A cell-based high-throughput screen to identify small molecular weight stimulators of the innate immune system revealed substituted pyrimido[5,4-b]indoles as potent NFκB activators. The most potent hit compound selectively stimulated Toll-like receptor 4 (TLR4) in human and mouse cells. Synthetic modifications of the pyrimido[5,4-b]indole scaffold at the carboxamide, N-3, and N-5 positions revealed differential TLR4 dependent production of NFκB and type I interferon associated cytokines, IL-6 and interferon γ-induced protein 10 (IP-10) respectively. Specifically, a subset of compounds bearing phenyl and substituted phenyl carboxamides induced lower IL-6 release while maintaining higher IP-10 production, skewing toward the type I interferon pathway. Substitution at N-5 with short alkyl substituents reduced the cytotoxicity of the leading hit compound. Computational studies supported that active compounds appeared to bind primarily to MD-2 in the TLR4/MD-2 complex. These small molecules, which stimulate innate immune cells with minimal toxicity, could potentially be used as adjuvants or immune modulators.
A cell-based high-throughput screen to identify small molecular weight stimulators of the innate immune system revealed substituted pyrimido[5,4-b]indoles as potent NFκB activators. The most potent hit compound selectively stimulated Toll-like receptor 4 (TLR4) in human and mouse cells. Synthetic modifications of the pyrimido[5,4-b]indole scaffold at the carboxamide, N-3, and N-5 positions revealed differential TLR4 dependent production of NFκB and type I interferon associated cytokines, IL-6 and interferon γ-induced protein 10 (IP-10) respectively. Specifically, a subset of compounds bearing phenyl and substituted phenyl carboxamides induced lower IL-6 release while maintaining higher IP-10 production, skewing toward the type I interferon pathway. Substitution at N-5 with short alkyl substituents reduced the cytotoxicity of the leading hit compound. Computational studies supported that active compounds appeared to bind primarily to MD-2 in the TLR4/MD-2complex. These small molecules, which stimulate innate immune cells with minimal toxicity, could potentially be used as adjuvants or immune modulators.
The innate immune response
is the first line of defense against
microbial pathogens such as viruses, bacteria, fungi, and protozoa.
A critical component of the innate immune response is the NFκB
family of transcription factors.[1,2] The Toll-like receptors
(TLRs) are critical components of the innate immune system that regulate
NFκB activation. In general, the TLRs recognize macromolecules
that are associated with pathogens and with cell stress. These pathogen-associated
molecular patterns (PAMPs) and their corresponding TLRs include: lipopeptides
(TLR2), double-stranded RNA (TLR3), lipopolysaccharide (LPS, TLR4),[3] bacterial flagellin (TLR5), guanine and uridine-rich
single-stranded RNA (TLR7, 8), and hypo-methylated CpG rich DNA (TLR9).[4] Innate immune cells use pattern recognition receptors
(PRRs) such as TLRs to promote a rapid response to perceived threats
before pathogen-specific adaptive immunity can be established. Indeed,
this rapid response, involving multiple components of the innate immune
system, has been recognized to guide the type of adaptive immune response
that is most effective for the specific pathogenic threat.The
discovery that certain small molecules could serve as specific
innate immune receptor ligands has expanded the possibility of designing
drugs that may modulate the immune response at its earliest stages.
Notable examples of such small molecules are the TLR7 agonists imiquimod,[5] isotorabine,[6] and
the more recent 8-oxo-9-benzyladenines,[7] as well as the TLR7/8 agonist resiquimod.[8] A useful application of this concept is the incorporation of adjuvants
into vaccines. Adjuvants are added to antigens in a vaccine setting
to provide enhanced response to poorly immunogenic antigens, to increase
seroconversion rates in populations with reduced responsiveness due
to age (infants and the elderly) or disease (diabetes, renal failure),
to facilitate the use of smaller doses of antigen, and to permit immunization
with fewer doses of vaccine.[9] Macromolecular
activators of the innate immune system, such as monophosphoryl lipid
A (MPLA), a TLR4 ligand, are utilized as adjuvants,[10] but development of small molecular weight nonlipid ligands
might have several advantages to address different immunological requirements
for vaccines or immune therapeutics.As part of our studies
on small molecules that can activate TLRs,
we conducted a high through-put screening (HTS) campaign in which
a commercially available library of compounds was screened in a human
cell-based NFκB activation assay. Many innate signaling pathways
converge on the transcription factor NFκB. Hence it was used
in the primary screen as a broad indicator of small molecule agonism
of innate immunity. The specific innate receptor for the lead compounds
was determined using genetically modified cells and primary mouse
and human blood or bone marrow mononuclear cells. Here we report our
discovery and initial structure–activity relationship (SAR)
studies of the pyrimido[5,4-b]indole class of ligands
and their characterization as TLR4/MD-2 agonists.
Results and Discussion
High-Throughput
Screen Study Design
A library of compounds
was acquired from the University of California, San Francisco, Small
Molecule Discovery Center consisting of about 170000 compounds from
eight suppliers (https://smdc.ucsf.edu). The library was
screened in three phases at a commercial HTS screening facility using
the THP-1human monocytic leukemia cell line, which contained a β-lactamase
reporter gene under the control of the NFκB response element
that had been stably integrated into the cells. All screens were performed
in activator mode using LPS as a positive control, achieving typical Z′ values above 0.75. The three-phase screening process
consisted of (1) a pilot screen of about 10000 compounds selected
as representative of the entire primary library, (2) the primary screen
of the entire library, and (3) a confirmation screen of about 2000
hits found in the primary screen. Compounds identified as active in
two screens were considered to be confirmed hits. An analysis of the
cluster enrichment methods for hit selection has been recently reported.[11]
Discovery of Pyrimido[5,4-b]indoles as Activators
of NFκB
Following the cluster enrichment analysis,
225 compounds were selected for further in vitro biological evaluation
involving cytokine induction assays in primary cells, including human
peripheral blood mononuclear cells (hPBMC), mouse splenocytes, mouse
bone marrow derived dendritic cells (mBMDC), and mouse bone marrow
derived macrophages (mBMDM). These cells were incubated in triplicate,
with each of the 225 compounds at a single concentration (1 μM
for splenocytes and 5 μM for all other mouse cells and human
cells), and the supernatants were tested for the presence of NFκB
dependent cytokines, IL-8 or IL-6, released from the human or mouse
cells, respectively. Thirty-nine of the 225 compounds stimulated the
human and mouse cells to secrete IL-8 or IL-6 above the detectable
limit. To further confirm activity, these compounds were repurchased
and retested by stimulating hPBMC and mBMDC with titrated doses and
assaying for IL-8 and IL-6.A few structurally diverse library
scaffolds were identified in these cytokine assays as having reproducible
responses in mouse and human stimulation assays. Among these scaffolds,
the pyrimido[5,4-b]indoles emerged as the most potent
and diverse class of compounds in the mouse cell assays with clear
evidence of SAR. Within this scaffold cluster, the leading hit from
the initial primary and secondary screens was a substituted acetamide
attached to the pyrimidoindole ring system through a thioether linkage
(Figure 1). Thus, compound 1 provided
a starting point for structure–activity studies.
Figure 1
Structure of
hit compound 1.
Structure of
hit compound 1.
Target Receptor Identification
To identify the target
receptor, we used HEK293 cells stably transfected with individual
human (h) TLRs: TLR2, TLR3, TLR4/MD-2/CD14, TLR5, TLR7, TLR8, or TLR9
expressing HEK293 cells, along with an NFκB activation reporter
producing secreted embryonic alkaline phosphatase (SEAP). Among the
tested TLR transfected cells, only those expressing TLR4/MD-2/CD14
responded to pyrimidoindoles, as shown in Figure 2A for compound 1. Because TLR4/MD-2/CD14 was
the receptor complex for the active compounds in this series, it was
important to rule out the possibility that activity might have been
caused by LPScontamination. Therefore, compound 1 (and
all active derivatives) was assayed for LPS (endotoxin) levels using
a commercially available detection system and found to contain less
than 10 endotoxin units (EU)/μmol compound. To further exclude
contamination, compound 1 was resynthesized according
to Scheme 1. Both samples of compound 1 displayed indistinguishable physicochemical and biological
properties, indicating that the positive biological activity was not
due to LPS or another contaminant. Full titration curves of the resynthesized
compound were performed using the SEAP assay with mouse and humanTLR4 transfected HEK293 cell lines, confirming dose-dependent activation
(Figure 2B, C).
Figure 2
Target identification
of compound 1 using human TLR
HEK293 reporter cell lines and genetically deficient cells. (A) Human
TLR2, TLR3, TLR4/MD-2/CD14, TLR5, TLR7, TLR8, and TLR9 HEK 293 Blue
cells or NF-κB/SEAPorter cells were incubated with compound 1 (10 μM) for 20–24 h, and activation was evaluated
by SEAP secretion in the culture supernatants by colorimetric assay
at OD405. Data shown are mean ± SEM of triplicates
and representative of two to three independent experiments showing
similar results. p < 0.05 was considered significant
compared to the vehicle control using Student’s t test. (B,C) Mouse (B) or human (C) TLR4 HEK transfectomas were incubated
with graded concentrations of compound 1. TLR4 mediated
NFκB activation was measured by SEAP secretion in the culture
supernatant. (D) WT and Tlr4–/– mBMDC were incubated with compound 1 (10 μM)
for18 h. IL-6 in the culture supernatants was measured by ELISA. (E)
Mouse TLR4 transfectoma cells were incubated with 2.5 μM compound 1 in the presence or absence of TLR4 antagonist LPS-RS (12,
111, 1000 ng/mL). Activation of the TLR4/NFκB pathway was evaluated
by SEAP secretion in the culture supernatants. * denotes p < 0.05 considered as significant compared to vehicle using one
way ANOVA with Dunnett’s post hoc testing. (F,G) WT and Cd14–/– mBMDC were incubated with
compound 1 (3.7 μM) overnight. IL-6 in the culture
supernatants was measured by ELISA (F) and type I IFN (IFN-β)
measured by luciferase release in from an ISRE reporter cell line
(G). p < 0.05 was considered as significant compared
to vehicle using Student’s t test. NS denotes
“not significant”. Data shown are mean ± SEM of
triplicates and representative of two independent experiments showing
similar results.
Target identification
of compound 1 using human TLR
HEK293 reporter cell lines and genetically deficient cells. (A) HumanTLR2, TLR3, TLR4/MD-2/CD14, TLR5, TLR7, TLR8, and TLR9HEK 293 Blue
cells or NF-κB/SEAPorter cells were incubated with compound 1 (10 μM) for 20–24 h, and activation was evaluated
by SEAP secretion in the culture supernatants by colorimetric assay
at OD405. Data shown are mean ± SEM of triplicates
and representative of two to three independent experiments showing
similar results. p < 0.05 was considered significant
compared to the vehicle control using Student’s t test. (B,C) Mouse (B) or human (C) TLR4HEK transfectomas were incubated
with graded concentrations of compound 1. TLR4 mediated
NFκB activation was measured by SEAP secretion in the culture
supernatant. (D) WT and Tlr4–/– mBMDC were incubated with compound 1 (10 μM)
for18 h. IL-6 in the culture supernatants was measured by ELISA. (E)
MouseTLR4 transfectoma cells were incubated with 2.5 μM compound 1 in the presence or absence of TLR4 antagonist LPS-RS (12,
111, 1000 ng/mL). Activation of the TLR4/NFκB pathway was evaluated
by SEAP secretion in the culture supernatants. * denotes p < 0.05 considered as significant compared to vehicle using one
way ANOVA with Dunnett’s post hoc testing. (F,G) WT and Cd14–/– mBMDC were incubated with
compound 1 (3.7 μM) overnight. IL-6 in the culture
supernatants was measured by ELISA (F) and type I IFN (IFN-β)
measured by luciferase release in from an ISRE reporter cell line
(G). p < 0.05 was considered as significant compared
to vehicle using Student’s t test. NS denotes
“not significant”. Data shown are mean ± SEM of
triplicates and representative of two independent experiments showing
similar results.
AUC values normalized
to compound 1. Stimulation with LPS (10 ng/mL), and 5
μM of 1 released an average of 20.6 ± 4.8
and 10.5 ± 1.3
ng/mL of IL-6, respectively.
All compounds tested at 5 μM,
values normalized to 1. Stimulation with LPS 10 ng/mL
and 5 μM of 1 released averages of 234.8 ±
8.6 and 590 ± 11 pg/mL of IP-10, respectively.
AUC values normalized to compound 1. TLR4 cell activation stimulated with 10 ng/mL LPS and 5
μM compound 1 resulted in OD630 of 1.69
± 0.08 and 1.52 ± 0.03, respectively.
All compounds tested at 10 μM,
values normalized to 1. Stimulation with LPS 10 ng/mL
and 10 μM of 1 released averages of 14.5 ±
0.6 and 7.8 ± 0.8 ng/mL of IL-8, respectively.
All compounds tested at 10 μM,
values normalized to 1. TLR4 cell activation stimulated
with 10 ng/mL LPS and 10 μM compound 1 resulted
in OD630 of 1.90 ± 0.04 and 0.88 ± 0.02, respectively.
To confirm that TLR4 was indeed the receptor, compound 1 was assayed for IL-6 production in mBMDCs from wild-type
and TLR4
deficient mice (Figure 2D). Genetic disruption
of TLR4completely abrogated IL-6 secretion induced by compound 1. The binding of compound 1 to the TLR4/MD-2/CD14complex was further confirmed using a competitive antagonist for the
TLR4 binding complex, LPS-RS (LPS from Rhodococcus
sphaeroides).[12] LPS-RS
inhibited the activation by compound 1 in a dose-dependent
manner, indicating that compound 1 bound to the TLR4complex (Figure 2E).TLR4 signals through
two distinct pathways, leading respectively
to NFκB dependent cytokines and type I interferon (IFN) production.
Several naturally occurring TLR4 ligands and MPLA have been reported
to require CD14 to activate the type I IFN regulatory pathway.[13]The hTLR4 transfected HEK293 cell line
(Figure 2A) also overexpresses MD-2 and CD14,
which are TLR4 accessory
proteins.[14,15] Compound 1, however, was not
dependent on CD14 for either IL-6 or type I IFN production, as demonstrated
using CD14 deficient cells (Figure 2F, G).
The supernatants from mBMDCs stimulated with graded doses of compound 1 were also tested for IP-10 as a surrogate marker of type
I IFN release. Results showed a dose-dependent response for type I
IFN (Figure 3A) production, which paralleled
that of IP-10 (Figure 3B).
Figure 3
Type I IFN induction
by compound 1. Wild-type mBMDCs
were incubated with graded concentrations of compound 1 overnight. DMSO 0.5% served as the vehicle control. The levels of
type I IFN were determined using an L929-ISRE luciferase reporter
cell line and a mIFNβ standard (A). IP-10 levels were determined
by ELISA (B). Data shown are mean ± SEM of triplicates and representative
of two independent experiments showing similar results. * denotes p < 0.05 compared to vehicle using one way ANOVA with
Dunnett’s post hoc testing.
Type I IFN induction
by compound 1. Wild-type mBMDCs
were incubated with graded concentrations of compound 1 overnight. DMSO 0.5% served as the vehicle control. The levels of
type I IFN were determined using an L929-ISRE luciferase reporter
cell line and a mIFNβ standard (A). IP-10 levels were determined
by ELISA (B). Data shown are mean ± SEM of triplicates and representative
of two independent experiments showing similar results. * denotes p < 0.05 compared to vehicle using one way ANOVA with
Dunnett’s post hoc testing.The above assays were utilized to compare the derivatives
of the
lead pyrimidoindole for their ability to activate mouse and humanTLR4. MouseTLR4 activation was assessed using primary mBMDC and IL-6
release (Figure 4) and confirmed using mTLR4
transfected HEK293 cells (Figure 5A, C). The
values shown in the Tables 1–3 for IL-6 release, and mTLR4 activation are area
under the curve (AUC) values for titrated doses of compounds from
312 nM to 10 μM. Each cytokine induction curve was first converted
to a percent activity curve, and then the AUC of the percent activity
curve was calculated. The process of converting to a percent activity
curve allowed subtracting background and adjusting for plate-to-plate
variation. Finally, the AUC values were normalized to the activity
of compound 1 within each experiment, set at 100. HumanTLR4 activation is shown for stimulation of PBMC (IL-8) and hTLR4HEK293 transfectomas (Figure 5B,D) at 10 μM,
as these assays were not sufficiently sensitive at lower concentrations
to make AUC comparisons. The levels of hTLR4 activation by SAR derivatives
are expressed relative to compound 1, set at 100.
Figure 4
Representative
data of SAR compound screening using mouse primary
dendritic cells. Biological screening of SAR compounds was conducted
using primary mBMDC. The cells were incubated with graded concentrations
of the indicated compounds 1, 36, and 39 (A) or 11 and 12 (B) for 18 h.
DMSO 0.5% served as the vehicle control. IL-6 levels in culture supernatants
were measured by ELISA. Data shown are mean ± SEM of triplicate.
Figure 5
Representative data of SAR compound screening
using TLR4 transfectomas.
Mouse TLR4 (A,C), and hTLR4 (B,D) HEK transfectomas were incubated
with graded concentrations of the indicated compounds 1, 36, and 39 (A) or 11 and 12 (B) for 18 h. DMSO 0.5% served as the vehicle control.
The specific activation of the reporter cell lines was measured by
SEAP activity in the supernatant by absorption at 630 nm. Data shown
are mean ± SEM of triplicate data.
Table 3
N-5 Derivatives
AUC values normalized
to compound 1. Stimulation with LPS (10 ng/mL) and 5
μM of 1 released an average of 20.6 ± 4.8
and 10.5 ± 1.3
ng/mL of IL-6, respectively.
All compounds tested at 5 μM,
values normalized to 1. Stimulation with LPS 10 ng/mL
and 5 μM of 1 released averages of 234.8 ±
8.6 and 590 ± 11 pg/mL of IP-10, respectively.
AUC values normalized to compound 1. TLR4 cell activation stimulated with 10 ng/mL LPS and 5
μM compound 1 resulted in OD630 of 1.69
± 0.08 and 1.52 ± 0.03, respectively.
All compounds tested at 10 μM,
values normalized to 1. Stimulation with LPS 10 ng/mL
and 10 μM of 1 released averages of 14.5 ±
0.6 and 7.8 ± 0.8 ng/mL of IL-8, respectively.
All compounds tested at 10 μM,
values normalized to 1. TLR4 cell activation stimulated
with 10 ng/mL LPS and 10 μM compound 1 resulted
in OD630 of 1.90 ± 0.04 and 0.88 ± 0.02, respectively.
Representative
data of SAR compound screening using mouse primary
dendritic cells. Biological screening of SAR compounds was conducted
using primary mBMDC. The cells were incubated with graded concentrations
of the indicated compounds 1, 36, and 39 (A) or 11 and 12 (B) for 18 h.
DMSO 0.5% served as the vehicle control. IL-6 levels in culture supernatants
were measured by ELISA. Data shown are mean ± SEM of triplicate.Representative data of SAR compound screening
using TLR4 transfectomas.
MouseTLR4 (A,C), and hTLR4 (B,D) HEK transfectomas were incubated
with graded concentrations of the indicated compounds 1, 36, and 39 (A) or 11 and 12 (B) for 18 h. DMSO 0.5% served as the vehicle control.
The specific activation of the reporter cell lines was measured by
SEAP activity in the supernatant by absorption at 630 nm. Data shown
are mean ± SEM of triplicate data.
SAR Studies
The primary HTS library contained a family
of 452 compounds in the pyrimido[5,4-b]indole class
and therefore represented a valuable initial indication of structural
features important for activation of NFκB. Upon inspection of
the NFκB activation values in the initial HTS relative to the
various structural features, several trends were apparent. There was
a requirement for a hydrophobic moiety in the region of the cyclohexyl
group of compound 1. In this same region, the carboxamide
function was essential. Furthermore, a hydrophobic group at N-3, preferably
a phenyl group, was also required for activity. Substitutions on the
N-3 phenyl, other than fluorine atoms, resulted in loss of activity.
Removal of the benzo ring of the indole portion of the scaffold also
resulted in loss of activity. When the C-4 oxo was replaced by NH
to form a triazine ring, loss of activity was observed. Finally, exchange
of the entire acetamide moiety on the C-2thiol with the N-3 phenyl
group, such that the acetamide was attached at N-3 and the phenyl
was attached at the C-2thiol, resulted in loss of activity. With
these SAR features as an initial guide, we elected to investigate
modifications of hit compound 1 at three regions while
maintaining the core pyrimido[5,4-b]indole ring system:
the N-substitution of the S-acetamide, the N-3 substituent,
and the N-5 substituent (Figure 6).
Figure 6
SAR regions
of modification of hit compound 1.
SAR regions
of modification of hit compound 1.N-Substitutions at the carboxamide moiety were first undertaken
to probe the limitations of the hydrophobic group requirement at this
position with respect to optimization of cytokine induction (Table 1). While keeping all
other structural features of hit compound 1 constant,
we prepared a series of carboxamides substituted with various alkyl,
cycloalkyl, aromatic, and heteroaromatic groups. The synthesis began
with construction of the appropriately substituted pyrimido[5,4-b]indole ring system as shown in Scheme 1. 2-Aminobenzonitrile (2) was reacted with ethyl
bromoacetate to yield ethyl 2-((2-cyanophenyl)amino)acetate (3) followed by base catalyzed ring closure to the aminoindole[16,17] (4). At this point in the synthesis, a variety of substituents,
represented by R1 in the scheme, may be introduced that
will determine the respective N-3 substituent following annulation
of the pyrimidine ring. Thus, reaction of compound 4 with
an isothiocyanate, such as phenylisothiocyanate, provided the substituted
thioureidoindole[16,18] (5), where R1was phenyl in this example. Ring closure of 5 using polyphosphoric acid yielded the pyrimido[5,4-b]indole (6) bearing the N-3 substituent R1. Alkylation of the 2-thioxo function with chloroacetic acid provided
the versatile intermediate 7, which was then used to
prepare the final test compounds (8) bearing a variety
of N-substitutions at the acetamide moiety, designated as R2 in Scheme 1. Compounds 1, 13, 14, 15, 16, 17, as well as the N-3 derivatives 37 and 38, discussed later, were registered with Chemical Abstracts
Service, but no literature references were found for any of these
compounds. The hit compound 1 was resynthesized according
to Scheme 1, as discussed above. Thus, by this
synthetic method, compounds shown in Table 1, with R1 held constant as a phenyl group, were prepared
and evaluated in a cytokine induction assay for IL-6 in mBMDC and
compared to compound 1. Examples of the data obtained
from typical cytokine induction assays and NFκB activation assays
are shown in Figures 4 and 5, wherein compounds 1, 11, 12, 36, and 39 are compared for
IL-6 production by mBMDC and hPBMC.AUC values normalized
to compound 1. Stimulation with LPS (10 ng/mL), and 5
μM of 1 released an average of 20.6 ± 4.8
and 10.5 ± 1.3
ng/mL of IL-6, respectively.All compounds tested at 5 μM,
values normalized to 1. Stimulation with LPS 10 ng/mL
and 5 μM of 1 released averages of 234.8 ±
8.6 and 590 ± 11 pg/mL of IP-10, respectively.AUC values normalized to compound 1. TLR4 cell activation stimulated with 10 ng/mL LPS and 5
μM compound 1 resulted in OD630 of 1.69
± 0.08 and 1.52 ± 0.03, respectively.All compounds tested at 10 μM,
values normalized to 1. Stimulation with LPS 10 ng/mL
and 10 μM of 1 released averages of 14.5 ±
0.6 and 7.8 ± 0.8 ng/mL of IL-8, respectively.All compounds tested at 10 μM,
values normalized to 1. TLR4 cell activation stimulated
with 10 ng/mL LPS and 10 μM compound 1 resulted
in OD630 of 1.90 ± 0.04 and 0.88 ± 0.02, respectively.Inspection of the IL-6 AUC
values relative to the hydrophobic group
R2 revealed that in general, compounds bearing the larger
cycloalkyl groups, such as cyclooctyl (9), cycloheptyl
(10), and cyclohexyl (1), were the most
active, followed by branched alkyls and then straight chain alkyls
and aromatic and heteroaromatic groups. A notable exception would
be the p-fluorophenyl (14) and o-fluorophenyl (15) compounds. For the R2 group, there appeared to be a strict “hydrophobic
volume” requirement for activity. Hence, the 3,3-dimethylbutylcompound (28) was among the most active of the alkyls,
with isopentyl (27), butyl (23), and isobutyl
(26) being somewhat less active. Interestingly, when
the alkyl chain length of R2 was extended to more than
five carbons or reduced to three or fewer carbons, whether branched
or not, significant loss of activity was observed (Table 1).Encouraged by the SAR trends at the carboxamide
moiety, we then
addressed the N-3 position (Table 2). As mentioned above, data from the NFκB
primary screen indicated that a hydrophobic group at N-3, preferably
an unsubstituted phenyl, was required for activity. However, to draw
that conclusion with confidence, more examples of compounds bearing
similar hydrophobic groups than were represented in the primary compound
collection were needed. We therefore elected to prepare and evaluate
a few additional derivatives of hit compound 1 with variation
only at N-3. Thus, the R1 group in compound 5 was varied by reaction of 4 with the appropriate isothiocyanate,
and the remaining three steps for each derivative were completed as
outlined in Scheme 1 while maintaining the
R2 group as cyclohexyl in each case. Table 2 compares the IL-6 inducing activity of compound 1 with the N-3 derivatives. As the IL-6 data indicate, replacing the
N-3 phenyl in compound 1 with a cyclohexyl group (36) caused some loss of activity, while substitution on the
N-3 phenyl (37 and 38) resulted in greater
loss of activity. Replacing N-3 phenyl with a larger aromatic, such
as naphthyl (39), or extending the phenyl with an alkyl
chain, as in phenethyl (40), yielded total loss of activity.
The combined substitution of R1 with cyclohexyl and R2 with cyclopentyl (41) resulted in partial loss
of activity.
Table 2
N-3 Derivatives
AUC values normalized to compound 1. Stimulation with LPS (10 ng/mL) and 5 μM of 1 released an average of 20.6 ± 4.8 and 10.5 ± 1.3
ng/mL of IL-6, respectively.
All compounds tested at 5 μM,
values normalized to 1. Stimulation with LPS 10 ng/mL
and 5 μM of 1 released averages of 234.8 ±
8.6 and 590 ± 11 pg/mL of IP-10, respectively.
AUC values normalized to compound 1. TLR4 cell activation stimulated with 10 ng/mL LPS and 5
μM compound 1 resulted in OD630 of 1.69
± 0.08 and 1.52 ± 0.03, respectively.
All compounds tested at 10 μM,
values normalized to 1. Stimulation with LPS 10 ng/mL
and 10 μM of 1 released averages of 14.5 ±
0.6 and 7.8 ± 0.8 ng/mL of IL-8, respectively.
All compounds tested at 10 μM,
values normalized to 1. TLR4 cell activation stimulated
with 10 ng/mL LPS and 10 μM compound 1 resulted
in OD630 of 1.90 ± 0.04 and 0.88 ± 0.02, respectively.
AUC values normalized to compound 1. Stimulation with LPS (10 ng/mL) and 5 μM of 1 released an average of 20.6 ± 4.8 and 10.5 ± 1.3
ng/mL of IL-6, respectively.All compounds tested at 5 μM,
values normalized to 1. Stimulation with LPS 10 ng/mL
and 5 μM of 1 released averages of 234.8 ±
8.6 and 590 ± 11 pg/mL of IP-10, respectively.AUC values normalized to compound 1. TLR4 cell activation stimulated with 10 ng/mL LPS and 5
μM compound 1 resulted in OD630 of 1.69
± 0.08 and 1.52 ± 0.03, respectively.All compounds tested at 10 μM,
values normalized to 1. Stimulation with LPS 10 ng/mL
and 10 μM of 1 released averages of 14.5 ±
0.6 and 7.8 ± 0.8 ng/mL of IL-8, respectively.All compounds tested at 10 μM,
values normalized to 1. TLR4 cell activation stimulated
with 10 ng/mL LPS and 10 μM compound 1 resulted
in OD630 of 1.90 ± 0.04 and 0.88 ± 0.02, respectively.The third area of interest
for modification was the N-5 position.
None of the compounds in the original HTS library had modifications
at this position to guide the SAR. We elected to prepare a few simple
N-5 alkyl derivatives of compound 1. The N-5 indole-like
nitrogen can be alkylated if the proton is first removed by a strong
base, such as sodium hydride. Starting from compound 1, Scheme 2 shows the preparation of the N-5
methyl derivative (42) that also produced a dimethyl
side product (43) wherein the second methylation occurred
at the carboxamide function, yielding the N,N-disubstituted derivative. The N-5 methyl butyrate ester
(44) of compound 1 was also prepared by
this method. Table 3 compares the IL-6 inducing activity of the N-5 derivatives relative
to compound 1. Interestingly, simple methylation at N-5
did not abrogate the immune stimulatory activity of compound 1. Moreover, methylation at N-5 decreased the toxicity of
compound 1 (Figure 7) as measured
by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)
viability assay. Dimethylation, on the other hand, resulted in total
loss of IL-6 activity (compound 43). To further probe
the N-5 position, we elected to study a few N-5 alkylated derivatives
of compound 1 while avoiding dialkylation in the process.
Accordingly, we prepared the t-butyl ester of 7a (compound 45) and then alkylated the N-5 position
with several primary alkyl halides, as depicted in Scheme 3. The resulting t-butyl esters
(46a–d) were easily hydrolyzed to
the corresponding free carboxylic acids (47a–d) and then converted to the N-cyclohexylcarboxamides
(48–51) by the same method as described
for compound 1. Thus, using this strategy, the N-5 n-propyl, n-pentyl, n-dodecyl,
and cyanomethyl derivatives were prepared. Finally, the cyanomethyl
derivative (51) was converted to the N-5 acetamide derivative
(52). All N-5 alkyl derivatives were found to be less
active than the N-5 methyl (42), with the n-propyl (48) being the next most active of the series.
The trend in the toxicity profile for these N-5 alkyl derivatives
was confirmed, at least for the shorter chain alkyls, in that the
N-5 methyl (42), cyanomethyl (51), and n-propyl (48) derivatives did not reduce cell
viability at concentrations up to 10 μM (Figure 7, shown for 42, data not shown for 48 and 51).
Scheme 2
N-5 Derivatives
(a)
NaH, DMF, room temp, then
CH3I; (b) NaH, DMF, then Br(CH2)3COOCH3
Figure 7
Assessment of cytotoxicity of compounds 1 and 42. Cytotoxicity of the compounds was evaluated
by MTT assay
as a measure of viability. mBMDC (105/well) were incubated
with graded concentrations of the compounds for 18 h and compared
with vehicle (0.5% DMSO) and MPLA (1 μg/mL). (A) IL-6 levels
in the culture supernatants were determined by ELISA. (B) The cells
were lysed after the overnight incubation with MTT reagents, and absorbance
at 570 nm was measured, subtracting the reference absorbance at 650
nm. Data shown are mean ± SEM of triplicates and are representative
of two independent experiments showing similar results.
Scheme 3
N-5 Alkyl Derivatives
(a) ClCH2COO t-Bu, KOH, DMA,H2O; (b) NaH, DMF, then R3X (iodopropane or iodopentane or bromododecane or bromoacetonitrile);
(c) TFA, DCM or CH3CN; (d) cyclohexyl-NH2, HATU,
DMF, room temp; (e) compound 51, H2SO4, H2O.
AUC values normalized
to compound 1. Stimulation with LPS (10 ng/mL) and 5
μM of 1 released an average of 20.6 ± 4.8
and 10.5 ± 1.3
ng/mL of IL-6, respectively.All compounds tested at 5 μM,
values normalized to 1. Stimulation with LPS 10 ng/mL
and 5 μM of 1 released averages of 234.8 ±
8.6 and 590 ± 11 pg/mL of IP-10, respectively.AUC values normalized to compound 1. TLR4 cell activation stimulated with 10 ng/mL LPS and 5
μM compound 1 resulted in OD630 of 1.69
± 0.08 and 1.52 ± 0.03, respectively.All compounds tested at 10 μM,
values normalized to 1. Stimulation with LPS 10 ng/mL
and 10 μM of 1 released averages of 14.5 ±
0.6 and 7.8 ± 0.8 ng/mL of IL-8, respectively.All compounds tested at 10 μM,
values normalized to 1. TLR4 cell activation stimulated
with 10 ng/mL LPS and 10 μM compound 1 resulted
in OD630 of 1.90 ± 0.04 and 0.88 ± 0.02, respectively.Assessment of cytotoxicity of compounds 1 and 42. Cytotoxicity of the compounds was evaluated
by MTT assay
as a measure of viability. mBMDC (105/well) were incubated
with graded concentrations of the compounds for 18 h and compared
with vehicle (0.5% DMSO) and MPLA (1 μg/mL). (A) IL-6 levels
in the culture supernatants were determined by ELISA. (B) The cells
were lysed after the overnight incubation with MTT reagents, and absorbance
at 570 nm was measured, subtracting the reference absorbance at 650
nm. Data shown are mean ± SEM of triplicates and are representative
of two independent experiments showing similar results.
N-5 Derivatives
(a)
NaH, DMF, room temp, then
CH3I; (b) NaH, DMF, then Br(CH2)3COOCH3
N-5 Alkyl Derivatives
(a) ClCH2COO t-Bu, KOH, DMA,H2O; (b) NaH, DMF, then R3X (iodopropane or iodopentane or bromododecane or bromoacetonitrile);
(c) TFA, DCM or CH3CN; (d) cyclohexyl-NH2, HATU,
DMF, room temp; (e) compound 51, H2SO4, H2O.The TLR4 transfectoma
lines transmit their reporter signal after
activation of the proinflammatory transcription factor, NFκB.
As one might expect, the IL-6 secretion induced by the compounds in
primary mBMDCs correlated well with their respective activities in
mouseTLR4 transfectoma cell lines. With few exceptions, the most
active compounds in the IL-6 assays were also those that showed the
highest activity in the mTLR4 transfectoma line (Tables 1–3). In addition to stimulating
the production of inflammatory cytokines (e.g., IL-6) by NFκB
activation, TLR4 signaling can induce the production of type I IFN
via the TRIF pathway.[19,20] This pathway results in the activation
of interferon regulatory factors (IRF), which are nuclear transcription
factors that promote the transcription of IFNs. Type I IFN is important
for the activation of antigen presenting dendritic cells, leading
to good adjuvant activity, and also promotes cellular defenses against
a variety of pathogens, particularly RNA viruses. This host defense
is aided by circulating interferon inducible factors such as IP-10.
The TLR4 transfectoma reporter lines in this study did not utilize
an interferon reporter system; hence, the IP-10 data does not correlate
as closely as the IL-6 data.Interestingly, several of the compounds,
namely 12, 13, 15, 16, 17, 29, 30, and 33, induced
diminished IL-6 release compared to compound 1 but had
nearly equivalent induction of IP-10 (Tables 1 and 2). These results suggest that there
may be a possibility of separating the two activities on the basis
of structural features. The correlation of IL-6 and IP-10 production
of the selected compounds is shown graphically in Figure 8. Two distinct groups of compounds are observed,
shown as circled regions of the plot. Thus, the group at the upper
right area (dotted circle) comprises compounds that induce high IL-6
and high IP-10 production, including compounds 1, 9, 10, 28, and 42.
The other group is composed of compounds that induce low IL-6 but
relatively higher IP-10 production. These two clusters were selected
based on mean ± SD of the normalized IL-6 and IP-10 values of
all SAR compounds. For IL-6, the mean was 32 ± 37. For IP-10,
the mean was 37 ± 33. The cluster in the dotted circle includes
those compounds whose values were at least one SD above the mean for
both cytokines. The other cluster includes those compounds whose values
were below the mean for IL-6 and were above the mean for IP-10. It
was notable that when the carboxamide substituent was phenyl or substituted
phenyl (13, 15, 16, 17), with the exception of p-fluorophenyl (14), IL-6 release was reduced while IP-10 production was maintained
at a relatively higher level. A few compounds bearing branched aliphatic
carboxamide substituents showed a similar trend, as noted above for 29, 30, and 33.
Figure 8
Correlation plot of IL-6
versus IP-10 induced by SAR derivatives.
IL-6 (AUC) and IP-10 values in Tables 1–3 were plotted. Two distinct groups, compounds in
the dotted circled (upper right) and in solid (left) circled areas,
were selected based on the induction of IL-6 and IP-10. Numbers in
the legends indicate compound ID. These clusters were selected based
on mean ± SD of the normalized IL-6 and IP-10 values of all SAR
compounds. For IL-6. the mean ± SD was 32 ± 37. For IP-10,
the mean ± SD was 37 ± 33. The group at the upper right
area comprises compounds that induce high IL-6 and high IP-10 production,
including compounds 1, 9, 10, 28, and 42, whose values were at least
one SD above the mean for both cytokines. The other cluster is composed
of compounds (13, 16, 17, 29, 30, and 33) that induce low
IL-6 but relatively higher IP-10 production, whose values were below
the mean for IL-6 and were above the mean for IP-10.
Correlation plot of IL-6
versus IP-10 induced by SAR derivatives.
IL-6 (AUC) and IP-10 values in Tables 1–3 were plotted. Two distinct groups, compounds in
the dotted circled (upper right) and in solid (left) circled areas,
were selected based on the induction of IL-6 and IP-10. Numbers in
the legends indicate compound ID. These clusters were selected based
on mean ± SD of the normalized IL-6 and IP-10 values of all SAR
compounds. For IL-6. the mean ± SD was 32 ± 37. For IP-10,
the mean ± SD was 37 ± 33. The group at the upper right
area comprises compounds that induce high IL-6 and high IP-10 production,
including compounds 1, 9, 10, 28, and 42, whose values were at least
one SD above the mean for both cytokines. The other cluster is composed
of compounds (13, 16, 17, 29, 30, and 33) that induce low
IL-6 but relatively higher IP-10 production, whose values were below
the mean for IL-6 and were above the mean for IP-10.In general, the active compounds exhibited greater
activity in
the murine cell systems than in human cells. Although the innate immune
system is highly conserved among species, there are differences between
human and mouseTLR4.[21−23] Similar to the active compounds reported here nonsynthetic
ligands have been noted to have species specific behavior. Notably,
tetraacylated lipidIVa, a synthetic lipid A precursor, has been reported
to act as a weak agonist to mouseTLR4/MD-2 but as an antagonist to
humanTLR4/MD-2.[24,25] Recently mouse and humanTLR4/MD2
crystal structures with this ligand have suggested that the different
charge distributions of mouse and humanTLR4/MD-2 affect the positions
of the phosphate groups of lipidIVa, orienting them in a manner that
would limit receptor dimerization by humanTLR4.[22,26]
Computational Studies
TLR4, in association with MD-2,
is responsible for the physiological recognition of LPS.[27,28] The structural basis of receptor specificity and of the mechanism
of activation by LPS have recently been elucidated by determining
the crystal structure of the TLR4/MD-2-LPScomplex at 3.1 A resolution.[29] Binding of agonistic ligands such as LPS causes
dimerization of the extracellular domains to form a TLR4/MD-2-LPS
macromolecular complex. Like the extracellular domains of other TLRs,
TLR4contains leucine-rich repeats and adopts a characteristic horseshoe-like
shape. MD-2 is noncovalently bound to the side of the horseshoe ring
and also directly interfaces with the ligand. MD-2 has a β-cup
fold structure composed of two antiparallel β-sheets forming
a large hydrophobic pocket for ligand binding. LPS binds to this pocket
and directly mediates dimerization of the two TLR4/MD-2complexes.
TLR4 can be activated by structurally diverse LPS molecules, which
have been predicted to occupy this pocket in MD-2.Small molecule
hit compounds discovered in the present study were also thought to
bind to the TLR4/MD-2complex in such a way as to facilitate dimerization.
Accordingly, we examined the predicted binding mode(s) of one of the
most active compounds to the murineTLR4/MD-2complex by conducting
molecular docking of compound 28 to the crystal structure
of the mousecomplex (PDB 2Z64) using the programs HEX[30] and AMPAC.[31] We selected the best configurations
of 28 bound to this complex based on molecular surface
shape complementarity and the most favorable intermolecular energy
of interactions. It is noteworthy that the best docking position for 28 was within the LPS pocket.[29] Figure 9 shows the predicted binding mode
of compound 28 in the TLR4/MD-2 model, while the set
of binding interactions that may keep the compound in the MD-2 pocket
bound to both TLR4 and MD-2 is depicted in Figure 10. There is a set of favorable electrostatic interactions resulting
in possible hydrogen bonds formed by the residues Glu439 of TLR4 and
Arg90 of MD-2 with compound 28, although the former would
approach a hydrogen bond interaction due to the flexibility of the
Glu439 side chain. There are also multiple potential hydrophobic interactions
with MD-2 and TLR4. Such interactions of the compound with two proteins
can improve the free energy of complex formation by approximately
8–10 kcal/mol.
Figure 9
Predicted binding mode of compound 28 to
mouse TLR4/MD-2
complex.
Figure 10
Predicted binding interactions of compound 28 with
mouse TLR4/MD-2 complex (PDB 2Z64).
Predicted binding mode of compound 28 to
mouseTLR4/MD-2complex.Predicted binding interactions of compound 28 with
mouseTLR4/MD-2complex (PDB 2Z64).Interestingly, the predicted
binding model for the pyrimidoindoles
is similar to that proposed for tricyclic antidepressant, amitriptyline.[32] This molecule also has some TLR4/MD-2 binding
and has a similar three-ring scaffold. Other small molecules such
as paclitaxel,[33] opioids,[34] and a peptide[35] have also been
reported to bind to the MD-2 binding pocket.
Conclusion
In the course of an HTS designed to identify activators of innate
immunity, a series of substituted pyrimido[5,4-b]indoles
were discovered as selective TLR4 ligands. Small molecules of this
class are unique among TLR4 activators in that they are “non-lipid-like”.
Structure–activity evaluation in both mouse and human cells
revealed that, to maintain activity, the carboxamide region of this
scaffold must contain a hydrophobic moiety of significant “volume”.
Interestingly, a subset of the compounds bearing phenyl and substituted
phenyl carboxamides induced lower NFκB-dependent inflammatory
cytokine release while maintaining interferon-dependent IP-10 production.
Varying the substituents at N-3 indicated an even greater restriction
for a hydrophobic moiety at this position, with a phenyl group being
preferred for activity. Finally, N-5 substitution revealed that short
alkyl substituents at this position attenuate cell toxicity relative
to the corresponding nonsubstituted derivative while maintaining TLR4
activity. Both the inflammatory cytokine and type I IFN inducing activities
of these compounds were CD14-independent. Computational studies with
one of the active compounds predicted binding primarily to MD-2 in
the murineTLR4/MD-2complex. Lead optimization studies to improve
the activity of the compounds using computational methods are currently
underway in our laboratories. To date, no single adjuvant can elicit
the optimal immune response to all pathogens in vaccines. Here we
have described a panel of small molecules that stimulate immune cells
to produce distinct profiles of NFκB and interferon associated
cytokines. This panel of molecules may allow for differential analysis
of the relative pathway induction needed for adequate immunoprotection
and immunotherapy for diverse pathogens.
Experimental
Section
Chemistry
Materials
Reagents were purchased
as at least reagent
grade from Sigma-Aldrich (St. Louis, MO) unless otherwise specified
and used without further purification. Solvents were purchased from
Fischer Scientific (Pittsburgh, PA) and were either used as purchased
or redistilled with an appropriate drying agent. HTS compound library
was obtained from UCSF Small Molecule Discovery Center (San Francisco,
CA). Commercial HTS service was provided by Invitrogen (Grand Island,
NY). Compounds 13, 14, 15, 16, 17, 37, and 38 were
purchased from Life Chemicals (Burlington, ON, Canada). All synthesized
compounds, intermediates, and purchased compounds from Life Chemicals
were determined to be >95% pure by HPLC utilizing an Agilent 1100
LC/MSD. Endotoxin levels of active compounds were measured with Endosafe-PTS
(Charles River, Wilmington, MA) and found to have less than 10 EU/μmol.
Reaction room temperature was maintained between 22 and 24 °C.
Instrumentation
Analytical TLC was performed using
precoated TLC silica gel 60 F254 aluminum sheets purchased
from EMD (Gibbstown, NJ) and visualized using UV light. Flash chromatography
was carried out on EMD silica gel 60 (40–63 μm) system
or with a Biotage Isolera One (Charlotte, NC) using the specified
solvent. Reactions were monitored using an Agilent 1100 LC/MSD (Santa
Clara, CA) with either a Supelco Discovery HS C18column (Sigma-Aldrich)
or an Onyx Monolithic C18 (Phenomenex, Torrance, CA) with purity above
98% by percent area. All synthesized compounds and intermediates were
analyzed by high-resolution MS using an Agilent 6230 ESI-TOFMS (Santa
Clara, CA). Selected compounds were analyzed by IR (KBr window method)
to confirm presence of functional groups having useful diagnostic
frequencies using a Perkin-Elmer 1600 series. 1H NMR spectra
were obtained on a Varian Mercury 300 or 400 (Varian, Inc., Palo Alto,
CA). 13C spectra were obtained on Varian 500 with Xsens
probe. The chemical shifts are expressed in parts per million (ppm)
using suitable deuterated NMR solvents in reference to TMS at 0 ppm.
The 3D structures were prepared and optimized using the AMPAC semiempirical
quantum chemistry program (Accelrys, San Diego, CA).
General Procedure
A for the Synthesis of Compound 5
To a solution
of compound 4 (1 equiv) in warm
EtOH was added the appropriate isothiocyanate (1.1 equiv) dropwise
with stirring. The reaction was refluxed for 6 h and cooled overnight.
Solids were filtered, washed with EtOH, and dried overnight in vacuo
to give compound 5.
General Procedure B for
the Synthesis of Compound 6b–d
Compound 5 was dissolved
in polyphosphoric acid and stirred at 110 °C for 3.5 h. Solution
was added to ice-cold water and extracted with EtOAc and dried over
MgSO4. The solid was then dried in vacuo overnight to give
compound 6.
General Procedure C for the Synthesis of
Compound 7
In a flame-dried flask, compound 6a (1 equiv)
and KOH (2 equiv) were dissolved in anhydrous EtOH with heat. In a
separate flame-dried flask, chloroacetic acid (1 equiv) was added
to anhydrous EtOH. Chloroacetic acid solution was then added to the
reaction mixture and refluxed for 6 h. The reaction was concentrated
by half and acidified with 3 M HCl to pH 4. The solids were collected,
washed with water, and dried in vacuo to give compound 7.
General Procedure D for the Synthesis of Compound 8
Compound 7 (1 equiv), triethylamine (2 equiv),
and the appropriate amine (1.1 equiv) were dissolved in anhydrous
DMF. To this solution, HATU (1.1 equiv) dissolved in DMF was added
and stirred until complete and concentrated in vacuo. The crude material
was then recrystallized in MeOH to give compound 8.
General Procedure E for the Synthesis of Compound 46
NaH (1.1 equiv) was added to a solution of compound 45 (1 equiv) in DMF. The reaction mixture was stirred for
5 min, and then the appropriate alkyl halide (1.1 equiv) was added
and stirred until complete. The crude product was extracted with EtOAc
and dried over MgSO4. Crude material was finally recrystallized
with MeOH to give compound 46.
General Procedure F for
the Synthesis of Compound 47
Compound 46 was dissolved in 1:1 acetonitrile/triflouroacetic
acid and stirred at room temperature overnight. Depending on the N-alkyl substitution, crude material either precipitated
as pure product or was purified by chromatography to give compound 47.
Compound 7a (50 mg, 0.14 mmol),
triethylamine (40 μL, 0.28 mmol), and cyclohexylamine (18 μL,
0.16 mmol) were dissolved in anhydrous DMF (1 mL). HATU (59.5 mg,
0.16 mmol) dissolved in 0.2 mL of DMF was added to the reaction mixture
and stirred for 20 min and then concentrated in vacuo. The crude material
was recrystallized with MeOH to give 61 mg in near quantitative yield. 1H NMR (300 MHz, DMSO-d6) δ
ppm 0.97–1.33 (m, 4 H), 1.51 (d, J = 8.53
Hz, 1 H), 1.56–1.80 (m, 4 H), 3.48 (m, 1 H), 3.87 (s, 2 H),
7.24 (t, J = 7.29 Hz, 1 H), 7.38–7.67 (m,
6 H), 8.05 (d, J = 7.98 Hz, 1 H), 8.15 (d, J = 7.70 Hz, 1 H), 12.09 (s, 1 H). HRMS calcd for C24H24N4O2SNa (M + Na)+, 455.1512; found, 455.1511.
Ethyl 2-((2-Cyanophenyl)amino)acetate
(3)
Anthranilonitrile (30.29 g, 256 mmol), ethyl
bromoacetate (29.46
mL, 267 mmol), and sodium bicarbonate (25.6 g, 300 mmol) were combined
in anhydrous EtOH (90 mL) and refluxed for 42 h. After cooling slightly,
solution was decanted from precipitate into a prewarmed flask. Further
cooling of the decantate yielded crystals, which were then filtered
and washed with cold water to give 23.6 g of white crystalline product. 1H NMR (500 MHz, DMSO-d6) δ
ppm 1.13–1.27 (m, 3 H), 4.04 (d, J = 6.41
Hz, 2 H), 4.12 (q, J = 7.22 Hz, 2 H), 6.36 (t, J = 6.25 Hz, 1 H), 6.64 (d, J = 8.54 Hz,
1 H), 6.69 (t, J = 7.47 Hz, 1 H), 7.41 (t, J = 7.78 Hz, 1 H), 7.49 (dd, J = 7.93,
1.53 Hz, 1 H). HRMS calcd for C11H12N2O2Na (M + Na)+, 227.0791; found, 227.0794.
Ethyl 3-Amino-1H-indole-2-carboxylate (4)
In a flame-dried flask, a suspension of potassium t-butoxide (7.124 g, 63.5 mmol) in anhydrous THF was stirred
and maintained below 30 °C under argon. To this solution was
added a solution of compound 3 (16 g, 63.5 mmol) in anhydrous
THF over 45 min and stirred for an additional 2 h. The reaction was
then poured into ice water, extracted with EtOAc, and dried over MgSO4. The solid was then dissolved in minimal EtOH, and water
was added dropwise until just cloudy and allowed to precipitate at
room temperature. Precipitate was filtered and washed with cold EtOH
to give 6.6 g of compound 4 in 51% yield. 1H NMR (500 MHz, DMSO-d6) δ ppm
1.33 (t, J = 7.02 Hz, 3 H), 4.29 (q, J = 7.12 Hz, 2 H), 5.68 (s, 2 H), 6.88 (ddd, J =
8.08, 5.80, 1.98 Hz, 1 H), 7.15–7.25 (m, 2 H), 7.74 (d, J = 7.93 Hz, 1 H), 10.34 (s, 1 H). HRMS calcd for C11H12N2O2 (M + H)+, 205.0972; found, 205.0973.
Compound 4 (50 mg, 0.28 mmol)
was reacted with 1-naphthyl isothiocyanate (49.9 mg, 0.27 mmol) in
warm EtOH (750 μL) according to general procedure A to give
58.6 mg of compound 5c in 61.5% yield. HRMS calcd for
C22H19N3O2SNa (M + Na)+, 412.1090; found, 412.1091.
To a flame-dried flask with cold
anhydrous EtOH (75 mL) was added
cold acetyl chloride (7 mL, 98.5 mmol) under argon with stirring.
In a separate flame-dried flask charged with argon, compound 5a (6.5 g, 19 mmol) was dissolved in anhydrous EtOH (25 mL)
and added to the acetyl chloride solution. The reaction was refluxed
for 12 h and cooled upon completion. Precipitate was filtered and
recrystallized with EtOH to give 3.77 g of compound 6a in 75% yield. 1H NMR (400 MHz, DMSO-d6) δ ppm 7.09–7.31 (m, 3 H), 7.31–7.61
(m, 6 H), 8.20 (d, J = 8.07 Hz, 1 H), 12.17 (br s,
1 H). HRMS calcd for C16H12N3OS (M
+ H)+, 294.0696; found, 294.0698.
Compound 7b (25 mg, 0.07 mmol),
triethylamine (19.5 μL, 0.14 mmol), and cyclopentylamine (8.82
μL, 0.08 mmol) were dissolved in anhydrous DMF (0.5 mL). HATU
(29.3 mg, 0.08 mmol) dissolved in 0.2 mL of DMF was added to the reaction
mixture and stirred for 20 min and concentrated in vacuo. The crude
material was recrystallized in MeOH to give 9.3 mg in 30.3% yield. 1H NMR (400 MHz, DMSO-d6) δ
ppm 0.97–2.09 (m, 18 H), 2.61–2.83 (m, 4 H), 4.02 (br
s, 1 H), 4.17–4.37 (m, 1 H), 7.20 (t, J =
7.30 Hz, 1 H), 7.36–7.56 (m, 2 H), 8.00 (d, J = 8.06 Hz, 1 H), 8.32 (d, J = 7.33 Hz, 1 H), 11.86
(br s, 1 H). 13C NMR (126 MHz, DMSO-d6) δ ppm 23.97, 25.25, 26.34, 28.78, 32.73, 37.02, 51.21,
61.99, 113.23, 120.31, 120.54, 120.67, 120.79, 127.66, 136.48, 139.24,
151.93, 155.94, 166.85. HRMS calcd for C23H28N4O2SNa (M + Na)+, 447.1825; found,
447.1827.
N-Cyclohexyl-2-((5-methyl-4-oxo-3-phenyl-4,5-dihydro-3H-pyrimido[5,4-b]indol-2-yl)thio)acetamide
(42) and N-Cyclohexyl-N-methyl-2-((5-methyl-4-oxo-3-phenyl-4,5-dihydro-3H-pyrimido[5,4-b]indol-2-yl)thio)acetamide (43)
To a solution of compound 1 (50
mg, 0.12 mmol) in DMF (1 mL) in a flame-dried flask was added NaH
60% dispersion in mineral oil (9.24 mg, 0.24 mmol) and stirred at
room temperature for 10 min. Methyl iodide (14.3 μL, 0.24 mmol)
was then added to the solution and stirred overnight. Reaction mixture
was concentrated and purified by preparative thin-layer chromatography
(40:60 ethyl acetate:hexane) to give 18.18 mg in 33.9% yield of compound 42 and 5.3 mg in 9.6% yield of compound 43. 42: 1H NMR (400 MHz, DMSO-d6) δ ppm 0.93–1.37 (m, 5 H), 1.40–1.90
(m, 5 H), 3.44–3.58 (m, 1 H), 3.88 (s, 2 H), 4.15 (s, 3 H),
7.31 (t, J = 7.33 Hz, 1 H), 7.39–7.53 (m,
2 H), 7.53–7.78 (m, 5 H), 8.05–8.27 (m, 2 H). 13C NMR (126 MHz, DMSO-d6) δ ppm
24.90, 25.62, 31.47, 32.82, 37.14, 48.41, 111.31, 119.12, 120.09,
120.74, 120.92, 128.03, 129.99, 130.34, 136.35, 137.47, 140.40, 153.27,
155.75, 166.14. HRMS calcd for C25H27N4O2S (M + H)+, 447.1849; found, 447.1852. 43: 1H NMR (400 MHz, DMSO-d6) δ ppm 0.76–1.98 (m, 10 H), 2.10 (s, 3 H), 3.97
(s, 2 H), 4.23 (s, 3 H), 7.16 (t, J = 7.52 Hz, 1
H), 7.26–7.85 (m, 8 H), 8.07 (d, J = 7.70
Hz, 1 H). 13C NMR (126 MHz, DMSO-d6) δ ppm 24.85, 25.60, 31.43, 32.28, 38.02, 50.24, 62.03,
111.44, 119.63, 120.34, 120.85, 121.08, 128.20, 129.74, 130.25, 135.43,
136.73, 139.34, 155.99, 156.50, 169.99. HRMS calcd for C26H29N4O2S (M + H)+, 461.2006;
found, 461.2007.
Compound 6a (100 mg, 0.34
mmol) and KOH (38.2 g, 0.68 mmol) were suspended in 2 mL of dimethylacetamide
in a flame-dried flask with stirring. H2O was added dropwise
until KOH was completely dissolved. t-Butyl chloroacetate
(49 μL, 0.34 mmol) was immediately added to the reaction mixture
and stirred at room temperature monitoring with thin-layer chromatography
(1:99 MeOH:DCM). Upon completion, reaction mixture was extracted with
ethyl acetate (20 mL) and water (40 mL), dried over MgSO4, and concentrated in vacuo. EtOH (5 mL) was added to the resulting
viscous liquid, and pure product was filtered to give 55 mg in 40%
yield. IR: 3192 (NH), 1732 (CO ester), 1674 (CO amide) cm–1. 1H NMR (400 MHz, DMSO-d6) δ ppm 1.32–1.48 (m, 9 H), 3.91 (s, 2 H), 7.26 (t, J = 7.31 Hz, 1 H), 7.42–7.55 (m, 4 H), 7.56–7.69
(m, 3 H), 7.96 (d, J = 8.29 Hz, 1 H), 12.13 (s, 1
H). HRMS calcd for C22H21N3O3SNa (M + Na)+, 430.1196; found, 430.1197.
A KOH (6
mg, 0.11 mmol) solution in water (50 μL) was added to a solution
of compound 51 in dimethylacetamide (200 μL) and
stirred at rt overnight. Reaction mixture was acidified with 3 M HCl,
extracted with ethyl acetate and water, dried over MgSO4, and concentrated to dryness in vacuo. Further recrystallization
in MeOH gives 20 mg in 37% yield. 1H NMR (400 MHz, DMSO-d6) δ ppm 0.94–1.36 (m, 5 H), 1.43–1.89
(m, 5 H), 3.43–3.62 (m, 1 H), 3.89 (s, 2 H), 5.24 (s, 2 H),
7.12–7.22 (m, 1 H), 7.30 (t, J = 7.33 Hz,
1 H), 7.42 (m, J = 7.30, 2.20 Hz, 2 H), 7.48–7.67
(m, 5 H), 8.10 (d, J = 7.70 Hz, 1 H), 8.18 (d, J = 7.70 Hz, 1 H). 13C NMR (126 MHz, DMSO-d6) δ ppm 24.90, 25.59, 32.82, 37.05, 46.85,
48.48, 111.36, 119.30, 120.36, 120.91, 120.95, 128.13, 129.87, 130.05,
130.41, 136.14, 137.82, 140.64, 153.47, 155.65, 166.26, 169.77. HRMS
calcd for C26H27N5O3SNa
(M + Na)+, 512.1727; found, 512.1726.
Biological
Studies
Animals
Seven–nine-week-old C57BL/6 (wild-type,
WT) and Cd14 (C57BL/6 background) were purchased from the Jackson Laboratories
(Bar Harbor, MA). Tlr4mice were a gift from Dr. Shizuo Akira (Osaka University, Japan)
and backcrossed for 10 generations onto the C57BL/6 background at
University of California, San Diego (UCSD). All animal experiments
were approved by the UCSD Institutional Animal Care and Use Committee.
In Vitro Cytokine Induction in Bone Marrow Derived Dendritic
Cells (BMDC)
BMDC were prepared from C57BL/6 mice as described.[36] BMDC (105cells per well) were plated
in 96-well plates in 200 μL of complete RPMI1640 supplemented
with 10% fetal calf serum (FCS; Sigma Aldrich), 100 U/mL penicillin,
and 100 μg/mL streptomycin (Invitrogen). The cells were incubated
with graded concentrations of the compounds for 18 h at 37 °C,
5% CO2. After 18 h incubation, the cell culture supernatants
were collected. LPS (purified LPS, Invivogen, San Diego, CA) or MPLA
(1 μg/mL synthetic MPLA, Invivogen, San Diego, CA) were used
as positive controls. The levels of IL-6 in the culture supernatants
were determined by ELISA (BD Biosciences, La Jolla, CA).[37] The AUC was calculated from BMDC dose–response
curves using Prism 5 (GraphPad, San Diego). Each cytokine induction
curve was first converted to a percent activity curve, and then the
AUC of the percent activity curve was calculated. The process of converting
to a percent activity curve allowed subtracting background and adjusting
for plate-to-plate variation. Finally, the AUC values were normalized
to the activity of compound 1 within each experiment,
set at 100. The cultures stimulated with LPS 10 ng/mL, and 5 μM
compound 1 released an average of 20.6 ng/mL ± 4.7
SD and 10.5 ng/mL ± 1.3 SD of IL-6, respectively.
In Vitro
Assays Using TLR Reporter Cell Lines
Murine
or humanTLR4HEK Blue cells (Invivogen, 2.5 × 104 cells
per well of a 96-well plate), or NFκB/SEAPorter HEK 293 cells
(Imgenex, San Diego, CA) for humanTLR2, TLR3, TLR5, TLR7, TLR8, or
TLR9 (5 × 104 cells per well of 96 well plate) were
incubated with graded doses of compound 1. The culture
supernatants were harvested after a 20–24 h incubation period.
SEAP activity in the supernatants was determined by a colorimetric
assay, using either the SEAPorter Assay Kit (Imgenex), with absorbance
read at 405 nm, or QuantiBlue (Invivogen), with absorbance read at
630 nm. Stimulation of the humanTLR4 cells with 10 ng/mL LPS resulted
in OD630 of 1.90 relative to 0.88 of cells stimulated with
10 μM compound 1. The relative reporter activation
for the murineTLR4 cells for cells stimulated with 10 ng/mL LPS and
5 μM compound 1 was 1.70 ± 0.08 and 1.52 ±
0.03, respectively.
In Vitro Activities in hPBMC
Human
PBMC were isolated from buffy coats obtained from the San Diego Blood
Bank (San Diego, CA) as described previously.[37,38] PBMC (1 × 106/mL) were incubated with various compounds
in complete RPMI for 18 h at 37 °C, 5% CO2, and culture
supernatants were collected. The levels of IL-8 in the supernatants
were determined by ELISA (BD Biosciences, La Jolla, CA). Human cell
cultures were treated with 10 ng/mL LPS as the positive control. Cultures
treated with LPS 10 ng/mL and compound 1 10 μM
released averages of 14.5 ± 0.6 ng/mL SD and 7.8 ± 0.8 ng/mL
SD of IL-8, respectively.
Type I IFN Assay
L929 cells stably
expressing an interferon
sensitive response element (ISRE) luciferase reporter construct were
kindly provided by Dr. B. Beutler (UT Southwestern, Texas).[39] The bioactivity of type I IFN in mBMDC supernatants
was measured by luciferase assay using L929-ISRE cells as described
previously.[39] L929-ISRE cells were plated
at 5× 104 cells per well in Dulbecco’s Modified
Eagle Medium (Invitrogen) supplemented with 10% FCS, 100 U/mL penicillin,
and 100 μg/mL streptomycin (DMEM-10) in a 96-well white-walled
clear-bottom plate. Thus, 50 μL of mBMDC supernatant was incubated
with L929-ISRE cells in 50 μL of DMEM for 6 h. Mu-IFN Beta Standard
(PBL Interferon Source, Piscataway, NJ) was used as a standard. The
luciferase activities were measured by Steady-Glo luciferase assay
buffer (Promega, Madison, WI).In addition, the levels of IP-10,
a surrogate marker of type I IFN, in the supernatants were determined
by ELISA (R&D Systems, Minneapolis, MN). The cultures stimulated
with LPS 10 ng/mL, and 5 μM compound 1 released
an average of 234.8 ± 8.6 and 590.0 ± 10.5 pg/mL of IP-10
respectively.
Statistical Analysis
The data are
represented as mean
± standard error of the mean (SEM). Areas-under-curve (AUCs)
were calculated using the trapezoid method. Prism 5 (GraphPad Software)
statistical software was used to obtain p-values
for comparison between groups (p < 0.05 was considered
significant). For the in vitro studies, two-tailed Student’s t test was used to compare two groups, and one-way ANOVA
Dunnett’s test was used to compare multiple groups.
Authors: Minya Pu; Tomoko Hayashi; Howard Cottam; Joseph Mulvaney; Michelle Arkin; Maripat Corr; Dennis Carson; Karen Messer Journal: Stat Med Date: 2012-07-05 Impact factor: 2.373
Authors: T Hayashi; S P Rao; K Takabayashi; J H Van Uden; R S Kornbluth; S M Baird; M W Taylor; D A Carson; A Catanzaro; E Raz Journal: Infect Immun Date: 2001-10 Impact factor: 3.441
Authors: Jongdae Lee; Tsung-Hsien Chuang; Vanessa Redecke; Liping She; Paula M Pitha; Dennis A Carson; Eyal Raz; Howard B Cottam Journal: Proc Natl Acad Sci U S A Date: 2003-05-08 Impact factor: 11.205
Authors: Peter H Goff; Tomoko Hayashi; Luis Martínez-Gil; Maripat Corr; Brian Crain; Shiyin Yao; Howard B Cottam; Michael Chan; Irene Ramos; Dirk Eggink; Mitra Heshmati; Florian Krammer; Karen Messer; Minya Pu; Ana Fernandez-Sesma; Peter Palese; Dennis A Carson Journal: J Virol Date: 2015-01-07 Impact factor: 5.103
Authors: Matthew D Morin; Ying Wang; Brian T Jones; Lijing Su; Murali M R P Surakattula; Michael Berger; Hua Huang; Elliot K Beutler; Hong Zhang; Bruce Beutler; Dale L Boger Journal: J Med Chem Date: 2016-04-25 Impact factor: 7.446
Authors: Rajesh Rajaiah; Darren J Perkins; Derek D C Ireland; Stefanie N Vogel Journal: Proc Natl Acad Sci U S A Date: 2015-06-23 Impact factor: 11.205
Authors: Peter H Goff; Tomoko Hayashi; Wenqian He; Shiyin Yao; Howard B Cottam; Gene S Tan; Brian Crain; Florian Krammer; Karen Messer; Minya Pu; Dennis A Carson; Peter Palese; Maripat Corr Journal: J Virol Date: 2017-09-12 Impact factor: 5.103
Authors: Matthew D Morin; Ying Wang; Brian T Jones; Yuto Mifune; Lijing Su; Hexin Shi; Eva Marie Y Moresco; Hong Zhang; Bruce Beutler; Dale L Boger Journal: J Am Chem Soc Date: 2018-10-16 Impact factor: 15.419
Figure 1
Figure 2
Scheme 1
Figure 3
Figure 4
Figure 5
Figure 6
Scheme 2
Figure 7
Scheme 3
Figure 8
Figure 9
Figure 10