(Received for publication, May 22, 1996, and in revised form, December 26, 1996)
From the Laboratoire d'Enzymologie, INSERM
Unité 392, Université Louis Pasteur de Strasbourg,
F-67400 Illkirch, France and § Unité de Pharmacologie
Cellulaire, Unité Associée I P/INSERM U 285, Institut
Pasteur, F-75015 Paris, France
Suramin, a hexasulfonated naphtylurea recently
used as an anti-tumor drug, is a potent inhibitor of human neutrophil
elastase, cathepsin G, and proteinase 3. The complexes it forms with
these enzymes are partially active on synthetic substrates, but full inhibition takes place when elastase activity is measured with fibrous
elastin or when cathepsin G activity is measured using platelet
aggregation. One molecule of elastase binds four molecules of suramin
with a Ki of 2 × 107
M as determined by enzyme inhibition or intrinsic
fluorescence enhancement of suramin. The binding curves show no sign of
cooperativity or anticooperativity. The Ki for the
complexes with cathepsin G and proteinase 3 are 8 × 10
8 and 5 × 10
7 M,
respectively. Ionic strength increases the Ki of the elastase-suramin complex in a way that suggests that four of the
six sulfonate groups of suramin form ionic interactions with basic
residues of the enzyme and that at saturation almost all arginines of
elastase form salt bridges with suramin. The neutrophil
proteinase-inhibitory activity of suramin might be used to prevent
tissue destruction and thrombus formation in diseases where massive
infiltration and activation of neutrophils take place.
Suramin, a hexasulfonated naphtylurea with a symmetrical structure (e.g. Ref. 1), was synthesized 75 years ago and has been used in the treatment of trypanosomiasis and onchocerciasis (2). In the last decade, many new and therapeutically significant properties of this compound were described (for a review, see Ref. 3). It has thus been tried, although unsuccessfully, in patients with acquired immunodeficiency syndrome (4). Suramin has, however, proven to be useful as an anti-tumor drug in patients with corticosurrenal carcinoma (5), metastatic adrenal cancer (6), carcinomas of the kidneys (7), and carcinomas of the prostate gland (8).
Suramin has multiple modes of action. Although the mechanism of its trypanocide action is not well understood, it is probably based on the polyanion's ability to inhibit glycolytic enzymes of the parasite, thus interfering with its metabolism. Its use in acquired immunodeficiency syndrome was based on its potential inhibition of retroviral reverse transcriptase (9) and on its ability to decrease the binding of human immunodeficiency virus to T cells (10). The polysulfate is thought to exert its anti-tumoral action by impairing the binding of growth factors to the surface receptors of malignant cells (11, 12) and by inhibiting angiogenesis (13). It may also have anti-invasive properties, due to the inhibition of heparanases (1).
Here we report on a new property of suramin: its inhibitory action on three neutrophil serine proteinases, namely NE,1 cat G, and PR3, which are thought to play a pathogenic function in pulmonary emphysema, cystic fibrosis, and rheumatoid arthritis (14) and whose inhibition is therefore of therapeutic importance.
NE and cat G were isolated from purulent sputum and active
site-titrated as described by Boudier and Bieth (15). PR3 was obtained
from Merck through the courtesy of Dr. P. Davies and Dr. J. L. Humes
and active site-titrated with titrated 1PI (16). Bovine
pancreatic trypsin (Choay, France) and chymotrypsin (Boehringer, Germany), titrated as described previously (15), were found to be 60 and 85% active, respectively, and were used without further purification. Suramin (hexasodium
sym-bis[m-aminobenzoyl-m-amino-p-methylbenzoyl-1-naphtylamino-4,6,8-trisulfonate] carbamide) was a gift from Dr. Möller and Dr. Mardin (Bayer, Germany). The chemical structure of this compound may be found in Ref
1. [14C]Serotonin was from Amersham (United Kingdom).
Remazol brilliant blue elastin was from Elastin Products Company
(Owensville, MO). The p-nitroanilide and thiobenzylester
substrates came from Bachem (Switzerland) and Enzyme Systems Products
(Dublin, CA), respectively. Fibrinogen, cytochalasin B, prostacyclin,
and 4,4
-dithiodipyridine were from Sigma.
Kinetics were monitored with a thermostated UVICON 941 spectrophotometer equipped with a PC microcomputer. The progress curves were recorded for 0.3-3 min, depending upon the reaction velocity.
The computer checked the linearity of the progress curves using linear regression analysis and calculated the initial reaction rates. Due to the great sensitivity of the spectrophotometer, less than 2% of the substrate was hydrolyzed during the rate measurements. The hydrolysis of p-nitroanilides was recorded at 410 nm, whereas that of the thiobenzylester was monitored at 324 nm in the presence of dithiodipyridine (17). All substrates except Suc-Ala3-pNA were dissolved in dimethylformamide whose final concentration in the reaction medium was 2% (v/v). Suc-Ala3-pNA was dissolved in N-methylpyrrolidone whose final concentration was 2.5% (v/v) during the measurement of Ki as a function of ionic strength and 1% (v/v) in all other experiments. The buffer was 50 mM Hepes, 100 mM NaCl, pH 7.4, unless otherwise stated. To determine the kinetic constants kcat and Km, initial rates were measured as a function of substrate concentration and the data were fitted to the Michaelis-Menten rate equation.
Nonlinear regression analysis (Enzfitter software) was used to calculate the best estimates of the enzymatic and fluorescence titration parameters.
ElastolysisThe effect of suramin on the elastolytic activity of NE was assessed as follows. Constant concentrations of NE were reacted with increasing concentrations of suramin in a total volume of 250 µl of 50 mM Hepes, 100 mM NaCl, pH 7.4. After 4 min, 200-µl aliquots were withdrawn from these mixtures and added to 2.3 ml of continuously stirred remazol brilliant blue elastin suspensions. After 15 and 30 min at 25 °C, 450-µl aliquots were withdrawn from these suspensions, diluted with 550 µl of ice-cold 0.75 M acetate buffer, pH 4.5, and centrifuged at 14,000 × g during 5 min. The absorbances of the supernatants were read at 595 nm to assay the concentration of soluble elastin peptides. The final concentrations of NE and remazol brilliant blue elastin were 2.5 µM and 3 mg/ml, respectively.
Preparation of Washed Human PlateletsHuman blood from
healthy volunteers was collected over citrate-phosphate-dextrose. The
platelet-rich plasma obtained after blood centrifugation (180 × g, 20 min, 37 °C) was incubated with 1 µM
[14C]serotonin (5 × 102 mCi/ml) for
30 min. Platelets were then isolated by successive centrifugations in
the presence of prostacyclin (0.5 µM) and finally resuspended at 4 × 108 cells/ml in Tyrode's buffer
composed of 137 mM NaCl, 2.68 mM KCl, 11.9 mM NaHCO3, 0.42 mM
NaH2PO4, 2 mM CaCl2, 1 mM MgCl2, 5.5 mM glucose, 5 mM Hepes, 0.35% BSA, pH = 7.4.
Two volumes of blood, collected as for platelet preparation, were mixed with one volume of 3% (w/v) dextran in saline and allowed to sediment for 30 min. This step was followed by Ficoll-Hypaque gradient and hypotonic lysis of erythrocytes. After they had been washed, the neutrophils were resuspended in a final volume of Hanks' balanced salt solution such that the cell concentration was 107 cells/ml. The viability of recovered neutrophils was 98.7 ± 2.5%, as measured by Trypan blue dye exclusion. Their purity, evaluated using Türk's stain, was 96.2 ± 2.1%.
Effect of Suramin on Platelet ActivationPlatelets were reacted with cat G or with formyl-Met-Leu-Phe-treated neutrophils in the presence or absence of suramin, and their activation was measured by their ability to aggregate and to release serotonin. A siliconized cell containing 250 µl of platelet suspension (108 cells), 250 µl of Hank's solution, and 0.7 mg/ml fibrinogen was placed in a Dual Aggro-meter (Chrono-Log Corp., Hevertown, PA) and preincubated at 37 °C for 90 s under continuous stirring (1,100 rpm). Suramin was then mixed with the suspension. Following a further 30 s, 0.25 µM cat G was added, and the resulting aggregation was monitored for 3 min by changes in light transmission. The reactions were terminated by the addition of 125 µl of a "stopping solution" (2.05% formaldehyde, 38.5 mM EDTA, 68.8 mM NaCl) immediately followed by centrifugation (13,000 × g, 3 min). The release of [14C]serotonin by platelets, measured on 400 µl of supernatants, was determined by scintillation counting (Counter 1212 Rackbeta; LKB, Wallac, Stockholm, Sweden). Controls without suramin or cat G were run in an identical way. Aggregation was expressed as a percentage of the maximal light transmission, and serotonin release was expressed as a percentage of the total [14C]serotonin content of the platelets. For the neutrophil-platelet interaction study, the Hanks' solution was replaced by neutrophils (5 × 106 cells/ml final concentration), and cells were preincubated for 5 min at 37 °C in the presence of fibrinogen and 5 µg/ml cytochalasin B. Suramin was then added. Following a further 30 s, aggregation was started with 0.5 µM formyl-Met-Leu-Phe.
Fluorescence SpectroscopyFluorescence emission spectra were recorded with a Shimadzu RF-5000 spectrofluorimeter equipped with a thermostated cell holder. The effect of increasing concentrations of NE on the fluorescence intensity of 1 µM suramin was measured at pH 7.4 (50 mM Hepes, 100 mM NaCl) and 25 °C using excitation and emission wavelengths of 315 and 401 nm, respectively.
Fig.
1A shows the effect of increasing
concentrations of suramin on the kinetics of
Suc-Ala3-pNA hydrolysis by NE. While suramin depresses the rate in a concentration-dependent manner, it
does not fully inhibit the enzyme, since curves 7 and 8 are barely distinguishable from each other. All progress curves were linear and
thus allowed initial rates to be calculated. Linear progress curves
were also observed for the four other proteinases. The initial rates
derived from these curves were expressed as relative rates so that they
could be compared on the same graph. Fig. 1B shows that the
activities of NE, cat G, and PR3 sharply decrease with the
concentration of suramin and then level off. With large excesses of
suramin, the residual activities of PR3, cat G, and NE were about 60, 20, and 10%, respectively. In contrast, trypsin and chymotrypsin were
insensitive to suramin when used at nanomolar concentrations. To favor
suramin binding, we used 103-fold higher concentrations of
trypsin and chymotrypsin. In turn, poor substrates were used to avoid
substrate depletion during the rate
measurements.2 Despite these conditions,
trypsin activity was only slightly affected, and chymotrypsin was not
inhibited at all.
Incomplete inhibition implies that the enzyme-inhibitor complex EI is able to bind and turn over the substrate S (or that ES is able to bind I). This special case of noncompetitive inhibition may be analyzed using Scheme I (18),
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
Scheme I leads to a simple (19) or a complex (20) steady-state rate
equation depending upon whether the inhibition is "classical"
(total I = free I) or "tight binding" (total I = free I + bound I). Classical inhibition does not depend upon the total enzyme
concentration, whereas tight binding inhibition does (21). We have
therefore measured the inhibition of NE and cat G by suramin using
different enzyme concentrations. Fig. 2 shows that the
inhibition of NE by suramin decreases with the enzyme concentration;
IC50 = 0.5, 0.8, and 1.47 µM for
[E]o = 0.1, 0.3, and 0.6, respectively. This
diagnoses tight binding inhibition (21). On the other hand, at high
enzyme concentration a tight binding inhibitor titrates the enzyme
(21). Indeed at [E]o = 2.5 µM, the
inhibition was linear up to 80% inhibition, and the extrapolated curve
intercepted the abscissa at [I]o = 9.4 µM,
indicating that about 4 mol of suramin are required to inhibit 1 mol of
NE, i.e. n = 4 (see inset to Fig. 2).
Similar experiments demonstrated tight binding inhibition of cat G by suramin and a 3:1 suramin:cat G binding stoichiometry (data not shown).
To calculate the equilibrium dissociation constants Ki of the enzyme-suramin complexes we therefore used the complex steady-state rate equation derived by Szedlacsek et al. (20), which takes into account both the tight binding character and the incompleteness of inhibition,
![]() |
(Eq. 1) |
![]() |
(Eq. 2) |
![]() |
(Eq. 3) |
|
Table I summarizes the kinetic parameters describing the inhibition of
leukoproteinases by suramin. It can be seen that partial inhibition
occurs because suramin decreases kcat ( < 1 but
0). PR3, whose complex with suramin has the highest
residual activity (Fig. 1), has also the highest
value. On the
other hand, suramin has comparable affinities for the free
enzymes. The substrates significantly lower these affinities
(
> 1).
The concentration of NaCl was varied
from 0.2 to 0.5 M so that we could study the effect of
Na+ counterions on the association of suramin to NE. For
each NaCl concentration we first determined Km,
kcat, , and
to calculate
Ki using Equation 1. Ionic strength was found to
affect the inhibition of NE by suramin in several ways. First,
decreases by a factor of about 2, while
increases about 5-fold when
[NaCl] rises from 0.2 to 0.5 M; ionic strength thus tends
to enhance the residual activity of the NE·suramin complex. Second,
Ki rises by a factor of 67 when [NaCl] increases from 0.2 to 0.5 M. The latter effect was assumed to be due
to the electrostatic effect of Na+ counterions on the ionic
interactions involved in the suramin·NE association. If we consider
the binding of suramin to NE as the substitution of the Na+
counterions of the polyanion by the positively charged amino acid
residues of the protein, we may describe the interaction by the
following equilibrium.
![]() |
![]() |
![]() |
(Eq. 4) |
Free suramin has a very
low intensity of fluorescence when excited at 315 nm. Upon the addition
of NE, this intensity increases very strongly (see inset to
Fig. 3). NE, however, does not change the maximum
emission wavelength of suramin ( = 401 nm). Use was made of this
fluorescence enhancement to measure the NE·suramin affinity in a
nonenzymatic way. Increasing concentrations of NE were added to
constant concentrations of suramin, and the fluorescence intensity of
each mixture was measured at
ex = 315 nm, and
em = 401 nm. Ki was calculated by a
nonlinear least squares fit of the data to the following
relationship,
![]() |
(Eq. 5) |
Ki was found to be (2.1 ± 0.4) × 107
M, which, within the limits of experimental error, is
identical to the enzymatically determined constant.
Fibrous elastin is the best known biological substrate of NE
(14). Fig. 4 shows that suramin is a very potent
inhibitor of the NE-catalyzed elastin solubilization. Most of the
inhibition curve is linear, confirming the tight binding inhibition
diagnosed with the synthetic substrate. Extrapolation of the linear
part of this curve to the abscissa indicates that ~10
µM suramin is required to inhibit 2.5 µM
NE, thus confirming the 4:1 binding stoichiometry evidenced with the
synthetic substrate. Interestingly, the NE·suramin complex has almost
no residual activity on elastin in contrast with what has been observed
with the synthetic substrate.
Enzymatically active cat G is able to activate platelets through
proteolytic cleavage (25, 26). We made use of this property to study
the effect of suramin on the biological activity of cat G. Constant
amounts of platelets were reacted with increasing amounts of suramin,
after which constant amounts of cat G were added. Fig. 5
shows that suramin inhibits both platelet aggregation and serotonin
release in a concentration-dependent manner. The concentration of suramin required to get 50% platelet activation (4 µM) is 3.3-fold higher than the IC50 observed
with the cat G/Suc-Ala2-Phe-pNA system (Fig. 1)
despite the fact that similar enzyme concentrations were used in the
two experiments. This might be due to nonspecific binding of part of
suramin by fibrinogen and the platelet surface. Alternatively, the
peptide cleaved by cat G on platelets might have a much higher affinity
for the enzyme than Suc-Ala2-Phe-pNA so that
higher suramin concentrations might have been required to overcome this
competition effect.
To investigate a more physiologically relevant system, we have also studied the effect of suramin on the platelet activation mediated by formyl-Met-Leu-Phe-stimulated neutrophils (see "Experimental Procedures"). Suramin was again able to prevent platelet aggregation and serotonin release in this biological system (data not shown).
Suramin has been shown to inhibit a great number of parasite and
human enzymes including trypanosoma glycolytic enzymes (3), human DNA
and RNA polymerases (25), human reverse transcriptase (9), ATPase, and
protein kinase C (5). It also inhibits human lysosomal enzymes such as
-glucuronidase (1), but its action on lysosomal proteinases has not
heretofore been described. This paper reports its inhibitory activity
on NE, cat G, and PR3, the lysosomal proteinases of human
neutrophils.
The inhibition of the leukoproteinases by suramin is incomplete,
i.e. even with a large excess of suramin there is a
substantial rate of substrate hydrolysis (e.g. Fig. 1). This
indicates that the proteinase-suramin complexes behave like enzymes
that are able to bind substrate and to hydrolyze it. The inhibition
data could be satisfactorily fitted to Equation 1, indicating that they
are described by Scheme I. Thus, the inhibition of the leukoproteinases by suramin depends upon four parameters: (i) n, the number
of binding sites for suramin on the proteinase; (ii)
Ki, the equilibrium dissociation constant of the
proteinase-suramin complex; (iii) , the factor that affects the
binding of substrate to the enzyme-suramin complex or that of suramin
to the enzyme-substrate complex; and (iv)
, the factor that affects
kcat. Theoretically, incomplete inhibition may
be observed if 0 <
1 and 1
<
(see also
Equation 3). In our investigation, one of the two limiting cases (
= 1 or
= 1) was observed for the cat
G/suramin/Suc-Ala2-Phe-pNA system for which
= 1. For all other systems, incomplete inhibition resulted from the
combined effects of
> 1 and
< 1 (Table I).
The magnitude of and
for the cat G-suramin system depends upon
the substrate used (Table I). The same is probably true for the
NE-suramin pair for which v
(a function of
and
; see Equation 3) is about 10% of vo
with the synthetic substrate (Fig. 2), whereas it is virtually 0 with
the substrate elastin (Fig. 4). Also, cat G-induced platelet
aggregation, a proteolytic event (25, 26), is fully inhibited by
suramin (Fig. 5). The observation that full inhibition is observed when large protein substrates are used suggests either that these substrates are unable to bind EI so that the ESI complexes
do not form (
=
) or that the ESI complexes have
little or no enzymatic activities (
0). With protein substrates,
the inhibition of proteolysis by suramin may therefore be simply
described by n and Ki. These parameters
are substrate-independent at least for NE, since titration of this
enzyme with suramin in the presence of a synthetic substrate (Fig. 2)
and fluorescence titration of suramin with NE (Fig. 3) yield similar
n and Ki values.
The effect of suramin on NE and cat G is reminiscent of that observed
with heparin, another polysulfated molecule that is also a partial
tight binding inhibitor of these enzymes (27, 28). The two polyanions
mainly differ in their ability to depress the enzymatic activities;
while the NE·heparin complex has a residual activity of 30 and 20%
on synthetic substrates and on elastin, respectively, the NE·suramin
complex is only 10% active on a synthetic substrate and is quite
inactive on elastin. Similar differences are also observed with cat G. DNA is another polyanion that reacts with NE. At physiological ionic
strength, unfractionated DNA yield 25% inhibition of NE activity (29),
whereas in vitro selected DNA sequences with high affinity
for NE have no significant inhibitory effect on the enzyme (30). Tight
enzyme-ligand binding with lack of inhibition might be explained by
assuming = 1 and
= 1, for instance (see Equation 3). DNA and
heparin are therefore less efficient elastase inhibitors than
suramin.
The four suramin binding sites of NE are probably equivalent and independent, since the saturation curves show no sign of cooperativity or anticooperativity whether NE is titrated with suramin (Fig. 1 and 2) or suramin is titrated with NE (Fig. 3). The linear titration curve obtained with elastin as a substrate (Fig. 4) also confirms that the binding sites are equivalent and independent.
There is a good relationship between the basicity of the proteinases and their affinity for suramin. (i) PR3 whose pI is 9.1 has a significantly lower affinity for suramin than NE or cat G whose pI is greater than 9.5 (31). (ii) Trypsin and chymotrypsin, which are much less basic proteins, are virtually not inhibited by suramin. This indicates that the enzyme-suramin binding involves electrostatic interactions between the negatively charged sulfonate groups of the polyanion and the positively charged basic amino acid residues of the proteinases. It is worthwhile noticing that the binding of suramin to trypanosoma glycolytic enzymes is also governed by ionic interactions (3, 24). The electrostatic nature of the suramin·NE binding is confirmed by the deleterious effect of ionic strength on Ki, which suggests that an average of four ionic interactions take place per molecule of suramin bound, i.e. that four out of six sulfonate groups participate in binding. NE lacks lysine and has 19 arginine residues, 17 of which are fully exposed to solvent (32). Since four molecules of suramin are bound per molecule of NE, it can be concluded that suramin neutralizes 16 of these guanidino residues.
NE, cat G, and PR3 may cause tissue destruction in lung emphysema, cystic fibrosis, and rheumatoid arthritis because they cleave extracellular matrix proteins (14, 31). Sulfated glycosaminoglycans prevent NE-induced lung emphysema in animal models (31, 33). These ligands are partial inhibitors of the elastolytic activity of NE (27, 34, 35). Suramin, which fully inhibits elastolysis and has already been used in human cancer therapy (5-8) should therefore prove to be an efficient drug in diseases characterized by massive neutrophil recruitment and activation. In addition, as an inhibitor of cat G-induced platelet aggregation, this drug should also be a powerful antithrombotic agent.
We thank Dr. P. Davies and Dr. J. L. Humes from Merck for the gift of PR3 and Dr. Möller and Dr. Mardin from Bayer for the gift of suramin.