3-O-desacyl monophosphoryl lipid A derivatives: Synthesis and immunostimulant activities

Article abstract of DOI:10.1021/jm990222b

The synthesis of a series of novel analogues of lipid A, the active principle of lipopolysaccharide is reported. In these compounds, the 1-O- phosphono and (R)-3-hydroxytetradecanoyl moieties of native Salmonella minnesota R595 lipid A have been replaced

Full text of DOI:10.1021/jm990222b

4640
J . Med. Chem. 1999, 42, 4640-4649
3-O-Desa cyl Mon op h osp h or yl Lip id A Der iva tives: Syn th esis a n d
Im m u n ostim u la n t Activities†
David A. J ohnson,* David S. Keegan, C. Gregory Sowell, Mark T. Livesay, Craig L. J ohnson, Lara M. Taubner,
Annalivia Harris, Kent R. Myers, J ennifer D. Thompson, Gary L. Gustafson, Michael J . Rhodes, J . Terry Ulrich,
J on R. Ward, Yvonne M. Yorgensen, J ohn L. Cantrell, and Valerie G. Brookshire
Pharmaceutical Discovery Division, Ribi ImmunoChem Research, Inc., 553 Old Corvallis Road, Hamilton, Montana 59840
Received May 4, 1999
The synthesis of a series of novel analogues of lipid A, the active principle of lipopolysaccharide,
is reported. In these compounds, the 1-O-phosphono and (R)-3-hydroxytetradecanoyl moieties
of native Salmonella minnesota R595 lipid A have been replaced with hydrogen and the length
of the normal fatty acyl residues has been systematically varied. Normal fatty acid chain length
in the 3-O-desacyl monophosphoryl lipid A (MLA) series is shown to be a critical determinant
of iNOS gene expression in activated mouse macrophages and the induction of proinflammatory
cytokines in human peripheral monocytes. Examination of pyrogenicity in rabbits and lethal
toxicity in D-galactosamine-treated mice shows that toxic effects in the MLA series can be
ameliorated by modifying fatty acid chain length. When used as an adjuvant for tetanus toxoid
vaccines, certain MLA derivatives enhance the production of tetanus toxoid-specific antibodies
in mice.
In tr od u ction
enhances both antibody and T-cell responses and is an
effective adjuvant in prophylactic and therapeutic vac-
cines.5 However, due to the inherent heterogeneity of
the cognate LPS and incomplete chemoselectivity in the
hydrolytic steps, MPL immunostimulant comprises
several less highly acylated compounds in addition to
hexaacyl component 2.5 Biological evaluation of the
major MPL constituents (either prepared synthetically
or isolated from the naturally derived material) indi-
cates that hexaacylation is important for maximal
immunostimulant activity in the 3-O-desacyl mono-
phosphoryl lipid A (MLA) series.5 This observation
corroborates other studies showing that a â-(1f6)
diglucosamine moiety bearing six pendent fatty acids
is prerequisite to the full expression of endotoxic activi-
ties.6
Gram-negative bacteria have long been known to
possess cell surface components capable of eliciting a
number of physiological and pathological responses in
higher animals.1 The main cell surface component,
lipopolysaccharide (LPS, endotoxin), and its endotoxic
principle, lipid A, are potent adjuvants for protein and
carbohydrate antigens and can markedly enhance both
humoral and cell-mediated immunity.2 However, the
profound pyrogenicity and lethal toxicity of LPS and
lipid A have precluded their use as adjuvants in human
vaccines.2
A few years ago, we demonstrated that the toxic
effects of Salmonella minnesota R595 lipid A (1) could
be ameliorated by selective hydrolysis of the 1-O-
phosphono3 and (R)-3-hydroxytetradecanoyl groups.4
Known as monophosphoryl lipid A (MPL) immuno-
stimulant, S. minnesota lipid A modified in this way
Fatty acid chain length also appears to be an impor-
tant determinant of bioactivity in lipid A molecules.7
However, the structural variability within individual
lipid A or LPS preparations with respect to the number,
type, and position of fatty acyl groups often makes it
difficult to draw definite conclusions about the biological
significance of fatty acid chain length. For example,
while the low endotoxicity observed for certain helico-
bacter and pseudomonas LPS has been attributed
primarily to the presence of underacylated (tetra- and
pentaacyl) lipid A components, significant amounts of
hexaacyl components containing fatty acyl chains which
differ in length from those found in toxic salmonella
lipid A have also been identified in these LPS prepara-
tions.8,9
Although the relationship between acyl chain length
and bioactivity has been investigated to some extent
with synthetic subunit analogues10,11 of lipid A contain-
ing up to three fatty acids and tetraacyl disaccharide
analogues12 of lipid IVa (a biosynthetic precursor of lipid
A exhibiting LPS antagonist properties), to our knowl-
† Dedicated to the memory of Dr. Edgar E. Ribi and in honor of his
seminal contributions to the chemistry and immunochemistry of
bacterial adjuvants.
* Corresponding author. Tel: (406) 363-6214. Fax: (406) 363-6129.
E-mail: dajohn@corixa.com.
10.1021/jm990222b CCC: $18.00 © 1999 American Chemical Society
Published on Web 10/16/1999

3-O-Desacyl MLA Derivatives
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 22 4641
Sch em e 1a
a
Reagents: (a) 5a -f, EDC‚MeI, CH2Cl2; (b) Troc-Cl, DMAP, pyridine, CH2Cl2; then 80% aq AcOH, 60 °C.
Ch a r t 1. (R)-3-n-Alkanoyloxytetradecanoic Acids
12a -f and diols 7a -f. Remarkably, very few silver-
promoted condensation reactions of N-Troc- and other
N-alkoxycarbonyl-protected 2-amino-2-deoxyglycosyl chlo-
rides have been reported.18,19 The TMSEt group was
selected for anomeric protection in the synthesis of the
glycosyl chlorides 12a -f because of the facility with
which it can be converted directly into 1-chloro deriva-
tives.20
edge no systematic study has ever been conducted with
the basic immunostimulatory pharmacophore of lipid A
comprising the obligatory â-(1f6) diglucosamine back-
bone possessing six fatty acids. Accordingly, to clarify
the importance of normal fatty acid chain length within
the MLA series and provide structural information for
the design of synthetically simpler derivatives, we
prepared and evaluated the immunostimulant activity
of a series of chain length homologues of MPL constitu-
ent 2. Compounds 3a -f are derivatives of 2 in which
the (R)-3-hexa-, do-, and tetradecanoyloxytetradecanoyl
residues on the 2, 2’, and 3’ positions, respectively, have
been replaced with identical (R)-3-n-alkanoyloxytet-
radecanoyl residues possessing even-numbered normal
fatty acyl chains containing between 4 and 14 carbon
atoms.
Glycosyl acceptors 7a -f were synthesized as shown
in Scheme 1. N-Acylation of amino alcohol 4 with (R)-
3-n-alkanoyloxytetradecanoic acids 5a -f in the presence
of 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide meth-
iodide (EDC‚MeI) provided amides 6a -f. Hydroxyl
protection with 2,2,2-trichloroethyl chloroformate (Troc-
Cl) and acetonide hydrolysis during workup gave diols
7a -f.
The synthesis of glycosyl donors 12a -f and the
glycosylation of diols 7a -f are presented in Scheme 2.
N-Protection of amino alcohol 8 under Schotten-Bau-
mann conditions with Troc-Cl afforded compound 9 in
nearly quantitative yield. EDC‚MeI-mediated 3-O-acy-
lation of 9 with 5a -f and acetonide hydrolysis provided
compounds 10a -f. The greater primary selectivity of
1,1-dimethyl-2,2,2-trichloroethyl chloroformate (TCBOC-
Cl) vis-a`-vis Troc-Cl17 was then exploited for the one-
pot synthesis of 11a -f by successive treatment of diols
10a -f with TCBOC-Cl and diphenyl chlorophosphate
to give donor progenitors 11a -f in >90% yield. In
comparison, the two-step introduction of Troc and
diphenyl phosphate moieties in 10f proceeds in <60%
yield and entails a tedious chromatographic separation
of regioisomeric Troc carbonates (cf. ref 17).21 Trans-
formation of the TMSEt glycosides 11a -f into the
glycosyl chlorides 12a -f was accomplished directly with
20
R,R-dichloromethyl methyl ether/ZnCl2 in chloroform
Resu lts a n d Discu ssion
in >85% isolated yield. Unlike the corresponding bro-
mides, which are often used without purification in
glycosylation reactions,17,22 the N-Troc-protected glyco-
syl chlorides 12a -f could be chromatographed on silica
gel and stored without any special precautions.
Ch em istr y. The target MLA derivatives 3a -f were
prepared in a highly convergent manner (Schemes 1 and
2) from the known benzyl and 2-(trimethylsilyl)ethyl
(TMSEt, SE) â-glycosides 413 and 8,14 respectively, and
(R)-3-n-alkanoyloxytetradecanoic acids 5a -f 15 (Chart
1). Our strategy, based on our recently reported total
synthesis of the major constituents (e.g., compound 2)
of S. minnesota MPL,16 utilizes the N-2,2,2-trichloroeth-
oxycarbonyl (Troc) method17,18 and silver ion catalysis18
to assemble the key â-(1f6) disaccharide intermediates
13a -f in a stereospecific manner from glycosyl chlorides
Koenigs-Knorr coupling of chlorides 12a -f with
glycosyl acceptors 7a -f was effected in the presence of
silver triflate to give exclusively the â-disaccharides
13a -f. The â-(1f6) linkage was assured by 300 MHz
1H-1H COSY experiments and by analogy with syn-
thetic 2.16 It is noteworthy that this silver-promoted
glycosylation with N-Troc-protected 2-amino-2-deoxy-

4642 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 22
J ohnson et al.
Sch em e 2a
a
Reagents: (a) Troc-Cl, 1 N aq NaHCO3, CH2Cl2; (b) 5a -f, EDC‚MeI, 4-pyrrolidinopyridine, CH2Cl2; then 90% aq AcOH, 60 °C; (c)
TCBOC-Cl, pyridine, CH2Cl2; then (PhO)2P(O)Cl, 4-pyrrolidinopyridine, (i-Pr)2NEt; (d) Cl2CHOMe, ZnCl2 (cat.), CHCl3; (e) 7a -f or 15,
AgOTf, 4 Å molecular sieves, ClCH2CH2Cl; (f) Zn, AcOH, 60 °C; (g) 5a -f, EEDQ, CH2Cl2; (h) 5f, EEDQ or EDC‚MeI, CH2Cl2; (i) Pd black,
AcOH, THF, 70 psig H2; then PtO2, 70 psig H2.
glucopyranosyl chlorides proceeds in 70-80% yield at
room temperature with only a slight excess (1.1 equiv)
of the glycosyl donor 12, whereas the corresponding
glycosyl bromide/Hg(CN)2 protocol17 requires a larger
excess of the (less-stable bromide) donor and elevated
temperatures to achieve comparable yields (see also
preparation of 16f in Experimental Section). Reductive
cleavage of the trichloroethyl-based protecting groups
of 13a -f with Zn/AcOH and N-acylation with (R)-3-n-
alkanoyloxytetradecanoic acids 5a -f in the presence of
2-ethoxy-1-ethoxycarbonyl-1,2-dihydroquinoline (EEDQ)
furnished disaccharides 14a -f in 50-65% yield.
Because the N-acyl groups of the two glucosamine
units are the same in the target compounds 3a -f, we
also examined the simultaneous introduction of the two
amide-linked acyl groups after disaccharide formation
in the synthesis of 14f. Glycosylation of the known bis-
Troc derivative 1523 with glycosyl chloride 12f (or the
corresponding bromide) gave disaccharide 16f. Depro-
tection of 16f and bis-N-acylation with (R)-3-tetrade-
canoyloxytetradecanoic acid (5f) in the presence of
EEDQ, however, provided the hexaacylated derivative
14f in only 25-30% yield. Since prolonged reaction
times (>16 h) in the bis-acylation and the use of
carbodiimide and other coupling protocols led to even
lower yields of 14fspartly because of 4,6-cyclic phos-
phate byproduct formation24sthis synthetic approach
was not pursued.
reverse-phase HPLC analysis25 of O-3,5-dinitrobenzyl
oxime derivatives. The 1H NMR spectra for 3a -f
showed the anomeric protons of the â-glycosides at ca.
δ 4.6 (d, J ∼ 8 Hz) and those of the reducing sugars at
ca. δ 5.1 (d, J ∼ 3.5 Hz) and 4.5 (d, J ∼ 8 Hz) in a 2:1
ratio, respectively, indicating the R-D-pyranose form of
3 predominates in solution.
Biology. Identifying analogues of lipid A possessing
more favorable activity/toxicity profiles requires suitable
models for evaluating toxicity and efficacy. As repre-
sentative efficacy models for immunostimulation, the
MLA derivatives 3a -f were compared to MPL immu-
nostimulant and major MPL component 2 for their
ability to induce nitric oxide synthase (iNOS) in murine
macrophages and cytokine production in human periph-
eral monocytes (Table 1). The induction of iNOS in
activated macrophages correlates directly with mac-
rophage cytotoxicity presumably, in part, via the de-
structive action of the radical nitric oxide (NO) on
microbial DNA and membranes.26 The production of the
proinflammatory cytokines tumor necrosis factor-R (TNF-
R) and interleukin-1â (IL-1â) by activated monocytes is
known to enhance normal host resistance to infection
and induce a cascade of other mediators whose functions
include T-cell activation and antibody synthesis.27 The
ex vivo stimulation of human whole blood provides a
simple model of local cytokine production which has
relevance to the microenvironment of an administered
vaccine.28 Certain MLA derivatives were further evalu-
ated for their ability to enhance IgG antibody responses
to tetanus toxoid vaccines in mice. The expression of
the complement fixing IgG2a and IgG2b subclasses is
thought to be important in antibody-dependent cellular
cytotoxicity and the clearance of infectious pathogens.29
Two-step hydrogenolytic deprotection17 of the benzyl
and phenyl protecting groups in 14a -f and lyophiliza-
tion of the free acids from 1-2% aqueous triethylamine
gave the desired MLA derivatives 3a -f as triethylam-
monium salts. These compounds were characterized by
NMR, negative ion FAB-MS, elemental analysis, and

3-O-Desacyl MLA Derivatives
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 22 4643
Ta ble 1. Immunostimulant Activities of MLA Derivativesa
pyrogenicity
relative ex vivo cytokine inductionc
TNF-R IL-1â
iNOS
fever
responsed
(°C)
lethal
inductionb
ED50 (ng/mL)
MPDe
(µg/kg)
toxicityf
LD50 (µg)
compd
n
3a
3b
3c
3d
3e
3f
2
4
6
8
10
12
>10000g
NDh
NDh
NDh
NDh
7.3 (4.1)
2.9 (3.6)
3.4 (3.1)
1.4 (2.1)
1.0
1.5
1.7
3.9
4.1
5.3
0.3
0.6
2.0
10
10
4.5
4.0
0.2
0.02
0.04
1.0
1.0
0.1
>10000g
NDh
0.3
0.05
1
5
10
2
NDh
(e2.5)i
(e0.06)i
(e2.5)i
15
3.0 (5.0)
4.6 (3.9)
1.7 (0.8)
1.6 (2.1)
1.0
2
15
5
MPL
a
b
All compounds were tested as triethylammonium salts. See the Experimental Section for details. ED50 values represent the
concentration of the test article required to produce a half-maximal response. c Human whole blood was stimulated with 10 µg/mL of the
test sample and analyzed for TNF-R and IL-1â by a sandwich ELISA; values in parentheses are relative responses at a concentration of
5 µg/mL. Total temperature rise for three rabbits at a dose of 10 µg/kg. e Minimum pyrogenic dose. f Simultaneous intravenous
d
administration of test article and D-galactosamine. g No response at this highest dose tested. ND, not detected; relative cytokine induction
h
i
e0.05. Pyrogenic at this dose; MPD not determined.
Febrile responses in rabbits and lethal toxicity in
D-galactosamine-sensitized mice are two models in
which the toxicity of immunostimulatory agents related
to lipid A has been evaluated.4 Accordingly, febrile
responses for MLA derivatives 3a -f, synthetic 2, and
naturally derived MPL were determined in a standard
three-rabbit pyrogen test at a dose of 10 µg/kg, and the
minimal pyrogenic dose (MPD) for each was assessed
by multiple-dose testing. Due to the relative insensitiv-
ity of normal mice to lipid A-like materials, lethal
toxicity was measured by intravenous injection of the
test substances admixed with D-galactosamine.30
sessing significantly different chain lengths but similar
acylation patterns and hydrophobic character make it
difficult to draw definite conclusions about the underly-
ing importance of hydrophobicity. In general, however,
increasing or decreasing the number of acyl residues
present in disaccharide lipid A derivatives from the
optimum of six (three in monosaccharides) reduces
endotoxic activity.6,11 Consistent with this observation,
underacylated components present in MPL immuno-
stimulant (and also presumably generated from 2 in
vivo by acyloxyacyl hydrolase-mediated hydrolysis32 of
the normal fatty acids) exhibit less iNOS5 and cytokine
activity33 than parent compound 2. Additionally, the
known34 monosaccharide GLA-47, corresponding to the
(tetraacylated) nonreducing sugar portion of compound
3f, is devoid of activity in both these models.35 These
data support the concept that the hydrophobe-hydro-
phile balance within the homologous MLA series is
involved in the expression of iNOS and proinflammatory
cytokines.
As shown in Table 1, the induction of both iNOS and
proinflammatory cytokines shows a profound bimodal
dependence on the length of the normal fatty acid
chains, reaching a maximum when n ) 8 (3d ) in each
case. The iNOS response is more sensitive than cytokine
induction to variations in chain length, showing a 100-
fold difference in potency between 3d (which possesses
a lower ED50 or greater macrophage-stimulating ability)
and 3f, but both models show a similar threshold chain
length for activity: iNOS and cytokine responses are
abolished when n ) 4 (3b) and n ) 6 (3c), respectively,
and for shorter chain derivatives. Thus, 3a and 3b are
inactive in both models, whereas 3c induces iNOS in
murine macrophage but not cytokines in human mono-
cytes. The precipitous decline in bioactivity for 3b and
3c in these two models is indicative of strict but slightly
different (and possibly species-related) structural and/
or conformational requirements for the two bioactivi-
ties.31
Lethal toxicity in the D-galactosamine model and
pyrogenicity in rabbits generally parallel iNOS and
cytokine responses (Table 1). The highly immunoactive
3d , which shows the greatest iNOS, TNF-R, and IL-1â
inducing ability, also exhibits the greatest lethal toxicity
(lowest LD50) in D-galactosamine-sensitized mice as well
as high pyrogenicity (MPD e 0.06 µg/kg) in comparison
to 3f and other MLA derivatives. Intermediate chain
compounds 3c and 3e also exhibit greater iNOS and
pyrogenic activity than 3f. But short chain derivatives
3a and 3b, which elicit no measurable activity in the
iNOS and cytokine models, show only about a 4-fold
decrease in lethal toxicity compared to 3f and are also
more pyrogenic. Given the role of TNF-R and IL-1â in
inflammatory and fever responses,36,37 the potent cy-
tokine-stimulating ability of 3f is particularly notewor-
thy in view of its reduced lethal toxicity and pyrogenicity
relative to those of 3d , 3e, and MPL immunostimulant,
suggesting that the structural requirements for benefi-
cial and detrimental endotoxic activities are signifi-
cantly different for the MLA series of compounds.38
Similarly, a comparison of the immunostimulant activi-
ties of MPL and MPL component 2 indicates that the
beneficial activity of MPL can be attributed primarily
to this major, hexaacyl congener and that the increased
relative toxicity of MPL is probably due to the presence
Interestingly, 3f and the major MPL constituent 2,
which contain the same absolute number but different
distribution of carbon atoms in the lipid side chains,
exhibit comparable activities, suggesting that bioactivity
of MLA derivatives in these two models may be gov-
erned more by the overall carbon count (i.e., hydropho-
bicity) than by the position of a particular acyloxyacyl
residue on the â-(1f6) diglucosamine moiety. The
hydrophobe-hydrophile balance in certain monosac-
charide lipid A analogues is known to play a more
important role than fatty acid position in NO and
cytokine production by murine macrophages.11 Never-
theless, the small structural differences between 3f and
2 coupled with the lack of bioactivity data for either
naturally derived or synthetic MLA derivatives pos-

4644 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 22
J ohnson et al.
Ta ble 2. Adjuvant Activity of MLA Derivativesa
side chains of lipid A-like compounds induce peculiar
conformations which dramatically affect cellular activa-
tion and the expression of endotoxic activities.7
tetanus toxoid vaccineb
compd
n
total Ig
IgG1
IgG2a
IgG2b
In conclusion, we have shown that normal fatty acid
chain length in the 3-O-desacyl MLA series of com-
pounds is a critical determinant of immunostimulant
activity and that toxic effects can be ameliorated by
modifying fatty acid chain length. The nearly identical
immunostimulation profiles obtained for 3f and major
MPL component 2 indicate that fatty acid dissimilitude
present in natural lipid A is not essential for immune
stimulation or favorable activity/toxicity profiles in the
MLA series of lipid A derivatives. Due to the potent
adjuvant activity of 3d and 3f in tetanus toxoid vaccines,
the adjuvanticity of these and other MLA derivatives
is currently being evaluated in infectious models. Cur-
rent efforts are also being directed toward the synthesis
and biological testing of synthetically simpler MLA
derivatives comprising three identical (R)-3-n-alkanoyl-
oxytetradecanoyl residues and structural mimetics of
the â-(1f6) diglucosamine moiety.35
3d
3f
MPL
control
8
12
6.2
8.2
5.7
1.0
2.4
4.5
2.5
1.0
6.0
14.6
3.4
4.3
4.5
1.6
1.0
1.0
a
All compounds were tested as triethylammonium salts. See
b
the Experimental Section for details. Mice were immunized on
day 0 and again on day 21 with 2.5% oil-water emulsions
containing test articles admixed with tetanus toxoid. Serum
samples were collected 14 days after the second immunization and
evaluated by ELISA analysis; values given represent experimental
test titers divided by 2.5% oil-water control titer.
of less highly acylated components.5 This latter observa-
tion is somewhat surprising in view of the postulate that
selective deacylation of the normal fatty acids in struc-
turally diverse (3-O-acyl) lipid A molecules in vivo by
human acyloxyacyl hydrolase has evolved as a defense
mechanism to reduce lipid A toxicity.32
Since the release of mediators such as TNF-R, IL-1â,
and NO is integral to the initiation of immune re-
sponses,26,27,35 the ability of mediator inducers 3d , 3f,
and MPL to enhance antibody responses to tetanus
toxoid vaccines was compared in a murine model (Table
2). Both 3d and 3f generally induce higher tetanus
toxoid-specific antibodies of all classes of immunoglo-
bulins tested than MPL. But compound 3f, which is 100
times less active than 3d in the iNOS model, exhibits
the highest total Ig response and produces more comple-
ment fixing IgG2a and IgG2b antibodies than either 3d
or MPL. (In a separate experiment, compound 2 exhib-
ited adjuvant activity similar to 3f, eliciting IgG2a
antibody titers to tetanus toxoid 4 times greater than
MPLsdata not shown). Although normal fatty acid
chain length clearly plays an indispensable role in the
expression of various endotoxic activities, the structural
requirements for adjuvant activity in tetanus toxoid
vaccines appear to be slightly different than those for
iNOS and cytokine induction. Nevertheless, it is note-
worthy that both relative cytokine induction and teta-
nus toxoid adjuvant activity for 3d and 3f are of the
same order of magnitude, whereas the toxicity of 3f as
measured by LD50 and MPD is 50 and >250 times lower,
respectively, indicating that toxic and beneficial endo-
toxic activities can at least be partly dissociated in the
MLA series.
Exp er im en ta l Section
Ch em istr y. Gen er a l Meth od s. Melting points were de-
termined with a Mel-Temp capillary melting-point apparatus
and are uncorrected. Elemental analyses (C, H, N) were
performed by Atlantic Microlabs, Norcross, GA, or Galbraith
Laboratories, Knoxville, TN. Total phosphorus was determined
by the method of Bartlett.40 High-resolution FAB mass spectra
were recorded at the University of California Riverside (UCR)
Mass Spectrometry Facility. TLC was performed on silica gel
60 F254 glass-backed plates (Merck) with visualization under
UV light and/or by staining with phosphomolybdate, anisal-
dehyde, or vanillin reagent. Flash chromatography was per-
formed on silica gel 60 (Merck, 230-400 mesh). Optical
rotations were determined with a J asco DIP-140 or DIP-1000
digital polarimeter. 1H NMR spectra were recorded on a
Bruker HX-270 spectrometer (270 MHz) or a Varian Gemini
300 spectrometer (300 MHz) using Me4Si as an internal
standard and CDCl3 as the solvent unless otherwise indicated;
the NH and OH protons of 3a -f and adventitious H2O were
preexchanged with deuterium (methanol-d4) prior to obtaining
the spectra. 13C NMR spectra were recorded at 75 MHz in
CDCl3 on the Varian spectrometer. Compounds 3a -f were
derivatized with 3,5-dinitrobenzyloxyamine (DNBA) and ana-
lyzed with a Waters HPLC system on a Novak C18 (4 µm) 3.9
× 300-mm column according to published procedures.25 Samples
were eluted with one of two binary gradient systems at a flow
rate of 1.0 mL/min: HPLC-1, buffer A ) 5 mM TBAP in 95%
MeCN-H2O, buffer B ) 5 mM tetra-n-butylammonium phos-
phate (TBAP) in 95% i-PrOH-H2O, linear gradient of 10% B
to 80% B over 45 min; and HPLC-2, buffer A ) 5 mM TBAP
in 85% MeCN-H2O, buffer B ) 5 mM TBAP in 95% i-PrOH-
H2O, convex gradient (Waters gradient curve no. 5) of 0% B
to 90% B over 55 min.
All starting materials and reagents were commercially
available and used without purification unless otherwise
indicated. Solvents were dried according to standard methods.
Ben zyl 2-Deoxy-4,6-O-isop r op ylid en e-2-[(R)-3-tetr a d e-
can oyloxytetr adecan oylam in o]-â-D-glu copyr an oside (6f).
A solution of compound 413 (0.432 g, 1.40 mmol) and 5f15 (0.70
g, 1.54 mmol) in CH2Cl2 (15 mL) was treated with EDC‚MeI
(0.62 g, 2.10 mmol) and stirred overnight at room temperature.
The reaction mixture was concentrated and the resulting
residue purified by flash chromatography on silica gel (gradi-
ent elution, 35f45% EtOAc/hexanes) to give 0.89 g (85%) of
compound 6f as a colorless solid: mp 68-71 °C; [R]D27-47.6°
(c 0.84, CHCl3); 1H NMR δ 7.33 (br s, 5H), 5.97 (d, 1H, J ) 6.0
Hz), 5.06 (m, 1H), 4.88 (d, 1H, J ) 11.8 Hz), 4.69 (d, 1H, J )
8.3 Hz), 4.58 (d, 1H, J ) 11.8 Hz), 4.0-3.75 (m, 3H), 3.61 (∼t,
1H, J ∼ 9 Hz), 3.51 (m, 1H), 3.30 (td, 1H, J ) 10, 5 Hz), 2.42
The results from the present study and earlier work
also suggest that fatty acid structure may be more
important than the presence or absence of the 1-phos-
phate and 3-hydroxytetradecanoate groups in determin-
ing the biological activity of lipid A-like molecules. For
example, the reduced toxicity5 of MPL component 2
relative to that of S. minnesota R595 lipid A (1) cannot
be reconciled with the high pyrogenicity and lethal
toxicity of MLA derivative 3d on the basis of chemical
modification (deacylation and dephosphorylation) alone.
Further, it has been reported that certain rhodobacter
lipid A variants which possess both an anomeric phos-
phate and a â-hydroxy fatty acyl group on the 3-position
but unique fatty acid chains are not only low in toxicity
but effective antagonists of enterobacterial lipid A as
well.39 These observations are consistent with the
postulate that subtle modifications to the hydrophobic

3-O-Desacyl MLA Derivatives
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 22 4645
(m, AB type, 2H), 2.25 (t, 2H, J ) 7.5 Hz), 1.53 (s, 3H), 1.44
(s, 3H), 1.7-1.1 (m, 42H), 0.88 (∼t, 6H); 13C NMR δ 174.0,
171.3, 136.9, 128.5, 128.1, 99.8, 99.6, 74.1, 71.9, 71.4, 71.1, 67.3,
62.0, 59.1, 42.6, 34.5, 31.9, 29.7, 29.5, 29.4, 29.3, 29.2, 29.1,
25.3, 25.0, 22.7, 19.1, 14.2; positive FAB-MS calcd for [M +
Na]+ 768.5390, found 768.5427. Anal. (C44H75NO8‚0.25H2O) C,
H, N.
room temperature. The reaction mixture was filtered through
a short pad of Celite filter agent and concentrated. The crude
product obtained was used without further purification. Pu-
rification of a separate sample by flash chromatography on
silica gel (gradient elution, 8f12% EtOAc/hexanes) afforded
pure 2-(trimethylsilyl)ethyl 2-deoxy-4,6-O-isopropylidene-3-O-
[(R)-3-tetradecanoyloxytetradecanoyl]-2-(2,2,2-trichloroethoxy-
carbonylamino)-â-D-glucopyranoside (92%) as
a colorless
Ben zyl 2-Deoxy-2-[(R)-3-tetr a d eca n oyloxytetr a d eca n -
oyla m in o]-3-O-(2,2,2-t r ich lor oet h oxyca r b on yl)-â-D-glu -
cop yr a n osid e (7f). A solution of compound 6f (0.74 g, 0.99
mmol), 4-(dimethylamino)pyridine (DMAP; 0.012 g, 0.098
mmol), and pyridine (0.20 mL, 2.5 mmol) in anhydrous CH2-
Cl2 (6 mL) was treated with Troc-Cl (0.15 mL, 1.1 mmol),
stirred at room temperature for 3 h, and then concentrated.
The crude product obtained was used without further purifica-
tion. Purification of a separate sample by flash chromatogra-
phy on silica gel (gradient elution, 10f15% EtOAc/hexanes)
afforded pure benzyl 2-deoxy-4,6-O-isopropylidene-2-[(R)-3-tet-
radecanoyloxytetradecanoylamino]-3-O-(2,2,2-trichloroethoxy-
carbonyl)-â-D-glucopyranoside as a colorless solid: mp 108-
109.5 °C; [R]D27-19.8° (c 1.05, CHCl3); 1H NMR δ 7.4-7.2 (m,
5H), 5.88 (d, 1H, J ) 8.2 Hz), 5.21 (∼t, 1H, J ∼ 9.5 Hz), 4.98
(m, 1H), 4.92 (d, 1H, J ) 8.2 Hz), 4.87 (d, 1H, J ) 12 Hz), 4.77
(m, AB type, 2H), 4.57 (d, 1H,J ) 12 Hz), 3.98 (dd, 1H, J )
11, 5.5 Hz), 3.87-3.65 (m, 3H), 3.40 (m, 1H), 2.39 (dd, 1H, J
) 14.8, 6.0 Hz), 2.26 (dd, 1H partly obscured by following
triplet, J ) 14.8, 5.6 Hz), 2.24 (t, 2H, J ) 7.5 Hz), 1.47 (s, 3H),
1.36 (s, 3H), 1.65-1.15 (m, 42H), 0.88 (∼t, 6H); 13C NMR δ
173.6, 169.8, 153.8, 136.8, 128.4, 127.9, 99.7, 99.6, 94.5, 71.9,
71.3, 71.1, 71.1, 70.9, 66.9, 62.0, 55.8, 41.9, 34.5, 34.1, 32.0,
29.7, 29.6, 29.4, 29.2, 29.0, 25.3, 25.0, 22.7, 19.0, 14.2. Anal.
(C47H77Cl3NO10‚1H2O) C, H, N.
gum: [R]D26-14° (c 0.89, CHCl3); 1H NMR δ 5.33-5.13 (m, 3H),
4.82-4.62 (m, 3H), 4.00-3.88 (m, 2H), 3.80 (∼t, 1H, J ∼ 10.5
Hz), 3.70 (∼t, 1H, J ∼ 9.5 Hz), 3.63-3.41 (m, 2H), 3.41-3.28
(m, 1H), 2.62 (dd, 1H, J ) 14, 6.9 Hz), 2.52 (dd, 1H, J ) 14,
5.7 Hz), 2.29 (t, 2H, J ) 7.5 Hz), 1.47 (s, 3H), 1.37 (s, 3H),
1.45-1.15 (m, 42H), 1.0-0.83 (m, 8H), 0.0 (s, 9H); 13C NMR δ
173.4, 169.9, 154.0, 101.0, 99.6, 74.5, 71.8, 71.5, 70.1, 67.8, 67.2,
62.1, 57.3, 39.4, 34.5, 33.9, 32.0, 29.7, 29.6, 29.4, 29.2, 29.0,
25.2, 25.0, 22.7, 19.0, 18.2, 14.2, -1.4. Anal. (C45H82Cl3NO10
Si) C, H, N.
-
A solution of the crude acetonide in 90% aq AcOH was
heated at 60 °C for 1 h and then concentrated. Flash chroma-
tography on silica gel (gradient elution, 30f40% EtOAc/
hexanes) gave 11.8 g (83%) of compound 10f as a colorless
gum: [R]D26-8.2° (c 1.14, CHCl3); 1H NMR δ 5.20 (d, 1H, J )
7.4 Hz), 5.15-5.0 (m, 2H), 4.77 (d, 1H, J ) 11 Hz), 4.65 (d,
1H, J ) 11 Hz), 4.57 (d, 1H, J ) 8.3 Hz), 4.03-3.78 (m, 3H),
3.72-3.36 (m, 4H), 2.52 (m, 2H), 2.30 (t, 2H, J ) 7.4 Hz), 1.7-
1.1 (m, 42H), 0.9 (m, 8H), 0.0 (s, 9H); 13C NMR δ 174.4, 171.0,
154.5, 100.6, 95.6, 75.5, 75.3, 74.3, 70.6, 68.7, 67.6, 61.7, 56.1,
39.4, 34.4, 34.2, 31.8, 29.5, 29.4, 29.2, 29.0, 25.0, 24.9, 22.6,
18.1, 14.0, -1.6; positive FAB-MS calcd for [M + Na]+
912.4358, found 912.4321. Anal. (C42H78Cl3NO10Si) C, H, N.
2-(Tr im eth ylsilyl)eth yl 2-Deoxy-4-O-d ip h en ylp h osp h o-
n o-3-O-[(R)-3-tetr a d eca n oyloxytetr a d eca n oyl]-6-O-(2,2,2-
t r ich lor o-1,1-d im et h ylet h oxyca r b on yl)-2-(2,2,2-t r ich lo-
r oeth oxyca r bon yla m in o)-â-D-glu cop yr a n osid e (11f). A
solution of compound 10f (10.9 g, 12 mmol) and pyridine (2
mL, 25 mmol) in CH2Cl2 (125 mL) at 0 °C was treated dropwise
over 15 min with a solution of TCBOC-Cl (3.17 g, 13.2 mmol)
in CH2Cl2 (25 mL). The reaction mixture was allowed to warm
slowly to ambient temperature over 3.5 h. 4-Pyrrolidinopyri-
dine (0.89 g, 6.0 mmol), N,N-diisopropylethylamine (10.5 mL,
60 mmol), and diphenyl chlorophosphate (3.7 mL,18 mmol)
were added sequentially and the resulting mixture was stirred
at room temperature for 5 h. The reaction mixture was diluted
with CH2Cl2, washed with cold 7.5% aq HCl and saturated aq
NaHCO3, dried (Na2SO4), and concentrated. Flash chroma-
tography on silica gel (12.5% EtOAc/hexanes) gave 15.1 g (95%)
A solution of the crude acetonide in 80% aq AcOH (15 mL)
was heated at 60 °C for 1 h. Volatiles were removed under
reduced pressure to give a syrup, which was purified by flash
chromatography (gradient elution, 50f55% EtOAc/hexanes)
to give 0.71 g (81%) of compound 7f as a colorless solid: mp
84-85 °C; [R]D23-21.0° (c 1.0, CHCl3); 1H NMR δ 7.4-7.2 (m,
5H), 5.94 (d, 1H, J ) 8.2 Hz), 5.17 (∼t, 1H, J ∼ 9.5 Hz), 5.01
(m, 1H), 4.93 (d, 1H, J ) 8.3 Hz), 4.86 (d, 1H, J ) 12 Hz), 4.77
(s, 2H), 4.62 (d, 1H, J ) 12 Hz), 3.98-3.65 (m, 3H), 3.47 (m,
1H), 2.64 (d, 1H, J ) 5 Hz), 2.42 (dd, 1H, J ) 14, 6 Hz), 2.30
(dd, 1H partly obscured by following triplet, J ) 14, 6 Hz),
2.25 (t, 2H, J ) 7.5 Hz), 1.98 (br s, 1H), 1.65-1.1, (m, 42H),
0.88 (∼t, 6H); 13C NMR δ 173.6, 169.9, 154.4, 136.9, 128.4,
128.0, 127.9, 99.3, 94.3, 79.8, 75.0, 71.3, 71.0, 69.6, 62.2, 55.2,
41.9, 34.5, 34.1, 32.0, 29.7, 29.6, 29.4, 29.2, 25.3, 25.0, 22.7,
14.2; positive FAB-MS calcd for [M + Na]+ 902.4120, found
902.4144. Anal. (C44H72Cl3NO10) C, H, N.
26
of compound 11f as a viscous oil: [R]D +5.3° (c 0.97, CHCl3);
1H NMR δ 7.8-7.2 (m, 10H), 5.70-5.54 (m, 2H), 5.21 (m, 1H),
5.03 (d, 1H, J ) 7.9 Hz), 4.86 (d, 1H, J ) 12 Hz), 4.66 (∼q,
1H, J ∼ 9 Hz), 4.58 (d, 1H, J ) 12 Hz), 4.34 (d, 1H, J ) 12
Hz), 4.27 (dd, 1H, J ) 12, 5 Hz), 3.95 (m, 1H), 3.81 (m, 1H),
3.37 (m, 1H), 3.34 (∼q, 1H, J ∼ 8 Hz), 2.45-2.15 (m, 4H), 1.90
(s, 3H), 1.83 (s, 3H), 1.65-1.1 (m, 42H), 1.0-0.8 (m, 8H), 0.0
(s, 9H); 13C NMR δ 173.7, 169.8, 153.9, 151.8, 150.2, 129.8,
125.6, 120.1, 120.0, 99.2, 90.0, 74.7, 74.6, 74.4, 71.7, 70.1, 67.7,
65.6, 57.1, 39.8, 34.4, 34.2, 31.9, 29.6, 29.5, 29.3, 29.2, 25.0,
22.7, 21.1, 21.0, 18.1, 14.1, -1.4; positive FAB-MS calcd for
Compounds 7a -e were prepared from 4 and 5a -e15 as
described above for the preparation of 7f from 4 and 5f.
2-(Tr im eth ylsilyl)eth yl 2-Deoxy-4,6-O-isopr opyliden e-2-
(2,2,2-t r ich lor oet h oxyca r b on yla m in o)-â-D-glu cop yr a n -
osid e (9). To a solution of compound 814 (6.46 g, 20.2 mmol)
in CHCl3 (300 mL) was added 1 N aq NaHCO3 (300 mL) and
Troc-Cl (8.5 g, 40 mmol). The resulting mixture was stirred
vigorously at room temperature for 3 h. The organic layer was
separated, dried (Na2SO4), and concentrated. Flash chroma-
tography on silica gel (gradient elution, 30f40% EtOAc/
hexanes) afforded 9.6 g (96%) of compound 9 as a colorless
solid: mp 69-70 °C; [R]D25-34° (c 1.40, CHCl3); 1H NMR δ
5.39 (d, 1H, J ) 7.4 Hz), 4.74 (m, 2H), 4.65 (d, 1H, J ) 8.3
Hz), 4.08-3.88 (m, 3H), 3.79 (∼t, 1H, J ∼ 10.5 Hz), 3.62-3.48
(m, 2H), 3.37-3.23 (m, 2H), 2.94 (br s, 1H), 1.52 (s, 3H), 1.44
(s, 3H), 0.94 (m, 2H), 0.0 (s, 9H); 13C NMR δ 154.4, 100.3, 99.8,
95.5, 74.6, 74.2, 71.2, 67.7, 67.0, 62.0, 59.3, 29.1, 19.1, 18.2,
-1.3; positive FAB-MS calcd for [M + Na]+ 516.0755, found
516.0760. Anal. (C17H30Cl3NO7Si) C, H, N.
[M + Na]+ 1346.4003, found 1346.3979. Anal. (C59H92Cl6NO15
PSi) C, H, N.
-
2-Deoxy-4-O-d ip h en ylp h osp h on o-3-O-[(R)-3-t et r a d e-
ca n oyloxytetr a d eca n oyl]-6-O-(2,2,2-tr ich lor o-1,1-d im eth -
yleth oxyca r bon yl)-2-(2,2,2-tr ich lor oeth oxyca r bon yla m i-
n o)-r-D-glu cop yr a n osyl Ch lor id e (12f). A solution of
compound 11f (6.5 g, 4.84 mmol) and dichloromethyl methyl
ether (2.18 mL, 24.2 mmol) in CHCl3 (60 mL) at 0 °C was
treated with ZnCl2 (1.0 M in ether; 2.41 mL, 2.41 mmol) and
then allowed to warm slowly and stir at room temperature
overnight. The reaction mixture was diluted with EtOAc,
washed with saturated aq NaHCO3, dried (Na2SO4), and
concentrated. Flash chromatography on silica gel (10% EtOAc/
hexanes) afforded 5.4 g (88%) of compound 12f as a colorless
gum: [R]D23-49.9° (c 1.0, CHCl3); 1H NMR δ 7.4-7.1 (m, 10H),
6.26 (d, 1H, J ) 3.6 Hz), 5.79 (d, 1H, J ) 8.0 Hz), 5.51 (∼t,
2-(Tr im eth ylsilyl)eth yl 2-Deoxy-3-O-[(R)-3-tetr adecan o-
yloxyt et r a d eca n oyl]-2-(2,2,2-t r ich lor oet h oxyca r b on yl-
a m in o)-â-D-glu cop yr a n osid e (10f). A solution of compound
9 (7.5 g, 15.2 mmol), 5f15 (7.58 g, 16.7 mmol), and 4-pyrroli-
dinopyridine (0.25 g, 1.7 mmol) in CH2Cl2 (95 mL) was treated
with EDC‚MeI (4.94 g, 16.7 mmol) and stirred overnight at

4646 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 22
J ohnson et al.
1H, J ∼ 10 Hz), 5.09 (m, 1H), 4.83 (∼q, 1H, J ∼ 9 Hz), 4.73 (d,
1H, J ) 12 Hz), 4.70 (d, 1H, J ) 12 Hz), 4.4-4.2 (m, 4H), 2.52-
2.34 (m, AB type, 2H), 2.18 (t, 2H, J ) 7.6 Hz), 1.88 (s, 3H),
1.78 (s, 3H), 1.6-1.1 (m, 42H), 0.88 (∼t, 6H); 13C NMR δ 173.1,
170.8, 154.2, 151.7, 150.2, 129.9, 125.7, 120.1, 95.1, 92.6, 90.2,
74.7, 72.9, 71.0, 71.0, 69.8, 69.6, 67.4, 64.4, 55.6, 39.0, 34.4,
34.1, 32.0, 29.7, 29.6, 29.4, 29.2, 25.1, 25.0, 22.7, 21.1, 21.0,
14.2; positive FAB-MS calcd for [M + Na]+ 1264.2956, found
1264.2896. Anal. (C54H79Cl7NO14P) C, H, N.
2H), 3.63 (∼q, 1H, J ∼ 9.5 Hz), 3.51-3.35 (m, 2H), 2.37 (m,
AB type, 2H), 2.24 (t, 2H, J ) 7 Hz), 1.90 (s, 3H), 1.84 (s, 3H),
1.6-1.05 (m, 42H), 0.88 (∼t, 6H). Anal. (C73H99Cl12N2O23P) C,
H, N.
Ben zyl 2-Deoxy-6-O-[2-deoxy-4-O-diph en yph osph on o-3-
[(R)-3-tetr a d eca n oyloxytetr a d eca n oyl]-2-[(R)-3-tetr a d e-
ca n oyloxyt et r a d eca n oyla m in o]-â-D-glu cop yr a n osyl]-2-
[(R)-3-tetr a d eca n oyloxytetr a d eca n oyla m in o]-â-D-glu co-
p yr a n osid e (14f). Meth od A. A solution of compound 13f
(4.35 g, 2.08 mmol) in AcOH (100 mL) at 60 °C was treated
with zinc dust (10 g, 0.16 mol) in three equal portions over a
1-h period. The cooled reaction mixture was sonicated, filtered
through a pad of Celite, and concentrated. The resulting
residue was partitioned between CH2Cl2 and saturated aq
NaHCO3, and the layers were separated. The organic layer
was dried (Na2SO4) and concentrated. A solution of the crude
triol obtained and compound 5f15 (1.13 g, 2.5 mmol) in CH2Cl2
(20 mL) was stirred with powdered 4 Å molecular sieves (0.2
g) for 0.5 h and then treated with EEDQ (0.76 g, 3.12 mmol).
The resulting mixture was stirred at room temperature for 5
h, filtered through Celite, and concentrated. Flash chroma-
tography (gradient elution, 1.5f2% MeOH-CHCl3) gave 2.05
g (50%) of compound 14f as a colorless syrup: [R]D26-23.4° (c
Compounds 12a -e were prepared from 9 and 5a -e15
following the sequence of reactions described above for the
preparation of 12f from 9 and 5f.
Ben zyl 2-Deoxy-6-O-[2-d eoxy-4-O-d ip h en ylp h osp h on o-
3-O-[(R)-3-tetr a d eca n oyloxytetr a d eca n oyl]-6-O-(2,2,2-tr i-
ch lor o-1,1-dim eth yleth oxycar bon yl)-2-(2,2,2-tr ich lor oeth -
oxycar bon ylam in o)-â-D-glu copyr an osyl]-3-O-(2,2,2-tr ich lo-
r oeth oxyca r bon yl)-2-[(R)-3-tetr a d eca n oyloxytetr a d eca n -
oyla m in o]-â-D-glu cop yr a n osid e (13f). A solution of com-
pound 12f (2.33 g, 1.85 mmol) and 7f (1.36 g, 1.54 mmol) in
1,2-dichloroethane (18.5 mL) was stirred with powdered 4 Å
molecular sieves (1 g) for 1 h and then treated with AgOTf
(1.43 g, 5.55 mmol) in one portion. After stirring 4 h at room
temperature in the dark, the reaction mixture was treated with
additional AgOTf (0.475 g, 1.85 mmol) and stirred overnight.
The resulting slurry was filtered through Celite and concen-
trated. Flash chromatography on silica gel (gradient elution,
20f25% EtOAc/hexanes) afforded 2.35 g (73%) of compound
13f as a colorless solid: mp 99.5-100.5 °C [R]D23-5.7° (c 1.0,
1
0.75, CHCl3); H NMR δ 7.4-7.15 (m, 15H), 6.35 (d, 1H, J )
7.3 Hz), 5.90 (d, 1H, J ) 4.7 Hz), 5.52 (∼t, 1H, J ∼ 9.5 Hz),
5.42-5.00 (m, 4H), 4.90 (d, 1H, J ) 12 Hz), 4.69 (∼q, 1H, J ∼
9 Hz), 4.60 (d, 1H, J ) 12 Hz), 4.54 (d, 1H, J ) 8.1 Hz), 4.13
(d, 1H, J ) 11 Hz), 3.86 (dd, 1H, J ) 11, 4 Hz), 3.8-3.35 (m,
8H), 2.54-2.16 (m, 12H), 1.7-1.0 (m, 126H), 0.88 (∼t, 18H);
13C NMR δ 174.0, 173.9, 173.6, 171.4, 170.9, 169.9, 150.4,
150.3, 150.1, 150.0, 137.3, 130.0, 128.6, 128.1, 126.0, 125.9,
120.2, 120.2, 120.1, 100.7, 99.6, 75.3, 75.1, 74.6, 74.4, 74.3, 72.5,
71.3, 70.9, 70.8, 70.3, 68.7, 60.5, 57.9, 56.2, 42.4, 41.6, 36.7,
34.6, 34.5, 32.1, 29.8, 29.7, 29.5, 29.4, 25.2, 22.9, 14.3. Anal.
(C115H195N2O21P) C, H, N.
1
CHCl3); H NMR δ 7.38-7.08 (m, 15H), 5.93 (d, 1H, J ) 7.7
Hz), 5.49 (t, 1H, J ∼ 9.5 Hz), 5.24-4.52 (m, 8H), 4.38 (d, 1H,
J ) 11 Hz), 4.24 (dd, 1H, J ) 11, 5 Hz), 4.17 (dd, 1H, J )
11.5, 2.5 Hz), 3.88 (dd, 1H, J ) 11.5, 3.9 Hz), 3.82-3.63 (m,
3H), 3.57-3.38 (m, 2H), 3.09 (br s, 1H), 2.46-2.15 (m, 8H),
1.90 (s, 3H), 1.84 (s, 3H), 1.65-1.1 (m, 84H), 0.88 (∼t, 12H);
13C NMR δ 174.0, 173.6, 169.8, 169.7, 154.7, 154.3, 151.8,
150.2, 137.1, 129.9, 128.4, 127.8, 125.7, 120.1, 100.6, 99.2, 95.2,
94.4, 90.2, 79.4, 77.2, 74.7, 74.3, 72.3, 72.2, 71.9, 71.0, 70.9,
70.3, 69.2, 69.0, 65.3, 56.9, 55.2, 41.9, 40.0, 34.5, 34.1, 32.0,
29.7, 29.6, 29.4, 29.2, 25.3, 25.1, 25.0, 22.7, 21.2, 21.1, 14.2;
positive FAB-MS calcd for [M + Na]+ 2111.7352, found
2111.7284. Anal. (C98H150Cl9N2O24P) C, H, N.
Compounds 14a -e were obtained from 13a -e and 5a -e15
following the procedure described for 14f in method A above.
Meth od B. Compound 16f (3.8 g, 2.08 mmol) was depro-
tected with zinc (15 g, 0.23 mol) and acylated with 5f (2.25 g,
4.94 mmol) as described above in method A to give 0.975 g
(25%) of compound 14f.
Compounds 13a -e were obtained from 12a -e and 7a -e
following the procedure described above for 13f.
2-Deoxy-6-O-[2-d eoxy-4-O-p h osp h on o-3-O-[(R)-3-tetr a -
d eca n oyloxyt et r a d eca n oyl]-2-[(R)-3-t et r a d eca n oyloxy-
tetr a d eca n oyla m in o]-â-D-glu cop yr a n osyl]-2-[(R)-3-tetr a -
d eca n oyloxytetr a d eca n oyla m in o]-D-glu cop yr a n ose Tr i-
eth yla m m on iu m Sa lt (3f). A solution of compound 14f (0.90
g, 0.46 mmol) in a mixture of AcOH (4.5 mL) and tetrahydro-
furan (40 mL) was hydrogenated in the presence of Pd black
(0.65 g) at room temperature and 70 psig for 24 h. After
removal of the catalyst by filtration, PtO2 (0.45 g) was added
and the hydrogenolysis was continued under the same condi-
tions for 18 h. The opalescent reaction mixture was diluted
with 2:1 CHCl3-MeOH (20 mL) to give a clear solution and
sonicated briefly. The catalyst was collected and washed with
2:1 CHCl3-MeOH, and the combined filtrate and washings
were concentrated. Flash chromatography on silica gel with
CHCl3-MeOH-H2O-Et3N (gradient elution, 90:10:1:1f87.5:
12.5:1:1) afforded 0.61 g of 3f as a colorless glass. The partially
purified product was dissolved in ice-cold 2:1 CHCl3-MeOH
(60 mL) and washed with ice-cold 0.1 N aq HCl (24 mL). The
organic phase was filtered and concentrated at 30 °C bath
temperature. The resulting free acid was lyophilized from 2%
aq Et3N (10 mL, pyrogen-free) to give 0.58 g (58%) of compound
3f triethylammonium salt (R:â ) 2:1) as a colorless powder:
mp 192-194 °C dec; [R]D25 +3.2° (c 0.72, CHCl3); 1H NMR (1:1
CDCl3-CD3OD) δ 5.28-5.08 (m, 5H), 5.05 (d, 0.67H, J ) 3.6
Hz), 4.62 (d, 1H, J ) 8.2 Hz), 4.51 (d, 0.33H, J ) 8.2 Hz), 4.22
(m, 1H), 4.12-3.91 (m, 3H), 3.88-3.54 (m, 5H), 3.46-3.32 (m,
1H), 3.25 (∼t, 1H, J ∼ 9 Hz), 3.17 (q, 6H, J ) 7.4 Hz), 2.76-
2.25 (m, 12H), 1.75-1.1 (m, 135H), 0.89 (∼t, 18H); 13C NMR
δ 174.1, 173.9, 171.2, 170.9, 170.6, 101.2, 91.3, 75.5, 73.4, 72.7,
71.6, 71.3, 71.2, 70.6, 70.3, 60.3, 54.0, 50.2, 50.0, 49.7, 49.4,
49.1, 48.8, 48.5, 45.9, 41.6, 41.3, 39.1, 34.6, 34.3, 32.0, 29.7,
29.4, 29.3, 25.3, 25.1, 22.7, 14.1, 8.5; negative FAB-MS calcd
Ben zyl 2-Deoxy-6-O-[2-d eoxy-4-O-d ip h en ylp h osp h on o-
3-O-[(R)-3-tetr a d eca n oyloxytetr a d eca n oyl]-6-O-(2,2,2-tr i-
ch lor o-1,1-dim eth yleth oxycar bon yl)-2-(2,2,2-tr ich lor oeth -
oxycar bon ylam in o)-â-D-glu copyr an osyl]-3-O-(2,2,2-tr ich lo-
roethoxycarbonyl)-2-(2,2,2-trichloroethoxycarbonylamino)-
â-D-glu cop yr a n osid e (16f). A solution of compound 11f (3.37
g, 2.54 mmol) in dry toluene (18 mL) was treated with Ac2O
(3.9 g, 38 mmol), followed by BF3‚OEt2 (0.29 g, 2.0 mmol)
dropwise over 5 min, and then stirred at room temperature
for 6 h. The reaction mixture was diluted with CH2Cl2, washed
with 5% aq NaHCO3, dried (Na2SO4), and concentrated. A
solution of the resulting acetate in CH2Cl2 (250 mL) at 0 °C
was saturated with HBr(g) and then allowed to warm slowly
and stir at room temperature overnight. Nitrogen was bubbled
through the solution for 10 min and the reaction mixture was
concentrated at 30 °C bath temperature. To a solution of the
resulting glycosyl bromide in dry CHCl3 (10 mL) were added
powdered 4 Å molecular sieves (2 g), anhydrous CaSO4 (4.5 g,
34 mmol), and compound 15 (0.675 g, 1.09 mmol). The
resulting mixture was stirred for 20 min at room temperature,
treated with Hg(CN)2 (2.2 g, 8.7 mmol), and heated to reflux
for 16 h in the dark. The cooled reaction mixture was diluted
with CHCl3 and filtered through a pad of Celite. The filtrate
was washed with 1 N aq KI, dried (Na2SO4), and concentrated.
Flash chromatography on silica gel (30% EtOAc/hexanes)
afforded 1.65 g (83%) of compound 16f as a colorless solid: mp
118-120 °C; [R]D26-6.3° (c 0.51, CHCl3); H NMR δ 7.4-7.1
1
(m, 15H), 5.49 (∼t, 1H, J ∼ 9.5 Hz), 5.24-4.94 (m, 3H), 4.89
(d, 1H, J ) 12 Hz), 4.87 (d, 1H, J ) 12 Hz), 4.74-4.54 (m,
6H), 4.39 (d, 1H, J ) 11 Hz), 4.23 (dd, 1H, J ) 11, 5 Hz), 4.16
(dd, 1H, J ) 11, 2.5 Hz), 3.90 (dd, 1H, J ) 12, 4 Hz), 3.78 (m,

3-O-Desacyl MLA Derivatives
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 22 4647
for [M - H]- 1728.2815, found 1728.2880; HPLC-1, tR ) 41.5
min, 97%. Anal. (C102H196N3O21P‚5H2O) C, H, N, P.
Compounds 3a -e were prepared from 14a -e by the pro-
cedure described above for compound 3f.
immunosorbent assay (ELISA) using Quantikine immunoas-
say kits (R & D Systems, Minneapolis, MN) for the quantita-
tion of human TNF-R and IL-1â.
P yr ogen icity. Febrile responses were determined in a
standard three-rabbit pyrogen test following current USP by
North American Science Associates, Inc. (NAMSA), North-
wood, OH. Stock solutions of the test compounds were pre-
pared at 100 µg/mL in 10% aq EtOH, sterile filtered, and then
diluted with 5% dextrose in water (D5W) to the desired
concentrations immediately prior to testing and injected into
the marginal ear vein of three New Zealand white rabbits
(female, 1.5-5.0 kg) at a dosage volume of 3 mL/kg of body
weight. Rabbit temperatures were recorded rectally every 30
min for 3 h after the injection. The maximum temperature
rise (as compared to baseline) was determined for each rabbit.
The minimum pyrogenic dose or MPD was determined as the
dose at which the total temperature rise for three rabbits was
g1.2 °C.
2-Deoxy-6-O-[2-deoxy-4-O-ph osph on o-3-O-[(R)-3-bu tan -
oyloxyt et r a d eca n oyl]-2-[(R)-3-b u t a n oyloxyt et r a d eca n -
oyla m in o]-â-D-glu cop yr a n osyl]-2-[(R)-3-bu ta n oyloxytet-
r a d eca n oyla m in o]-D-glu cop yr a n ose tr ieth yla m m on iu m
sa lt (3a ): mp 165-166 °C dec; negative FAB-MS calcd for [M
- H]- 1307.8121, found 1307.8146; HPLC-1, tR ) 9.4 min, 98%.
Anal. (C72H136N3O21P‚2H2O) C, H, N, P.
2-Deoxy-6-O-[2-deoxy-4-O-phosphono-3-O-[(R)-3-hexanoyl-
oxytetradecanoyl]-2-[(R)-3-hexanoyloxytetradecanoylamino]-
â-D-glu cop yr a n osyl]-2-[(R)-3-h exa n oyloxytetr a d eca n oyl-
a m in o]-D-glu cop yr a n ose tr ieth yla m m on iu m sa lt (3b): mp
179-182 °C dec; negative FAB-MS calcd for [M - H]-
1391.9060, found 1391.9021; HPLC-1, tR ) 17.1 min, 98%.
Anal. (C78H148N3O21P‚2H2O) C, H, N, P.
Let h a l Toxicit y. Lethal toxicity in D-galactosamine-
sensitized mice was determined by the method of Galanos.30
Stock solutions prepared at 100 µg/mL in 10% aq EtOH were
diluted in D5W to the desired concentrations and admixed with
an equal volume of D5W containing 150 mg/mL D-galac-
tosamine. Groups of 10 ICR mice (female, 4-8 weeks) were
injected via the tail vein with 0.2 mL of the admixtures, and
mortality was monitored for 48 h. A control group of 40 mice
were injected with D5W only. The 50% lethal dose or LD50 was
calculated by the Reed-Muench method.43
Ad ju va n t Activity. Adjuvant activity was assessed by
measuring tetanus toxoid-specific antibody responses in the
serum of mice immunized with tetanus toxoid and the test
compound in a 2.5% oil-water emulsion. The oil-water
emulsions were prepared by dissolving 10 mg of the test
compounds in 2 mL of 12% lecithin-squalene (“oil”) by heating
and sonicating at 56 °C and emulsifying 50 µL of the resulting
solutions in 2 mL of 0.1% Tween 80-saline (“water”) contain-
ing 2 µg of tetanus toxoid. Groups of 8 (C57BL/6 × DBA/2)F1
mice (female, 6-8 weeks) were immunized subcutaneously
with 0.2 mL of the 2.5% oil-water emulsions. The final mouse
dosages of tetanus toxoid and the test compounds were 0.2
and 50 µg, respectively. A second immunization was admin-
istered 21 days after the primary. Control mice received
tetanus toxoid in vehicle consisting of 2.5% oil-water. Serum
samples were collected 14 days after the second injection via
the orbital plexus, diluted 2-fold for 11 dilutions starting with
an initial dilution of 1:200, and individually tested for anti-
tetanus toxoid activity by ELISA using tetanus toxoid-coated
microtiter plates. The endpoint antibody titers of each mouse
were evaluated as total Ig as well as IgG1, IgG2a, and IgG2b
isotypes specific for tetanus toxoid.
2-Deoxy-6-O-[2-d eoxy-4-O-p h osp h on o-3-O-[(R)-3-octa n -
oyloxyt et r a d eca n oyl]-2-[(R)-3-oct a n oyloxyt et r a d eca n -
oyla m in o]-â-D-glu cop yr a n osyl]-2-[(R)-3-octa n oyloxytet-
r a d eca n oyla m in o]-D-glu cop yr a n ose tr ieth yla m m on iu m
sa lt (3c): mp 185-187 °C dec; negative FAB-MS calcd for [M
- H]- 1475.9999, found 1476.0001; HPLC-1, tR ) 24.4 min,
96%. Anal. (C84H160N3O21P‚5H2O) C, H, N, P.
2-Deoxy-6-O-[2-deoxy-4-O-ph osph on o-3-O-[(R)-3-decan -
oyloxyt et r a d eca n oyl]-2-[(R)-3-d eca n oyloxyt et r a d eca n -
oyla m in o]-â-D-glu cop yr a n osyl]-2-[(R)-3-d eca n oyloxytet-
r a d eca n oyla m in o]-D-glu cop yr a n ose tr ieth yla m m on iu m
sa lt (3d ): mp 155-156 °C dec; negative FAB-MS calcd for [M
- H]- 1560.0938, found 1560.0925; HPLC-1, tR ) 30.7 min
(cf. HPLC-2, tR ) 37.5 min), 96%. Anal. (C90H172N3O21P‚2H2O)
C, H, N, P.
2-Deoxy-6-O-[2-d eoxy-4-O-p h osp h on o-3-O-[(R)-3-d od e-
can oyloxytetr adecan oyl]-2-[(R)-3-dodecan oyloxytetr ade-
ca n oyla m in o]-â-D-glu cop yr a n osyl]-2-[(R)-3-d od eca n oyl-
oxytetr adecan oylam in o]-D-glu copyr an ose tr ieth ylam m o-
n iu m sa lt (3e): mp 159-162 °C dec; negative FAB-MS calcd
for [M - H]- 1644.1877, found 1644.1908; HPLC-2, tR ) 42.7
min, 97%. Anal. (C96H184N3O21P‚4H2O) C, H, N, P.
Biologica l Assa ys. MPL immunostimulant was produced
at Ribi ImmunoChem Res., Inc. from the LPS of S. minnesota
R595 following procedures which have been described previ-
ously.5
iNOS In d u ction . The ability of test compounds to induce
iNOS gene expression was assessed in an in vitro system
utilizing peritoneal exudate cells (PEC) from Propionibacte-
rium acnes-primed mice.41 In this assay groups of 10 ICR mice
(female, 4-8 weeks) were injected intraperitoneally on day 0
with 1 mg of P. acnes heat-killed cells in 0.2 mL of sterile
saline. PEC were harvested on day 4, washed, and allowed to
adhere in 96-well microtiter plates (2 × 105 cells/well). Non-
adherent cells were removed by washing, and the test materi-
als at various concentrations (in quadruplicate) were added
to the wells in RPMI-1640 media containing 10% fetal calf
serum, 50 µg/mL gentamicin, and 250 µg/mL amphotericin B.
Cells were incubated in the presence of the test compounds
for 24 h at 37 °C in a CO2 incubator. The conditioned
supernatant fractions were then assayed for nitrite content
by the Greiss reaction.42 For compounds inducing iNOS, the
half-maximal response (expressed as 50% effective dose or
ED50) was determined with SlideWrite Plus for Windows
version 3.0 (Advanced Graphics Software, Encinitas, CA) using
a nonlinear curve-fitting algorithm. In all cases, correlation
coefficients (r) of the fitted lines were >0.95.
Ex Vivo Cytok in e P r od u ction . In the ex vivo assay for
TNF-R and IL-1â production, human whole blood was collected
from a healthy human volunteer into heparinized tubes and
0.45 mL of whole blood was admixed with 0.05 mL of
phosphate-buffered saline (PBS, pH 7.4) containing the test
compounds. The tubes were incubated for 4 h at 37 °C on a
shaker apparatus, diluted with 1.5 mL of sterile PBS, and
centrifuged. The supernatants were removed and analyzed (in
duplicate) for TNF-R and IL-1â by a sandwich enzyme-linked
Ack n ow led gm en t. The authors are grateful to Ms.
Kara D. Looysen for assisting with the NMR measure-
ments and to the Department of Chemistry, University
of Montana, for the use of its NMR facility during the
early stages of this project. The authors also thank Dr.
J ory R. Baldridge for his critical reading of the manu-
script and helpful discussions.
Refer en ces
(1) Lu¨deritz, O.; Freudenberg, M. A.; Galanos, C.; Lehmann, V.;
Rietschel, E. T.; Shaw, D. H. Lipopolysaccharides of Gram-
Negative Bacteria. In Current Topics in Membranes and Trans-
port; Razin, S., Rottem, S., Eds.; Academic Press: New York,
1982; pp 79-151.
(2) J ohnson, A. G. Molecular Adjuvants and Immunomodulators:
New Approaches to Immunization. Clin. Microbiol. Rev. 1994,
7, 277-289.
(3) Qureshi, N.; Mascagni, P.; Ribi, E.; Takayama, K. Monophos-
phoryl Lipid A Obtained from Lipopolysaccharides of Salmonella
minnesota R595. J . Biol. Chem. 1985, 260, 5271-5278.
(4) Myers, K. R.; Truchot, A. T.; Ward, J .; Hudson, Y.; Ulrich, J . T.
A Critical Determinant of Lipid A Endotoxic Activity. In Cellular
and Molecular Aspects of Endotoxin Reactions; Nowotny, A.,
Spitzer, J . J ., Ziegler, E. J ., Eds.; Endotoxin Research Series,
Vol. 1; Elsevier: Amsterdam, 1990; pp 145-156.

4648 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 22
J ohnson et al.
(5) Ulrich, J . T.; Myers, K. R. Monophosphoryl Lipid A as an
Adjuvant. In Vaccine Design: The Subunit and Adjuvant Ap-
proach; Powell, M. F., Newman, M. J ., Eds.; Plenum: New York,
1995; pp 495-524.
(6) Quereshi, N.; Takayama, K. Structure and Function of Lipid A.
In The Bacteria; Iglewski, B. H., Clark, V. L., Eds.; Academic:
New York, 1990; Vol. 11, pp 319-338.
(7) Rietschel, E. T.; Kirikae, T.; Schade, F. U.; Ulmer, A. J .; Holst,
O.; Brade, H.; Schmidt, G.; Mamat, U.; Grimmecke, H.-D.;
Kusumoto, S.; Za¨hringer, U. The Chemical Structure of Bacterial
Endotoxin in Relation to Bioactivity. Immunobiology 1993, 187,
169-190.
(8) Kulshin, V. A.; Za¨hringer, U.; Lindner, B.; J a¨ger, K.-E.; Dmitriev,
B. A.; Rietschel, E. T. Structural Characterization of the Lipid
A Component of Pseudomonas aeruginosa Wild-Type and Rough
Mutant Lipopolysaccharides. Eur. J . Biochem. 1991, 198, 697-
704.
(9) Moran, A. P.; Lindner, B.; Walsh, E. J . Structural Characteriza-
tion of the Lipid A Component of Helicobacter pylori Rough- and
Smooth-Form Lipopolysaccharides. J . Bacteriol. 1997, 179,
6453-6463.
(10) Ishida, H.; Fujishima, Y.; Ogawa, Y.; Kumazawa, Y.; Kiso, M.;
Haegawa, A. Synthesis and Mitogenic Activity of Nonreducing-
Sugar Subunit Analogs of Bacterial Lipid A Composed only of
3-Hydroxytetradecanoic Acid and Its Homologs. Biosci. Biotech-
nol. Biochem. 1995, 59, 1790-1792.
(11) Funatogawa, K.; Matsuura, M.; Nakano, M.; Kiso, M.; Hasegawa,
A. Relationship of Structure and Biological Activity of Monosac-
charide Lipid A Analogues to Induction of Nitric Oxide Produc-
tion by Murine Macrophage RAW264.7 Cells. Infect. Immun.
1998, 66, 5792-5798 and references therein.
(25) Hagen, S. R.; Thompson, J . D.; Snyder, D. S.; Myers, K. M.
Analysis of a Monophosphoryl Lipid A Immunostimulant Prepa-
ration from Salmonella minnesota R595 by High-Performance
Liquid Chromatography. J . Chromatogr. A 1997, 767, 53-61.
(26) Hibbs, J . B., J r.; Taintor, R. R.; Vavrin, Z.; Rachlin, E. M. Nitric
Oxide: A Cytotoxic Activated Macrophage Effector Molecule.
Biochem. Biophys. Res. Commun. 1988, 157, 87-94.
(27) DeForge, L. E.; Kenney, J . S.; J ones, M. L.; Warren, J . S.;
Remick, D. G. Biphasic Production of IL-8 in Lipopolysaccharide
(LPS)-Stimulated Human Whole Blood. Separation of LPS- and
Cytokine-Stimulated Components using Antitumor Necrosis
Factor and Anti-IL-1 Antibodies. J . Immunol. 1992, 148, 2133-
2141.
(28) Nerad, J . L.; Griffiths, J . K.; Van der Meer, J . W. M.; Endres,
S.; Poutsiaka, D. D.; Keusch, G. T.; Bennish, M.; Salam, M. A.;
Dinarello, C. A.; Cannon, J . G. Interleukin-1â (IL-1â), IL-1
Receptor Antagonist, and TNF-R Production in Whole Blood. J .
Leukocyte Biol. 1992, 52, 687-692.
(29) Snapper, C. M.; Finkelman, F. D. Immunoglobulin Class Switch-
ing. In Fundamental Immunology, 4th ed.; Paul, W. E., Ed.;
Lippincott-Raven: Philadelphia, PA, 1999; pp 831-861.
(30) Galanos, C.; Freudenberg, M. A.; Reutter, W. Galactosamine-
Induced Sensitization to the Lethal Effects of Endotoxin. Proc.
Natl. Acad. Sci. U.S.A. 1979, 76, 5939-5943.
(31) The iNOS and cytokine experimental data presented in Table 1
are typical of a minimum of three separate experiments per-
formed in quadruplicate and duplicate, respectively. The lack
of cytokine induction in human monocytes with octanoyl deriva-
tive 3c in the face of its murine iNOS activity and lethal toxicity
in the D-galactosamine mouse model may be attributable both
to an NO contribution44 to toxicity in D-galactosamine-sensitized
mice as well as to species differences: Certain lipid A-like
compounds show endotoxic activity in mice but are inactive or
act as antagonists to LPS and lipid A in human systems.7
Further, an activity profile similar to that of 3c was observed
recently for a closely related monosaccharide lipid A derivative
possessing three (R)-3-octanoyloxytetradecanoyl groups.35
(32) Erwin, A. L.; Munford, R. S. Deacylation of Structurally Diverse
Lipopolysaccharides by Human Acyloxyacyl Hydrolase. J . Biol.
Chem. 1990, 265, 16444-16449.
(33) Cantrell, J . L.; Brookshire, V. G. Unpublished results.
(34) Kiso, M.; Tanaka, S.; Fujita, M.; Fujishima, Y.; Ogawa, Y.;
Ishida, H.; Hasegawa, A. Synthesis of the Optically Active 4-O-
Phosphono-D-glucosamine Derivatives Related to the Nonreduc-
ing-Sugar Subunit of Bacterial Lipid A. Carbohydr. Res. 1987,
162, 127-140.
(35) J ohnson, D. A.; Sowell, C. G.; J ohnson, C. L.; Livesay, M. T.;
Keegan, D. S.; Rhodes, M. J .; Ulrich, J . T.; Ward, J . R.; Cantrell,
J . L.; Brookshire, V. G. Synthesis and Biological Evaluation of
a New Class of Vaccine Adjuvants: Aminoalkyl Glucosaminide
4-Phosphates (AGPs). Bioorg. Med. Chem. Lett. 1999, 9, 2273-
2278.
(12) Fukase, K.; Fukase, Y.; Oikawa, M.; Liu, W.-C.; Suda, Y.;
Kusumoto, S. Divergent Synthesis and Biological Activities of
Lipid A Analogues of Shorter Acyl Chains. Tetrahedron 1998,
54, 4033-4050.
(13) Kiso, M.; Tanaka, S.; Tanahashi, M.; Fujishima, Y.; Ogawa, Y.;
Hasegawa, A. Synthesis of 2-Deoxy-4-O-phosphono-3-O-tetrade-
canoyl-2-[(3R)- and (3S)-3-Tetradecanoyloxytetradecanamido]-
D-glucose:
A Diastereomeric Pair of 4-O-Phosphono-D-glu-
cosamine Derivatives (GLA-27) Related to Bacterial Lipid A.
Carbohydr. Res. 1986, 148, 221-234.
(14) Ogawa, Y.; Kitagawa, M.; Fujishima, Y.; Kiso, M.; Hasegawa,
A.; Ishida, H.; Azuma, I. Synthesis and Immunoadjuvant Activ-
ity of the Conjugates of 1-Thio-N-acetyl-muramoyl Dipeptide
with Lipid A Subunit Analogs. Agric. Biol. Chem. 1989, 53,
1025-1036.
(15) (R)-3-n-Alkanoyloxytetradecanoic acids 5a -f were prepared from
p-bromophenacyl (R)-3-hydroxytetradecanoate (>99.5% ee) by
the acylation/deprotection sequence described previously for 5f:
Keegan, D. S.; Hagen, S. R.; J ohnson, D. A. Efficient Asymmetric
Synthesis of (R)-3-Hydroxy- and Alkanoyloxytetradecanoic Acids
and Method for the Determination of Enantiomeric Purity.
Tetrahedron: Asymmetry 1996, 7, 3559-3564.
(16) J ohnson, D. A.; Sowell, C. G.; Keegan, D. S.; Livesay, M. T.
Chemical Synthesis of the Major Constituents of Salmonella
minnesota Monophosphoryl Lipid A. J . Carbohydr. Chem. 1998,
17, 1421-1426.
(17) Imoto, M.; Yoshimura, H.; Shimamoto, T.; Sakaguchi, N.;
Kusumoto, S.; Shiba, T. Total Synthesis of Escherichia coli Lipid
A, the Endotoxically Active Principle of Cell-Surface Lipopolysac-
charide. Bull. Chem. Soc. J pn. 1987, 60, 2205-2214.
(18) For a review on the glycosylation of 2-amino-2-deoxy carbohy-
drates including silver-promoted glycosylations of N-alkoxycar-
bonyl-protected glucopyranosyl chlorides, see: Banoub, J .; Boul-
langer, P.; Lafont, D. Synthesis of Oligosaccharides of 2-Amino-
2-deoxy Sugars. Chem. Rev. 1992, 92, 1167-1195.
(19) Driguez, H.; Vermes, J . P.; Lessard, J . The Chromous Chloride
Promoted Addition of N-Haloamides to Olefins. V. The Addition
of N-Chloroamides to Enol Ethers: Synthesis of Acyloxy and
Acyl Derivatives of R-Amino Acetals and Ketals (Aldehydes and
Ketones) and of 2-Amino Sugars. Can. J . Chem. 1978, 56, 119-
130.
(20) J ansson, K.; Noori, G.; Magnusson, G. 2-(Trimethylsilyl)ethyl
Glycosides. Transformation into Glycopyranosyl Chlorides. J .
Org. Chem. 1990, 55, 3181-3185.
(21) J ohnson, D. A.; Harris, A. Unpublished results.
(22) Ellervik, U.; Magnusson, G. Glycosylation with N-Troc-Protected
Glycosyl Donors. Carbohydr. Res. 1996, 280, 251-260.
(23) Nakamoto, S.; Takahashi, T.; Ikeda, K.; Achiwa, K. Efficient
Synthesis of Novel Monosaccharide Analogs of Lipid A. Chem.
Pharm. Bull. 1985, 33, 4098-4101.
(24) J ohnson, D. A.; Livesay, M. T. Unpublished results. This side
reaction has been observed in a closely related system possessing
a 4-O-diphenylphosphoryl substituent: Imoto, M.; Kusunose, N.;
Kusumoto, S.; Shiba, T. Synthetic Approach to Bacterial Li-
popolysaccharide, Preparation of Trisaccharide Part Structures
Containing KDO and 1-Dephospho Lipid A. Tetrahedron Lett.
1988, 29, 2227-2230.
(36) Beutler, B.; Cerami, A. The Biology of Cachectin/TNFsA
Primary Mediator of the Host Response. Annu. Rev. Immunol.
1989, 7, 625-655.
(37) Durum, S. K.; Schmidt, J . A.; Oppenheim, J . J . Interleukin 1:
An Immunological Perspective. Annu. Rev. Immunol. 1985, 3,
263-287.
(38) It has been reported that the GLA-47, the nonreducing sugar
portion of 3f, exhibits no toxicity in D-galactosamine-sensitized
mice at
a dose of 1 µg: Takada, H.; Kotani, S. Structural
Requirements of Lipid A for Endotoxicity and Other Biological
Activities. Crit. Rev. Microbiol. 1989, 16, 477-523. This result
together with the known degradation of lipid A molecules by
acyloxyacyl hydrolase32 suggests that free (R)-3-n-alkanoyloxy-
tetradecanoic acids are neither generated in vivo nor responsible
for the observed toxicities of 3a -f in D-galactosamine-sensitized
mice. In point of fact, (R)-3-decanoyloxytetradecanoic acid, the
fatty acid component of highly immunoactive 3d , was found to
be nontoxic in the D-galactosamine model at a dose of 50 µg
(highest dose tested). This dose is 2500 times (>10000 times on
a molar basis) the LD50 of 3d , indicating that the toxicity of 3d
and presumably other MLA derivatives is not due to the fatty
acid domain (J . T. Ulrich, J . R. Ward, unpublished results).
(39) Holst, O. Endotoxin Antagonists: Possible Candidates for the
Treatment of Gram-Negative Sepsis. Angew. Chem., Int. Ed.
Engl. 1995, 34, 2000-2002 and references therein.
(40) Bartlett, G. R. Phosphorus Assay in Column Chromatography.
J . Biol. Chem. 1959, 234, 466-468.
(41) Gustafson, G. L.; Rhodes, M. J .; Hegel, T. Monophosphoryl Lipid
A as a Prophylactic for Sepsis and Septic Shock. In Bacterial
Endotoxins: Lipopolysaccharides from Genes to Therapy; Levin,
J ., Alving, C. R., Munford, R. S., Redl, H., Eds; Progress in
Clinical and Biological Research, Vol. 392; Wiley-Liss: New
York, 1995; pp 567-579.

3-O-Desacyl MLA Derivatives
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 22 4649
(42) Green, L. C.; Wagner, D. A.; Glogowski, D. L.; Skipper, P. L.;
Wishnok, J . S.; Tannenbaum, S. R. Analysis of Nitrate, Nitrite,
and [15N]Nitrate in Biological Fluids. Anal. Biochem. 1982, 126,
131-138.
(43) Reed, L. J .; Muench, H. A Simple Method for Estimating Fifty
Per Cent Endpoints. Am. J . Hyg. 1938, 27, 493-497.
(44) Morikawa, A.; Kato, Y.; Sugiyama, T.; Koide, N.; Chakravortty,
D.; Yoshida, T.; Yokochi, T. Role of Nitric Oxide in Lipopolysac-
charide-Induced Hepatic Injury in D-Galactosamine-Sensitized
Mice as an Experimental Endotoxic Shock Model. Infect. Immun.
1999, 67, 1018-1024.
J M990222B