibitors that
specifically inhibit the strand transfer (ST) step [10]. Their development represents a
significant milestone in the field of HIV-1 drug discovery and establishes IN as an important
target for the antiretroviral therapy [11]. However, the steady emergence of resistance against
and cross-resistance within this inhibitor class demands novel anti-HIV drugs that are
structurally and mechanistically different [11, 12]. Therefore, the development of allosteric
IN inhibitors targeting the protein-protein interactions between IN and a cellular cofactor
such as the lens epithelium-derived growth factor/p75 (LEDGF/p75) is a new promising
approach to inactivate the enzyme and to limit the emergence of drug-resistance [13, 14].
LEDGF/p75 is the most studied cellular cofactor that is essential for tethering the IN
pre-integration complex to host chromatin and also for the recruitment of other cellular
factors to the pre-integration complex, thereby facilitating effective integration [15].
Additionally, LEDGF/p75 was shown to protect IN from proteolytic degradation, modulate
3

ACCEPTED MANUSCRIPT
IN oligomerization and stimulate its catalytic activity [16, 17]. Inhibition of the
LEDGF/p75-IN interaction can seriously inhibit viral replication [18]. This is also supported
by the recent finding that polymorphisms in the PSIP-1 gene that codes for LEDGF/p75 can
affect rates of HIV disease progression [19].
An impressive panel of small molecules have been rationally developed (Figure 1)
[20-26] to disrupt the interactions between the IN binding domain (IBD) of LEDGF/p75 and
the IN catalytic core domain (CCD), based on the co-crystal structure of the LEDGF/p75
IBD-IN CCD [27]. These IN-LEDGF/p75 inhibitors are mainly categorized into two
chemical classes: benzylindol-3-yl diketoacid (e.g. Fig.1, CHIBA-3053) [20, 22] and the
2-(quinolin-3-yl) acetic acid (e.g. Fig.1, CX14442, BI-224436) [13, 21], with the latter
displaying more potent antiretroviral activity. A compound from this class (BI-224436) was
progressed into Phase I clinical development [28, 29]. Notably, the antiviral potency of the
LEDGF/p75-IN inhibitor was unaffected by the IN ST inhibitors resistance mutations [16,
21]. Unfortunately, the clinical development of BI-224436 was put on hold without disclosing
the reason, probably due to the enterohepatic recirculation-related PK issues [30]. Therefore,
a diverse structural class of IN-LEDGF/p75 inhibitors will be needed for the advancement of
the field.
Cl
OH
O
O
O
O
N
F
N
H
O
OH
N
N
HO
F
O
O
O
F
F
N
H
O
H
N
O
N
N
H
HO
O
N
N
Raltegravir
Elvitegravir
Dolutegravir
O
N
H
O
N
C
N
O
N
O
OH
OH
OH
OH
O
N
COOH
OH
S
OH
O
O
O
S
N
CX14442
O
CHIBA-3053
BI-224436
LNZ-91
Figure 1. Representative HIV-1 integrase inhibitors
Most of the effective IN-LEDGF/p75 inhibitors contain a hydrophobic moiety and a
4

ACCEPTED MANUSCRIPT
carboxylic acid functionality in common, which mimic the hot spot residues Ile365 and
Asp366 within the interface of LEDGF/p75 IBD. The H-bonding interactions between the
carboxylic acid and the backbone amide protons of residues E170 and H171 of IN was
critical to antiviral potency and that there was no tolerated isosteric replacement for the acid
[29]. Interestingly, our previous mechanism-based IN inhibitor research identified another
chemotype (Fig. 1, LNZ-91) bearing a 2,3-dihydroxybenzamide scaffold as dual inhibitors
targeting both IN catalytic site and the IN-LEDGF/p75 interaction [23]. The SAR study on
LNZ-91 delivered a selective IN-LEDGF/p75 inhibitor LZ-14 (Figure 2), which served as a
good starting point to design new IN-LEDGF/p75 inhibitors. Herein, we report our discovery
of a novel class of 3-hydroxypicolinamide-based integration inhibitors through scaffold
hopping from LZ-14 by merging the oxidation-liable aniline amino group into the phenyl
ring (Fig. 2). Interestingly, these 3-hydroxypicolinamide derivatives block HIV replication
via a dual mode of action; inhibiting both IN-LEDGF/p75 interaction and IN dimerization.
They have no significant activity on IN-catalyzed integration reactions, i.e. 3’-processing and
strand transfer at 20 ꢀM. Furthermore, these non-catalytic site inhibitors exhibited low
micromolar EC50 values against HIV replication in a cell-based assay. We observed a direct
correlation for selective IN-LEDGF/p75 inhibition and the IN dimerization inhibition due to
the IN dimer interface engaged in the interaction with LEDGF/p75 IBD. These findings offer
a path for designing improved inhibitors for potential clinical development.
Figure 2. Design and SAR study of the novel pyridinylcarboxamide-based LEDGF/p75-IN
inhibitors
5

ACCEPTED MANUSCRIPT
Chemistry
Starting from the hit LZ-14, a series of picolinamide derivatives were designed and
synthesized with various substitution patterns to determine an optimal scaffold (Figure 2).
Initially, the substituted position of the hydroxyl group on the picolinamide core was
investigated. The synthesis of substituted N-(4-fluorobenzyl)picolinamide derivatives (1a-g)
was conveniently achieved by the condensation of 4-fluorobenzylamine with corresponding
commercially available picolinic acid (Scheme 1). For the synthesis of the
N-(4-fluorobenzyl)-3-methoxypicolinamide (1d), the starting material, 3-methoxypicolinic
acid (6) was prepared from 3-hydroxypicolinic acid via alkylation and hydrolysis.
Scheme 1. Reagents and conditions: a) 4-fluorobenzyamine, HOBt, DIPEA, EDCI, DMF,
microwave, 50℃, 2h; b) CH3I, K2CO3, acetone, refluxing; c) NaOH, MeOH, H2O, r.t.
Abbreviations: HOBt: hydroxybenzotriazole, DIPEA: N,N-diisopropylethylamine, EDCI:
1-ethyl-3-(3-dimethylaminopropyl)carbodiimide, DMF: dimethylformamide.
Then,
the
substitution
on
amide
N-atom
was
investigated.
N-substituted-3-hydroxy-picolinamide derivatives (2a-p) were prepared via condensation
according to the same procedure above. Treatment of the N-alkyl substituted
3-hydroxy-picolinamide derivatives (2m-p) with methoxymethyl chloride in anhydrous
dichloromethane followed by a nucleophilic substitution on benzylbromide generated the
corresponding N,N’-disubstituted-3-OMOM-picolinamide derivatives (8m-p). Removal of
the MOM group from the 3-hydroxy with acetyl chloride gave the corresponding
N,N’-disubstituted-3-hydroxy-picolinamide derivatives (3m-p).
6

ACCEPTED MANUSCRIPT
Scheme 2. Reagents and conditions: a) RNH2, HOBt, DIPEA, EDCI, DMF, microwave, 50℃,
2h; b) MOMCl, DIPEA, anhydrous DCM, 0℃-rt; c) NaH, 1-(bromomethyl)-4-fluorobenzene,
anhydrous DMF, 0℃-rt; d) acetyl chloride, MeOH, 0℃-rt; e) CDI, THF, refluxing.
Abbreviations: MOMCl: chloro(methoxy)methane, CDI: 1,1'-carbonyldiimidazole, THF:
tetrahydrofuran, DCM: dichloromethane.
For the construction of conformationally constrained analogs of the
3-hydroxy-picolinamides (4a, 4c), the N-aryl-3-hydroxypicolinamide substrates (2a, 2c)
were refluxed with 1,1'-carbonyldiimidazole (CDI) in THF, affording the
3-aryl-2H-pyrido[2,3-e][1,3]oxazine-2,4(3H)-diones in good yields.
Results and Discussion
Recently, development of allosteric IN inhibitors targeting the interactions of IN with the
cellular cofactor LEDGF/p75 has gained significant attention in the field of anti-HIV drug
design [11, 31]. However, most of the IN-LEDGF/p75 inhibitors reported in the literature
possess an aryl acetic acid pharmacophore [13], inhibiting both interaction with LEDGF/p75
and catalytic activities of IN [20, 23, 26]. The carboxylic acid was critical to the antiviral
potency [29], but the presence of a carboxylic acid can complicate the development of a lead
series into a marketed drug for a number of reasons [30, 32]. Therefore, our goal in this study
was to develop selective IN-LEDGF/p75 inhibitors with a new scaffold, providing structural
probes to dissect the multifaceted nature of the IN protein throughout the course of HIV-1
7

ACCEPTED MANUSCRIPT
replication.
Hit identification with 3-hydroxypicolinamide scaffold as selective IN-LEDGF/p75
inhibitors. In our initial search for novel IN inhibitors [23], we found a moderately selective
IN-LEDGF/p75 inhibitor LZ-14, which had no effect on the integration reactions at 100 ꢀM
(Figure 2). Considering the oxidation reliability of the anilino amino group, we merged the
amino group into the phenyl ring to generate a pyridine core. However, the resulting
N-(furan-2-ylmethyl)-2-hydroxynicotinamide (LZ-15) lost inhibitory activity against the
interaction of LEDGF/p75-IN. Fortunately, tuning the substituted position of the
carboxamide group along the pyridine ring resulted in potent and selective LEDGF/p75-IN
inhibitors. For example, the N-(furan-2-ylmethyl)-3-hydroxypicolinamide (LZ-21)
effectively inhibited IN-LEDGF/p75 with an IC50 value of 2.4 ꢀM, but did not inhibit the
3’-processing and strand transfer steps at 10 ꢀM.
Based on this observation, a series of picolinamide derivatives were synthesized. Firstly,
we investigated the optimal position of the hydroxyl group on the pyridine ring by using the
N-(4-fluorobenzyl)-picolinamide as the template (Table 1). Rotation of the hydroxyl group
around the phenyl ring indicated that only the N-(4-fluorobenzyl)-3-hydroxypicolinamide (1a)
showed an IC50 value of 8.6 ꢀM to inhibit the IN-LEDGF/p75 interaction. Compounds with
hydroxyl group substituted at 4-, 5- or 6-position were inactive (Table 1, compound 1e-g).
Moreover, when the 3-hydroxyl group was removed (compound 1b), or displaced with a
fluoro atom (compound 1c), or substituted by methyl group (compound 1d)，the resulting
compounds were all inactive in the AlphaScreen protein-protein interaction assay. Obviously,
the 3-hydroxy group on the picolinamide ring was essential for the selective IN-LEDGF/75
inhibition. Notably, these 3-hydroxypicolinamide-based new structure IN-LEDGF/p75
inhibitor also blocked the IN dimer formation.
Further structural modification was conducted with respect to various substitutions on the
optimal scaffold of 3-hydroxypicolinamide as a new class of IN-LEDGF/p75 inhibitors. The
hit compound 1a was used as the starting template for the 2nd round of SAR study below,
which was also renumbered as 2a accordingly for a convenient comparison.
8

ACCEPTED MANUSCRIPT
Table 1. The HIV-1 IN-LEDGF/p75 interaction, as well as IN dimerization and catalytic
function (3'-Processing and Strand transfer) inhibitory activities of picolinamides with
various substitutions on the pyridine ring
IN Enzyme Inhibition
IC50 (ꢀM)a
(IC50, µM)
Compd.
R
IN-LEDGF/p75
Dimer
Inhibition
6.1± 0.9
>20
3'-Pb
STc
Inhibition
8.6 ± 0.4
>20
3-OH
H
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
>20
1a
1b
1c
1d
1e
1f
3-F
>20
>20
3-OMe
4-OH
5-OH
6-OH
>20
>20
>20
>20
>20
>20
1g
>20
>20
a Concentration required to inhibit 50%; b 3'-Processing activity; c Strand transfer activity
SAR study with favorable N-substitution on carboxamide conferring dual HIV-1 IN
inhibitors targeting IN-LEDGF/p75 and IN dimer. Based on the preliminary SAR study
above, further structural optimization on the 3-hydroxypicolinamide scaffold was focused on
the substitution on the amino group of the carboxamide portion. According to the interface
structure revealed by the X-ray crystal structure of the IN CCD dimer in complex with
LEDGF/p75 IBD [21], we proposed that the substituents on the amide portion might provide
the interactions with the hydrophobic pocket of IN. Initially, we tried various hydrophobic
groups including substituted aromatic, heterocyclic and aliphatic groups within the hit
compound 2a (Table 2). Most of the compounds with only single substituent on the amide
moiety significantly inhibited IN-LEDGF/p75 interaction. For the N-aromatic ring substituted
derivatives, the inhibitory activities were tolerant for both heteroaromatic ring and the phenyl
9

ACCEPTED MANUSCRIPT
ring with different substituents (2a-c, 2g-l). Exceptions were found on the
3-hydroxy-N-(naphthalen-1-ylmethyl)
picolinamide
(2e)
and
3-hydroxy-N-(4-(trifluoromethyl)benzyl)picolinamide (2f), which lacked inhibitory activity
in the AlphaScreen assay. It indicated that the strong electron-withdrawing group and bulky
group at this position were unfavorable. However, when nitro group was installed at the
4’-position of the N-benzyl ring, it might provide additional H-bonding interaction as an
H-bond acceptor with the protein pocket, thus compensating the activity loss caused by its
electronic effect (2j, IC50 = 4 ꢀM). On the contrary, compounds with electron donating group
on
the
phenyl
ring
displayed
improved
potency.
For
example,
and
N-(4-aminobenzyl)-3-hydroxypicolinamide
(2h)
3-hydroxy-N-(3,4,5-trimethoxybenzyl)picolinamide (2i) inhibited IN-LEDGF/p75 with an
IC50 value of 1.7 and 2.5 ꢀM, respectively. More significantly, those potent IN-LEDGF/p75
inhibitors bearing N-aryl-3-hydroxypicolinamide displayed in vitra therapeutically selective
anti-HIV effect in cellular assays. Because the cellular potency is collectively affected by
enzyme inhibition and cell membrane permeability, the antiviral activity is not strictly
consistent with the protein-protein interaction inhibitory activity. On the other hand,
compared to a potent HIV-1 IN strand transfer inhibitor L-870810 (targeting the catalytic site
of the enzyme, an investigational anti-HIV drug)[33], these allosteric inhibitors were less
potent in the protecting T-cells from HIV infection (data in Table 2). However, these
non-catalytic site inhibitors of HIV-1 IN could provide new mechanism and new structure to
inhibit HIV IN and thus function cooperatively and synergistically with the active
site-directed enzyme inhibitors to achieve highly active antiviral efficacy and address the
drug-resistance issues.
However, aliphatic substituent on the amino group of the carboxamide moiety caused an
obvious decrease in inhibitory activity, indicated by comparing the parent compound
N-(4-fluorobenzyl)-3-hydroxypicolinamide
(2a)
with
N-(cyclohexylmethyl)-3-hydroxypicolinamide (2d). Further replacing the cyclohexylmethyl
substituent with butyl group (2m) led to the complete loss of the inhibitory activity.
In addition, the effect of the membrane permeability of the small molecule inhibitor on
the antiviral activity was further discussed. n-Octanol-water partition coefficient (P) in case
10

ACCEPTED MANUSCRIPT
of neutral drugs or distribution coefficient (D) in case of ionizable drugs are assumed to be
good indicators of cell permeability. Drugs with LogP and LogD values of 2-4 are generally
predicted to absorb completely by cell membrane [34]. We calculated the LogP and LogD at
pH 7.4 for compounds of series 2 (Table 2) using MarvinView 6.1.4 from ChemAxon [35].
Compound 2h with low LogP and LogD values (1.46 and 1.23, respectively) is most active in
enzyme inhibition assay, but not active in antiviral assay. On the other hand, compounds 2j,
2k, 2l with LogP and LogD values of >2 and moderate enzyme inhibitory activities possesses
good in vitro anti-HIV activities on the T-lymphocyte C8166 cell lines.
Table 2. The HIV-1 IN-LEDGF/p75 interaction, as well as IN dimerization and catalytic
function (3'-Processing and Strand transfer) inhibitory activities and antiviral potency of
3-hydroxypicolinamides with various substitution on the carboxamide portion
IC50 (ꢀM)a
b
c
EC50
CC50
Compd.
R
TId LogPe LogDf
(ꢀM)
(ꢀM)
IN-LEDGF
8.6 ± 0.4
Dimer 3’-P ST
2a
(1a)
6.1
>20 >20 77.6±5.7 114
1.5
2.43
2.2
2b
2.4 ± 0.3
7.5 ± 1.5
19 ± 1.4
10
8
>20 >20 269±141 575
>20 >20 75.6±6.0 218
>20 >20 66.2±4.7 137
2.1
2.9
2.1
1.35
2.2
1.12
1.97
2.45
(LZ-21)
2c
>20
2.68
2d
2e
>20
>20
>20 >20 58.2±4.7 88.8
1.5
3.28
3.05
11

ACCEPTED MANUSCRIPT
>20
>20
>20 >20 43.2±4.7 31.4
< 1
3.16
2.93
2f
4.6 ± 0.7
1.7 ± 0.5
2.5 ± 0.5
8
>20 >20 70.0±3.7 138
2.0
2.3
> 9
2.39
1.46
1.81
2.16
1.23
1.58
2g
2h
2i
4
>20 >20 260±60
>596
9.8
>20 >20 58.7±3.8 >500
NO2
4.0 ± 0.6
>20
>20
>20 >20 26.4±2.7 127
4.8
2.23
3.03
2
2j
>20 >20 19.2±2.5 310
19.4 ±
2.8
2k
4.2 ± 0.7
16
8.0 ± 1.4
>20
>20
>20
>20 >20
>860
>45 2.07
1.89
1.71
1.66
2l
4.4
>20 >20 >500
>1000
2m
5.4± 0.8±
2.33 ±
0.8
0.8
123
52790
L-870810
0.07 nM
nM
nM
a Concentration required to inhibit 50%; b Effective concentration required to reduce HIV-1-induced cytopathic
effect by 50% in lymphocyte C8166 cells; c Cytotoxic concentration to reduce lymphocyte C8166 cells viability
by 50%; d Therapeutic index: ratio CC50/EC50; e, f Predicted log value of partition-coefficient (P) or
distribution-coefficient (D) at pH 7.4
Compounds with two substituents on the nitrogen atom of the carboxamide moiety
(supplementary Table S1, 3m-p), only N-butyl-N-(4-fluorobenzyl)-3-hydroxypicolinamide
(3m) showed moderate inhibition in the AlphaScreen assay, while the rest were inactive. This
result suggested that a free hydrogen on the amino group was important for the integration
inhibition of the 3-hydroxypicolinamide-based IN-LEDGF/p75 inhibitors. Not surprisingly,
the conformationally constrained analogs (supplementary Table S1, 4a, 4c) were inactive
against the integration process, probably due to the deficiency of 3-hydroxyl of pyridine ring
and the N-H of amide group.
It is worth noting that most of our 3-hydroxypicolinamide-based IN-LEDGF/p75
12

ACCEPTED MANUSCRIPT
inhibitors displayed low micromolar inhibition against IN dimerization as well. It has been
reported that 2-(quinolin-3-yl)-acetic acid-derived allosteric IN inhibitors have been shown to
bind to the IN dimer interface at the LEDGF/p75 binding pocket, stabilize interacting IN
subunits, and promote aberrant, higher order IN multimerization [36, 37]. Our results suggest
a similar multimodal nature of the IN-LEDGF/p75 inhibitors, targeting both the
IN-LEDGF/p75 interaction and the IN dimerization. However, the dual inhibition mechanism
is not clear in details. According to the co-crystal structure of IN CCD dimer in complex with
LEDGF IBD [27], the IN dimer interface is involved in the interaction with LEDGF/p75 IBD.
Therefore, we speculate that the inhibition of the IN dimerization by the
3-hydroxypicolinamide-based LEDGF inhibitors might result from the conformational
changes at the dimerization interface induced by the inhibitor binding to the LEDGF binding
site on the HIV-IN. Further study is underway to elucidate the multimode mechanism of
action of these new scaffold HIV-1 IN-LEDGF/p75 inhibitors.
Binding mode interpretation and pharmacophore model establishment. The
structure-activity relationship of the hydroxypicolinamide-based IN-LEDGF/p75 inhibitors
could be well rationalized by the binding mode predicted by molecular modeling. According
to the X-ray structure of the IN CCD dimer in complex with the LEDGF IBD [27], the
IN-LEDGF/p75 interface comprises several hydrogen bonds and hydrophobic interactions.
The side chain of LEDGF/p75 residue Ile365 projects into a hydrophobic pocket formed by
the IN B-chain residues Leu102, Ala128, Ala129, and Trp132 and the A-chain residues
Thr174 and Met178. The LEDGF/p75 Asp366 makes a bi-dentate hydrogen bond to the
main-chain amides of IN residues Glu170 and His171 in the chain A. The backbone amide of
LEDGF/p75 residue Ile365 makes a hydrogen bond to the backbone carbonyl group of
Gln168 in chain A.
13

ACCEPTED MANUSCRIPT
Figure 3. HIV-IN-LEDGF/p75 interactions within the crystal structure PDB: 2B4J, where
LEDGF IBD is bound with IN CCD dimer. Color code: IN chain A – cyan; IN chain B –
green; LEDGF/p75 – brown.
We predicted the binding mode of these 3-hydroxypicolinamide derivatives by docking
into the allosteric site of IN at the IN-LEDGF/p75 interface. Active compounds, such as 2h and
2k were docked in a similar fashion (Figure 4A-D) into the allosteric site of IN and formed
several hydrogen bonds and hydrophobic interactions with the important residues important for
IN-LEDGF/p75 interactions. The amide carbonyl group of these compounds formed H-bond
with backbone NH of His171, while 3-hydroxyl group formed H-bond with side chain
carboxylate of Glu170. Another hydrogen bond interaction was observed between pyridine
nitrogen of picolinamide and OH of Thr174. Pyridine ring was also involved in the pi-cation
interaction with Lys173. Phenyl ring of 2h and 2k were projected towards the hydrophobic
pocket formed by Gln168, and Met178 of chain A and Leu102, Ala128, Ala129, Trp131, and
Trp132 of chain B. Molecular docking also suggests that inactive compounds, 1d and 4a
(Figure 4E-H) failed to form some interactions such as hydrogen bond with Glu170.
Compound 4a also lacked pi-cation interaction with Lys173. Although molecular docking was
able to explain inhibition of IN-LEDGF/p75 interactions for this class of compounds, but
14

ACCEPTED MANUSCRIPT
inhibition of dimerization activity by compounds 2h and 2k could not be explainable by our
models. The inhibition of dimerization is perhaps caused by the conformational changes at the
dimerization interface induced by 2h upon binding to the LEDGF binding site on the HIV-IN.
Figure 4. Predicted binding mode of 3-hydroxypicolinamide. Three dimensional (3D) docking poses as well as
two dimensional (2D) protein-ligand interactions plot of actives 2h (A, B) and 2k (C, D), and inactives 1d (E, F)
15

ACCEPTED MANUSCRIPT
and 4a (G, H). Yellow surface is chain A, whereas, blue surface is chain B. 3D docking figures are generated
using PyMOL (www.pymol.org) and 2D interactions plots are generated using Discovery Studio Visualizer 3.5
Client and redrawn in ChemBioDraw Ultra 14.0 (PerkinElmer).
To further validate this binding mode of 3-hydroxypicolinamide analogs, a shape-based
pharmacophore was generated by aligning docked conformations of 2a, 2b, 2h and 2k (Figure
5A). The pharmacophore consisted of 5 features - two hydrogen bond acceptors, two rings and
one hydrogen bond donor, and represented the exact molecular docking interactions. Of the
two rings, one ring, which is mapped onto pyridine ring, matches pi-cation interactions, while
the other ring, which is mapped onto other aryl ring, matches hydrophobic interactions. The
pharmacophore model was used to distinguish the active and inactive compounds of this study
with AUC value of 0.647 (Figure 6). Within the top 10 hits, six are actual active hits (true
positives) and four are inactive compounds (false positives). The pharmacophore was also able
to identify active hits with high accuracy (AUC value of 0.997) when these active compounds
were mixed with a set of 500 decoys randomly collected from the NCI compounds library.
Another ROCS pharmacophore model was generated based on the docking conformation of an
inactive compound 1d with a hydrogen bond donor feature, which is responsible for hydrogen
bond interactions with Glu170. This particular hydrogen bond is missing in that
pharmacophore (Figure 5B). Also this inactive pharmacophore failed to distinguish between
active and inactive compounds (AUC value = 0.488).
16

ACCEPTED MANUSCRIPT
Figure 5. Shape-based ROCS pharmacophore of 3-hydroxypicolinamide is surrounded by diagrammatic
representation of the binding site. A. Pharmacophore of active compounds. Docking poses of three active
3-hydroxypicolinamides, 2a, 2b, 2h and 2k are aligned and mapped onto the shape-based pharmacophore model.
Yellow diagram is chain A, whereas, blue diagram is chain B. Green balls represent ring features, while blue and
red meshes represent hydrogen bond donor and acceptor, respectively. Red dotted lines indicate hydrogen
bonding interactions and purple dotted line indicates π ˗ cation interaction. B. Pharmacophore of the inactive
compound 1d. The pharmacophore is missing one HBD feature, which is responsible for hydrogen bond
interaction with Glu170.
Figure 6. Validation of Shape-based ROCS pharmacophore generated from aligned docked conformation of
active compounds. Blue dotted line represents random model with AUC value of 0.5, green curve represents the
screening of actives and inactives of this study using the ROCS pharmacophore model with moderate accuracy
(AUC = 0.647) and red curve represents the screening of a set of decoys randomly selected from the NCI mixed
with actives of this study with higher accuracy (AUC = 0.997).
Conclusions
Starting from our previously identified selective IN-LEDGF/p75 inhibitor, we
17

ACCEPTED MANUSCRIPT
discovered a novel class of 3-hydroxypicolinamide-based inhibitors through scaffold hopping.
These 3-hydroxypicolinamide derivatives block HIV replication in T-lymphocyte C8166 cell
lines in micromolar concentrations via a dual mode of action, inhibiting both IN-LEDGF/p75
interaction and IN dimerization. They have no significant effects on IN-catalyzed integration
reactions. Molecular modeling predicted the plausible binding mode of these compounds.
These 3-hydroxypicolinamide derivatives formed several H-bonds and hydrophobic
interactions with important allosteric site residues, such as Glu170, His171, Thr174, Trp132,
Ala128, thus, disrupting IN-LEDGF/p75 interactions. Furthermore, these non-catalytic site
inhibitors exhibited micromolar EC50 values against HIV replication in a cell-based assay.
Our findings highlight the correlation of selective IN-LEDGF/p75 inhibition and the IN
dimerization inhibition due to the IN dimer interface involved in the interaction with
LEDGF/p75 IBD. Though the protein-protein interaction inhibitors are less potent than the
active site-directed enzyme inhibitors, it provides new mechanism and new structure to
address the drug-resistance issue and thus holds big promise for the second generation IN
inhibitors development. Further optimization of our lead compounds and the mechanism
study are in progress to develop a clinical candidate.
Experimental Section
General Synthetic Methods. The 1H NMR spectra were recorded on a Varian 300-MHz
spectrometer. The data are reported in parts per million relative to TMS and referenced to the
solvent. Elemental analyses were obtained using a Vario EL spectrometer. Melting points
(uncorrected) were determined on a Buchi-510 capillary apparatus. The MS spectra were
obtained on an APEXIII 7.0 T FTMS mass spectrometer. The flash column chromatography
was performed on silica gel H (10-40
procedure. The purity of all target compounds was determined by analytical HPLC on Agilent
Technologies 1260 Infinity with Aglient ZORBAX 3.5 M SB-phenyl column (4.6 × 75 mm)
ꢀm). Anhydrous solvents were obtained by standard
ꢀ
in one solvent system (solvent A, 0.025% TFA in water; solvent B, 0.025% TFA in
acetonitrile). with gradient indicated below; flow rate, 1.5 mL/min; UV detector, 300 nm.
Method: 5% B to 95% B in 12 min, 95% B for 3 min. All melting points were measured
using a BÜCHI 510 melting point apparatus.
18

ACCEPTED MANUSCRIPT
N
-(4-Fluorobenzyl)-3-hydroxypicolinamide (1a, 2a). To
a
stirred solution of
3-hydroxypicolinic acid (100 mg, 0.72 mmol), hydroxybenzotriazole (abbreviated HOBt, 194
mg, 1.44 mmol), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDCI, 276 mg, 1.44
mmol) in dimethylformamide (DMF, 5 mL) at rt was added
N,N-diisopropylethylamine
(DIPEA, 0.25 mL, 1.44 mmol) and (4-fluorophenyl)methanamine (0.18 mL, 1.44 mol). The
reaction mixture was stirred at 50°C in microwave reactor for 2h. After removal of most of
the DMF, the residue which was purified by chromatography using ether/ethyl acetate (3:1)
as eluent to give compound 1a (2a) as a white solid (166 mg
(300 MHz, CDCl3) 12.05 (s, 1H), 8.33 (s, 1H), 8.04 (dd,
4H), 7.04 (t, = 8.7 Hz, 2H), 4.61 (d,
= 6.2 Hz, 2H). . 13C NMR (151 MHz, CDCl3)
163.1, 161.5, 157.8, 139.5, 133.3, 131.3, 129.5 (d, = 112.1 Hz), 126.2, 115. 7 (d,
，
94%), m.p. 71-73 oC. 1H NMR
J = 3.9, 1.8 Hz, 1H), 7.38-7.27 (m,
δ
J
J
δ
168.7,
J
J
= 21.6
Hz), 42.3. EI-MS m/z: 246 (M)+. HRMS (ESI) Calcd for C13H11FN2O2 [M+H]+: 247.0877,
found 247.0873. Purity: 99.8% ( R = 6.84 min).
-(4-Fluorobenzyl)picolinamide (1b). Compound 1b was prepared from picolinic acid and
t
N
(4-fluorophenyl)methanamine according to the same procedure described for 1a. Purification
via column chromatography afforded 1b as colorless oil (311 mg, 94%). 1H NMR (300 MHz,
CDCl3)
Hz, 1H), 7.48-7.40 (m, 1H), 7.38-7.29 (m, 2H), 7.02 (t,
2H). EI-MS m/z: 230 (M)+. Purity: 100% ( R
3-Fluoro- -(4-fluorobenzyl)picolinamide (1c). Compound 1c was prepared from
3-fluoropicolinic acid and (4-fluorophenyl)methanamine according to the same procedure
described for 1a. Purification via column chromatography afforded 1c (322 mg 90%) as a
= 4.4 Hz, 1H), 8.17 (s,
= 8.5, 5.4 Hz, 2H), 7.03 (t,
= 6.1 Hz, 2H). EI-MS m/z: 248 (M)+. Purity: 100% ( R
= 5.60
δ
8.53 (d,
J
= 4.8 Hz, 1H), 8.42 (s, 1H), 8.24 (d,
J
= 7.8 Hz, 1H), 7.88 (td,
J
= 7.7, 1.6
J
= 8.7 Hz, 2H), 4.63 (d,
J
= 6.2 Hz,
t
= 5.59 min).
N
，
o
1
white solid, m.p. 81-83 C. H NMR (300 MHz, CDCl3)
1H), 7.62-7.52 (m, 1H), 7.48 (dt, = 8.3, 4.0 Hz, 1H), 7.35 (dd,
= 8.7 Hz, 2H), 4.61 (d,
min).
-(4-Fluorobenzyl)-4-hydroxypicolinamide (1e). Compound 1e was prepared from
δ 8.36 (d, J
J
J
J
J
t
N
4-hydroxypicolinic acid and (4-fluorophenyl)methanamine according to the same procedure
described for 1a. Purification via column chromatography afforded 1e (166 mg, 94%) as
colorless oil. 1H NMR (300 MHz, CDCl3)
δ
8.76 (s, 1H), 8.35 (d, J = 5.9 Hz, 1H), 7.83 (d, J
19

ACCEPTED MANUSCRIPT
= 2.6 Hz, 1H), 7.40-7.31 (m, 2H), 7.02 (t,
J = 8.7 Hz, 3H), 4.63 (d, J = 6.2 Hz, 2H). EI-MS
m/z: 246 (M)+.
N
-(4-Fluorobenzyl)-5-hydroxypicolinamide (1f). Compound 1f was prepared from
5-hydroxypicolinic acid and (4-fluorophenyl)methanamine according to the same procedure
described for 1a. Purification via column chromatography afforded 1f (164 mg, 93%) as
1
colorless oil. H NMR (300 MHz, CD3OD)
δ
8.14 (d,
J
= 2.8 Hz, 1H), 7.95 (d,
J = 8.6 Hz,
1H), 7.36 (dd,
J
= 8.4, 5.4 Hz, 2H), 7.25 (dd,
J
= 8.6, 2.8 Hz, 1H), 7.03 (t,
J
= 8.8 Hz, 2H),
4.55 (s, 2H). EI-MS m/z: 246 (M)+. Purity: 98.2% ( R
t
= 5.25 min).
N
-(4-Fluorobenzyl)-6-hydroxypicolinamide (1g). Compound 1g was prepared from
6-hydroxypicolinic acid and (4-fluorophenyl)methanamine according to the same procedure
described for 1a. Purification via column chromatography afforded 1g (321 mg，91%) as
1
colorless oil. H NMR (300 MHz, CD3OD)
δ
7.69 (s, 1H), 7.40-7.31 (m, 2H), 7.17 (d,
J =
12.2 Hz, 1H), 7.05 (t,
(M)+. Purity: 95.8% ( R
Methyl 3-methoxypicolinate (5). To a stirred suspension of 3-hydroxypicolinic acid (350 mg,
2.52 mmol), K2CO3 (871 mg, 6.3 mmol) in acetone (25 ml) at rt was added CH3I (345 l,
J
= 8.7 Hz, 2H), 6.75 (d,
J = 9.0 Hz, 1H), 4.54 (s, 2H). EI-MS m/z: 246
t
= 4.59 min).
ꢀ
5.54mmol) during 10 min. The mixture was refluxed for 12h and the excess acetone was
removed under reduced pressure. The residue was diluted with dichloromethane (DCM) and
the organic phase was washed with water (2 times) and brine, dried over anhydrous Na2SO4
and concentrated under vacuum. The residue was chromatographed on silica gel [petroleum
ether/EtOAc (PE/EA) = 1:2] to provide compound 5 (261 mg, 62%) as colorless oil. 1H NMR
(300 MHz, CDCl3)
3.93 (s, 3H).
δ 8.29 (d, J = 4.4 Hz, 1H), 7.42 (dt, J = 16.5, 6.5 Hz, 2H), 3.98 (s, 3H),
3-Methoxypicolinic acid (6). The reaction mixture of compound 5 (261 mg, 1.56 mmol),
NaOH (281mg, 7.03mmol) in MeOH (9 ml) and water (6 ml) was stirred at rt for 1.5h. The
solvent was removed under reduced pressure and the residue was diluted with water. The
solution was extracted with EtOAc (5 times). The organic phases were combined and
1
evaporated under vacuum to give 6 as a light yellow solid. H NMR (300 MHz, CD3OD)
δ
8.18 (s, 1H), 7.78 (d, J = 8.3 Hz, 1H), 7.67 (s, 1H), 3.96 (s, 3H).
N
-(4-Fluorobenzyl)-3-methoxypicolinamide (1d). Compound 1d was prepared from
20

ACCEPTED MANUSCRIPT
compound 6 and (4-fluorophenyl)methanamine according to the same procedure described
for 1a. Purification via column chromatography afforded 1d (61 mg, 81%) as light brown oil.
1H NMR (300 MHz, CDCl3)
2H), 6.99 (t, = 8.6 Hz, 2H), 4.62 (d,
Purity: 100% ( R = 6.74 min).
-(Furan-2-ylmethyl)-3-hydroxypicolinamide (LZ-21, 2b). Compound 2b was prepared
δ
8.45 (d,
J
= 2.9 Hz, 1H), 7.75 (t,
J = 9.4 Hz, 2H), 7.40-7.31 (m,
J
J
= 5.6 Hz, 2H), 4.08 (s, 3H). EI-MS m/z: 260 (M)+.
t
N
from 3-hydroxypicolinic acid and furan-2-ylmethanamine according to the same procedure
described for 2a. Purification via column chromatography afforded 2b (148 mg, 94%) as a
white solid, m.p. 68-70 oC. 1H NMR (300 MHz, CDCl3):
δ
12.01 (s, 1H), 8.32 (s, 1H), 8.05 (d,
= 0.9 Hz, 1H), 7.35-7.30 (m, 2H), 6.33 (d, = 5.9 Hz, 2H), 4.64 (d,
= 6.0 Hz, 2H). 13C NMR (151 MHz, CDCl3)
168.6, 157.8, 150.5, 142.5, 139.6, 131.3,
J
J
= 4.0 Hz, 1H), 7.39 (d,
J
J
δ
128.7, 126.1, 110.5, 107.9, 35.9.EI-MS m/z: 218 (M)+. HRMS (ESI) Calcd for C11H10N2O3
[M+H]+: 219.0764, found 219.0759. Purity: 99.0% ( R
= 5.75 min).
3-Hydroxy- -(thiophen-2-ylmethyl)picolinamide (2c). Compound 2c was prepared from
t
N
3-hydroxypicolinic acid and thiophen-2-ylmethanamine according to the same procedure
described for 2a. Purification via column chromatography afforded 2c (152 mg, 90%) as a
o
1
white solid, m.p. 79-80 C. H NMR (300 MHz, CDCl3):
(dd, = 3.9, 1.8 Hz, 1H), 7.33 (t, = 3.0 Hz, 2H), 7.25 (d,
1H), 7.01-6.95 (m, 1H), 4.81 (d,
139.9, 139.6, 131.3, 128.8, 127.0, 126.4, 126.1, 125.5, 37.7. EI-MS m/z: 234 (M)+. HRMS
(ESI) Calcd for C11H10N2O2S [M+H]+: 235.0536, found 235.0529. Purity: 99.8% ( R
= 6.42
min).
-(Cyclohexylmethyl)-3-hydroxypicolinamide (2d). Compound 2d was prepared from
δ
12.01 (s, 1H), 8.34 (s, 1H), 8.04
J
J
J
= 1.2 Hz, 1H), 7.06 (d, J = 2.6 Hz,
J
= 6.1 Hz, 2H). 13C NMR (101 MHz, CDCl3)
δ 168.4, 157.8,
t
N
3-hydroxypicolinic acid and cyclohexylmethanamine according to the same procedure
described for 2a. Purification via column chromatography afforded 2d (147 mg, 87%) as a
o
1
white solid, m.p. 70-72 C. H NMR (300 MHz, CDCl3):
1.8 Hz, 1H), 7.34-7.27 (m, 2H), 4.83 (s, 1H), 3.29 (t, = 6.6 Hz, 2H), 1.76 (dd,
Hz, 5H), 1.23 (dd, = 21.0, 9.6 Hz, 4H), 1.02 (dd,
= 23.3, 10.6 Hz, 2H). 13C NMR (101
MHz, CDCl3) 168.7, 157.7, 139.4, 131.5, 128.5, 126.1, 45.2, 38.0, 30.8, 26.3, 25.8. EI-MS
δ
12.27 (s, 1H), 8.04 (dd, J = 4.0,
J
J
= 27.3, 12.4
J
J
δ
m/z: 234 (M)+. HRMS (ESI) Calcd for C13H18N2O2 [M+H]+: 235.1441, found 235.1435.
21

ACCEPTED MANUSCRIPT
Purity: 99.9% (tR = 7.61 min).
3-Hydroxy- -(naphthalen-1-ylmethyl)picolinamide (2e). Compound 2e was prepared
N
from 3-hydroxypicolinic acid and naphthalen-1-ylmethanamine according to the same
procedure described for 2a. Purification via column chromatography afforded 2e (163 mg,
82%) as a white solid, m.p. 108-110 oC. 1H NMR (300 MHz, CDCl3)
1H), 8.10 (d, = 8.5 Hz, 1H), 8.02-7.97 (m, 1H), 7.87 (dd, = 15.2, 8.3 Hz, 2H), 7.52 (dq,
= 15.1, 6.9 Hz, 4H), 7.35 (d, = 3.5 Hz, 2H), 5.10 (d,
= 5.9 Hz, 2H). EI-MS m/z: 278 (M)+.
Purity: 99.1% ( R = 7.87 min).
3-Hydroxy- -(4-(trifluoromethyl)benzyl)picolinamide (2f). Compound 2f was prepared
δ 12.28 (s, 1H), 8.50 (s,
J
J
J
J
J
t
N
from 3-hydroxypicolinic acid and (4-(trifluoromethyl)phenyl)methanamine according to the
same procedure described for 2a. Purification via column chromatography afforded 2f (173
o
1
mg, 81%) as a white solid, m.p. 100-102 C. H NMR (300 MHz, CDCl3)
8.61 (s, 1H), 8.07 (dd, = 3.8, 2.0 Hz, 1H), 7.62 (d, = 8.1 Hz, 2H), 7.49 (d,
7.39 (dd, = 2.8, 2.2 Hz, 2H), 4.71 (d,
R = 7.69 min).
-((1 -Indol-5-yl)methyl)-3-hydroxypicolinamide (2g). Compound 2g was prepared from
3-hydroxypicolinic acid and (1 -indol-5-yl)methanamine according to the same procedure
described for 2a. Purification via column chromatography afforded 2g (161 mg, 84%) as a
δ
12.06 (s, 1H),
J
J
J
= 8.0 Hz, 2H),
J
J
= 6.3 Hz, 2H). EI-MS m/z: 296 (M)+. Purity: 99.9%
(t
N
H
H
o
1
white solid, m.p. 119-120 C. H NMR (300 MHz, CDCl3)
δ 7.95 (s, 1H), 7.53 (s, 1H),
7.42-7.19 (m, 4H), 7.15 (s, 1H), 7.09 (s, 1H), 4.62 (s, 2H). 13C NMR (151 MHz, CDCl3)
δ
168.4, 157.8, 139.4, 135.3, 131.6, 128.8, 128.5, 128.1, 126.0, 124.9, 122.4, 120.3, 111.4,
102.7, 43.6. EI-MS m/z: 267 (M)+. HRMS (ESI) Calcd for C15H13N3O2 [M+H]+: 268.1081,
found 268.1074. Purity: 98.2% ( R
-(4-Aminobenzyl)-3-hydroxypicolinamide (2h). Compound 2h was prepared from
3-hydroxypicolinic acid and 4-(aminomethyl)aniline according to the same procedure
described for 2a. Purification via column chromatography afforded 2h (149 mg 85%) as a
white solid, m.p. 132-134 oC. 1H NMR (300 MHz, CD3OD)
8.08 (dd, = 4.4, 1.3 Hz, 1H),
7.45-7.38 (m, 1H), 7.31 (d, = 8.5 Hz, 1H), 7.16 (d, = 8.6 Hz, 2H), 6.77 (d, = 8.5 Hz, 2H),
4.46 (s, 2H). 13C NMR (151 MHz, CDCl3)
168.4, 157.8, 145.9, 139.4, 131.6, 129.2, 128.6,
t = 6.61 min).
N
，
δ
J
J
J
J
δ
127.2, 126.0, 115.3, 42.7.EI-MS m/z: 243 (M)+. HRMS (ESI) Calcd for C13H13N3O2 [M+H]+:
22

ACCEPTED MANUSCRIPT
244.1081, found 244.1075. Purity: 98.4% (tR = 3.10 min).
3-Hydroxy- -(3,4,5-trimethoxybenzyl)picolinamide (2i). Compound 2i was prepared from
N
3-hydroxypicolinic acid and (3,4,5-trimethoxyphenyl)methanamine according to the same
procedure described for 2a. Purification via column chromatography afforded 2i (192 mg,
84%) as a white solid, m.p. 120-122 oC. 1H NMR (300 MHz, CDCl3)
δ
12.54 (s, 1H), 8.99 (s,
1H), 8.09 (s, 1H), 7.50 (s, 2H), 6.62 (s, 2H), 4.58 (d,
J = 5.9 Hz, 2H), 3.87 (s, 6H), 3.83 (s,
3H). EI-MS m/z: 318(M)+.
3-Hydroxy-N-(4-nitrobenzyl)picolinamide (2j). Compound 2j was prepared from
3-hydroxypicolinic acid and (4-nitrophenyl)methanamine according to the same procedure
described for 2a. Purification via column chromatography afforded 2j (98 mg, 84%) as a
white solid, m.p. 129-130 oC. 1H NMR (300 MHz, CDCl3)
(d, = 8.8 Hz, 2H), 8.08 (dd, = 4.0, 1.8 Hz, 1H), 7.54 (d,
2H), 4.75 (d,
= 6.4 Hz, 2H). 13C NMR (126 MHz, CDCl3)
140.9, 138.0, 129.3, 128.5, 125.9, 124.1, 42.8. ESI-MS: m/z 274.2[M+H]+. HRMS (ESI)
Calcd for C13H11N3O4 [M+H]+: 274.0822, found 274.0817. Purity: 99.3% ( R
= 6.79 min).
-(3-Chloro-4-fluorobenzyl)-3-hydroxypicolinamide (2k). Compound 2k was prepared
δ
11.92 (s, 1H), 8.65 (s, 1H), 8.22
= 8.9 Hz, 2H), 7.42-7.37 (m,
167.7, 158.7, 147.5, 144.7,
J
J
J
J
δ
t
N
from 3-hydroxypicolinic acid and (3-chloro-4-fluorophenyl)methanamine according to the
same procedure described for 2a. Purification via column chromatography afforded 2k (122
mg, 94%) as a white solid, m.p. 90-92 oC. 1H NMR (300 MHz, CDCl3)
(s, 1H), 8.12-8.05 (m, 1H), 7.45 (d, = 2.6 Hz, 2H), 7.41 (s, 1H), 7.23 (s, 1H), 7.12 (t,
Hz, 1H), 4.59 (d, 168.8, 158.4, 157.4, 156.7,
= 6.1 Hz, 2H). 13C NMR (151 MHz, CDCl3)
139.6, 134.7, 131.2, 129.4 (d, = 156.0 Hz), 127.5 (d, = 7.3 Hz), 126.2, 121.3 (d, = 17.9
Hz), 116.8 (d,
= 21.3 Hz), 41.8. ESI-MS: m/z 281.1[M+H]+. HRMS (ESI) Calcd for
C13H10ClFN2O2 [M+H]+: 281.0488, found 281.0481. Purity: 98.3% ( R
= 7.48 min).
-(4-Fluorophenyl)-3-hydroxypicolinamide (2l). Compound 2l was prepared from
δ
12.23 (s, 1H), 8.85
= 8.6
J
J
J
δ
J
J
J
J
t
N
3-hydroxypicolinic acid and 4-fluoroaniline according to the same procedure described for 2a.
Purification via column chromatography using PE/EA=1:1 afforded 2l (91mg, 54%) as a pale
o
1
yellow solid, m.p. 138-140 C. H NMR (300 MHz, CDCl3)
8.15 (dd, = 3.7, 2.1 Hz, 1H), 7.72 (dd, = 8.4, 4.8 Hz, 2H), 7.45 (s, 2H), 7.10 (t,
2H).EI-MS: m/z232 (M)+. Purity: 99.9% ( R
δ
12.05 (s, 1H), 10.16 (s, 1H),
J
J
J
= 8.6 Hz,
t
= 6.93 min).
23

ACCEPTED MANUSCRIPT
N
-Butyl-3-hydroxypicolinamide
(2m).
Compound
2m
was
prepared from
3-hydroxypicolinic acid and butan-1-amine according to the same procedure described for 2a,
purified as colorless oil (246 mg, 88%).
N
-Butyl-3-(methoxymethoxy)picolinamide (7m). To the solution of 2m (246 mg, 1.27
mmol) in anhydrous DCM (10 ml) was added DIPEA (0.332 ml, 1.9 mmol), stirred for 15min
at 0°C under N2. Then chloro(methoxy)methane (MOMCl, 0.115 ml, 1.52 mmol) which was
dissolved in DCM (1 ml) was added to above system during 10min at 0°C. The reaction
mixture was stirred for 36h at r.t. After removal of most of DCM, the concentration provided
the residue which was purified by chromatography using petroleum ether/ethyl acetate (3:1)
1
as eluent to give compound 7m as light yellow oil (220 mg, 73%). H NMR (300 MHz,
CD3OD)
δ
8.21 (d,
J
= 4.7 Hz, 1H), 7.71 (d,
= 0.8 Hz, 3H), 3.40 (t,
= 14.7, 7.5 Hz, 2H), 0.97 (t, = 7.2 Hz, 3H).
-Butyl- -(4-fluorobenzyl)-3-(methoxymethoxy)picolinamide (8m). To a stirred solution
J
= 8.5 Hz, 1H), 7.45 (dd,
J = 8.6, 4.6 Hz, 1H),
5.30 (s, 2H), 3.48 (d,
J
J = 7.0 Hz, 2H), 1.67 – 1.55 (m, 2H), 1.44 (dd,
J
J
N
N
of 7m (217 mg, 0.91 mmol) in anhydrous DMF (10 mL) was added NaH (73 mg, 1.82 mmol)
and 1-(bromomethyl)-4-fluorobenzene (0.36 mL, 2.89 mmol) at 0°C. The above mixture was
stirred for 8h at rt under N2, the excess DMF was removed under reduced pressure. The
residue was diluted with DCM (10 mL) and the organic phase was washed with brine, dried
with anhydrous Na2SO4, filtered, concentrated under vacuum. Purification via column
chromatography using PE/EA=2:1 afforded compound 8m as colorless oil (252 mg, 80%). 1H
NMR (300 MHz, CD3OD)
7.33-7.24 (m, 1H), 7.07 (dt,
δ
8.27-8.17 (m, 1H), 7.75-7.64 (m, 1H), 7.51-7.39 (m, 2H),
= 21.4, 8.8 Hz, 2H), 5.29 (d, = 14.0 Hz, 2H), 4.79 (s, 1H),
= 15.2, 7.6 Hz,
= 14.8, 7.5 Hz, 1H), 0.95 (t, J
J
J
4.35 (s, 1H), 3.45 (dd,
1H), 1.44 (m, 2H), 1.25 (dd,
= 7.3 Hz, 1H), 0.69 (t, = 7.4 Hz, 2H).
-Butyl- -(4-fluorobenzyl)-3-hydroxypicolinamide (3m). To the solution of 8m (250 mg,
0.722 mmol) in anhydrous MeOH (10 ml) was added acetyl chloride (670 l, 19.8 mmol)
J
= 5.8, 1.2 Hz, 3H), 3.10-2.99 (m, 1H), 1.61 (dd,
J
J
= 14.7, 7.5 Hz, 1H), 1.07 (dd,
J
J
N
N
ꢀ
during 10min at 0°C. The reaction mixture was stirred at 0°C for another 10min and then for
12h at r.t, the excess MeOH was removed under reduced pressure. The residue was purified
via column chromatography using PE/EA=10:1 afforded compound 3m as colorless oil (170
24

ACCEPTED MANUSCRIPT
mg, 78%). 1H NMR (300 MHz, CD3OD)
= 8.6, 5.3 Hz, 2H), 7.19 (t, = 8.7 Hz, 2H), 4.19 (s, 2H), 3.08-2.98 (m, 2H), 1.75-1.62 (m,
2H), 1.44 (dt, = 14.9, 7.3 Hz, 2H), 0.98 (t,
= 7.3 Hz, 3H). ESI-MS: m/z 303.2 [M+H]+.
3-Hydroxy- -isobutylpicolinamide (2n). Compound 2n was prepared as the same
δ 8.16 (s, 1H), 7.86 (d, J = 23.1 Hz, 2H), 7.53 (dd, J
J
J
J
N
procedure described for 2a, purified as colorless oil (223 mg, 80%).
N-isoButyl-3-(methoxymethoxy)picolinamide (7n). Compound 7n was prepared as the
1
same procedure described for 7m. 7n was purified as light yellow oil (218 mg, 72%). H
NMR (300 MHz, CD3OD)
8.6, 4.6 Hz, 1H), 5.30 (s, 2H), 3.48 (d,
(m, 1H), 0.99 (d, = 6.7 Hz, 6H).
-(4-Fluorobenzyl)- -isobutyl-3-(methoxymethoxy)picolinamide (8n). Compound 8n
was prepared as the same procedure described for 8m. 8n was purified as colorless oil (220
mg, 81%). 1H NMR (300 MHz, CD3OD)
8.22 (t, = 4.3 Hz, 2H), 7.68 (t, = 8.1 Hz, 2H),
7.51-7.37 (m, 4H), 7.30-7.22 (m, 2H), 7.11 (t, = 8.8 Hz, 2H), 7.03 (t, = 8.7 Hz, 2H), 5.29
(s, 2H), 5.27 (s, 2H), 4.81 (s, 2H), 4.37 (s, 2H), 3.46 (s, 3H), 3.43 (s, 3H), 2.90 (d, = 7.5 Hz,
2H), 2.20-2.08 (m, 1H), 1.94-1.83 (m, 1H), 1.00 (d, = 6.7 Hz, 6H), 0.73 (d, = 6.7 Hz, 6H).
-(4-Fluorobenzyl)-3-hydroxy- -isobutylpicolinamide (3n). Compound 3n was prepared
as the same procedure described for 3m and purified as colorless oil (168 mg, 77%). 1H NMR
(300 MHz, CDCl3) 11.90 (s, 2H), 8.12 (d, = 27.1 Hz, 2H), 7.82 (d, = 29.2 Hz, 2H), 7.48-
7.29 (m, 3H), 7.23 (s, 2H), 7.00 (t, = 7.7 Hz, 4H), 5.08 (s, 2H), 4.75 (s, 2H), 3.57 (s, 2H),
3.28 (d, = 7.1 Hz, 2H), 2.09 (t, = 12.4 Hz, 1H), 1.93 (t, = 10.1 Hz, 1H), 0.94 (d, = 6.5
Hz, 6H), 0.81 (d, = 6.18 min).
= 6.3 Hz, 5H).ESI-MS: m/z 303.2 [M+H]+. Purity: 100% ( R
3-Hydroxy- -isopentylpicolinamide (2o). Compound 2o was prepared as the same
procedure described for 2a, purified as light yellow oil (237 mg, 82%).
δ
8.22 (d,
J
= 4.6 Hz, 1H), 7.72 (d,
J
= 8.6 Hz, 1H), 7.46 (dd,
J =
J
= 0.5 Hz, 3H), 3.22 (d,
J = 6.9 Hz, 2H), 2.00 – 1.85
J
N
N
δ
J
J
J
J
J
J
J
N
N
δ
J
J
J
J
J
J
J
J
t
N
N
-isoPentyl-3-(methoxymethoxy)picolinamide (7o). Compound 7o was prepared as the
1
same procedure described for 7m. 7o was purified as light yellow oil (220 mg, 77%). H
NMR (300 MHz, CD3OD)
8.6, 4.6 Hz, 1H), 5.30 (s, 2H), 3.48 (d,
1H), 0.99 (d, = 6.7 Hz, 6H).
-(4-Fluorobenzyl)-
δ
8.22 (d,
J
= 4.6 Hz, 1H), 7.72 (d,
= 0.5 Hz, 3H), 3.22 (d,
J
= 8.6 Hz, 1H), 7.46 (dd,
J =
J
J = 6.9 Hz, 2H), 2.00-1.85 (m,
J
N
N
-isopentyl-3-(methoxymethoxy)picolinamide (8o). Compound 8o
25

ACCEPTED MANUSCRIPT
was prepared as the same procedure described for 8m. 8o was purified as colorless oil (271
mg, 81%). 1H NMR (300 MHz, CD3OD)
7.48 (t, = 2.9 Hz, 2H), 7.29 (dd, = 8.3, 5.3 Hz, 1H), 7.11 (t,
4.79 (s, 2H), 3.45 (s, 3H), 3.11-3.02 (m, 2H), 1.51 (dd, = 14.9, 7.3 Hz, 1H), 1.39 (dd,
15.2, 7.2 Hz, 2H), 0.62 (d, = 6.5 Hz, 6H).
-(4-Fluorobenzyl)-3-hydroxy-
δ
8.22 (t,
J
= 4.9 Hz, 1H), 7.71 (d,
= 8.8 Hz, 2H), 5.32 (s, 2H),
J = 8.6 Hz, 1H),
J
J
J
J
J =
J
N
N
-isopentylpicolinamide (3o). Compound 3o was prepared
1
as the same procedure described for 3m, purified as colorless oil (155 mg, 79%). H NMR
(300 MHz, CD3OD)
δ 8.16 (s, 1H), 7.87-7.84 (m, 2H), 7.53 (dd, J = 8.4, 5.2 Hz, 2H), 7.19 (t,
J = 8.5 Hz, 2H), 4.19 (s, 2H), 3.10-3.02 (m, 2H), 1.74-1.64 (m, 1H), 1.59 (dd, J = 7.2, 1.2 Hz,
2H), 0.96 (d, J = 6.3 Hz, 6H). ESI-MS: m/z 317.2 [M+H]+.
3-Hydroxy-
N-(2-methylallyl)picolinamide (2p). Compound 2p was prepared as the same
procedure described for 2a and purified as light yellow oil (215 mg, 78%).
3-(Methoxymethoxy)-
N
-(2-methylallyl)picolinamide (7p). Compound 7p was prepared as
1
the same procedure described for 7m. 7p was purified as light yellow oil (141 mg, 54%). H
NMR (300 MHz, CD3OD)
8.6, 4.6 Hz, 1H), 5.31 (s, 2H), 4.98 (s, 2H), 3.96 (s, 2H), 3.48 (s, 3H), 1.81 (s, 3H).
-(4-Fluorobenzyl)-3-(methoxymethoxy)- -(2-methylallyl)picolinamide (8p). Compound
δ 8.23 (d, J = 4.6 Hz, 1H), 7.72 (d, J = 8.6 Hz, 1H), 7.47 (dd, J =
N
N
8p was prepared as the same procedure described for 8m. 8p was purified as colorless oil
(165 mg, 80%).
N
-(4-Fluorobenzyl)-3-hydroxy-
N-(2-methylallyl)picolinamide (3p). Compound 3p was
prepared as the same procedure described for 3m, purified as colorless oil (126 mg, 88%). 1H
NMR (300 MHz, CDCl3)
7.31 (dd, = 8.4, 4.5 Hz, 4H), 7.01 (t,
4.82 (d, = 9.2 Hz, 2H), 4.70 (d,
ESI-MS: m/z 301.1 [M+H]+. Purity: 99.8% ( R
δ
12.10 (s, 2H), 8.06 (d,
J
= 7.9 Hz, 2H), 7.44 (d,
J
J
= 8.2 Hz, 2H),
= 17.3 Hz, 2H),
J
J
= 8.9 Hz, 4H), 5.29 (s, 2H), 4.94 (d,
J
J
= 8.9 Hz, 4H), 4.02 (s, 2H), 1.76 (s, 3H), 1.70 (s, 3H).
= 6.20 min).
t
3-(4-Fluorobenzyl)-2H-pyrido[2,3-e][1,3]oxazine-2,4(3H)-dione (4a). To the solution of 2a
(102 mg, 0.42 mmol) in anhydrous tetrahydrofuran (THF, 10 ml) was added
1,1'-carbonyldiimidazole (CDI, 135 mg, 0.84 mmol). The reaction mixture was refluxed for
24h. The excess THF was removed under reduced pressure and the residue was purified via
column chromatography using PE/EA = 1:1 afforded compound 4a as a white solid (103 mg,
26

ACCEPTED MANUSCRIPT
91%), m.p. 175-177 oC. 1H NMR (300 MHz, CDCl3)
7.57 (dd, = 8.3, 5.6 Hz, 2H), 7.01 (t, = 8.5 Hz, 2H), 5.21 (s, 2H). ESI-MS: m/z 273.1
[M+H]+. Purity: 90.1% ( R
= 6.00 min).
δ 8.75 (s, 1H), 7.66 (d, J = 2.7 Hz, 2H),
J
J
t
3-(Thiophen-2-ylmethyl)-2H-pyrido[2,3-e][1,3]oxazine-2,4(3H)-dione (4c). Compound 4c
was prepared as the same procedure described for 2a, purified as colorless oil (93 mg, 91%),
m.p. 176-178 oC. 1H NMR (300 MHz, CDCl3)
7.30 (d, = 3.5 Hz, 1H), 7.27 (s, 1H), 6.99-6.93 (m, 1H), 5.41 (s, 2H). ESI-MS: m/z 261.0
[M+H]+. Purity: 96.4% ( R
= 5.55 min).
δ 8.78 – 8.73 (m, 1H), 7.66 (d, J = 3.6 Hz, 2H),
J
t
Expression and purification of recombinant HIV-1 integrase and LEDGF/p75 proteins.
The plasmids for His-IN, GST-IN, and LEDGF/p75 were expressed in Escherichia coli
BL21(De3) pLysS expression strain (Invitrogen) after induction by IPTG (1 mM) at 0.75
optical density absorbance (595 nm). The culture was grown for an additional 3-4 h at 37 °C,
followed by centrifugation at 3000 rpm in a bucket rotor centrifuge (Beckman) for 20 min.
Cell lysis, Ni-affinity chromatography purification, and dialysis were performed as detailed
in the literature[38].
IN−LEDGF/p75 AlphaScreen. The AlphaScreen assay was performed according to the
manufacturer’s protocol (PerkinElmer, Benelux). Reactions were performed in a 25 ꢀL final
volume in 384-well Optiwell microtiter plates (Perkin Elmer). The reaction buffer contained
25 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1 mM MgCl2, 0.01% (v/v) Tween-20, and 0.1%
(w/v) bovine serum albumin. His6-tagged integrase (300 nM final concentration) was
incubated with compound at a final concentration of 20
ꢀM for 30 min at 4 °C. FLAG-tagged
LEDGF/p75 protein was then added at 300 nM final concentration, and the mixture was
incubated for an additional 60 min at 4 °C. Subsequently, 5 ꢀL of Ni-chelate-coated donor
27

ACCEPTED MANUSCRIPT
beads and 5
ꢀL of anti-FLAG antibody coated acceptor beads were added to a final
concentration of 20
ꢀg/mL of both beads. Proteins and beads were incubated for 1 h at 30 °C
in order to allow association to occur. Exposure of the reaction to direct light was omitted as
much as possible, and the emission of light from the acceptor beads was measured in the
EnVision plate reader (PerkinElmer) and analyzed using the EnVision manager software.
Quench Counterscreen. Quench counterscreening was performed identically to
IN−LEDGF/p75 AlphaScreen, but a Flag-6X-histidine construct was used in place of HIV-1
IN and LEDGF/p75.
AlphaScreen Dimerization Assay. IN extracts were dissolved in reaction buffer (150 mM
NaCl, 25 mM Tris-HCl pH 7.3, 1 mM MgCl2, 0.1% Tween 20, 0.1% BSA). A sample volume
of 25 µL was used in the 384 well format. GST-tagged WT IN was incubated in equimolar
concentration with 6xHis-tagged mutant IN at 4 °C overnight. The next morning,
glutathione-coated donor beads and Ni2+-coated acceptor beads (PerkinElmer) were added at
a final concentration of 1 mg/mL. The plate was incubated at 30 °C for one hour and then
scanned on an EnVision™ plate reader (PerkinElmer).
Preparation of Oligonucleotide Substrate. The HIV-1 IN catalytic activity assay uses a
21’mer top strand: (5
′-GTGTGGAAAATCTCTAGCAGT-3
′), and a 21’mer bottom strand:
(5′-ACTGCTAGAGATTTTCCACAC-3
′
). The top strand was labelled at the 5
′end with
γ
[32P]-ATP by T4 polynucleotide kinase (Epicenter, Madison, WI). The mixture was then
incubated at 95 °C for 15 min to inactivate the kinase and the bottom strand was added in 1.5
molar excess. The strands were allowed to anneal by cooling the mixture slowly to room
temperature. Any unincorporated material was subsequently removed by centrifuging the
mixture through a Spin-25 mini-column (USA Scientific, Ocala, FL).
Integrase Catalytic Activity Assay
analyzed by preincubating recombinant wild-type HIV-1 IN, at a final concentration of
200 nM, with the inhibitor in reaction buffer (50 mM NaCl, 1 mM HEPES, pH 7.5, 50
EDTA, 50 M dithiothreitol, 10% glycerol (w/v), 7.5 mM MnCl2, 0.1 mg/mL bovine serum
. The extent of 3′-processing and strand transfer was
ꢀM
ꢀ
28