Selective inhibition of the interaction of C1q with immunoglobulins and the classical pathway of complement activation by steroids and triterpenoids sulfates

Article abstract of DOI:10.1016/j.bmc.2007.03.002

Since undesirable activation of the complement system through the classical pathway is associated with tissue damage and other pathologic proinflammatory consequences at ischemia/reperfusion injury, autoimmune diseases, and rejection of allo- and xenografts, creation of selective inhibitors of the classical pathway leaving the alternative pathway intact is of great importance. Classical pathway is triggered by binding of its recognizing unit, protein C1q, to a number of targets like antibodies, pentraxins, apoptotic cells, and others. In order to obtain inhibitors blocking the first step of the classical cascade, synthesis of sulfates of steroids (Δ⁵-3β-hydroxycholenic, Δ⁵-3β-hydroxyetiocholenic, deoxycholic, and cholic acids) and triterpenoids (betulin, 20,29-dihydro-20,29-dichloromethylenbetulin, betulinic, ursolic, and oleanolic acids) has been performed. Testing of the compounds in classical pathway inhibition assay has displayed derivatives of triterpenoid betulin (betulin disulfate and betulinic acid sulfate) to be the most potent inhibitors. Further studies of the two compounds established that their activity to inhibit the classical pathway had been due to their capability to block the interaction of C1q with antibodies. Betulin disulfate and betulinic acid sulfate have shown weak inhibition of the alternative route of activation, what makes them promising inhibitors for the selective suppression of the classical complement pathway at the earliest possible level as well as perspective agents for blocking the interaction of C1q with its other targets.

Full text of DOI:10.1016/j.bmc.2007.03.002

Bioorganic & Medicinal Chemistry 15 (2007) 3489–3498
Selective inhibition of the interaction of C1q with immunoglobulins
and the classical pathway of complement activation by
steroids and triterpenoids sulfates
a,
a
a
b
*
Svetlana Bureeva, Julian Andia-Pravdivy, Andrey Symon, Anna Bichucher,
a
c
d
a
Vera Moskaleva, Vladimir Popenko, Alexey Shpak, Vitaly Shvets,
b
a
Leonid Kozlov and Alexander Kaplun
a
M.V. Lomonosov State Academy of Fine Chemical Technology, Biotechnology Department,
Vernadskogo str., 86, 119571 Moscow, Russia
G.N. Gabrichevsky Research Institute of Epidemiology and Microbiology, Admiral Makarov str., 10, 125212 Moscow, Russia
b
c
V.A. Engelhard Institute of Molecular Biology RAS, Vavilov str., 32, 119991 Moscow, Russia
M.V. Lomonosov Moscow State University, Chemistry Department, GSP-2, Vorob’evy Hills, 119992 Moscow, Russia
d
Received 13 August 2006; revised 17 February 2007; accepted 2 March 2007
Available online 6 March 2007
Abstract—Since undesirable activation of the complement system through the classical pathway is associated with tissue damage
and other pathologic proinflammatory consequences at ischemia/reperfusion injury, autoimmune diseases, and rejection of allo-
and xenografts, creation of selective inhibitors of the classical pathway leaving the alternative pathway intact is of great importance.
Classical pathway is triggered by binding of its recognizing unit, protein C1q, to a number of targets like antibodies, pentraxins,
apoptotic cells, and others. In order to obtain inhibitors blocking the first step of the classical cascade, synthesis of sulfates of ste-
5
5
roids (D -3b-hydroxycholenic, D -3b-hydroxyetiocholenic, deoxycholic, and cholic acids) and triterpenoids (betulin, 20,29-dihydro-
0,29-dichloromethylenbetulin, betulinic, ursolic, and oleanolic acids) has been performed. Testing of the compounds in classical
2
pathway inhibition assay has displayed derivatives of triterpenoid betulin (betulin disulfate and betulinic acid sulfate) to be the most
potent inhibitors. Further studies of the two compounds established that their activity to inhibit the classical pathway had been due
to their capability to block the interaction of C1q with antibodies. Betulin disulfate and betulinic acid sulfate have shown weak inhi-
bition of the alternative route of activation, what makes them promising inhibitors for the selective suppression of the classical com-
plement pathway at the earliest possible level as well as perspective agents for blocking the interaction of C1q with its other targets.
ꢀ
2007 Elsevier Ltd. All rights reserved.
1
–3
1. Introduction
mation, opsonization, and direct cell lysis.
Thirty
plasma and membrane components, factors, regulators,
and receptors of the complement system are linked in
biochemical cascades and can be activated via three
pathways. The classical pathway of the complement is
usually activated when the first complement component
C1q binds to a complex of antigen and IgM or IgG anti-
body. Interaction of C1q with pentraxins, lipopolysac-
charides, viral particles, apoptotic cells, and some
other agents can also trigger the classical pathway. Acti-
vation of the lectin pathway is initiated by mannose-
binding lectin (MBL) recognizing mannose on bacteria
surface, by IgA or by structures exposed on damaged
endothelium. The alternative pathway is activated by
polysaccharides on microbial surfaces and complex
polysaccharides, for example, yeast cell walls, endotox-
ins, and viral particles. Activation of the complement
One of the most ancient parts of immunity is the com-
plement system, which triggers a powerful, coordinated
repertoire of antimicrobial reactions, including inflam-
Abbreviations: BSA, bovine serum albumin; ClogP, hydrophobicity;
EA, antibody-sensitized erythrocytes; EAC1q, antibody-sensitized
erythrocytes with bound C1q; ED50, concentration of a compound
causing 50% hemolytic effect; Er, rabbit erythrocytes; IC50, concen-
i
tration of a compound causing 50% inhibition; K , inhibition constant;
R1q, human serum depleted of C1q; TEBA, triethylbenzylammonium
chloride.
Keywords: Complement system; Inhibitors; Betulin; Steroid; Triterpe-
noid; Sulfate.
*
Corresponding author. Tel.: +7 495 9368892; fax: +7 495 4348233;
e-mail: sveta407@mail.ru
0
968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.03.002

3490
S. Bureeva et al. / Bioorg. Med. Chem. 15 (2007) 3489–3498
system via the alternative and the lectin pathways is a
major mechanism in innate immunity and takes place
upon direct contact with microorganisms.
inhibition of the classical complement pathway (IC50
2–6 lM) but no inhibition of the alternative comple-
ment pathway. A new series of potent and selective
inhibitors of serine protease C1s described recently
includes thiopheneamidine derivatives inhibiting the
classical pathway of complement activation at IC
The wide array of proinflammatory consequences of
4
complement activation makes the complement system
5
0
1
0
an attractive target for therapeutic intervention and
has led to the isolation, design, and synthesis of numer-
ous complement inhibitors targeting in most cases all
0.3 lM.
This work focuses on studying of negatively charged low
molecular weight inhibitors of protein C1q. The protein
has six heterotrimeric globular domains (gC1q), which
are responsible for binding of C1q to most of its targets:
1
,5–7
three complement pathways.
However, at many
pathological conditions complement-mediated damage
is initiated by only one complement route. Activation
of the classical complement pathway is involved in tis-
sue injury resulting from deposition of autoantibodies
and immune complexes, which may occur in autoim-
mune diseases such as systemic lupus erythematosus,
rheumatoid arthritis, myasthenia gravis, and hyper-
1
1
12
13
antibodies, C-reactive protein, b-amyloid peptide,
and other molecules. Recent studies dealing with mutant
recombinant forms of C1q fragments and three-dimen-
sional models of the C1q globular domain in complex
with two of its physiological ligands, C-reactive protein
and IgG, have revealed a central role for positively
charged amino acid residues in the interaction of C1q
8
,9
acute xenograft rejection. Thus, the selective inhibi-
tion of the classical pathway would result in
suppression of the injury, but not affect the alternative
and lectin pathways, playing a key role in innate
immunity against pathogens. That is why inhibitors
of the first classical pathway component C1 (compris-
ing C1q, C1r, and C1s) have particular perspectives
for the therapeutic complement inhibition. Roos has
described peptide 2J (CEGPFGPRHDLTFCW)
selected from phage-displayed peptide library on the
1
1,14
with its targets.
It likely contributes to high effective-
ness of negatively charged compounds in blocking the
C1q functions.
Natural and synthetic polymers carrying sulfate groups
(heparin, carrageenin, dextran sulfate, chondroitin sul-
fate, polyvinyl sulfate, and others) have been known to
have strong complement-inhibiting properties for more
15,16
8
basis of binding to protein C1q. It showed effective
than 65 years.
The reason for the fact that they are
Cl
Cl
H
H
H
OSO H
OSO H
3
3
H
H
H
H
COOH
COOH
H
H
H
H
H
H
HO SO
HO SO
HO SO
3
3
3
1
3
2
HO SO
3
COOH
COOH
H
H
H
H
H
H
H
H
H
H
HO SO
HO SO
5
HO SO
6
3
4
3
3
H
HO SO
3
COOH
H
H
H
H
H
COOH
COOH
H
H
H
H
HO SO
OSO H
3
3
HO SO
HO SO
3
3
H
7
8
9
5
Figure 1. Studied compounds: betulin disulfate (1), 20,29-dihydro-20,29-dichloromethylenbetulin disulfate (2), betulinic acid sulfate (3), D -3b-
5
hydroxyetiocholenic acid sulfate (4), D -3b-hydroxycholenic acid sulfate (5), deoxycholic acid disulfate (6), cholic acid trisulfate (7), ursolic acid
sulfate (8), oleanolic acid sulfate (9).

S. Bureeva et al. / Bioorg. Med. Chem. 15 (2007) 3489–3498
3491
still not used as therapeutic complement inhibitors is
1
an anticoagulant effect of heparin-like compounds.
romethylenbetulin, betulinic, ursolic, and oleanolic
acids) sulfates (Fig. 1). Activity of the most potent inhib-
itors, betulin disulfate and betulinic acid sulfate, was
studied in more detail. They appeared to inhibit the first
step of the classical pathway activation, the interaction
of protein C1q with antibodies, what can be the basis
of their classical pathway-inhibiting properties. The
compounds showed no considerable activity with re-
spect to the alternative pathway of complement
activation.
7
We proposed that low molecular weight compounds
having at least two negatively charged groups would
display comparable complement-inhibiting activity
without mentioned side effects. A series of bisphenol
1
8
disulfates (IC50 40–250 lM) was obtained on the
basis of this hypothesis and here we describe a set of
sulfated steroids and triterpenoids as effective comple-
ment inhibitors. As steroids and triterpenoids have
been widely accepted to be low toxic compounds, their
sulfo- and phosphoesters are supposed to have good
therapeutic perspectives.
2. Results
2.1. Chemistry
Some steroids and triterpenoids have already been stud-
ied for possible complement-inhibiting properties. Pack-
ard and Weiler have found a set of sulfates and
phosphates of steroids inhibiting the alternative path-
20,29-Dihydro-20,29-dichloromethylenbetulin (14) was
2
4
prepared by Makosza reaction of dichlorocyclopropa-
nation of olefins using sodium trichloroacetate as a car-
bon source and triethylbenzylammonium chloride
(TEBA) as a phase-transfer catalyst (Scheme 1, Steps 1
and 2) followed by the deprotection of hydroxyl groups
1
9
way of activation at high concentrations (0.6 mg/ml).
Assefa and collaborators have discovered the classical
pathway-inhibiting properties of betulinic acid (IC50
5
77 lM), oleanolic acid (IC50 72.3 lM), and some nega-
tively mono-charged derivatives (the most active com-
2
5
(Scheme 1, Step 3).
2
0
pound had IC50 31.8 lM)
analogues (IC50 488–633 lM) of oleanolic acid. Ability
and A/B-ring partial
2
1
For the obtaining of sulfates 1–9 from correspondent
alcohols we developed method using sulfuric acid and
acetic anhydride in pyridine as reagents (Scheme 1, Step
of ursolic acid to inhibit the classical pathway (IC
5
0
2
2
5
4.7 lM) has been described by Master. Potent inhib-
itor targeting the classical pathway, clionasterol (IC50
.1 lM), has been found as a result of study focusing
on complement-inhibiting properties of a series of ste-
1
8
4). This way is a good alternative for the ‘classical’
method of sulfation of steroids, when pyridine sulfur tri-
4
2
6
oxide as a reagent is applied. In that case, bad quality
of pyridine sulfur trioxide often results in low conver-
sion of starting alcohol, difficult purification of a prod-
uct from polymeric byproducts, and the low yield of
di-substituted derivative. The new method allows
obtaining of sulfates of aliphatic and aromatic alcohols
with high yield.
2
3
roids. This compound does not inhibit the alternative
pathway at concentrations as high as 400 lM and has
been shown to exert its effect interfering with comple-
ment component C1.
Here we describe synthesis and classical pathway-inhib-
5
iting properties of some steroids (D -3b-hydroxycholen-
5
ic, D -3b-hydroxyetiocholenic, deoxycholic, cholic acids)
and triterpenoids (betulin, 20,29-dihydro-20,29-dichlo-
Betulin diphosphate was obtained using POCl in ace-
3
tone with following hydrolysis by water.
H
H
OH
OAc
Step 1
Ac O, Δ
H
H
Step 2
H
H
H
H
2
CCl COONa
3
HO
AcO
TEBA, Δ
11
1
2
Cl
Cl
Cl
Cl
Cl
Cl
H
H
H
Step 3
Step 4
H SO , Ac O
OAc
OH
OSO H
3
H
H
H
KOH
2
4
2
H
H
H
H
H
H
Py, Δ
AcO
HO
HO SO
3
1
3
14
2
Scheme 1.

3492
S. Bureeva et al. / Bioorg. Med. Chem. 15 (2007) 3489–3498
2
.2. Hemolytic activity of steroids and triterpenoids
turned out to be a thin suspension of nanocrystals
(Fig. 2C). Apparently, hemolytic activity of compound
9 is connected with these nanocrystals, which might be
able to damage the membranes of erythrocytes and
cause thereby hemolysis.
sulfates depends on the morphology of supramolecular
structures they form
Before the studying of complement-inhibiting properties
of compounds 1–10 in complement-hemolytic system we
determined their intrinsic hemolytic activity incubating
various concentrations of them with sheep erythrocytes.
Two of the compounds, 20,29-dihydro-20,29-dichlorom-
ethylenbetulin disulfate (2) and oleanolic acid sulfate (9),
showed strong lytic properties (Table 1). High hemolytic
activity of oleanolic acid sulfate was an unexpected re-
sult considering the fact that its close analogues, sulfates
of betulinic (3) and ursolic acids (8), did not display lytic
properties in diapason of concentrations below 900 lM.
Working with compounds under research, we observed
that in contrast to the rest of the inhibitors, sulfates 3,
2.3. Steroids and triterpenoids sulfates inhibit the classical
pathway of complement activation
Ability of compounds 1–8, 10 to inhibit the classical path-
way of complement activation was assessed in hemolytic
system using high-diluted guinea-pig serum as a comple-
ment source and antibody-sensitized sheep erythrocytes.
It was established that all the compounds exert comple-
ment-inhibiting properties (Table 1) and the activity de-
pends on the hydrophobicity of the inhibitors. Betulin
disulfate (1) (Fig. 3), 20,29-dihydro-20,29-dichlorome-
thylenbetulin disulfate (2), and betulinic acid sulfate (3),
compounds having the highest hydrophobicity, displayed
8
, 9 formed gels in veronal buffer. We determined the
critical micelle concentration (CMC) in veronal buffer
for these compounds (35, 276, 14 lM, correspondently)
and hypothesized that their molecules could assemble
into supramolecular structures of different morphology,
which could explain their different hemolytic properties.
In order to examine our hypothesis, we studied solutions
of compounds 3, 8, and 9 by microphotography (Fig. 2).
Interestingly, microphotos of structures derived by bet-
ulinic and ursolic acids sulfates (Fig. 2A and B) revealed
the presence of thin long fibers of 4 and 10–40 nm thick-
ness, correspondently. Solution of oleanolic acid sulfate
5
the best effect. More hydrophilic compounds, D -3b-
hydroxyetiocholenic acid sulfate (4), deoxycholic acid
disulfate (6), cholic acid trisulfate (7), were less active.
It was interesting to compare the complement-inhibiting
5
activity of sulfates of D -3b-hydroxycholenic acid (5)
(IC50 23.7 lM), deoxycholic acid (6) (IC50 185.6 lM),
and cholic acid (7) (IC50 100.8 lM) having the same
structure of steroid skeleton but different quantity of
charged groups. Surprisingly, the results of their testing
Table 1. Classical pathway-inhibiting and hemolytic properties of steroids and triterpenoids sulfates
a
b
No.
Compound
ClogP
Classical pathway IC50 , lM
Hemolytic activity ED50 , lM
1
2
3
4
5
6
7
8
9
Betulin disulfate
20,29-Dihydro-20,29-dichloromethylenbetulin disulfate
Betulinic acid sulfate
9.55
9.29
9.21
5.03
6.63
5.20
3.42
9.28
9.33
7.27
6.9 ± 3.1
6.8 ± 0.2
9.9 ± 1.5
850
85
>900
>1600
>400
>1600
>1400
>900
15
5
D -3b-Hydroxyetiocholenic acid sulfate
409.5 ± 16.5
23.7 ± 6.5
185.6 ± 6.5
100.8 ± 12.9
25.2 ± 0.3
—
5
D -3b-Hydroxycholenic acid sulfate
Deoxycholic acid disulfate
Cholic acid trisulfate
Ursolic acid sulfate
Oleanolic acid sulfate
Betulin diphosphate
1
0
42.6 ± 2.9
>800
a
To assess the capability of the compounds to inhibit the classical pathway, antibody-sensitized sheep erythrocytes were incubated with high-diluted
guinea-pig serum (600-fold diluted) in presence of different concentrations of the compounds.
To determine the hemolytic activity of the compounds, sheep erythrocytes were incubated with different concentrations of the compounds.
b
1
00 nm
250 nm
200 nm
Figure 2. Negatively stained nanostructures in water: (A) betulinic acid sulfate (480 lM); (B) ursolic acid sulfate (360 lM); (C) oleanolic acid sulfate
360 lM).
(

S. Bureeva et al. / Bioorg. Med. Chem. 15 (2007) 3489–3498
3493
1
1
20
00
A
B
6
Hemolytic method
ELISA
5
4
3
2
1
0
8
6
4
2
0
0
0
0
0
high diluted serum
middle-diluted serum
0
20
40
60
8 0
100 120 140
-50 -40 -30 -20 -10
0
10 20 30 40 50
Betulin disulfate, uM
Betulin disulfate, uM
Figure 3. (A) Effect of betulin disulfate on activation of the classical pathway initiated by incubation of antibody-sensitized sheep erythrocytes with
high-diluted serum (600-fold diluted guinea-pig serum) or middle-diluted serum (170-fold diluted human serum). (B) Effect of betulin disulfate on
C1q enzyme activity assessed by ELISA and hemolytic method.
Table 2. Complement-inhibiting properties of betulin disulfate and betulinic acid sulfate
No
Compound
IC50, lM
i
K , lM
Inhibition of the complement system activation
Inhibition of the C1q—
immunoglobulin interaction
a
b
c
Classical pathway
Alternative pathway
ELISA
Hemolytic
d
method
High-diluted
guinea-pig serum
Middle-diluted
human serum
1
3
Betulin disulfate
Betulinic acid sulfate
6.9 ± 3.1
9.9 ± 1.5
14.6 ± 2.1
53.83 ± 5.43
414.5 ± 13.5
640.0 ± 86.0
42.2 ± 2.6
57.3 ± 12.0
8.6 ± 1.8
14.6 ± 3.5
a
Classical pathway—inhibiting properties were assessed by incubation of antibody-sensitized sheep erythrocytes with high-diluted guinea-pig serum
600-fold diluted) or middle-diluted human serum (170-fold diluted) in presence of the compounds.
(
b
c
Activity to inhibit the alternative pathway was determined by incubation of rabbit erythrocytes with human serum in presence of the compounds.
Microtiter wells were coated with IgG3 and incubated with C1q preincubated with the compounds.
d
Cellular intermediates EAC1q were obtained by interaction of EA with C1q source in the presence of the inhibitors and incubated with serum
devoid of C1q to develop complement-mediated lysis of cells.
show that the increase in quantity of charged groups does
not provide more potent complement inhibition. Proba-
bly, the introduction of more than two charged groups
leads to considerable decrease in hydrophobic portion
of the inhibitors and as a consequence decreases the activ-
ity. Apparently, the reduction of the hydrophobicity is
also the reason for the lower activity of betulin diphos-
phate (10) with respect to that of betulin disulfate (1).
induced by incubation of rabbit erythrocytes with
human serum at the conditions excluding the classical
pathway’s activation. Sulfates of betulin and betulinic
acid exerted weak inhibitory properties with respect to
alternative pathway (Table 2) showing the possibility
of their use for the selective inhibition of the classical
complement pathway.
2.5. Betulin disulfate and betulinic acid sulfate inhibit the
interaction of C1q with immunoglobulins
Activity of the most promising inhibitors, betulin disul-
fate and betulinic acid sulfate, having the highest com-
plement-inhibiting effect and low hemolytic properties
was studied further in classical pathway-mediated
hemolytic reaction using middle-diluted human serum
Activation of the classical pathway of complement sys-
tem is triggered when recognizing protein C1q binds to
antibodies of immune complexes. To study the ability
of compounds 1 and 3 to block the interaction of C1q
with immunoglobulins, we used two different methods:
hemolytic method allowing to separate the first step of
complement activation and to assess the inhibitors’
activity with respect to C1q protein and enzyme-linked
immunosorbent assay based on the interaction of immo-
bilized IgG3 antibodies with C1q in the presence of the
inhibitors.
(
5
Table 2, Fig. 3). As expected, IC50 values were 2 and
.5 times higher than those required in high-diluted ser-
um assay for betulin disulfate and betulinic acid sulfate,
correspondently.
2.4. Betulin disulfate and betulinic acid sulfate weakly
inhibit the alternative pathway of complement activation
In order to estimate selectivity of the action of com-
pounds 1 and 3 on the classical pathway, we determined
their activity with respect to the alternative route of
complement activation. Alternative pathway was
2
7
According to the first method, EA are incubated with
an inhibitor and diluted human serum as a C1q source
at the conditions (30 ꢁC for 15 min) when complex C1

3494
S. Bureeva et al. / Bioorg. Med. Chem. 15 (2007) 3489–3498
2
8
5
is dissociated on 95% and C1q is capable to bind to
immune complexes, but complement cascade does not
develop further. The obtained EAC1q complexes are
isolated by centrifugation, supernatant is removed, and
cells are incubated with R1q serum (at 37 ꢁC for
and correspond to the results of some other researchers
pointing out the important role of the hydrophobicity
for the complement-inhibiting activity.
Betulinic acid sulfate (3) and ursolic acid sulfate (8) in
our assay showed more potent complement-inhibiting
properties than betulinic and ursolic acids studied by
other researchers (their activity is given in Introduction).
As a rule, IC50 values highly depend on complement-
inhibiting assay, source and concentration of serum used
for the activity determination. Unfortunately, we did
not find the possibility to assess the activity of betulinic
and ursolic acids in our assay to make the results more
comparable because of extremely poor solubility of the
compounds even in DMSO-containing water.
3
0 min). At these conditions the classical complement
pathway develops further, membrane attack complex
assembles on erythrocyte surface and causes lysis of
the cells to the extent corresponding to the quantity of
C1q bound to the EA in the presence of an inhibitor.
Using this method we established that betulin disulfate
and betulinic acid sulfate inhibit the C1q enzyme activity
(K 8.6 and 14.6 lM, correspondently) and C1q-immu-
noglobulin interaction (Table 2, Fig. 3).
i
K values obtained for these compounds using ELISA
i
(
42.2 and 57.3 lM) (Table 2 and Fig. 3) are worse than
Two compounds of the set, 2 and 9, turned out to have
unwanted strong hemolytic properties. Possible reason
for the high lytic effect of oleanolic acid sulfate is forma-
tion of nanocrystals, which were displayed by micropho-
tography (Fig. 2C). Low solubility and amphiphilic
properties can explain the appearance of fibrillar supra-
molecular structures of sulfates of betulinic and ursolic
acids (Fig. 2A and B), which most likely exist at acting
concentrations. It is well known that polymers, including
negatively charged polymers, have higher effectiveness at
the inhibition of the protein C1q functions as compared
with low molecular weight (or monomeric) compounds.
It can be explained by the presence of six identical glob-
ular domains in C1q structure and the opportunity of
polymers to interact with more than one globular head
of this protein simultaneously. For example, conjugates
those obtained by the first hemolytic method what can
be caused by differences in methods and concentrations
of the biological reagents used.
3. Discussion
Undesirable activation of the complement system
through the classical pathway is associated with tissue
damage and other proinflammatory consequences at
ischemia/reperfusion injury, autoimmune diseases, and
rejection of allo- and xenografts. Recent data suggest
that unwanted activation of the classical pathway may
involve binding of C1q to in situ deposited self proteins,
such as acute phase proteins of the pentraxin family, and
b-amyloid fibrils in brain lesions in Alzheimer’s dis-
8
,9
2
77
278
of Trp-Tyr (peptide imitating Trp -Tyr
C 2 domain of immunoglobulin, which were determined
residues of
2
9–32
ease.
That is why development of inhibitors capable
H
3
4
to block C1q functions and its interaction with the tar-
gets without affecting the alternative and lectin path-
ways is of great importance. A number of specific
molecules that can regulate the functional activity of
to be involved in C1q–IgG interaction ) coupled to BSA
and dextrans of different sizes were obtained by Ander-
son showed more than 100-fold enhancement in activity
3
5
as against monomeric Trp-Tyr peptide. Taking the
speculations into account it was logical to expect an
increase of complement-inhibiting activity of com-
pounds 3 and 8 with respect to other compounds’ activ-
ity. However, we do not observe such regularity: the
activity of the compounds is in the same diapason as
other inhibitors have. Thinking about possible explana-
tions for this, we can suggest that the fibrillar structures
considered here are much less flexible than usual poly-
mers or their charged groups are not exposed effectively
enough for the interaction with the protein. It should be
noticed that microphotographs of water solutions of
betulin disulfate did not reveal the presence of any supra-
molecular structures at concentrations up to 3 mM.
1
,29
C1q have been reviewed.
The natural C1q-binding
proteins and their functional parts, monoclonal antibod-
ies directed against C1q, several series of C1q-binding
peptides, and competitive inhibitors derived from the se-
quence of C1q are among them. Authors of the reviews
note that the main problems arising at the inhibition of
the C1q functions by these molecules are their short
half-life in circulation, possible immunogenic reactions,
low effectiveness of such inhibitors, and risk of trigger-
ing C1 activation by the inhibitors in case they bind to
the globular head of C1q (it especially concerns multi-
meric structures).
Effective and selective synthetic low molecular weight
inhibitors of non-protein nature could be a good alter-
native for the inhibition of the C1q functions. It
prompted us to conduct this study focusing on the syn-
thesis and investigation of the complement-inhibiting
properties of a set of steroids and triterpenoids sulfates
After testing of the obtained compounds with high di-
luted guinea-pig serum and determination of their
hemolytic properties, two most promising inhibitors,
betulin disulfate and betulinic acid sulfate, were sorted
out for the next experiments. They were tested in the
classical pathway assay using middle-diluted human ser-
um and showed expected less activity (Table 2, Fig. 3).
Weak inhibition of the alternative pathway found for
betulin disulfate and betulinic acid sulfate shows that
compounds 1 and 3 do not inhibit considerably the last
stages of complement cascade, common for the three
(
Fig. 1). Testing of compounds 1–8, 10 in the classical
pathway-induced assay showed that all the compounds
inhibited the classical pathway and the most hydropho-
bic structures, derivatives of triterpenoid betulin 1–3, ex-
erted the most potent effect (Table 1). These
1
8,33
observations are in line with our previous findings

S. Bureeva et al. / Bioorg. Med. Chem. 15 (2007) 3489–3498
3495
3
8
routes of complement activation, what makes possible
the selective blocking of the classical pathway of activa-
tion by these compounds. Activity of some negatively
charged compounds to inhibit more than one step of
complement activation was described more than once.
The bright example is suramin having six sulfo-groups,
as described previously. Oleanolic, ursolic, and cholen-
ic acids were obtained as a kind gift from Dr. E. Zvonk-
ova (Russian Institute of Aromatic Herbs, Moscow).
Deoxycholic and cholic acids are commercially available
from Aldrich. Sheep erythrocytes and rabbit antibodies
against them were obtained from ‘Biolek’ company
(Kharkov, Ukraine). Rabbit erythrocytes, serums,
reagent R1q, IgG3, C1q, anti-C1q antibodies were
obtained from G.N.Gabrichevsky Research Institute
of Epidemiology and Microbiology (Moscow, Russia).
5
which blocks most steps of the complement cascade.
Our further effort was put into determination of whether
betulin disulfate and betulinic acid sulfate inhibit the
interaction of C1q with antibodies preventing thereby
the development of the classical complement pathway.
1
The H NMR spectra were recorded at 300 MHz on
We attempted to assess K for the inhibition of the C1q
i
Bruker DPX-300 (Germany) instrument. Chemical
shifts are reported as d units (ppm), and signals are
expressed as s (singlet), d (doublet), m (multiplet). The
elemental analysis was performed by Thermo Finigan
EA1112 CHNS analyzer (Italy). The Shimadzu (Japan)
liquid chromatography single quadrupole mass-spec-
trometer (LCMS-2010A) with atmospheric pressure
electrospray ionization (API-ESI) source was used for
determination in a negative ionization mode. The oper-
ation conditions were: temperature of drying gas
(250 ꢁC), capillary voltage (4000 V), dry gas (nitrogen,
10 L/min), and nebulizing gas (nitrogen, 1.5 L/min),
respectively. Quadrupole full scan analyses were con-
ducted from m/z 200 to 1400. Mass-spectra of 10 ppm
solutions of tested compounds were obtained by direct
injection technique with methanol/water (4:1 w/w) as a
solvent with flow rate 0.2 ml/min. A 20-lL sample vol-
ume was injected.
functions by two different methods and found that inhibi-
tion constant values obtained by the hemolytic method,
supposing the formation of cellular intermediate EAC1q
in the presence of an inhibitor, better correlate with IC50
values of classical pathway inhibition assay than those
obtained by ELISA. However, the considerable inhibi-
tion of C1q functions found in these assays and weak
activity of betulin disulfate and betulinic acid sulfate with
respect to the alternative pathway suggest that the block-
ing of the interaction of C1q with immunoglobulins may
be the main mechanism of action of the compounds.
Evaluating the possible therapeutic future of steroids
and triterpenoids sulfates, it should be noticed that the
initial compounds, betulin, betulinic, oleanolic, and
ursulic acids, are low toxic compounds and were de-
3
6
scribed to have hepatoprotective effect. Their sulfation
unlikely changes the toxic properties, as the conjugation
with sulfate group is one of the physiologic processes of
biotransformation. The bioavailability of betulin is
For electron microscopy the specimens were placed onto
grids coated with freshly glow-discharged carbon-collo-
dion film, negatively stained with 1% aqueous uranyl
acetate for 1 min, and investigated using JEM-100CX
electron microscope (JEOL) at 80 kV.
3
0–50% (our unpublished data), however, that of
charged analogues might be lower. Formation of any
supramolecular structures by the compounds can unpre-
dictably change the bioavailability decreasing the prob-
ability of hydrolysis of sulfate groups in the acidic
medium and making possible the absorption of the
Buffers used were GVB, 5 mM sodium veronal, 145 mM
NaCl containing 0.1% gelatin, 0.02% NaN , pH 7.4;
3
3
7
2+
GVB , GVB containing 0.5 mM MgCl and 0.15 mM
nanoparticles through M cells of Peyer’s patch.
2
CaCl ; GVBE, GVB containing 10 mM EDTA; PBS,
2
In conclusion, in search of selective low molecular weight
inhibitors of recognition stage of the classical pathway of
the complement system activation a set of steroids and
triterpenoids sulfates have been studied. Betulin disulfate
and betulinic acid sulfate displayed the highest effect at
the inhibition of the classical pathway activation and
weak activity with respect to the alternative route. The
compounds likely suppress the classical pathway of com-
plement by blocking the first stage of its activation. The
results make betulin disulfate and betulinic acid sulfate
promising inhibitors for the selective blocking of the clas-
sical complement pathway at the earliest possible level as
well as perspective agents for the inhibition of the inter-
actions of C1q with its other targets.
10 mM Na HPO , 10 mM NaH PO , 145 mM NaCl,
2
4
2
4
pH 7.4; PBST, PBS containing 0.05% Tween 20, pH
7.4; carbonate buffer, 15 mM Na CO , 35 mM NaH-
CO , pH 9.6; citrate-phosphate buffer containing
2
3
3
0.006% H O , 38 mM citric acid, 67 mM Na HPO Æ12-
2
2
2
4
H O, pH 5.0.
2
4.2. Chemistry
4.2.1. 20,29-Dihydro-20,29-dichloromethylenbetulin (14).
Acetic anhydride (170 ml) was added to betulin (11)
(20.0 g, 45.2 mmol) and heated to the complete dissolu-
tion. The reaction mixture was cooled, and the precipi-
tated crystals were filtered, washed with water to pH
7, dried over P O , and recrystallized from isopropanol,
2 5
to give betulin diacetate (12) as white acicular crystals;
yield 15.3 g (64.3%); mp 223–224 ꢁC. Sodium trichloro-
acetate (3.67 g, 19.8 mmol) and TEBA (38 mg,
0.2 mmol) were added to a solution of betulin diacetate
4
. Materials and methods
4
.1. Reagents and materials
(5.20 g, 9.90 mmol) in chloroform (10 ml) at room tem-
Betulin was isolated from the birch bark and betulinic
acid was obtained by betulin oxidation in our laboratory
perature. The reaction mixture was boiled and inten-
sively stirred for 2 days, cooled to room temperature,

3496
S. Bureeva et al. / Bioorg. Med. Chem. 15 (2007) 3489–3498
twice washed with water, and evaporated. The residue
5.00 g) was dissolved without further purification in a
4.2.3. Betulin diphosphate (10). Solution of betulin (1 g,
2.26 mmol) in 6 ml of pyridine was dropped for 1 h to
(
mixture of THF (50 ml) and methanol (13 ml) and after
adding of 4 N NaOH (9 ml) was stirred at room temper-
ature for 17 h. The reaction mixture was neutralized
with 20% HCl to pH 7 and evaporated to dryness.
The obtained crystals were dissolved in chloroform
the mixture of POCl (1.68 ml, 0.02 mol) and 8 ml of
3
acetone at stirring and cooling to 0 ꢁC. After evaporat-
ing, 30 ml of dioxane and 2 ml of water were added to
the residue and the mixture was boiled for 1 h. The mix-
ture was concentrated to dryness under reduced pres-
sure, the residue was dissolved in 20 ml of mixture of
chloroform–methanol, 3:1, and crystals containing de-
sired product were precipitated by adding 15 ml of
water. The precipitate was filtered, dried, and chromato-
graphed in mixture of chloroform–methanol–pyridine–
formic acid, 1:1:0.15:0.15, giving 1.43 g (69%) of solid;
(
Na SO , and evaporated. The residue was recrystallized
100 ml), washed with water (3· 100 ml), dried with
2
4
from a dichloromethane–isopropanol mixture to give
betulin derivative (14) as white crystals; yield 3.50 g
1
(
NMR (CDCl ): 0.74, 0.81, 0.95, 0.98, 0.99, and 1.21
67.3%); mp 201–202.5 ꢁC; R 0.16 (chloroform);
H
f
3
(
all 3H, s, CH ), 2.30 (1H, m, H19), 3.19 (1H, m, H3),
3
R 0.59 (methanol–pyridine–phosphoric acid, 2:0.5:0.04);
f
H NMR (CD OD): 0.87, 0.94, 1.04, 1.07, 1.13, and
3
1.74 (all 3H, s, CH ), 2.48 (1H, m, H19), 3.77 (1H, d,
3
1
3.21 (1H, d, J 10.7 Hz, H28), 3.71 (1 H, d, J 10.7 Hz,
H28).
J10.0 Hz, H28), 3.89 (1H, m, H3), 4.22 (1H, d,
J10.0 Hz, H28).
4
.2.2. Steroids and triterpenoids sulfates (1–9). Steroids
and triterpenoids sulfates were obtained from correspon-
dent alcohols. Standard procedure for starting com-
pound carrying one hydroxyl group was followed.
Sulfuric acid (0.16 ml, 3 mmol) and acetic anhydride
Results of elemental analysis of all compounds were
satisfactory.
(
0.28 ml, 3 mmol) were mixed with dry pyridine (5 ml)
4.3. Classical pathway hemolytic assay
and after 5 min of stirring at 50–55 ꢁC a solution of alco-
hol (1 mmol) in 2 ml of pyridine was added. The mixture
was stirred for 30 min at the same conditions, cooled to
Capability of the compounds 1–8, 10 to inhibit the clas-
sical pathway-mediated hemolysis was measured using
antibody-coated sheep erythrocytes (EA, 1.5 · 10 /ml)
8
0
ꢁC, and 25% ammonia water (0.74 ml) was added.
After 15 min of stirring, the precipitate was filtered and
the filtrate was placed to a crystallizer for concentrating.
After escaping of ammonium acetate, the mixture was
chromatographed to yield ammonium salt of sulfate as
a solid. The sodium salt was obtained by adding equiva-
lent of sodium hydroxide in methanol and concentrating
to dryness (yield 80–95%). Twofold and threefold quan-
tity of sulfuric acid and acetic anhydride were used for
obtaining di- and trisulfates.
and guinea-pig serum as a complement source at a dilu-
tion previously determined to lyse 70–80% of cells
(resulting dilution ꢁ1:600). Two hundred microliters
of EA was mixed with 200 lL of guinea-pig serum and
600 lL of various concentrations of the inhibitors di-
2
+
luted in GVB . To determine the spontaneous lysis of
erythrocytes, 200 lL of EA was mixed with 800 lL of
2
+
GVB . Three uninhibited samples consisted of 200 lL
of EA, 200 lL of guinea-pig serum, and 600 lL of
GVB . The reaction mixtures were incubated for
2
+
2
+
The structures of the compounds obtained were con-
H
30 min at 37 ꢁC and diluted with 3 ml of cold GVB
1
firmed by H NMR- and mass-spectroscopy.
1
to stop the reaction. Cells were separated by centrifuga-
tion, and the absorbance at 414 nm of the supernatants
was measured to quantify hemoglobin release. Values
for the spontaneous lysis were substracted from sample
values, and the fractional inhibition was determined rel-
ative to the uninhibited (no added compounds) sample.
NMR spectra in general corresponded to those for
the starting compounds excepting the absence of sig-
nals of hydroxyl groups’ protons and paramagnetic
shifts of the protons’ signals of groups adjacent to
1
hydroxyl groups. H NMR-spectra information (d6-
DMSO); number of compound: chemical shift (inte-
gral, multiplicity, constant of spin–spin interaction (if
noted), number of carbon atom at standard numera-
tion of steroid/triterpenoid skeleton): 1: 3.75 (1H, d,
J 10.0 Hz, H28), 3.86 (1H, m, H3), 4.39 (1H, d, J
To perform the classical complement inhibition assay
with middle-diluted human serum, 10 lL of EA
9
(1 · 10 /ml) was mixed with 200 lL of various concen-
2
+
trations of the inhibitors diluted in GVB and 290 lL
2
+
1
0.0 Hz, H28); 2: 3.28 (1H, d, J 10.0 Hz, H28), 3.61
of NHS diluted 1:100 in GVB
(resulting dilution
(
(
1H, m, H3), 3.83 (1H, d, J 10.0 Hz, H28); 3: 3.59
1H, m, H3); 4: 3.83 (1H, m, H3); 5: 3.82 (1H, m,
ꢁ1:170). After 30 min of incubation at 37 ꢁC, the reac-
2
+
tion was stopped by adding 500 lL of cold GVB ,
and the results were determined as described above.
Data are given as an average of three repetitions ± stan-
dard deviation.
H3); 6: 3.93 (1H, m, H3); 4.34 (1H, m, H12); 7: 3.78
1H, m, H3); 4.12 (1H, m, H7), 4.37 (1H, m, H12),
: 3.64 (1H, m, H3), 9: 3.61 (1H, m, H3). For recording
(
8
mass-spectra negatively charged deprotonated molecu-
lar ions of obtained compounds were produced.
Mass-spectra information; number of compound: prod-
4.4. Alternative pathway hemolytic assay
+
uct ions/negative charge—m/z: 1: (Mꢀ2H )/2 ꢀ 300.0;
Inhibition of the alternative pathway-mediated hemoly-
sis was determined by measuring of lysis of rabbit eryth-
rocytes (Er) under the influence of human serum.
Different concentrations of inhibitors diluted in 78 lL
of GVB were mixed with 5 lL of 50 mM MgEGTA,
+
+
2
(
2
3
:
(Mꢀ2H )/2 ꢀ 342.0; 3: (MꢀH )/1 ꢀ 535.4; 4:
+
+
+
MꢀH )/1 ꢀ 397.3; 5: (MꢀH )/1 ꢀ 453.3; 6: (Mꢀ2H )/
+
+
ꢀ 275.0;
7:
(Mꢀ2H )/2 ꢀ 323.0,
(Mꢀ3H )/
+
+
ꢀ 215.0; 8: (MꢀH )/1 ꢀ 535.0; 9: (MꢀH )/1 ꢀ 535.0.

S. Bureeva et al. / Bioorg. Med. Chem. 15 (2007) 3489–3498
3497
Table 3. Content of the tubes for hemolytic assay for the inhibition of
the C1q-immunoglobulin interaction, lM
are given as an average of three repetitions ± standard
deviation.
Step 1
Step 2
2+
EA GVB
NHS Inhibitor
Acknowledgment
C
C
C
L
R
S
200 100
200 100
2
490 H O
490 R1q
490 R1q
490 R1q
Authors were supported financially by Russian Founda-
tion for Basic Research No. 06-04-49795-a.
200
90
10
10
Experimental tubes 200
90
9
7
lL of undiluted NHS, and 10 lL of Er (1 · 10 /ml).
References and notes
The reaction mixtures were incubated at 37 ꢁC for
0 min, and the reaction was stopped by adding
00 lL of cold GVBE. The tubes were centrifuged,
2
4
1
2
3
. Makrides, S. Pharm. Rev. 1998, 50, 59–87.
. Walport, M. J. N. Eng. J. Med. 2001, 344, 1058–1066.
. Mollnes, T. E.; Song, W. C.; Lambris, J. D. Trends
Immunol. 2002, 23, 61–64.
and the percentage of hemolysis was determined as de-
scribed above. Data are given as an average of three
repetitions ± standard deviation.
4. Morgan, B. P. Eur. J. Clin. Invest. 1994, 24, 219–228.
5
6
. Asghar, S. Pharm. Rev. 1984, 36, 223–244.
. Sahu, A.; Lambris, J. D. Immunopharmacology 2000, 49,
4.5. Hemolytic assay for the inhibition of the C1q-immuno-
globulin interaction
1
33–148.
7
. Bureeva, S.; Andia-Pravdivy, J.; Kaplun, A. Drug Discov-
ery Today 2005, 10, 1535–1542.
Antibody-sensitized sheep erythrocytes diluted in
GVB
8
. Roos, A.; Nauta, A. J.; Broers, D.; Faber-Krol, M. C.;
Trouw, L. A.; Drijfhout, J. W.; Daha, M. R. J. Immunol.
2
+
8
(1.5 · 10 cells/ml), NHS diluted 1:15 in
2
+
GVB , and various concentrations of the inhibitors
were added to experimental tubes according to
Table 3, column ‘Step 1’. Tubes CL and CR are the
controls of full, and R1q-reagent-mediated lysis of
2
. Walport, M. J. N. Eng. J. Med. 2001, 344, 1140–1144.
001, 167, 7052–7059.
9
10. Subasinghe, N. L.; Ali, F.; Illig, K. R.; Rudolph, J.; Scott,
K.; Khalil, E.; Soll, R. M.; Bone, R. F.; Spurlino, J. C.;
DesJarlais, R. L.; Crysler, C. S.; Cummings, M. D.;
Morris, P. E.; Kilpatrick, J. M., Jr.; Babu, Y. S. Bioorg.
Med. Chem. Lett. 2004, 14, 3043–3047.
1. Kojouharova, M. S.; Gadjeva, M. G.; Tsacheva, I. G.;
Zlatarova, A.; Roumenina, L. T.; Tchorbadjieva, M. I.;
Atanasov, B. P.; Waters, P.; Urban, B. C.; Sim, R. B.; Reid,
K. B.; Kishore, U. J. Immunol. 2004, 172, 4351–4358.
2. Roumenina, L. T.; Ruseva, M. M.; Zlatarova, A.; Ghai,
R.; Kolev, M.; Olova, N.; Gadjeva, M.; Agrawal, A.;
Bottazzi, B.; Mantovani, A.; Reid, K. B.; Kishore, U.;
Kojouharova, M. S. Biochemistry 2006, 45, 4093–4104.
erythrocytes, correspondently. Tubes C represent unin-
S
hibited samples, which were run in triplicate. The reac-
tion mixtures were incubated at 30 ꢁC for 15 min and
cells were separated by centrifugation. Supernatants
were carefully removed and ingredients were added
according to Table 3, column ‘Step 2’. Dilution of re-
agent R1q was previously determined to lyse 70–80%
of the erythrocytes. After 1 h of incubation at 37 ꢁC,
1
1
2
+
3
ml of cold GVB was added. To determine the extent
of hemolysis, reaction mixtures were centrifuged and
optical densities of supernatants were determined at
13. Tacnet-Delorme, P.; Chevallier, S.; Arlaud, G. J.
J. Immunol. 2001, 167, 6374–6381.
4
14 nm. Values z expressing the quantity of active C1q
1
4. Gaboriaud, C.; Juanhuix, J.; Gruez, A.; Lacroix, M.;
Darnault, C.; Pignol, D.; Verger, D.; Fontecilla-Camps, J.
C.; Arlaud, G. J. J. Biol. Chem. 2003, 278, 46974–46982.
5. Lauenstein, K.; Siedentopf, H. G.; Fischer, H. Z. Natur-
forsch. 1965, 20, 575–581.
molecules were calculated from formula (I) taking into
account the optical densities of supernatants of corre-
spondent tubes. Inhibition constants were determined
from Fig. 3B. Data are given as an average of three
repetitions ± standard deviation.
1
1
6. Raepple, E.; Hill, H.; Loos, M. Immunochemistry 1976, 13,
ꢀ
ꢁ
2
17. Wuillemin, W. A.; te Velthuis, H.; Lubbers, Y. T. P.; de
51–255.
C
L
ꢀ C
R
z ¼ ln
ðIÞ
C
L
ꢀ Exp
Ruig, C. P.; Eldering, E.; Hack, C. E. J. Immunol. 1997,
1
59, 1953–1960.
1
1
8. Bureeva, S.; Andia-Pravdivy, Y.; Petrov, G.; Igumnov,
M.; Kaplun, A.; Romanov, S.; Kolesnikova, E.; Kozlov,
L. Bioorg. Med. Chem. 2005, 13, 1045–1052.
9. Packard, B. D.; Weiler, J. M. Infect. Immun. 1983, 40,
1
4.6. ELISA for the inhibition of the interaction of C1q
with IgG3
Microtiter wells were coated with human IgG3 (0.5 lg/
well) in carbonate buffer and any residual binding sites
were blocked with 1% w/v BSA in PBS for 1 h at
011–1014.
20. Assefa, H.; Nimrod, A.; Walker, L.; Sindelar, R. Bioorg.
Med. Chem. Lett. 1999, 9, 1889–1894.
3
3
7 ꢁC. After washing, wells were incubated (1.5 h at
7 ꢁC) with C1q, previously preincubated with betulin
21. Assefa, H.; Nimrod, A.; Walker, L.; Sindelar, R. Bioorg.
Med. Chem. Lett. 2001, 11, 1619–1623.
2
2. Master, H. E.; Khan, S. I.; Poojari, K. A. Bioorg. Med.
Chem. Lett. 2003, 13, 1249–1251.
disulfate or betulinic acid sulfate in different concentra-
tions diluted in PBST. After washing, wells were incu-
bated with rabbit anti-C1q antibodies conjugated with
horseradish peroxidase in suitable dilution. To assess
the quantity of the bound antibodies, the color was
2
3. Cerqueira, F.; Watanadilok, R.; Sonchaeng, P.; Kijjoa, A.;
Pinto, M. Planta Med. 2003, 69, 174–176.
2
2
4. Dehmlow, E. V. Tetrahedron Lett. 1976, 17, 91–94.
5. Symon, A. V.; Veselova, N. N.; Kaplun, A. P.; Vlasenk-
ova, N. K.; Fedorova, G. A.; Liutik, A. I.; Gerasimova,
G. K.; Shvets, V. I. Bioorg. Khim 2007, 31, 320–325.
0
0
developed using 3,3 ,5,5 -tetramethylbenzidine in cit-
rate–phosphate buffer containing 0.006% H O . Data
2
2

3498
S. Bureeva et al. / Bioorg. Med. Chem. 15 (2007) 3489–3498
2
2
2
2
3
3
6. Sobel, A. E.; Spoerri, P. E. J. Am. Chem. Soc. 1941, 63,
33. Bureeva, S.; Andia-Pravdivy, J.; Bichucher, A.; Oris-
hchenko, D.; Kaplun, A. Mendeleev Commun. 2005, 15,
253–256.
34. Takada, A.; Shirahama, S. Immunopharmacology 1985, 9,
87–95.
35. Anderson, B.E.; Fryer, J.P. U.S. Patent Number
5,977,076, 1999.
36. Szuster-Ciesielska, A.; Kandefer-Szerszen, M. Pharmacol.
Rep. 2005, 57, 588–595.
37. Hussain, N.; Jaitley, V.; Florence, A. T. Adv. Drug Deliv.
Rev. 2001, 50, 107–142.
38. Son, L. B.; Kaplun, A. P.; Spilevskii, A. A.;
Andiia-Pravdivyi, I. u. E.; Alekseeva, S. G.; Gribor’ev,
V. B.; Shvets, V. I. Bioorg. Khim 1998, 24,
787–793.
1259–1261.
7. Kozlov, L. V.; Sizoi, M. N.; Zinchenko, A. A.; Levkovskii,
A. V. Biokhimiia 1986, 51, 707–718.
8. Ziccardi, R. J.; Tschopp, J. Biochem. Biophys. Res.
Commun. 1982, 107, 618–623.
9. Roos, A.; Ramwadhdoebe, T. H.; Nauta, A. J.; Hack, C.
E.; Daha, M. R. Immunobiology 2002, 205, 595–609.
0. Jiang, H.; Robey, F. A.; Gewurz, H. J. Exp. Med. 1992,
175, 1373–1379.
1. Griselli, M.; Herbert, J.; Hutchinson, W. L.; Taylor, K.
M.; Sohail, M.; Krausz, T.; Pepys, M. B. J. Exp. Med.
1999, 190, 1733–1740.
3
2. Jiang, H.; Burdick, D.; Glabe, C. G.; Cotman, C. W.;
Tenner, A. J. J. Immunol. 1994, 152, 5050–5059.