Ortho -Substituted lipidated Brartemicin derivative shows promising Mincle-mediated adjuvant activity

Article abstract of DOI:10.1039/c9ob02397f

The macrophage inducible C-type lectin (Mincle) is a pathogen recognition receptor (PRR) that is a promising target for the development of Th1-polarising vaccine adjuvants. We recently reported on the synthesis and evaluation of lipidated Brartemicin analogues that showed Mincle agonist activity, with our lead agonist exhibiting potent Th1 adjuvant activity that was greater than that of trehalose dibehenate (TDB). Herein, we report on the efficient synthesis and subsequent biological evaluation of additional lipidated Brartemicin analogues that were designed to determine the structural requirements for optimal Mincle signalling. While all the Brartemicin analogues retained their ability to signal through Mincle and induce a functional response, the o-substituted and m,m-disubstituted derivatives (5a and 5d, respectively) induced a potent inflammatory response when using cells of both murine and human origin, with this response being the greatest observed thus far. As the inflammatory response elicited by 5a was slightly better than that induced by 5d, our findings point to o-substituted Brartemicin analogues as the preferred scaffold for further adjuvant development.

Full text of DOI:10.1039/c9ob02397f

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
ortho-Substituted lipidated Brartemicin derivative
shows promising Mincle-mediated adjuvant
activity†
Cite this: DOI: 10.1039/c9ob02397f
Amy J. Foster,a,b Kristel Kodar, a,b Mattie S. M. Timmer
*
a,b and
a,b
Bridget L. Stocker
*
The macrophage inducible C-type lectin (Mincle) is a pathogen recognition receptor (PRR) that is a prom-
ising target for the development of Th1-polarising vaccine adjuvants. We recently reported on the syn-
thesis and evaluation of lipidated Brartemicin analogues that showed Mincle agonist activity, with our lead
agonist exhibiting potent Th1 adjuvant activity that was greater than that of trehalose dibehenate (TDB).
Herein, we report on the efficient synthesis and subsequent biological evaluation of additional lipidated
Brartemicin analogues that were designed to determine the structural requirements for optimal Mincle
signalling. While all the Brartemicin analogues retained their ability to signal through Mincle and induce a
functional response, the o-substituted and m,m-disubstituted derivatives (5a and 5d, respectively)
induced a potent inflammatory response when using cells of both murine and human origin, with this
response being the greatest observed thus far. As the inflammatory response elicited by 5a was slightly
better than that induced by 5d, our findings point to o-substituted Brartemicin analogues as the preferred
scaffold for further adjuvant development.
Received 6th November 2019,
Accepted 11th December 2019
DOI: 10.1039/c9ob02397f
rsc.li/obc
Since the discovery of Mincle, significant effort has been
directed toward the identification of self and foreign Mincle
Introduction
The macrophage inducible C-type lectin (Mincle, Clec4e, ligands.3,4 Trehalose glycolipids (TGLs), such as the
ClecSf9) is a pattern recognition receptor commonly expressed Mycobacterium tuberculosis cell wall glycolipid, trehalose dimy-
on the cell surface of innate immune cells such as macro- colate (TDM, 1, Fig. 1), and its synthetic analogue, trehalose
phages, dendritic cells (DCs), and neutrophils.1,2 Mincle plays dibehenate (TDB, 2), were among the first described Mincle
a pivotal role in innate immunity through the recognition of a ligands with the capacity to trigger a potent inflammatory
vast array of both endogenous and exogenous amphiphilic response.5–7 Given the ability of TGLs to induce an inflamma-
compounds, many of which incorporate a lipid-chain.3,4 The
activation of Mincle by an appropriate ligand leads to phos-
phorylation of the FcRγ adaptor protein, which in turn triggers
the recruitment of spleen tyrosine kinase (syk) to the cell mem-
brane and activation of the Card9-Bcl10-Malt-1 signalling
cascade.2,5,6 Translocation of the nuclear factor kappa-light
chain enhancer of activated B cells (NF-κB) to the cell nucleus
then promotes the expression of inflammatory cytokines and
chemokines, such as interleukin (IL)-1β, macrophage inflam-
matory protein (MIP)-2, and IL-6.
aSchool of Chemical and Physical Sciences, Victoria University of Wellington,
PO Box 600, Wellington, New Zealand. E-mail: bridget.stocker@vuw.ac.nz,
mattie.timmer@vuw.ac.nz
bCentre for Biodiscovery, Victoria University of Wellington, P. O. Box 600,
Wellington 6140, New Zealand
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
Fig. 1 Representative trehalose glycolipids investigated for their Mincle
c9ob02397f
agonist activity.
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
tory immune response, there has been substantial interest in
harnessing the inherent immunostimulatory properties of
TGLs for use as vaccine adjuvants. As such, TDB (2) has been
formulated into a variety of dimethyldioctadecylammonium
(DDA)-containing liposomes, including liposomes comprised
of DDA and TDB (DDA:TDB, termed CAF01) as well as DDA:
TDB liposomes supplemented with the TLR-3 ligand, poly-IC
(CAF05),8 or the TLR-4 ligand, monophosphoryl lipid (MPL)-A
(CAF06).9 All three formulations (CAF01, CAF05, and CAF06)
have been found to induce a strong T cell response,8–10 with
further evaluation of CAF01 highlighting promising adjuvant
activity in tuberculosis,11–15 malaria,15 influenza,16 and chla-
mydia15 vaccination models. CAF01 has been progressed to
human clinical trials, whereby healthy volunteers were vacci-
nated with the tuberculosis antigen Ag86B-ESAT-6 adjuvanted
with CAF01.14,17 During these trials, vaccination was found to
produce a potent and persisting T cell response and the
vaccine was found to be ‘safe and tolerable’.
In addition to TDM (1) and TDB (2), a vast array of lipid-
containing glucose-,18–23 mannose-,20,24 glycerol-,18,25 and tre-
halose-derived ligands26–34 have been identified as Mincle ago-
nists. Mincle recognises carbohydrate-derived ligands through
its carbohydrate recognition domain (CRD), which contains
the Glu-Pro-Asn (EPN) motif that mediates binding to glucose-
and mannose-containing structures.28,29 Here, Ca2+ found
within the EPN is thought to bind to the 3- and 4-hydroxyls of
one glucose-residue, while an adjacent hydrophobic groove is
thought to be able to accommodate the α-branched or linear
lipid portions commonly found in Mincle ligands.28,29,35
Additionally, it has been proposed that within bovine Mincle
(bMincle), Arg182 is suitably placed to interact with the
hydroxyls of treahalose.29
Fig. 2 Target lipidated Brartemicin analogues.
C18-lipid chain at the phenyl ortho (o)- and meta (m)-positions
(5a and 5b, respectively, Fig. 2) were proposed so that their
immunostimulatory activity could be compared to the pre-
viously described para (p)-alkylated analogue 5c (p-OC18),
which in our earlier studies was found to exhibit similar
agonist activity to C18dMeBrar (4).37 Given the potent Mincle
agonist activity of branched trehalose glycolipids,7,19,33,38 we
also proposed that the increased lipophilicity of the m,m-bis
(OC18) Brartemicin derivative 5d might enhance Mincle adju-
vant activity. As the interaction between C18dMeBrar (4) and
Arg183 is crucial for ligand binding and cellular activation,37
Brartemicin analogues that integrated either an ethylene (5e)
or an ethynyl (5f) spacer between the trehalose ester and the
aromatic scaffold were proposed so that further insight in to
the positioning of the aromatic moiety and Arg183 could be
gained. For these adducts, a shorter (C16) lipid chain would be
included to maintain the overall lipid length. Finally, carbon-
linked (5f), nitrogen-linked (5h), and sulfur-linked (5i) deriva-
tives would be prepared so as to investigate the importance of
the linkage between the aromatic scaffold and the lipid-chain.
Herein, we report on the synthesis of these derivatives and
their ability to activate Mincle expressing nuclear factor of acti-
vated T cell (NFAT)-green fluorescent protein (GFP) reporter
cells,2,39 bone marrow derived macrophages of murine
origin,40 and human peripheral blood monocytes.41
In 2015, the natural product Brartemicin (3) was deter-
mined to be a high-affinity ligand for bovine Mincle through
the use of competition binding assays and molecular simu-
lation.36 Furthermore,
a
crystal structure obtained of
Brartemicin (3) binding to bMincle suggested that, much like
trehalose, Brartemicin (3) binds the CRD of Mincle.35 Therein,
one of the aromatic esters of Brartemicin (3) was located near
Arg182, which suggests that π–cation interactions may be
important for the binding of aromatic Mincle ligands.35
Following these studies, we reported on the synthesis and bio-
logical evaluation of a variety of Brartemicin analogues and
demonstrated that Brartemicin analogues incorporating long
(C18) lipid chains, but not Brartemicin itself, were potent
Results and discussion
Synthesis of Brartemicin analogues
Mincle agonists.37 Moreover, Arg183 (hMincle) was deemed To synthesise the desired lipidated Brartemicin analogues, we
crucial for ligand binding of the C18-alkylated desmethyl- envisioned a strategy whereby the target analogues 5a–b and
brartemicin analogue (C18dMeBrar, 4), with 4 subsequently 5d–i could be accessed via the esterification of a suitably pro-
being found to possess superior adjuvant activity to TDB in a tected trehalose derivative (6 or 7) with a range of appropriately
delayed-type hypersensitivity (DHT) immunisation assay where functionalised carboxylic acids (8a–b, 8d, 9e–f, and 10g–i,
OVA was used as the model antigen.
Scheme 1).26 Here, it was proposed that benzylated trehalose
While lipidated Brartemicin analogues are known to be 6, which is readily accessible from α,α′-D-trehalose (11) in three
potent Mincle agonists, the exact structural requirements for steps,26 could be esterified with acids 8a–b, 8d, 9e, and 10g–i
Mincle signalling remain unknown. To this end, we sought to to give the corresponding protected glycolipids, which after
further explore the structure–activity relationships (SAR) of a debenzylation would provide target analogues 5a–b, 5d–e, and
library of Brartemicin analogues. Here, analogues containing a 5g–i. For those target glycolipids that would not tolerate the
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 1 Retrosynthetic analysis of the target lipidated Brartemicin analogues.
hydrogenolysis conditions required for global debenzylation,
namely 5f and 5i, it was proposed that the silyl ether protected
trehalose derivative 7, which is prepared from α,α′-D-trehalose
(11) in a single synthetic step,42 could be used for esterification
with acids 9f and 10i to give the corresponding protected gly-
colipids. Subsequent removal of the TMS protecting groups
would yield the desired glycolipids 5f and 5i. Acids 8a–b, 8d,
9e–f and 10i were envisioned to be prepared from commer-
cially available esters (12a–b, 12d, 13e–f, and 14i, respectively)
via alkylation with either 1-bromooctadecane37 or
1-bromohexadecane37,43 and subsequent base-catalysed hydro-
lysis of the ester.37 The carbon-linked benzoic acid 10g could
be prepared from aldehyde 14g through a Wittig reaction44
using octadecyltriphenylphosphonium bromide, which is
readily accessible from 1-bromooctadecane,45 and subsequent
hydrolysis of the methyl ester.37 Finally, we proposed that alky-
lamino benzoic acid 10h could be prepared from ethyl ester
14h through reductive amination with octadecanal followed by
base-mediated hydrolysis of the ester.37,46 Octadecanal is, in
turn, prepared through the oxidation of 1-octadecanol.47
With the restrosynthetic strategy in place, the syntheses of
carboxylic acids containing ether-linked lipids were under-
taken (Scheme 2). Here, methyl salicylate (12a), ethyl 3-hydro-
xybenzoate (12b), and methyl 3,5-dihydroxybenzoate (12d)
were alkylated with 1-bromooctadecane in the presence of
tetrabutylammonium iodide (TBAI) to give the corresponding
lipophilic esters. Treatment of the lipophilic esters with
sodium hydroxide in the appropriate alcohol then yielded the
target benzoic acids (8a, 8b, and 8d) in modest (26%) to excel-
lent (78%) yield over two steps. To prepare the carboxylic acids
incorporating the ethylene and ethenyl spacers (9e and 9f,
respectively), dihydrocinnamate 13e and cinnamate 13f were
first treated with 1-bromohexadecane to install the slightly
truncated lipid-chain. Subsequent hydrolysis of the methyl
Scheme 2 Synthesis of acids 8a–b, 8d, and 9e–f.
esters then afforded the desired acids 9e and 9f in good yields
(63–70% over two steps).
Next, we embarked on the synthesis of the carbon-linked
glycolipid (5f), whereby 1-bromooctadecane (15) was treated
with triphenylphosphine in toluene to give the corresponding
phosphonium bromide 16, according to a literature procedure
(Scheme 3).45 Subsequent reaction of phosphonium bromide
16 with BuLi afforded the corresponding ylide, which was
immediatly subjected to a Wittig reaction with methyl 4-for-
mylbenzoate (14g) to give methyl benzoates 17 as
a
3 : 1 mixture of the Z- and E-alkenes in good (76%) yield.
Subjection of benzoates 17 to 5 M sodium hydroxide in MeOH
then gave the desired benzoic acids 10g as an isomeric
mixture. As hydrogenation of the alkene was to occur at a later
stage in the synthesis, no attempt was made to separate the E-
and Z-isomers of 10g.
To prepare the alkylamino benzoic acid 18, octadecanol 19
was first converted to the corresponding aldehyde 19 using a
pyridinium chlorochromate mediated oxidation to yield octa-
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 1 Assembly of lipidated Brartemicin analogues
Yield (%)
6 → 21
or
7 → 22
21 → 5
or
22 → 5
Scheme 3 Synthesis of acid 10g.
Entry
1
Diester
Method
A
21a, 84
21b, 88
21d, 61
5a, 57
5b, 67
5d, 54
decanal (19) in excellent (85%) yield (Scheme 4).47 Reductive
amination of aldehyde 16 with ethyl 4-aminobenzoate in the
presence of sodium triacetoxyborohydride then gave the nitro-
gen-linked lipophilic ester 20, which was hydrolysed in the
presence of sodium hydroxide and ethanol to give
target alkylamino benzoate 10h in excellent (92%) yield.
Finally, the sulfur-linked acid 10i was prepared via the alkyl-
ation of methyl 4-mercaptobenzoate 14i with 1-bromooctade-
cane, followed by the base-mediated hydrolysis of the methyl
ester to give the target acid 10i in modest yield (40% over two
steps). Here, the reduced solubility of acid 10i in typical
organic solvents (e.g. EtOAc and CH2Cl2) necessitated the use
of hot solvent to extract the target compound during work-up.
In this way, the desired compound was isolated in excellent
yield (quant.).
2
3
A
A
4
5
6
7
8
A
B
A
A
B
21e, 44
22f, 34
21g, 34
21h, 38
22i, 77
5e, 64
5f, 40
5g, 58
5h, 28
5i, 52
With the carboxylic acids in hand, the target lipidated
Brartemicin derivatives were prepared via the esterification of
hexa-O-Bn-trehalose (6, Method A) or hexa-O-TMS-trehalose (7,
Method B) with the previously prepared carboxylic acids
(Table 1). Accordingly, α,α′-D-trehalose (11) was tritylated at the
6- and 6′-hydroxyls and per-benzylated, followed by removal of
the trityl groups according to a literature procedure (Method
A).26 Benzylated trehalose 6 was then subjected to 1-ethyl-3-(3-
dimethylaminopropyl)carbodiimide (EDCI) and 4-(dimethyl-
amino)pyridine (DMAP) mediated esterifications with acids
8a–b, 8d, 9e, and 10g–h to give the protected glycolipids (21a–
b, 21d–e, and 21g–i) in modest (34%) to excellent (88%) yields.
Here, Heteronuclear Multiple Bond Correlations (HMBCs)
between the carbonyl carbons of the lipophilic esters and H-6a
and H-6b of the trehalose scaffold confirmed that the aromatic
lipids had been successfully installed. Hydrogenolysis of the
benzylated glycolipids (21a–b, 21d–e, and 21g–i) using
Pearlman’s catalyst followed by silica-gel column chromato-
graphy provided the target glycolipids in 28–67% yield. To
prepare those target derivatives that are not compatible with
hydrogenolysis conditions, the synthetic strategy reported by
Toubiana et al.42 was employed to prepare 2,2′,3,3′,4,4′-hexa-O-
trimethylsilyl-α,α′-D-Trehalose (7) from α,α′-D-trehalose (11) in a
single synthetic step (Method B). More specifically, α,α′-D-tre-
halose (11) was treated with N,O-bis(trimethylsilyl)acetamide
(BSA) and tetra-N-butylammoniuim fluoride (TBAF) to give the
per-silylated adduct, which was subsequently treated with pot-
Scheme 4 Synthesis of benzoic acids 10h and 14i.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Organic & Biomolecular Chemistry
Paper
assium carbonate and methanol to selectively remove the the proportion of GFP positive cells as the p-OC18 analogue,
primary TMS ethers. Esterification of the TMS-protected 5c. These data suggest that while Mincle can accommodate o-,
derivative 7 with cinnamic acid 9f and sulfur-linked acid 10i m-, and p-substituted Brartemicin analogues (5a, 5b, and 5c,
yielded the TMS-protected glycolipids 22f and 22i in modest respectively), ligands substituted at the o-hydroxyl provide a
(34%) to good (77%) yield. Global TMS-ether deprotection powerful scaffold for the generation of potent Mincle ligands.
using H+-Dowex furnished the remaining target analogues (5f At the higher ligand concentration (1 nmol per well) however,
and 5i) in modest (40–52%) yields.
the response induced by 5a (o-OC18) decreased, particularly in
the hMincle reporter assay. We recently determined that glyco-
lipid presentation, which is influenced by ligand solubility,
could affect Mincle agonist activity.48 Thus, changes in the
Structure-activity relationship of lipidated Brartemicin
analogues
With the target lipidated Brartemicin analogues 5a–5i in hand, concentration of 5a (o-OC18) may have subtly affected the
the ability of these analogues to signal through Mincle was manner in which it was presented to Mincle. Of the remaining
assessed using NFAT-GFP reporter cells.2,39 Here, plate-bound Brartemicin analogues, 5d (m,m-bis[OC18]), 5e (p-DHCinOC16),
TDB (2) and analogues 5a–i were employed to stimulate 5f (p-CinOC16), 5g (p-CC18), 5h (p-NHC18), and 5i (p-SC18), all
NFAT-GFP reporter cells expressing murine Mincle (mMincle) induced the mMincle and hMincle reporter cells to express
+ FcRγ (Fig. 3a), human Mincle (hMincle) + FcRγ (Fig. 3b), or approximately equal quantities of GFP as TDB (2) and 5c
FcRγ only (Fig. 3). Mincle binding and cellular activation was (p-OC18). This further exemplifies the tolerance of Mincle and
monitored by the expression of GFP, as detected by flow cyto- its ability to accommodate alterations to the lipid structure in
metry. Consistent with our previous findings,37 the lipidated Brartemicin derivatives, including differences in the lipid-posi-
Brartemicin analogues 5a–i all strongly activated mMincle and tion, lipid-number, aromatic position, and linkage type.
hMincle expressing reporter cells, with the newly synthesised Moreover, when comparing this data to that obtained from our
analogues (5a–b, 5d–i) inducing the expression of comparable first generation library of lipidated-brartemicin derivatives,37 it
quantities of GFP to TDB (2) and the previously described lipi- can be seen that 5a exhibits similar mMincle- and hMincle-
dated Brartemicin analogue 5c (p-OC18) (Fig. 3).37 At the con- mediated signalling to that of C18dMeBrar 4 (see Fig. 1, ESI†).
centration of 0.1 nmol per well, 5a (o-OC18) and 5b (m-OC18)
While the reporter cells are a convenient tool for the identi-
best activated the hMincle and mMincle reporter cells, with fication of new Mincle ligands, we wanted to assess the ability
treatment with 5a (o-OC18) producing approximately double of the Brartemicin analogues to induce an immune response
Fig. 3 Lipidated Brartemicin analogues bind and signal through mMincle and hMincle. NFAT-GFP 2B4 reporter cells expressing mMincle + FcRγ (a),
hMincle + FcRγ (b), or FcRγ-only were stimulated using ligand-coated plates (0.1 or 1 nmol per well) for 20 hours. The 2B4 cells were then harvested
and examined for NFAT-GFP expression using flow cytometry. Data represent the mean of two independent experiments performed in duplicate
(mean SEM). Statistical significance was calculated in comparison to iPrOH only using 2-way ANOVA (Dunnett’s multiple comparison test), ****P ≤
0.0001, ***P ≤ 0.001.
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
in a system more akin to a physiological environment. stimulating similar levels of IL-1β production to both TDB (2)
Accordingly, the Brartemicin derivatives were screened for and 5c (p-OC18). Moreover, 5a (o-OC18) and, to a lesser extent
their ability to activate BMDMs to produce the pro- and anti- 5d (m,m-bis[OC18]) and 5h (p-NHC18), exhibited stronger
inflammatory cytokine IL-6, chemokine MIP-2, and the pro- agonist activity than C18dMeBrar
4 (see Fig. 2, ESI†).
inflammatory cytokine IL-1β. Here, the Brartemicin analogues Remarkably, at the concentration of 1 nmol per well, 5a
(5a–i) stimulated BMDMs to produce appreciable quantities of (o-OC18), 5d (m,m-bis[OC18]), and 5h (p-NHC18) induced
IL-6, MIP-2, and IL-1β, with all Brartemicin derivatives (5a–i) approximately a 2–3-fold increase in IL-1β production, as com-
and TDB (2) producing approximately equal amounts of IL-6 pared to TDB (2) and 5c (p-OC18) (Fig. 4). This striking increase
and MIP-2 (Fig. 4). At the concentration of 0.1 nmol per well in IL-1β production by BMDMs stimulated with 5a, 5d, and 5h
however, subtle differences were observed in the relative levels suggests that these Brartemicin derivatives, and in particular
of IL-6 produced by BMDMs following stimulation with the 5a (o-OC18), possess great promise as vaccine adjuvants.
newly prepared Brartemicin analogues. For example, 5d (m,m-
In addition to BMDMs, we also wanted to assess the ability
bis[OC18]) induced the BMDMs to produce comparatively more of the lipidated Brartemicin analogues to stimulate human
IL-6 than TDB (2) or 5c (p-OC18), while 5e (p-DHCinOC16) and peripheral blood monocytes to produce the inflammatory cyto-
5f (p-CinOC16) induced the production of less IL-6 than TDB kine, IL-8 (Fig. 5).49 Thus, human monocytes, which have pre-
(2) and 5c (p-OC18). In addition to IL-6 and MIP-2, the lipi- viously been shown to express Mincle,41,50,51 were isolated
dated Brartemicin analogues (5a–i) all stimulated the BMDMs from whole blood and incubated in plates coated with 0.1 or
to produce significant quantities of IL-1β, with 5b (m-OC18), 5e 1 nmol per well of TDB (2), p-OC18 (5c), C18dMeBrar, or
(p-DHCinOC16), 5f (p-CinOC16), 5g (p-CC18), and 5i (p-SC18) lipidated analogues 5a–b and 5d–i. For comparison, the ability
Fig. 4 Lipidated Brartemicin analogues induce inflammatory cytokine and chemokine production by BMDMs. Harvested WT BMDMs were incu-
bated in plates coated with Brartemicin derivatives (5a–5i, 0.1 and 1 nmol ligand per well), iPrOH, LPS (100 ng mL−1) or TDB (2, 0.1 or 1 nmol ligand
per well). IL-6, MIP-2, and IL-1β production was measured by ELISA of the supernatant collected after 24 hours. Data represents the mean of two
(IL-6) or three (MIP-2, IL-1β) experiments performed in duplicate (mean SEM). Statistical significance was calculated in comparison to iPrOH only
using 2-way ANOVA (Dunnett’s multiple comparison test), ****P ≤ 0.0001, ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 5 Lipidated Brartemicin analogues induce IL-8 production by human monocytes. Human monocytes negatively enriched from whole blood
were incubated in plates coated with Brartemicin derivatives (5a–5i, 0.1 and 1 nmol ligand per well), iPrOH, LPS (100 ng mL−1) or TDB (2, 0.1 or
1 nmol ligand per well). After 24 hours, the supernatant was collected and IL-8 levels were determined by ELISA. Data represents the mean of three
independent experiments performed in triplicate (mean SEM). Statistical significance was calculated in comparison to iPrOH only using 2-way
ANOVA (Dunnett’s multiple comparison test), ****P ≤ 0.0001, ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05.
of C18dMeBrar (4)37 to activate human monocytes was investi- goes a conformational change upon ligand binding and, by
gated as this data had not been previously obtained. Herein, doing so, is better able to accommodate ligands.
the o- and m,m-substituted analogues (5a and 5d) strongly acti- Notwithstanding, the strong adjuvant activity elicited by 5a
vated the monocytes to produce appreciable quantities of IL-8, (o-OC18) is noteworthy. We also observed that the addition of a
as compared to TDB (2), 5c (p-OC18), and C18dMeBrar (4). In second C18-lipid onto each of the aromatic diesters (i.e. 5d,
contrast, 5b (m-OC18), 5e (p-DHCinOC16), 5f (p-CinOC16), 5g m,m-bis[OC18]) led to an increase in IL-1β and IL-8 production
(p-CC18), 5h (p-NHC18), and 5i (p-SC18) induced the pro- by BMDMs and human monocytes, respectively.
duction of modest levels of IL-8. While we cannot conclusively
Taken together, these results further corroborate previous
say that the production of IL-8 by human monocytes in studies that determined that alterations to the lipid-structure
response to TDB or 5a and 5d is solely dependent on Mincle of trehalose diesters can influence ligand activity,33,34,37,38,53
signalling, the data from the reporter cell assays where Mincle- however, to date, it is difficult to predict how such subtle
dependence was observed indicates that IL-8 production, at changes will influence Mincle-mediated agonist activity.3
least in part, is due to the engagement of Mincle.
Herein, we found that more pronounced differences in the
Taken as a whole, our findings further support the growing immunomodulatory activity of the lipidated-brartemicin
body of evidence suggesting that trehalose diesters incorporat- derivatives were observed when changes were made to the
ing long-chain lipids are potent Mincle agonists that induce orientation of the lipid chain, as illustrated by 5a (o-OC18) and
an inflammatory immune response by antigen presenting 5d (m,m-bis[OC18]), which are strong agonists of both mMincle
cells.18,26,33,34,52 Furthermore, we have demonstrated that both and hMincle and are also able to induce inflammatory
mMincle and hMincle readily accommodate changes to the responses in in vitro assays using cells of murine and human
lipid structure of Brartemicin derivatives, with analogues of 5c origin. Moreover, the potential of species-specific differences
(p-OC18) that incorporate modifications to the position of the in the agonist activity of Mincle ligands,3 and the importance
lipid-chain and aromatic ring, the number of lipid-chains, and of employing a number of cellular assays so as to best investi-
the type of linkage between the aromatic ring and the lipid- gate the structure–activity relationships of Mincle ligands40,48
chain retaining their ability to signal through both mMincle is further illustrated in our work. As 5a (o-OC18) elicits a more
and hMincle (Fig. 3). Herein, the positioning of the lipid on potent immune response when using both murine and human
the Brartemicin scaffold appears to have more influence on in vitro assays, these findings cement o-substituted
Mincle agonist activity than the type of lipid-linker used, as Brartemicin analogues as the preferred scaffold for further
illustrated by the ability of the Brartemicin analogue substi- adjuvant development.
tuted at the ortho-hydroxyl (5a [o-OC18]) to activate the reporter
cells at low concentrations (Fig. 3) and induced a strong
immune response in both BMDMs and human monocytes
(Fig. 4 and 5). To explain this observation, attempts to dock 5a
Conclusions
(o-OC18) into the CRD of Mincle were made.37 However, a poor At present, TDB is the preferred Mincle agonist for adjuvant
docking score was observed. By considering our own work,37 development due to its structural simplicity and the promising
and that of others,36 it has been previously demonstrated that results achieved with the CAF01 liposome system. Due to the
Mincle binding does not directly correlate to a functional success of TDB, there has been much interest in the develop-
immune response. Thus, it is possible that 5a (o-OC18) binds ment of additional Mincle agonists that have enhanced
to a site other than the CRD on Mincle,3 or that Mincle under- activity. We recently reported on the synthesis of several lipi-
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
dated Brartemicin derivatives (e.g., 4 [C18dMeBrar] and 5c
[p-OC18]) that possess potent Mincle agonist activity and are
available in five steps from commercially available materials.
In addition, our lead agonist, 4 (C18dMeBrar), demonstrated
potent Th1 adjuvant activity, which was greater than that of
TDB. In this work, we have prepared an additional series of
lipidated Brartemicin analogues that feature structural modifi-
cations such as alterations to the positioning of the lipophilic
chain on the aromatic diesters, the total number of lipophilic
chains, the position of the aromatic group on the lipid-chain,
and the type of linkage incorporated between the aromatic
group and the lipid-tail. The target glycolipids were efficiently
prepared from commercially available materials (4–5 steps,
longest linear sequence) and in overall yields of 5–46%. While
all the synthesised ligands demonstrated an ability to signal
through mMincle and hMincle, the o-substituted derivative 5a
(o-OC18) and, to a lesser extent, the m,m-substituted derivative
5d (m,m-bis[OC18]) induced potent inflammatory responses in
both BMDMs and human monocytes and are therefore, potent
Mincle ligands that possess superior activity to TDB, 5c
(p-OC18), and C18dMeBrar (4).
2 S. Yamasaki, E. Ishikawa, M. Sakuma, H. Hara, K. Ogata
and T. Saito, Nat. Immunol., 2008, 9, 1179.
3 C. D. Braganza, T. Teunissen, M. S. M. Timmer and
B. L. Stocker, Front. Immunol., 2018, 8, 1940.
4 S. J. Williams, Front. Immunol., 2017, 8, 1662.
5 K. Werninghaus, A. Babiak, O. Groß, C. Hölscher,
H. Dietrich, E. M. Agger, J. Mages, A. Mocsai, H. Schoenen
and K. Finger, J. Exp. Med., 2009, 206, 89–97.
6 H. Schoenen, B. Bodendorfer, K. Hitchens, S. Manzanero,
K. Werninghaus, F. Nimmerjahn, E. M. Agger, S. Stenger,
P. Andersen, J. Ruland, G. D. Brown, C. Wells and R. Lang,
J. Immunol., 2010, 184, 2756–2760.
7 E. Ishikawa, T. Ishikawa, Y. S. Morita, K. Toyonaga,
H. Yamada, O. Takeuchi, T. Kinoshita, S. Akira, Y. Yoshikai
and S. Yamasaki, J. Exp. Med., 2009, 206, 2879–2888.
8 P. Nordly, F. Rose, D. Christensen, H. M. Nielsen,
P. Andersen, E. M. Agger and C. Foged, J. Controlled
Release, 2011, 150, 307–317.
9 P. Nordly, E. M. Agger, P. Andersen, H. M. Nielsen and
C. Foged, Pharm. Res., 2011, 28, 553–562.
10 J. Davidsen, I. Rosenkrands, D. Christensen, A. Vangala,
D. Kirby, Y. Perrie, E. M. Agger and P. Andersen, Biochim.
Biophys. Acta, 2005, 1718, 22–31.
11 L. Holten-Andersen, T. Doherty, K. Korsholm and
P. Andersen, Infect. Immun., 2004, 72, 1608–1617.
12 J. S. Woodworth, S. B. Cohen, A. O. Moguche,
C. R. Plumlee, E. M. Agger, K. B. Urdahl and P. Andersen,
Mucosal Immunol., 2017, 10, 555–564.
Conflicts of interest
There are no conflicts of interest to declare.
13 T. Lindenstrøm, E. M. Agger, K. S. Korsholm, P. A. Darrah,
C. Aagaard, R. A. Seder, I. Rosenkrands and P. Andersen,
J. Immunol., 2009, 182, 8047–8055.
Acknowledgements
The authors would like to thank The Royal Society of New
Zealand, Marsden Fund (VUW1401) and The Health Research 14 J. T. van Dissel, S. A. Joosten, S. T. Hoff, D. Soonawala,
Council New Zealand (Hercus Fellowship, B.L.S., 2013/33) for
funding. Additionally, we would like to acknowledge Professor
Sho Yamasaki (Laboratory of Molecular Immunology,
C. Prins, D. A. Hokey, D. M. O’Dee, A. Graves, B. Thierry-
Carstensen and L. V. Andreasen, Vaccine, 2014, 32, 7098–
7107.
Immunology Frontier Research center, Osaka University, Suita 15 E. M. Agger, I. Rosenkrands, J. Hansen, K. Brahimi,
565-0871, Japan) for providing us with the Mincle expressing
2B4 reporter cells and Amy T. Lynch (Immunoglycomic labora-
B. S. Vandahl, C. Aagaard, K. Werninghaus, C. Kirschning,
R. Lang and D. Christensen, PLoS One, 2008, 3, e3116.
tory, Victoria University of Wellington) for drawing blood from 16 I. Rosenkrands, C. Vingsbo-Lundberg, T. J. Bundgaard,
the donors.
T. Lindenstrøm, V. Enouf, S. van der Werf, P. Andersen and
All experimental mice were housed in the animal facility at
E. M. Agger, Vaccine, 2011, 29, 6283–6291.
the Malaghan Institute of Medical Research, Wellington, New 17 T. H. Ottenhoff, T. M. Doherty, J. T. van Dissel, P. Bang,
Zealand, and all murine experimental procedures were
approved by the Victoria University Animal Ethics Committee
K. Lingnau, I. Kromann and P. Andersen, Hum. Vaccines,
2010, 6, 1007–1015.
in accordance with their guidelines for the care of animals 18 P. L. van der Peet, C. Gunawan, S. Torigoe, S. Yamasaki and
(protocol nr 22371). The use of human leukocyte from healthy S. J. Williams, Chem. Commun., 2015, 51, 5100–5103.
donors with written informed consent was approved by New 19 P. L. van der Peet, M. Nagata, S. Shah, J. M. White,
Zealand Northern A Health and Disability Ethics Committee
(approval number 15/NTA/178).
S. Yamasaki and S. J. Williams, Org. Biomol. Chem., 2016,
14, 9267–9277.
20 A. Decout, S. Silva-Gomes, D. Drocourt, S. Barbe, I. André,
F. J. Cueto, T. Lioux, D. Sancho, E. Pérouzel, A. Vercellone,
J. Prandi, M. Gilleron, G. Tiraby and J. Nigou, Proc. Natl.
Acad. Sci. U. S. A., 2017, 114, 2675–2680.
References
1 M. Matsumoto, T. Tanaka, T. Kaisho, H. Sanjo, 21 M. Nagata, Y. Izumi, E. Ishikawa, R. Kiyotake, R. Doi,
N. G. Copeland, D. J. Gilbert, N. A. Jenkins and S. Akira,
S. Iwai, Z. Omahdi, T. Yamaji, T. Miyamoto and T. Bamba,
J. Immunol., 1999, 163, 5039–5048.
Proc. Natl. Acad. Sci. U. S. A., 2017, 114, E3285–E3294.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Organic & Biomolecular Chemistry
Paper
22 T. Ishikawa, F. Itoh, S. Yoshida, S. Saijo, T. Matsuzawa,
T. Gonoi, T. Saito, Y. Okawa, N. Shibata and T. Miyamoto,
Cell Host Microbe, 2013, 13, 477–488.
23 F. Behler-Janbeck, T. Takano, R. Maus, J. Stolper, D. Jonigk,
M. Tort Tarrés, T. Fuehner, A. Prasse, T. Welte,
K. Drickamer and T. B. Poulsen, MedChemComm, 2015, 6,
647–652.
37 A. J. Foster, M. Nagata, X. Lu, A. T. Lynch, Z. Omahdi,
E. Ishikawa, S. Yamasaki, M. S. M. Timmer and
B. L. Stocker, J. Med. Chem., 2018, 61, 1045–1060.
M. S. M. Timmer, B. L. Stocker, Y. Nakanishi, T. Miyamoto, 38 A. J. Smith, S. M. Miller, C. Buhl, R. Child, M. Whitacre,
S. Yamasaki and U. A. Maus, PLoS Pathog., 2016, 12,
R. Schoener, G. Ettenger, D. Burkhart, K. Ryter and
e1006038.
J. T. Evans, Front. Immunol., 2019, 10, 338.
24 S. Yamasaki, M. Matsumoto, O. Takeuchi, T. Matsuzawa, 39 M. Ohtsuka, H. Arase, A. Takeuchi, S. Yamasaki, R. Shiina,
E. Ishikawa, M. Sakuma, H. Tateno, J. Uno, J. Hirabayashi,
Y. Mikami, K. Takeda, S. Akira and T. Saito, Proc. Natl.
Acad. Sci. U. S. A., 2009, 106, 1897–1902.
T. Suenaga, D. Sakurai, T. Yokosuka, N. Arase and
M. Iwashima, Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 8126–
8131.
25 Y. Hattori, D. Morita, N. Fujiwara, D. Mori, T. Nakamura, 40 K. Kodar, J. L. Harper, M. J. McConnell, M. S. M. Timmer
H. Harashima, S. Yamasaki and M. Sugita, J. Biol. Chem.,
2014, 289, 15405–15412.
and B. L. Stocker, Immun., Inflammation Dis., 2017, 5, 503–
514.
26 A. A. Khan, S. H. Chee, R. J. McLaughlin, J. L. Harper, 41 J. Ostrop, K. Jozefowski, S. Zimmermann, K. Hofmann,
F. Kamena, M. S. M. Timmer and B. L. Stocker,
ChemBioChem, 2011, 12, 2572–2576.
E. Strasser, B. Lepenies and R. Lang, J. Immunol., 2015,
195, 2417–2428.
27 B. L. Stocker, A. A. Khan, S. H. Chee, F. Kamena and 42 R. Toubiana, B. C. Das, J. Defaye, B. Mompon and
M. S. Timmer, ChemBioChem, 2014, 15, 382–388. M.-J. Toubiana, Carbohydr. Res., 1975, 44, 308–312.
28 A. Furukawa, J. Kamishikiryo, D. Mori, K. Toyonaga, 43 D. S. Janni and M. K. Manheri, Langmuir, 2013, 29, 15182–
Y. Okabe, A. Toji, R. Kanda, Y. Miyake, T. Ose, S. Yamasaki 15190.
and K. Maenaka, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 44 I. Kuzmenko, M. Kindermann, K. Kjaer, P. B. Howes, J. Als-
17438–17443.
Nielsen, R. Granek, G. V. Kiedrowski, L. Leiserowitz and
29 H. Feinberg, S. A. F. Jégouzo, T. J. W. Rowntree, Y. Guan,
M. Lahav, J. Am. Chem. Soc., 2001, 123, 3771–3783.
M. A. Brash, M. E. Taylor, W. I. Weis and K. Drickamer, 45 S. Livi, J.-F. Gérard and J. Duchet-Rumeau, Chem.
J. Biol. Chem., 2013, 288, 28457–28465. Commun., 2011, 47, 3589–3591.
30 A. Huber, R. S. Kallerup, K. S. Korsholm, H. Franzyk, 46 R. Sugi, A. Yokoyama, T. Furuyama, M. Uchiyama and
B. Lepenies, D. Christensen, C. Foged and R. Lang, Innate
Immun., 2016, 22, 405–418.
31 A. A. Khan, F. Kamena, M. S. Timmer and B. L. Stocker,
Org. Biomol. Chem., 2013, 11, 881–885.
32 K. Kodar, S. Eising, A. A. Khan, S. Steiger, J. L. Harper,
M. S. Timmer and B. L. Stocker, ChemBioChem, 2015, 16,
683–693.
T. Yokozawa, J. Am. Chem. Soc., 2005, 127, 10172–10173.
47 A. A. Khan, S. H. Chee, B. L. Stocker and M. S. Timmer,
Eur. J. Org. Chem., 2012, 995–1002.
48 B. L. Stocker, K. Kodar, K. Wahi, A. J. Foster, J. L. Harper,
D. Mori, S. Yamasaki and M. S. M. Timmer, Glycoconjugate
J., 2019, 36, 69–78.
49 W. E. Holmes, J. Lee, W.-J. Kuang, G. C. Rice and
W. I. Wood, Science, 1991, 253, 1278–1280.
33 J. H. Bird, A. A. Khan, N. Nishimura, S. Yamasaki,
M. S. M. Timmer and B. L. Stocker, J. Org. Chem., 2018, 83, 50 T. Hupfer, J. Schick, K. Jozefowski, D. Voehringer, J. Ostrop
7593–7605. and R. Lang, Front. Immunol., 2016, 7, 423.
34 A. Khan, K. Kodar, M. S. M. Timmer and B. L. Stocker, 51 D. Vijayan, K. J. Radford, A. G. Beckhouse, R. B. Ashman
Tetrahedron, 2018, 74, 1269–1277. and C. A. Wells, Immunol. Cell Biol., 2012, 90, 889–895.
35 H. Feinberg, N. D. Rambaruth, S. A. Jégouzo, 52 J. Schick, P. Etschel, R. Bailo, L. Ott, A. Bhatt, B. Lepenies,
K. M. Jacobsen, R. Djurhuus, T. B. Poulsen, W. I. Weis,
M. E. Taylor and K. Drickamer, J. Biol. Chem., 2016, 291,
21222–21233.
C. Kirschning, A. Burkovski and R. Lang, Infect. Immun.,
2017, 85, e00075–e00017.
53 J. R. Al Dulayymi, M. S. Baird, M. Maza-Iglesias,
R. T. Hameed, K. S. Baols, M. Muzael and A. D. Saleh,
Tetrahedron, 2014, 70, 9836–9852.
36 K. M. Jacobsen, U. B. Keiding, L. L. Clement,
E. S. Schaffert, N. D. S. Rambaruth, M. Johannsen,
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem.