J . Med. Chem. 1993,36, 778-783
778
Neuraminidase-Resistant Hemagglutination Inhibitors: Acrylamide Copolymers Containing a C-Glycoside of N-Acetylneuraminic Acid] Michelle A. Sparks,2 Kevin W. Williams,3 and George M. Whitesides' Department of Chemistry, Harvard University, Cambridge, Massachusetts 02138 Received September 14, 1992
Copolymers consisting of a polyacrylamide backbone with side chains terminated in C-glycosidic analogs of N-acetylneuraminic acid were synthesized by free radical copolymerization of a-2-C[3-[[2-(N-acryloylamino)ethyl]thio] propyl] -N-acetylneuraminic acid (5) with acrylamide. Unlike natural and synthetic polyvalent materials that contain N-acetylneuraminic acid in 0-glycosidic form, these C-glycosidic copolymers resist neuraminidase-catalyzed cleavage of the neuraminic acid residue from the copolymer backbone. Examination of these C-glycosidic copolymers in a hemagglutination inhibition assay indicated that they are as effective in vitro as polyvalent 0-glycosidic copolymers in inhibiting agglutination of erythrocytes by influenza virus. The minimum value of the inhibition constant, calculated on the basis of the concentration of Neu5Ac groups in solution, is IQHA1) lo-7 M for both copolymers. The inhibitory potency of the C-glycosidebased copolymers becomes more significant at lower concentrations of Neu5Ac moieties in solution than does the inhibitory potency of the 0-glycoside-based copolymer.
-
the monomeric 2-C- [3-[ [2-(N-acryloylamino)ethyllthiolPolyvalency-the ability of a ligand to bind to a target propyl] -N-acetylneuraminic acidz1 (5, Scheme I). via multiple chemical interactions-is considered an The copolymers of 5 and acrylamide-poly(5-coimportant factor in many pathogen-host cell interactions acrylamidel-were prepared with a range of mole fractions that result in infe~tion.~ Influenza A, an orthomyxovirus of 5 attached to an acrylamide backbone. We have responsible for the most severe outbreaks of influenza, measured the ability of these copolymers to inhibit the adheres to the terminal N-acetylneuraminicacid (Neu5Ac) agglutinationof chicken erythrocytesinduced by influenza of glycolipids and glycoproteins on the surface of mamvirus (X-31)22in vitro by employing a standard hemagmalian epithelial cells via the viral surface lectin, hemaginhibition (HAI) assay.23 These copolymers glutinin (HA).5 Monomeric a - 2 - m e t h y l - N - a c e ~ ~ e ~ ~ i nglutination have the potential to compete with polyvalent cell surfaces ic acid (1, Neu5Aca2Me) binds weakly to HA.6 A variety for HA (and perhaps NA) binding sites on the viral surface of structurally modified, monomeric Neu5Ac derivatives and to inhibit virus-host cell binding, without being have been tested as inhibitors of hemaggl~tination;I-~ only susceptible to cleavage and release of the Neu5Ac groups recently have monomeric derivatives been prepared that by NA.17 are significantly more effective than Neu5Aca2Me, The synthesisof 5 began withglycosyl chloride !Az4 This The detailed mechanism of inhibition by these compounds compound was converted to an anomeric mixture of remains unclear. Neu5Ac C-glycosides 3 (a-2-C-allylNeu5Ac) and 4 (0-2An alternative approach to the prevention of infection C-allylNeu5Ac)under the allylation conditions developed is the development of polyvalent drugs that could, in by PaulsenZ5and BednarskiSz6Separation of 3 from 4,253 principle, bind to and block access to receptor sites of the followed by photolytic addition of 2-aminoethanethiol pathogen.1° A number of potent, naturally-occurring, hydrochloride to the double bond terminus of 3,14927was polyvalent compounds that inhibit the adhesion of ina convenientmethod for the installation of a linkage group fluenza virus to cells in vitro are kn0wn.l' In addition, with a terminal primary amine. This amine allowed for we12J3 and others14-16have developed inhibitors of influattachment of the acrylamide moiety needed for copoenza virus-induced hemagglutination that are polyvalent lymerization. Addition of the 2-aminoethanethiol adduct in Neu5Ac. In vitro, these polyvalent inhibitors prevent to a solution of N-(acryloy1oxy)succinimide in basic hemagglutination based on virus-cell adhesion at conMeOH/H20 (pH 10) produced the desired Neu5Accentrations of Neu5Ac groups between 10-6and lPM.12-16 acrylamide 5 in 85% purified yield. All Neu5Ac-based inhibitors, whether natural or synthetic The copolymers were synthesized by photochemically in origin, are susceptible to glycosidic cleavage by initiated copolymerization of acrylamide and 5. In all neuraminidase (NA, EC 3.2.1.l8),l7-l9a glycosidase that preparations, the initial concentration of acrylamide was is present on the surface of the influenza virus. Neuraminifixed at 1.0 M. The initial concentration of 5 was varied dase-catalyzed cleavage of the Neu5Ac groups present in from 4.0 mM to 4.0 M; this range corresponded to a mole the oligosaccharide moieties of glycoproteins is thought fraction of 5 (XNeu5Ac) in totalacrylamide moieties of 0.0050.80 (eq 1). We assume this mole fraction is preserved in to allow the virus particle to escape from naturally the final copolymer. occurring polyvalent inhibitors17-19 and aid in the release of newly formed virion from the infected 151 In order to circumvent the neuraminidase-catalyzed XNeu5Ac - [5] [acrylamidel cleavage of terminal Neu5Ac residues from synthetic, polyvalent copolymers presenting Neu5Ac groups as side These copolymers have not been fully characterized. c h a i n ~ , ~ we ~ J have ~ J ~ prepared random copolymers inPrevious experience in the preparation of acrylamide corporating C-glycosidic NeuBAc moieties derived from l.899
+
QO22-2623/93/1836-Q778$Q4.Q0/ Q
1993 American Chemical Society
Neuraminidase-Resistant Hemagglutination Inhibitors
Journal of Medicinal Chemistry, 1993, Vol. 36, No. 6 779
Scheme I. Synthesis of Copolymer 5 and Acrylamide
k# NeuSAc, 1 a OAc OAc OAc
9 0
OH
2
HlNCO
Id NeuSAc
NeuSAc
a. Procedure of Paulsen and Bednarski, ref 25,26; b. Aminoethanethiol hydrochloride, AICV,hv; c. N-acroyloxysuccinimide,Etfl, MeOH; d. Acrylamide, AICV,H20,hv.
copolymers of the parent 0-glycosides of Neu5Ac has, however, shown that copolymers produced by an analogous procedure are retained by a 100 000 MW cutoff dialysis membrane.12 Unlike their 0-glycosidic counterparts, the C-glycoside copolymers of Neu5Ac are resistant to neuraminidase. Incubation of a sample of C-glycoside copolymer (XNeuBAc = 0.20) with neuraminidase prior to an HA1 assay did not result in a decrease in inhibitory potency when compared to the HA1 results of the untreated copolymer sample. Under analogous conditions, the 0-glycosidic copolymer lost its inhibitory activity.lg Aliquots of the monomers 3 and 5, and of each poly(5co-acrylamide) (XNeuBAc = 0.005-0.8), were prepared for testing in an HA1 assay by diluting the potential inhibitor in sterile phosphate buffered saline solution (PBS)and adjusting to pH 7.2. The solutions were employed in a standard HA1 assay22y23 to evaluate the potential of these compounds to inhibit agglutination of chicken erythrocytes induced by influenza virus. This assay involves adding virus to a solution of serially diluted inhibitor in a 96-well titer plate, incubating the mixture for 30 min, adding a suspension of erythrocytes, and determining the lowest concentration of NeuBAc groups in solution for which hemagglutination is no longer observed under the specified assay conditions. The endpoint of the HA1 assay, Ki(HA1), is defined as this lowest concentration. This assay depends on a number of variables, not all of which are under complete control (especially the quality of the erythrocytes). The assay is reasonably reproducible (fl well) provided appropriate controls are followed. The relevance of this in vitro assay to in vivo infectivity remains to be established. Figure 1and Table I summarize the results of the HA1 assay obtained with the poly(5-co-acrylamide)copolymers containing C-glycosideNeu5Ac groups and compares these new results with those obtained with the copolymers
poly(5-co-acrylamide) PN~USAC liposomes" D-equine aZM" o-guinea
L NeuSAea2Me
102
pig aZM"
0
02
0.4
0.6
monovalent NeuSAc analogs6.9
08
1
v
h.
Neu5Ac
Figure 1. Inhibitory potency, IQHA1), of the C- and 0-glycosidic copolymers of Neu5Ac as a function of mole fraction, X N ~ ~ of~ A Neu5Ac covalently incorporated into an acrylamide copolymer. The data points presented for the C- and 0-glycosidiccopolymers are averagesof four to six seta of data obtained fromindependently prepared samples of copolymer. The data for the 0-glycosidic copolymers have been previouslypublished from this laboratory.12 K , . ( H M )is the minimum concentrationof copolymer-linkedNeu5Ac residues in solution that inhibits agglutination of chicken erythrocytes in vitro at 4 "C. Assay conditions are given in the Experimental Section. The horizontal line denotes the threshold antiagglutinationinhibitorypotency of monomericNeu5Aca2Me 1 (Ki(HAI) = 2.8 X M)6 assayed under identical conditions. The error bars represent *l well in the HA1 assay for (m) C-glycosidiccopolymer of NeuSAc, and (0) 0-glycosidiccopolymers of Neu5Ac. containing 0-glycoside Neu5Ac groups, reported previously.12 The data are expressed in terms of Ki(HA*) versus XNeuSAc. For both sets of copolymers, the antihemagglutination potency (measured by Ki(HA1) of the C- and 0-glycosidic copolymers of Neu5Ac) reaches a maximum value a t XNeu5ACm" = 0.3-0.4. We have attributed the bell shape of these curves to a combination of entropy and optimal glycoside spacing effects.12 The horizontal line in Figure 1represents Ki(HA1)for Neu5Aca2Me 1 (2.8 X
~ ,
780 Journal of Medicinal Chemistry, 1993, Vol. 36,No.6 Table I. AntihemagglutinationInhibitory Potency XN&Acrn" and XNeuSAc for the C-and O-Glycoside Acrylamide Copolymers of Neu5Ac copolymer KiHA1', PM XNeu5Acm" XNeu5Aco'5 C-glycosideNeu5Ac 0.4 0.3-0.4 0.01 O-glycoside Neu5Ac 0.2 0.3-0.4 0.09 XNeu5.4cmaX is the mole fraction of copolymer-linked Neu5Ac that yields the maximum antiagglutination potency. XNeu5.4c0'5 is the minimum mole fraction of copolymer-linkedNeu5Ac required to exert half of the maximum inhibitory potency above the threshold inhibitory level given by monomericNeu5Aca2Me in solution(KiHA1) = 2.8 x 10-3 M).
MI6 and is shown as a reference for the K P I ) of monomeric analogs of Neu5Ac. Figure 1permits two important comparisons between the C- and O-glycosidic copolymers. First, both copolymers have similar maximum inhibitory potency in vitro. Four identical, independently prepared poly(5-co-acrylamide) polymers having XNeu5Ac = 0.4 showed inhibition =2X to 6 X M of hemagglutination with in Neu5Ac groups. A parallel set of experiments conducted with copolymers containing O-glycosidic Neu5Ac gave WHA1) of 2 X M in NeuBAc groups.12 The to 3 X slightly weaker inhibition of hemagglutination for the C-glycosidic copolymer (at XNeu5Ac > 0.1) relative to the O-glycosidic copolymer suggested by Figure 1 is within the margin of experimental error in the hemagglutination assay (fl well). Second, the minimum mole fraction that is required to exert half of the maximum inhibitory potency = 2.5 X above the threshold inhibitory level (i.e. Ki(HA1) M for Neu5Aca2Me, :)1 X N ~ ~ ~ A is C "shifted ~, from 0.09 for the O-glycosidic copolymer to 0.01 for the C-glycosidiccopolymer. Together with XNeu5AcmaJ,the mole fraction of copolymer-linked NeuEiAc that yields the ''~ maximum antiagglutination potency, the X N ~ ~ S A Cindex provides a convenient means to compare the effectiveness of different types of copolymers containing neuraminic acid (Table I). The monomeric a-C-glycosides 3 and 5 were also tested in an HA1assay. For these monomers, Ki(HA1) fell between 2.5 and 5 mM; these values are similar to the values of Ki(HA1) obtained for most monomeric NeuBAc analogs.The @-analogs4, and the analogous Neu5Ac-acrylamide 5 that was prepared from 4, showed no inhibition at concentrations
